DOING THINGS WITH THINGS

This page intentionally left blank

Doing Things with Things
The Design and Use of Everyday Objects

Edited by

ALAN COSTALL
University of Portsmouth, UK

and

OLE DREIER
University of Copenhagen, Denmark
© Alan Costall and Ole Dreier 2006

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording
or otherwise without the prior permission of the publisher.

Alan Costall and Ole Dreier have asserted their moral right under the Copyright, Designs and
Patents Act, 1988, to be identified as the editors of this work.

Published by
Ashgate Publishing Limited Ashgate Publishing Company
Gower House Suite 420
Croft Road 101 Cherry Street
Aldershot Burlington, VT 05401-4405
Hampshire GU11 3HR USA
England

Ashgate website: http://www.ashgate.com

British Library Cataloguing in Publication Data

Doing things with things: the design and use of everyday
objects – (Ethnoscapes)
1. Material culture 2. Technology – Social aspects
3. Knowledge, Theory of 4. Knowledge, Theory of, in children
1. Costall, Alan II. Dreier, Ole
306.4'6

ISBN-10: 0 7546 4656 4

ISBN-13: 978-0-7546-4656-3

Library of Congress Cataloging-in-Publication Data

Doing things with things: the design and use of everyday objects / edited by
Alan Costall and Ole Dreier.
p. cm. -- (Ethnoscapes)
Includes index.
ISBN: 0-7546-4656-4
1. Material culture--Philosophy. I. Costall, Alan. II. Dreier, Ole. III.
Series: Ethnoscapes (Unnumbered)

GN406.D65 2006
302--dc22
2006007245

Printed and bound in Great Britain by Antony Rowe Ltd, Chippenham, Wiltshire.
 Contents

List of Contributors vii

List of Figures and Tables ix

Introduction 1
Alan Costall and Ole Dreier

Part I Intentionality and the Functionality of Things

1. The Case of the Recalcitrant Prototype 15

Beth Preston

2. Use Plans and Artefact Functions: An Intentionalist Approach to

Artefacts and their Use 29
Pieter E. Vermaas and Wybo Houkes

Part II Things in the World of the Child

3. Autism and Object Use: The Mutuality of the Social and Material
in Children’s Developing Understanding and Use of Everyday Objects 51
Emma Williams and Linda Kendell-Scott

4. Object Use in Pretend Play: Symbolic or Functional? 67

Ágnes Szokolszky

5. Culture, Language and Canonicality: Differences in the Use of

Containers between Zapotec (Mexican Indigenous) and Danish
Children 87
Kristine Jensen de López

Part III Transformation and Things

6. The Cognitive Biographies of Things 113

David de Léon
vi Doing Things with Things
7. The Woman who used her Walking Stick as a Telephone:
The Use of Utilities in Praxis 131
Hysse Birgitte Forchhammer

8. Politics of Things: The Interplay of Design and Practice in a Design

Workshop with Children 147
Estrid Sørensen

Part IV Organisation and Things

9. Working with Material Things: From Essentialism to

Material–Semiotic Analysis of Sociotechnical Practice 167
Finn Olesen and Randi Markussen

10. Words and Things: Discursive and Non-Discursive Ordering in a

Networked Organisation 193
Steven D. Brown and David Middleton

11. Learning to Do Things with Things: Apprenticeship Learning in

Bakery as Economy and Social Practice 209
Klaus Nielsen

12. Urban Makings and the Formalisation of Informal Settlements 225

Gustavo Ribeiro

Name Index 239

Subject Index 241
 List of Contributors

Steven D. Brown, Department of Human Sciences, Loughborough University,

England. s.d.brown@lboro.ac.uk

Alan Costall, Department of Psychology, University of Portsmouth, England. alan.

costall@port.ac.uk

David de Léon, Cognitive Science, Lund University, Lund, Sweden. david.deleon@

sonyericsson.com

Ole Dreier, Department of Psychology, University of Copenhagen, Denmark. ole.

dreier@psy.ku.dk

Hysse Forchhammer, Neurological Department, Glostrup University Hospital,

Glostrup, Denmark. hybf@glostruphosp.kbhamt.dk

Wybo Houkes, Department of Philosophy, University of Eindhoven, The Netherlands.

W.N.Houkes@tm.tue.nl

Kristine Jensen de López, Department of Communication, Aalborg University,

Denmark. kristine@hum.aau.dk

Linda Kendell-Scott, University of Winchester, England. linda.kendell-scott@

winchester.ac.uk

Randi Markussen, Dept. of Information and Media Studies, Aarhus University,

Denmark. rmark@imv.au.dk

David Middleton, Department of Human Sciences, Loughborough University,

England. d.j.middleton@lboro.ac.uk

Klaus Nielsen, Institute of Psychology, Aarhus University, Denmark. klausn@psy.

au.dk

Finn Olesen, Dept. of Information and Media Studies, Aarhus University, Denmark.
finno@imv.au.dk
viii Doing Things with Things
Beth Preston, Department of Philosophy, University of Georgia, USA. epreston@
uga.edu

Gustavo Ribeiro, Royal Danish Academy of Fine Arts, School of Architecture,

Copenhagen, Denmark. gustavo.ribeiro@karch.dk

Ágnes Szokolszky, Deparment of Psychology, University of Szeged, Hungary.

szokolszky@t-online.hu

Estrid Sørensen, Institute of Technology Studies, Technical University, Berlin,

Germany. estrid@gmx.net

Pieter E. Vermaas, Department of Philosophy, Delft University of Technology, The

Netherlands. p.e.vermaas@tbm.tudelft.nl

Emma Williams, Department of Psychology, University of Surrey, England.

e.i.williams@surrey.ac.uk
 List of Figures and Tables

Figures

Figure 4.1 Photographs of the experimental objects 76

Figure 4.2 Ranking of the fitness of the substitute objects 77
Figure 4.3 Frequency of action types 80
Figure 5.1 Illustration of the semantic differences expressed in Danish
and Zapotec 97
Figure 5.2 Imitation of the UNDER spatial relation to the Initial-Canonical
oriented basket 99
Figure 5.3 Imitation of the UNDER spatial relation to the Initial-Inverted
oriented basket 100
Figure 5.4 Responses by language group to the I/LÁANI configuring
linguistically-guided object manipulation for the Initial-Inverted
basket 103
Figure 6.1 The shelf (to the left of the door), sink and plate rack (far left of
picture) and corner of fridge (far right side of picture). Part
of the dining room can be seen through the kitchen door 116
Figure 6.2 Basic categorisation of the shelf 117
Figure 6.3 A schematic history of the shelf 121
Figure 9.1 A reproduction of one of the very first medicine lists of the
EMM to be used for measuring out medicine at the ward. 183
Figure 9.2 A medicine list in harmony with the situated needs of the nurses 186
Figure 12.1 Tung Pattana Boardwalk (Photo: Jørgen Andreasen) 231
Figure 12.2 Tung Pattana Boardwalk (Photo: Jørgen Andreasen) 231

Tables

Table 4.1 Target and substitute objects 75

Table 5.1 The three main response types to the imitative and the
linguistically-guided task by language group 98
Table 6.1 Summary of main cognitive differences between using the box
and using the shelf. 124
This page intentionally left blank
 Introduction
Alan Costall and Ole Dreier

It has been claimed that the natural sciences have abstracted for themselves a
‘material world’ set apart from human concerns, and the social sciences, in their turn,
constructed ‘a world of actors devoid of things’ (Joerges,1988, p. 220). A dualism
between people and things has certainly been institutionalized in the very distinction
between the natural and human sciences. Yet, the situation is more complicated and
in some respects even worse than that. There is great diversity among the human
sciences in their focus on objects. At one extreme, archaeology would seem obliged
by its very nature to take into account human tools and artefacts. At the other
extreme, mainstream psychology with its emphasis upon internal mental structures,
and commitment to stimulus-response thinking, has presented us with a world not
just devoid of things but also of agents.
As psychologists critical of the neglect of the material conditions of human
thought and action, we were keen to bring together a number of researchers in
the human sciences who have been taking things more seriously, and whose work
interconnects in several distinctive and important ways. This book stems from
two interdisciplinary meetings we organized at the University of Copenhagen. In
addition to the discussions held at these meetings, the later drafts of the chapters
were circulated among all of the contributors to make the individual contributions
speak more directly to each other and to common issues. We are grateful to Benny
Karpatschof and Kurt Keller, who also took part in the workshops, for their insightful
commentaries on the following chapters. We would also like to thank the Danish
Research Council for the Humanities for funding the workshops; the Department
of Psychology, Copenhagen University for hosting the meetings and providing a
visiting professorship for Alan Costall which led to our joint project.
This book reflects a growing interest within the human sciences about how
people relate to objects. However, the standard approaches assume that the meanings
of things can be localized – within the things themselves, or within our heads, or
within our ‘discourse’. Furthermore, most of the recent work in this field has also
been concerned with how people cope with new technologies, with an emphasis as
much on the novelty of these technologies as upon their highly technical nature.
In contrast, the contributors to the present book mainly focus on everyday objects
and how these objects enter into our activities over the course of time. Indeed,
even though we have included some chapters concerned with the use of ‘high-tech’
objects, these also approach such objects as parts of ongoing everyday activities. The
2 Doing Things with Things
chapters also reflect a distinctive and coherent combination of different theoretical
approaches, including actor network theory, ecological psychology, cognitive
linguistics, and science and technology studies. What unites the contributors to this
book is not just their rejection of the standard notion of objects and their properties
as inert, inherently meaningless, and quite alien to ourselves. All of the contributors
emphasize the need to understand the relations between people and objects in terms
of process and change.

I. Intentionality and the Functionality of Things

The first two chapters address a fundamental issue taken up again in many of the
later chapters. Although any particular object lends itself to a limitless number
of possible uses, many of the objects we encounter nevertheless present us with
their own unique meaning: an apple, for example, is ‘for eating’, even though it
can be thrown, or used as a paperweight, or even as a target for archery. Both Beth
Preston and also Pieter Vermaas and Wybo Houkes critically examine the standard
philosophical accounts of the ‘proper function’ of artefacts (notably, Ruth Millikan’s
writings) and present alternative analyses. As a whole the discussion is framed in
terms of three supposedly distinctive roles for human agents with regard to artefacts
– designer, maker, and user. While Preston argues that the intentions of the designer
are not sufficient for establishing proper function, Vermaas and Houkes defend such
an intentionalist approach to artefacts and artefact use.
The standard view of artefacts assumes that the design, making and use of
artefacts are quite distinct and separate activities. In her chapter, ‘The case of the
recalcitrant prototype’, Beth Preston points out that while users often use artefacts
creatively for purposes other than their designers had in mind – and, in so doing, also
modify them – standard theories of how artefact functions are established privilege
the intentions of designers. The use of artefacts is then said to rest on their ‘proper
function’ as opposed to a contingent, occasional ‘system function’. Preston argues
that a distinction between proper function and system function is necessary in order
to prevent the proper functions of artefacts from proliferating in an uncontrolled
and incoherent way. At the same time, however, this distinction makes the view
that artefact function is established by designers’ intentions problematic. To resolve
this dilemma, Preston proposes that a history of selection and reproduction, rather
than the intentions of designers, establishes the (direct and derived) proper functions
of an artefact. In accounting for how artefacts get their functions neither proper
function nor system function then require an appeal to intentions.
Pieter Vermaas and Wybo Houkes present an action-theoretical view of artefacts
in their chapter on ‘Use plans and artefact functions: an intentionalist approach
to artefacts and their use’. They reject the strong emphasis on functionality in the
standard view of artefacts, and endorse a version of intentionalism by describing
artefact design and use primarily in terms of plans: that is, structured sequences
of considered actions. Users, they argue, may take artefacts simply as a means of
 Introduction 3
attaining their own desired ends and thus be primarily interested in their own plans
for realising those ends without even supposing that artefacts have functions. This
question about what something is for, should, on their view, be understood in terms
of how and for what goals that thing could be used. In designing, designers construct
what Vermaas and Houkes call ‘use plans’ for the artefacts they are designing.
The concept of proper use may now be maintained by claiming that use plans for
artefacts constructed by designers hold a privileged position relative to other plans.
Consequently, the proper use of an artefact is the use that follows a plan that is
approximately identical to the use plan developed by its designers.
A further distinction can also be made between good (rational) and bad (irrational)
use, which is independent of the distinction between proper and non-standard use,
allowing us to recognise the rationality of non-standard uses of artefacts. Vermaas and
Houkes argue that action plans are incomplete ‘lightweight plans’ with respect to both
the necessary skills involved and the circumstances of their realisation. Furthermore,
they see the functions of an artefact as corresponding to those dispositions of the
artefact that serve to explain why the use plan for the artefact achieves its goal.
The arguments and differing views of these two chapters set the stage for the rest
of the book. In fact, designing, making, and using artefacts – and the relationships
between them – are addressed in various ways in all the other chapters, for example,
in relation to the significance of multi-purpose artefacts and the transformation of
artefacts in ongoing activities.

II. Things in the World of the Child

How do young children come to develop a sense of things having a proper function
– an inherent, ‘impersonal’ meaning – and how might this depend upon the particular
cultural context? These questions are only beginning to be addressed by developmental
psychologists. Indeed, the way children relate to things more generally has been
neglected within psychology. In part, this has been the legacy of the developmental
theorist, Jean Piaget. Despite treating the interactions between children and
objects as the foundation of development (‘sensori-motor’ intelligence’), Piagetian
theory, along with most of cognitive psychology, posits abstract thinking (‘formal
operations’) as the culmination of the developmental process. Even the social-cultural
theory of Vygotsky, widely regarded as a radical alternative to Piagetian theory, itself
posits a process of decontextualization, through a developmental transition from the
interpersonal to the intrapersonal (Wertsch, 1996).
In one important respect, matters have become even worse within developmental
theory in recent years. In going beyond Piaget, many developmental psychologists
now stress the importance of domain specific knowledge, and regard the developing
child as a ‘theorist’, forming and elaborating theories restricted to highly specific
domains. In particular, ‘the Theory of Mind’ approach, which has been dominant
within developmental psychology for the last decade, has largely removed the issue
of children’s use of objects from the research agenda, not only because of its emphasis
4 Doing Things with Things
upon ‘mind’, but also because of its explicit separation of children’s understanding
of other people from their understanding of things.
Autism, according to this approach, is a disorder stemming from a specific failure
to understand other minds, and, furthermore, one that affects the child’s ability to
relate to other people, but not to things. In their chapter, on ‘Autism and object use’,
Emma Williams and Linda Kendell-Scott present evidence of extensive delay and
disruption for autistic children in their dealings with things, not least in the ‘proper’
use of objects (the topic of the chapters in Section 1 of this book). As they point out,
even if the Theory of Mind account were correct, such disruption in object use should
have been expected, given the way we normally come to learn about novel objects,
by being introduced to them by other people, noticing other people using them,
and entering into the standard practices in which things have their place. However,
Williams and Kendell-Scott argue that the fundamental division between people
and things, assumed by the Theory of Mind approach, is untenable: the widespread
failure to use things properly in commonly understood ways must itself set powerful
obstacles to relating to other people and communicating with them.
The unwordliness of psychological theory is further challenged by Ágnes
Szokolszky. Her chapter on ‘Object use in pretend play’ is concerned with the
impressively resourceful way in which young children can incorporate objects into
their play activity, make-believing these things are something they are not. At one
of the workshops on which this book is based, Szokolszky used the example of her
young son announcing, somewhat unpromisingly, that his discarded sock was a ‘gun’.
But her son then held the sock taut between his two hands, and ‘aimed’ it at her. The
fact that things lend themselves to so many pretend uses (and indeed practical uses)
has encouraged theorists to suppose that the ability to engage in imaginative play
with objects is purely ‘cognitive,’ and little, if anything, to do with the play object
itself. Yet it is just not true that we can do anything with anything. The fact that any
object could in principle be used in an infinite variety of ways, does not mean that
their functionality is completely unconstrained. Socks may lend themselves to being
used as pretend guns, but they do not make good pretend houses, nor even swords.
The puzzle for researchers of cognitive development has been that pretend play
seems to be highly anomalous. After all, one might suppose that the developing child
should be concerned with finding out how things really work, rather than delighting
in treating them in such ‘unrealistic’ ways. Ágnes Szokolszky’s research has nicely
demonstrated that young children are well aware of the suitability – the ‘affordances’
– of different objects for particular kinds of play event, such as using those objects as
a pretend pillow. Her conclusion, therefore, is that children are not ‘escaping’ from
reality when they engage in pretend play, but rather they are, in an important sense,
getting things right. They are exploring the various action possibilities of objects,
over and above their proper functions.
These two chapters on object use and autism, and pretend play, were informed
by a broadly ‘ecological’ framework (Gibson, 1979; see also Costall, 1995). Kristine
Jensen de López’s research is situated within the new perspective of ‘cognitive
linguistics,’ an approach that takes the human body as a fundamental source of
 Introduction 5
linguistic meaning. Her chapter compares how Zapotek and Danish speaking
children comprehend spatial relations. Zapotek, a Mayan language, describes the
spatial relationships between objects in terms of body parts (head, stomach, etc),
rather than the prepositional forms used in Danish and English. When the children in
her study were asked to imitate the experimenter’s actions of putting items in, on, or
under a basket, there were striking differences between the two groups of children.
Even when the experimenter had placed the item on or under the basket, the Danish
children were much more likely to put the object in the basket than were the Mayan
children. Kristine Jensen de López proposes, however, that these differences cannot
solely be understood in terms of linguistic differences, but also in terms of how things
are normally used differently in the two groups. The human body itself is configured
differently according to these different uses of objects. In the Mayan culture, objects
are routinely used in multiple ways, as when a basket is turned upside down to serve
as a cage to confine a chicken about to lay its eggs.
Each of the chapters in this section is concerned with the question of the proper
function of objects, and raises the further questions (already addressed in the first
section) of how the proper function of an object is to be construed theoretically, and
how children (at least under certain cultural conditions) come to develop a sense
that objects have such a privileged, proper meaning. Proper function is defined
culturally, and its definition depends on the ways of designing and using things
within a particular culture.

III. Transformation and Things

From the perspective of the human sciences, an emphasis upon the importance
of things in human affairs can often seem like a retreat into reductionism and
materialistic determinism. However, in the present book, objects are far from
considered in isolation, but, instead, as participating in human activities. The three
chapters in this section are concerned with things in transformation: how things are
incorporated into our activities over the course of time, and, how, in that process,
they come into new relations with other things and other people, and thus take on
new significance.
As we have already noted, much of the existing research on how people relate
to things has concerned information technology and, furthermore, information
technologies relatively new to the users. It has also typically focused on team-
work. Seldom has the research taken a longer-term historical perspective, tracing
the processes by which the material resources for various practices themselves both
become shaped and, in turn, shape those practices (but see Hutchins, 1995). And,
most importantly, there has been little research pitched at the level of everyday
engagements with ‘low-tech’ objects used primarily by single individuals.
David de Léon has been studying people in the course of preparing and cooking
meals in their own homes. As he explains in his chapter on the ‘Cognitive biographies
of things’, there is no single locus of order in such activities, but a continual interplay
6 Doing Things with Things
between the human agents and the things they are using. Order is ‘bounced’, as
it were, between agent and the objects in this process, as in the deliberate acts of
‘maintenance’, where time is set aside to rearrange things so that they can support
the next phase of the activity. But this can also come about unintentionally, as when
frequently used objects end up being conveniently close to hand precisely because
they are likely to have been recently used, and thus left drying, for example, on the
plate rack.
The chapter focuses on the history of a spice rack over the course of almost thirty
years, and how it was transformed over this period according to the circumstances
and interests of its main user. Initially, the collection of spices was very small and not
even kept in a rack but in a box; as it grew larger, it was then set out on a deep shelf,
and then, finally, on an extensive set of shallow, one-jar deep, shelves.
The ‘cognitive congeniality’ of any object, should be regarded not as an intrinsic
property of that object, but, as David puts it, concerns ‘the combination of techniques,
procedures, and / or habits, with particular artefacts and task environments, in relation
to specific tasks’. For example, only if the spice jars are removed one at a time can
the shelves serve to ‘remind’ the user about where a jar that has just been removed
needs to be returned. Furthermore, from a much broader temporal perspective, this
thing, the spice rack, far from being a pre-existing, fixed resource for cookery, has
itself been undergoing continual transformation and elaboration within the activities
of cookery.
Hysse Birgitte Forchhammer’s chapter, ‘The woman who used her walking stick
as a telephone,’ also takes up the issue, raised in Section 1, of the proper function
of things. Her research explores the many different ways that the aids provided to
stroke patients fail to be used ‘as intended’ – as a result of poor design, unanticipated
changes in the circumstances of the user, and the fact that such aids can serve
to trap the user into an identity of a disabled person. However, her chapter also
presents many examples of improvisation, as when one woman used a walking stick
to communicate with her neighbour, by tapping on the ceiling. The analysis she
presents not only ranges across the different levels of definition of function (the
‘mass’ or impersonal, the interpersonal, and the individual), but has, as she points
out, also important practical implications for the design of disability aids.
Many of the contributors to this book highlight the multiple ways a single object
can be used, in improvisation and in pretend play. And, as Kristine Jensen de López
argues in her chapter, there are cultural differences in the extent to which things are
supposed to have a single, privileged function. Yet certain objects even in Western
culture are distinguished by their multifunctionality. One kind of multifunctionality
is exemplified by the Swiss army knife with its numerous blades and gadgets, but
this is no more than a concatenation of specific tools with definite functions. The
computer is an example of multifunctionality in a much more interesting sense. Now
the digital computer was envisaged, from the outset, as a ‘general purpose machine’,
that is, a machine whose function was to be constrained solely by the computer
program and not by the machine itself. Even so, the computer has still managed
 Introduction 7
to enter into our lives in a whole diversity of ways unanticipated by its original
designers, not least as a resource for play.
Estrid Sørensen’s chapter, ‘The politics of things’, is about the ‘open’ nature of
computer technology, and describes the design and use of virtual reality software
for groups of young children to create their own micro-worlds. What constitutes the
design, she argues, is not just the computer interface provided by the software but also
the surrounding pedagogical practices. Although Sørensen is specifically concerned
with an open-ended design, it is never the case that a design can completely prescribe
practices, any more than rules fully define their application. Thus, it is necessary
in general, Sørensen argues, to study the continual interplay between design and
practices, and how this interplay leads to orderings in practice and of practice (Mol
& Law, 1994). Fluidity and regionality are examples of such orderings. By drawing
upon two instances from her study of the design workshop, Sørensen shows that these
two orderings may either co-exist harmoniously or conflict in ongoing practices.
Sørensen traces the transformation of designs in ongoing practice through the
introduction of new objects, the children’s ideas, their discussions, and division of
labour. In that sense, Sørensen proposes that we should regard design as politics,
in which a design is proposed, picked up, continued, and either changed or left
unchanged. Studies of such continuing processes of design in practice may lead to a
better understanding of how best to design for ongoing, open-ended practices within
both the school setting and beyond.

IV. Organisation and Things

The chapters in this book are primarily about how things are used by individual users
or informal groups. The four chapters in the final section, however, examine the active
role of things in the organisation of practices and in the practices of organisations,
and bring out the wider cultural, social, organisational, economic, and educational
aspects of things and how they are used. In most theoretical approaches to things,
classification is assumed to be the most natural way of understanding objects. Yet,
‘sorting things out’ (Bowker & Star, 1999) is not the only way we engage with
objects. Once we consider things, not as isolated objects or artefacts, but as already
involved in ongoing practices, we can no longer take their classification for granted.
Classificatory practices are themselves rather special kinds of activities, often
subservient to other ends in practice. Indeed, the relevant ‘properties’ of things, and
the proper functions of those things, are transformed within those practices.
Furthermore, once we situate things within practices, the stability of things and
the stability of practices can be seen as inseparable, and as mutually affording each
other. Things are involved in both maintaining and changing the order of social
practices. They are ordered and reordered, and they are arranged and rearranged in
and for particular social practices (Dreier, 2005, Schatzki, 2002). Thus, in order to
make sense of the human practices of organisations and of organising, it is necessary
to include the various things involved.
8 Doing Things with Things
In their chapter, ‘Working with material things: From essentialism to material-
semiotic analysis of sociotechnical practice’, Finn Olesen and Randi Markussen
describe their field study of socio-technical practice in a hospital ward concerning
the introduction of a an electronic module for dispensing medication. Their
approach is based in science and technology studies, in which people and things are
regarded as intertwined in heterogeneous sociotechnical arrangements. Rather than
regarding things as somehow pre-given objects with definite known qualities, they
are studied as dynamic elements in a continuous flow of activity. Similarly, design
itself should be understood as an evolving process in an ongoing practice. A practice
(even one involving conflict) may attain stability for some time, due to a dynamic,
interdependent evolution of things and humans which stabilises particular relations
between them. This was the case for the medication list in their study, so that it
became a ‘stable actor’ within the hypercomplex, sociotechnical arrangements of
the ward.
Following Haraway (1997), Olesen and Markussen maintain that social
relationships may become congealed into and located within objects as
‘decontextualized things’. Material and semiotic dimensions of reality are thus
interwoven, and material artefacts, such as the medication list, also comprise
‘fictional’ constructions (compare the chapters by Szokolszky and Forchhammer).
Olesen and Markussen are engaged, as they put it, in an investigation into the
‘mundane philosophy of things’ in everyday practice. They criticize the notion that
a thing is the sum of its properties since properties are not inherent to the object, but
ascribed to it through a process of classification and comparison with other, similar
things. In the case of the electronic medication module this becomes very evident
in practice, since there was a struggle over its definition and use, and thus no single
classification scheme was able to unify the competing classifications and exhaust the
essence of the thing.
Despite the intentions of its original designers, the electronic medication record
does not in fact function as a ‘classification machine’ that ensures widespread order
within the organization. It can only be understood in terms of how it is embedded
in the ongoing practices in the ward, such as the closely linked practices of writing
and reading and measuring medication. Olesen and Markussen’s chapter serves to
expand a conception of writing as a practice linking people and things in the material
fabric of the world and things.
This theme is also central to Steven D. Brown and David Middleton’s chapter,
‘Words and things: Discursive and non-discursive ordering in a networked
organization’. They focus on the active role of email archives in a company producing
a complex product in a process distributed across different parts of the world, and
upon the relationship between words and things within the unfolding organisational
practices. In part, they are attempting to move critically beyond a background of
discursive psychology and studies of communication in the work place which, they
argue, consider things simply as formulated and described by speakers. Within that
standard approach, the social is the standard against which to consider the role of
things, and the things themselves are thus regarded as external to sociality.
 Introduction 9
However, instead of just asking how we talk about things, Brown and Middleton
argue that we should also ask how things become fit objects for talking about, and
how things make us talk with each other in particular ways, perform activities in
particular ways, manage in particular ways, and so forth. In accordance with this
alternative point of view, they ask how the archival practices concerning the emails
in the company order the socialities of its organisational practice. Organisational
order, they contend, is produced not only by discursive but also by non-discursive
means.
Brown and Middleton consider the archives as a means of creating and maintaining
order and stability in the practice of the organisation by creating interdependencies
between persons and things. Furthermore, this organisational order stretches across
different times and places, drawing together a large number of people, materials and
activities around the entire world. The creation of schedules requiring everyone to
adhere to the same strict programme of events is a distinct feature of this ordering
across times and places by means of the archives. We might add that the creation of
schedules leads to the creation of sequences (Abbott, 2001), that is, a sequential order
(Dreier, 2005). A further means of creating and maintaining order across places is
the modularisation of the complex process of production across diverse production
sites, ordered into a coherent process of production by means of the scheduling and
sequencing accomplished through the email archives. That, in turn, has created new
problems of order in the organisational practice of the company.
In his chapter, ‘Learning to do things with things – Apprenticeship learning
in bakery as economy and social practice’, Klaus Nielsen also examines ongoing
practices in which things and people participate in dynamic, transforming and
changing ways. However, since doing things with things is mostly something
we have to learn to do, his research has been examining how the arrangement of
education in work practices mediates the way we learn to do things and thus come to
do things with things. For this purpose, he adopts a social practice theory of learning
as situated in particular ongoing practices and in particular arrangements of learning
and education (Dreier, 2003; Lave, 1993).
In his interviews with apprentice bakers about their training, Nielsen found
they were concerned with economic issues rather than ‘learning’. Yet, according
to the apprentices, all issues relating to economy were neglected in the training
provided at the trade school. While learning in school is arranged by dissecting the
topic into elements to be learned sequentially as accumulated discrete categories
and skills, practical considerations about the organisation of work and economical
considerations led to a quite different arrangement of learning in the bakery itself.
Here ‘economy’ is to be understood in both the sense of market economy and also
the maximisation of efficiency in the use of time, materials, resources, and so
forth. According to the ideal of school learning, however, such economic issues,
so central to apprenticeship, are dismissed as incidental, and a distraction from the
proper business of learning. In contrast, in the bakery, these economic issues are an
essential part of what needs to be learnt. The apprentices, journeymen and masters
alike emphasized this positive, enabling influence of economy on learning in the
10 Doing Things with Things
educational arrangement in the bakery. It is also the case that the form taken by
this learning process – the sequence of close initial supervision gradually leading
towards increased ‘individual autonomy’ – was clearly economically motivated.
However, as Nielsen explains, there are aspects of the economic realities of the
actual work situation that are far from benign. The craft of baking is challenged
from two sides. There is the deskilling of the baker apprentices promoted by the
growing industrialisation of baking (Braverman, 1974; Projektgruppe Automation
und Qualifikation, 1987). Second, there is the issue of ‘commodity aesthetics’
(Haug, 1971), since the bakers have to learn to adopt the perspective of the potential
customer, and place a higher priority on the ‘look’ of the product, than how it actually
tastes. Indeed, as Nielsen pointed out at one of the workshops that gave rise to this
book, the bakers are now required even to dress up in special clothing in order to
look more like bakers.
A recurrent theme in this book concerns the need to understand things not in
isolation but in relation to the human activities in which they are involved, and the
different, even conflicting, meanings things can have for the people involved with
them. Another main theme has concerned the need to think of design not as a separate
stage prior to the use of things, but as a continuing process within the context of their
actual use. Gustavo Ribeiro’s chapter, ‘Urban makings: formalisation of informal
settlements in Thailand’ is about changes in the landscapes and infrastructures of
small third-world low-income settlements such as the development of walkways or
bridges, either with or without the assistance of architects and other experts external
to the local community. As far as the first- and third-world experts are concerned,
such developments should be regarded primarily as ‘sites for social activities’ with
the paradoxical consequence that the material products of such construction projects
– the things themselves – are regarded as merely incidental to the business of
empowering and unifying communities in the process of planning and construction.
Ribeiro draws a telling contrast with his earlier studies on the creation of urban
landscapes and patterns of urban change in Latin American shanty-towns. Here,
simply by moving between one place to another across open ground in the course of
their daily lives, the people created beaten tracks on the ground which came to define
a network of footpaths; these, in turn, became the source of the layout of the shanty
towns that were eventually constructed on the previously open sites. The formations
of physical environments/structures and of social relations are inseparable aspects
of the organisation and change of ongoing everyday practices. Ribeiro contrasts
such long-term, unplanned, urban making through ‘direct action’ with ‘mediated
urban making’ brought about through the intervention of external actors and the
introduction of formal planning procedures. Such ‘mediated urban making’ creates
a form of practice in which experts are positioned in such a way that they come to
frame the practice they are involved in making solely in terms of organisation and
education (much in the way that, as Brown and Middleton in their chapter argue,
things in social theory can become relegated to mere subjects of talk).
Ribeiro points to a serious consequence of the definition of the process of
construction not just as educational but also as ceasing with the creation of the
 Introduction 11
‘final structure’. The project-mediated form of making downplays the importance
of the maintenance of the constructed object through use, and the incorporation
of physical intervention into the daily life of the community. It thus threatens the
very sustainability of any project. Such externally imposed organisation of projects
complicates and distorts the interrelations between the material and social aspects of
practices so that the material landscapes themselves tend to be reduced to mere ‘sites
of the social’ (Schatzki, 2002). Ribeiro argues that the role of material transformations
in daily social practices must be brought to the fore, and the understanding of the
role of projects must be re-anchored in such ongoing daily practices.
Finally, Ribeiro also addresses the issue of stability and change that has also
been raised in many of the earlier chapters. Compared to consolidated urban areas,
squatter settlements have an inherent power to change and adapt. In contrast to the
‘permanence and persistence’ of urban structures, we are presented with changes
– ‘instead of morphology, patterns of use’. Such settlements are transitory, and any
stability that might arise must be understood as an accomplishment of ongoing
practices.

Conclusion

The beaten paths studied by Ribeiro in his research on shanty towns provide a vivid
example of the kind of understanding of things presented in this book – things not as
fixed and independent of human beings, but as themselves transformed, even coming
into being, within ongoing human practices, and which, in turn, transform those
practices. We learn much more about both people and things by studying them as
worldly, though not just as in the world, but as incorporated in practices in the world
(Holzkamp, 1996). However, there also needs to be an emphasis upon process, upon
change. Stability is not the ‘natural’ state of things. Objects cannot be understood as
static entities with fixed categories. Even stability must be understood as processual,
one stable pattern giving rise to another. So stability is never final but open to further
unanticipated change. Things are never entirely resolved, once and for all.

References

Abbott, A. (2001). Time matters. On theory and method. Chicago: University of

Chicago Press.
Bowker, G.C. and Star, S.L. (1999). Sorting things out: Classification and its
consequences. Cambridge, MA.: MIT Press.
Braverman, H. (1974). Labor and monopoly capital. The degradation of work in the
twentieth century. New York: Monthly Review Press.
Costall, A. (1995). Socializing affordances. Theory and Psychology, 5, 467–481.
Dreier, O. (2003). Learning in personal trajectories of participation. In N. Stephenson,
L.H. Radtke, R.J. Jorna & H.J. Stam (Eds.), Theoretical psychology. Critical
contributions (pp. 20–29). Concord, ON: Captus Press.
12 Doing Things with Things
Dreier, O. (2005). The social practice of psychotherapy. Theory, structure, critique.
In A. Gülerce, A. Hofmeister, I. Staeuble, G. Saunders & J. Kaye (Eds.),
Contemporary theorizing in psychology: Critical perspectives (pp. 162–170).
Concord: Captus University Publishers.
Haraway, D. (1997). Modest_Witness@Second_Millennium. Female Man©_Meets_
OncoMouse™– Feminism and Techno science. New York: Routledge.
Haug, W. F. (1986). Critique of commodity aesthetics : appearance, sexuality, and
advertising in capitalist society. (R. Bock, trans.) Minneapolis: University of
Minnesota Press, c1986.
Holzkamp, K. 1996. Manuskripte zum Arbeitsprojekt ‘Lebensführung’. Forum
Kritische Psychologie 36, 7–112.
Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press.
Joerges, B. (1988). Technology in everyday life: Conceptual queries. Journal for the
Theory of Social Behaviour, 18, 221–237.
Kirsh, D. (1995). The intelligent use of space. Artificial Intelligence, 73, 31–68.
Lave, J. (1993). The practice of learning. In S. Chaiklin & J. Lave (Eds.). Understanding
practice. Perspectives on activity and context. New York: Cambridge University
Press.
Mol, A. & Law, J. (1994). Regions, networks and fluids: Anaemia and social
topology. Social Studies of Science, 24(4), 641–671.
Projektgruppe Automation und Qualifikation (1987). Widersprüche der
Automationsarbeit. Berlin: Argument Verlag.
Schatzki, Theodore R. (2002). The site of the social. A philosophical account of the
constitution of social life and change. University Park, PA.: Pennsylvania State
University Press.
Wertsch, J.V. (1996). The role of abstract rationality in Vygotsky’s image of mind. In
A. Tryphon & J. Voneche (Eds.), Piaget-Vygotsky: The social genesis of thought
(pp. 25–43). Hove, East Sussex, UK: Psychology Press.
 PART I
Intentionality and the Functionality
of Things
This page intentionally left blank
 Chapter 1

The Case of the Recalcitrant Prototype

Beth Preston

In the presumed absence of a creator god, the functions of biological organs are
not functions any agent intended them to have. Consequently, theories of biological
function must start from the assumption that function can be established without
appeal to the intentions of intelligent agents. But the opposite assumption operates
with regard to artifact function. The intentions of human agents are commonly
regarded as a necessary condition for establishing function. Although it is difficult to
show that the intentions of any particular agent are required, it does seem that without
human designers, makers, and users in general, there would be no human artifacts,
let alone artifact functions (McLaughlin 2001, p. 48). But many theorists also regard
the intentions of the designer, specifically, as a sufficient condition for establishing
function. In other words, if an agent intends an artifact she designs to have a specific
function, then that is the function it has, at least initially. It is this latter claim I
wish to examine here. I shall show that regarding designers’ intentions as sufficient
generates serious problems for a theory of artifact function. Moreover, there is a
viable theory of artifact function that does not appeal directly to the intentions of
designers as sufficient conditions. Thus there is ample reason to reject the claim that
designers’ intentions are sufficient. But I shall begin by explaining why this claim
has so much appeal for function theorists in the first place.

Proper Function and System Function

There are three distinguishable roles human agents may play with regard to artifacts
– designer, maker, and user. As commonly understood, the designer specifies the
characteristics of the artifact to be made; the maker does the actual construction
work; and the user subsequently puts the artifact to use. These roles are not mutually
exclusive. A single individual may play any combination or all of them with regard
to a specific artifact. But they can also be played by completely different individuals
– indeed, by individuals who do not communicate or are unaware of each others’
existence, and whose intentions are thus independent.1 Unfortunately, writers on

1 The distinction between these roles does not appear to be contingent on the historical
mode of production. Plato (1992, 272/Republic 601c–602a) distinguishes between ‘the one
who knows’ how an artifact should be (the designer), the maker who is instructed by the one
who knows, and the user. He then argues that in fact the user has the most knowledge about
16 Doing Things with Things
artifact function typically use terms like ‘creator,’ ‘producer,’ or ‘maker’ in a way
that is ambiguous with regard to the designer and maker roles. But as we shall see, it
is clear that in fact it is the intentions of designers in particular that are in question in
theories of how artifact functions are established. Thus sorting out this issue depends
on distinguishing these roles.
On the other hand, an important feature of these roles, and one that will come
into play in the course of the argument, is that many activities are borderline or fuzzy
with regard to classification in terms of them. For example, designers often make
scale models, or make and try out specific parts of the artifact during the design
process in order to see whether certain materials are suitable, or certain devices safe
and effective, or whatever. More importantly for present purposes, users often use
artifacts creatively. This may involve using something for a novel purpose – using
a screwdriver to remove staples, for instance. Or it may involve some modification
of the artifact to suit it for the novel purpose – filing down a key to make it more
effective as a cutting instrument, for instance. Similarly, makers often construct
creatively, departing from the instructions of designers in order to improve the artifact,
or to continue construction in the absence of specified materials or needed tools.
For example, cooks continually tinker with recipes to suit them to their own tastes,
or substitute ingredients when specified ingredients are unavailable, prohibitively
expensive, and so on. Phenomena like these show that although the roles of designer,
maker, and user are distinguishable – even clearly and unequivocally distinguished
in many cases – they are not rigorously distinct.
There is another important feature of these roles I will be leaving out of
account in this paper. It is also possible – indeed, common – for groups rather than
individuals to fill these roles. This is clearly the case for users, who typically form
loose communal groups. But production, too, is typically collaborative, even in non-
industrial contexts, and design is often collaborative, perhaps particularly in industrial
contexts. Moreover, there are important social and communicative relations among
the individuals or groups filling the various roles. For example, designers may seek
user input or maker advice during the design process, or users or makers may offer
or even demand this. In short, these roles as actually practiced are thoroughly social
in character. This raises a different problem for the intentionalist view of artifact
function because it requires an account of collective intentions. But I will abstract
from this important issue in framing my argument, because the question of whether
designer intention is sufficient for the establishment of artifact function does not
depend on whether the intention in question is individual or collective, but only on
what relation it bears to the artifact. So when I speak of ‘the designer,’ for instance, I
will be referring to the role in abstraction from the concrete collective or individual
ways in which it may be carried out in practice.

the artifact, and should instruct the maker, i.e., the user should also be the designer. But the
very fact that an argument was called for here shows that the roles of user and designer were
not routinely identified.
 The Case of the Recalcitrant Prototype 17
It is also important to distinguish between proper functions and system functions
(also sometimes called causal role functions). A proper function is what a thing is
supposed to do – usually understood in the case of artifacts as what it is designed to
do. A system function is a contingent purpose a thing may serve on occasion without
having been designed to do so. For example, the proper functions of spoons are
stirring and transporting food. Among the system functions a spoon may serve are
use as a percussive musical instrument, as a unit of measure for cooking ingredients,
as a dibble for transplanting seedlings, as an opener for cocoa tins, and so on.
Biological traits also have both proper functions and system functions. For example,
the human hand has the proper functions of grasping and gripping, but it can be used
in various ways as a weapon.
The best-known account of how system functions are established comes from
an early paper on functional analysis by Robert Cummins (1975). According to
Cummins, a function is established in terms of the capacity or disposition a thing
has to perform a specific role in the context of a system. A system is a configuration
of components interacting in an orderly way such that the performance of the whole
can be explained in terms of the performances of the components. In the case of
biology, the system is usually the whole organism, or one of its subsystems, such as
the circulatory system. In the case of artifacts the system is usually a suite of related
artifacts and the practices involved in using them. For example, the proper function
of beer is as a beverage for humans. But there is a practice among home gardeners of
pouring beer into saucers and then setting them out in the garden as slug traps. The
system function of the beer in this context of organized social practices and artifacts
is slug bait.
It is important to note that this account of how system functions are established
does not require an appeal to the intentions of intelligent agents. A function is
established solely in virtue of having a role in a containing system, whether or
not that role was ever intended by anybody. This is an important positive feature
of Cummins’ account in its application to biological function, where an appeal to
intentions is ruled out. But it also suggests an account of artifact function that does not
require an appeal to intentions. In this connection it should be noted that Cummins’
account can allow for indirect appeal to intentions of agents in the case of artifacts.
For example, in the case of beer as slug bait there are in fact intentions of beer users
involved in sustaining the relevant gardening practices. But the important point is
that were these practices the practices of a race of intentionless garden robots, the
beer would still have the system function of slug bait. It is the system and the role of
the artifact in the system that is essential.
The best known account of how proper functions are established comes from
the work of Ruth Millikan (1984, 1993) – indeed, ‘proper function’ is her term. On
Millikan’s theory, proper functions can be established either directly or derivatively.
A thing has a specific performance as its direct proper function if it is reproduced
from ancestors which successfully engaged in that performance, and which survived
and proliferated because of this (1984, p.28), i.e., if there is a selection history for
the performance. Thus the human hand has the direct proper functions of grasping
18 Doing Things with Things
and gripping because the successful use of their hands for these purposes contributed
to the ability of our ancestors to survive and reproduce. Similarly, the direct proper
functions of a spoons are stirring and transporting food, since historically it is in
virtue of such successful stirring and transporting that spoons have been maintained
and reproduced.
When something has a direct proper function that it accomplishes by producing
some other thing, the produced thing inherits a derived proper function from the
producing thing, even though it has not itself undergone selection for this function
(1984, pp.41-3). Thus if a chameleon contains a mechanism the direct proper function
of which is to change the chameleon’s skin color to camouflage it from predators, the
skin coloration this mechanism produces on any particular occasion has the derived
proper function of camouflaging the chameleon. Similarly, if a spark plug has the
proper function of initiating combustion by producing a spark, the spark has the
derived proper function of initiating combustion.
As in the case of system function, this account of direct and derived proper
functions does not require any appeal to intentions of agents. Direct proper functions
and the derived proper functions that depend on them are established by a history
of selection and reproduction. In the biological realm selection and reproduction are
causal processes so an appeal to the intentions of directing intelligent agents is neither
required nor desired. In the case of artifacts, the intentions of human agents in fact
implement both selection and reproduction indirectly. But if a race of intentionless
robots were responsible for the same history of selection and reproduction, the
artifacts would have the same direct and derived proper functions. It is the selection
history that is essential, not the intentions of the agents or the robots involved.
The relationship between proper functions and system functions is a controversial
one in the recent philosophical literature on function. They are regarded by some as
rival theories, one of which is wrong and must be discarded. Davies (2001) argues,
for instance, that all proper functions are really system functions. Attempts are made
by others to unify these two concepts of function either by subordinating one of them
to the other (Walsh and Ariew 1996) or by subordinating both to an overarching
conceptual framework (Kitcher 1993). Yet other, more pluralistically minded
theorists insist that these are distinct concepts of function with different domains or
ranges of application (Godfrey-Smith 1993, Millikan 1989, Preston 1998). I hold this
last, pluralist view, and I assume it for the purposes of this paper. The central point
to be borne in mind is that on the pluralist view proper function and system function
are distinct but equally important conceptions of function, applicable to distinct but
equally important functional phenomena. In this connection it should also be noted
that on the pluralist view, as well as on some of the other views, system functions
may become proper functions over the course of time. For example, if breweries
start to make and market beer specifically for use as slug bait, and it is regularly
bought and used by gardeners for that purpose, it will acquire slug bait as a proper
 The Case of the Recalcitrant Prototype 19
2
function in addition to its original proper function as a beverage. But what I have to
say does not depend on settling any of these issues. The basis for my argument is just
that if you recognize a theoretical distinction between proper function and system
function, then holding the view that artifact function is established by designers’
intentions is problematic.3

The Recalcitrant Prototype

Because direct proper functions are established through an historical process of

selection and reproduction, the first instance of what will later become a functional
trait itself has no proper function. For example, suppose a chance mutation causes a
single moth to have a color that is particularly good as camouflage from predators.
This first moth – I shall refer to it as the prototype moth – has an advantage in
survival and breeding, as do its offspring which (let us say) inherit the camouflaging
coloration. On Millikan’s theory, the mutated color of the prototype moth does not
have the direct proper function of camouflaging it from predators; that is only the
effect it has at this initial stage. Only after some indeterminate number of generations
of selection and reproduction can we say the coloration has the direct proper function
of camouflaging these moths from predators.
Some theorists (e.g., McLaughlin 2001) regard this inability to ascribe proper
functions to apparently fully operational prototypes as a serious failing of Millikan’s
theory and other theories like it. But there are a couple of considerations that take
the edge off this criticism. First, even if prototypes do not have proper functions, if
the operations they perform fit successfully into roles in existing systems, they have
system functions (Millikan 1989, 175). For example, the coloration of the prototype
moth in our example above has a camouflaging, protective role in the predator-prey
system. Similarly, a can opener of novel design that succeeds in opening cans plays
a successful role in our food storage and retrieval system. It is important to note

2 My arguments for the pluralist view, as well as further discussion of the ways system
functions and proper functions are related to each other, may be found in Preston (1998).
3 It should also be noted here that there may be cross-cultural and developmental
differences in the way proper functions and system functions of artifacts are handled or
evaluated. For example, modern industrial culture in the West has resulted in a material culture
oriented towards special purpose artifacts. There are dedicated devices for opening paint cans,
for instance, even though a screwdriver or a knife or a dozen other things readily available to
the average painter work perfectly well. In other cultures this is not the case, and this suggests
that in these cultures system functions may be more important to people, both practically and
in terms of what they might value. Similarly, within cultures there may be developmental
differences with regard to the perception of proper function and of the possibilities for system
functional (non-proper) uses stemming from the fact that children have to learn that artifacts
have proper functions, and not just what proper functions various artifacts have. It is clearly
very important to investigate these cross-cultural and developmental aspects of artifact
function. But since nothing in my argument depends on them, I will have to leave these
aspects aside for the purposes of this paper.
20 Doing Things with Things
that not all prototypes will have such a system function, though. Having a system
function depends on having a current capacity to fill a system-defined role. So if a
mutation produces a prototype trait which does not fit into such a role, or which does
not fulfill it successfully, the prototype will have no system function. For example, if
the sole predators of our moths are bats hunting entirely by echolocation, there will
be no role in the predator-prey system for the new coloration to play. On the other
hand, if the predators are sighted and the new coloration does nothing to camouflage
the moth from them, the new coloration will not have the relevant capacity. In either
case, the new coloration will fail to have a system function. And of course this can
happen with artifacts as well – imagine a machine invented by a visionary medieval
machinist to open cans, although there are none to open; or a modern prototype can
opener which simply does not work.
So appealing to system function to solve the prototype problem will still leave
some prototypes with neither direct proper functions nor system functions. This is
acceptable to some (although by no means all) theorists in the case of biological
function, where some mutations produce changes that are indifferent or harmful to
the possessor from the point of view of selection, and to which we might therefore be
reluctant to assign any kind of function at all. But in the case of artifact prototypes,
even the anachronistic or non-working ones were intended by their producers to
have a function, so there is a strong impetus to ascribe that intended function to
them. Moreover, there is a strong impetus to ascribe proper function rather than
merely system functions to prototypes because the intentions of their producers are
construed as establishing what the prototype is supposed to do, whether or not it
actually does it. So most theorists will simply not be satisfied with the appeal to
system function in the case of artifact prototypes.
There is a specific application of Millikan’s theory of proper functions which
offers an alternative. Millikan holds that intentional states of agents have direct proper
functions – desires have the direct proper function of getting themselves fulfilled, for
example. Some desires are desires to accomplish this result by producing something
else. For example, I may desire to open cans by producing a machine to do the job,
and this machine then has the derived proper function of opening cans (1984, 49;
1999, 205). This is true even if it is a unique prototype – a can opener of completely
novel design, for instance. And it is true even if the prototype does not work. Thus an
artifact may have its proper function established either directly by a selection history
that does not require an appeal to intentions, or derivatively through its producer’s
intentions, or in both ways in parallel. If in both ways, the direct and derived proper
functions usually coincide. But where they do not, there is no problem in principle
with an artifact having more than one proper function, or even conflicting proper
functions (1999, 205). For example, as a matter of selection history beer is brewed
as a beverage for humans. This is its direct proper function. But suppose I buy up a
brewery with the express intention of brewing beer for use by gardeners as slug bait.
The beer I brew now has both its original direct proper function (beverage), and a
new derived proper function (slug bait) layered over it. But the most important point
here is that when the concept of derived proper function is applied to intentional
 The Case of the Recalcitrant Prototype 21
states in this way, it renders the intentions of producers sufficient for establishing
proper functions of artifacts.
Millikan’s theory of derived proper function thus provides for prototypes to have
proper functions if they derive those functions from a producing mechanism. For
example, the chameleon already contains a mechanism with the direct proper function
of changing its skin color to camouflage it from predators. So any skin coloration this
device produces, even if it is unique in the history of chameleons, has the derived
proper function of camouflage. Similarly, and most importantly for our purposes,
because all intentional states of agents have direct proper functions on Millikan’s
view, all prototype artifacts produced as a consequence of those intentions have
derived proper functions. This will still leave some biological prototypes without a
function, but as we noted above, this may be acceptable in the biological domain.
And it will yield what seems like the intuitively right result for artifact prototypes. So
Millikan provides a theoretical underpinning for the common idea that the intentions
of designers are a sufficient condition for the establishment of artifact function. Even
theorists who do not necessarily subscribe to Millikan’s theory of proper function
agree with her leading intuition here (Dipert 1993, McLaughlin 2001, Vermaas and
Houkes 2003). For if the intentions of designers are not a sufficient condition for the
establishment of function, prototype artifacts will have no proper functions – there
will be nothing they are supposed to do. So the view that designers’ intentions are
sufficient seems eminently well motivated. When someone designs an artifact, they
do so with some purpose for it in mind, and intuitively that just is what it is for
– its purpose. This intuition is reinforced by the theoretically generated problem of
prototype artifacts that have as yet no selection history, and thus no possible source
of proper function other than derivation from their designer’s intentions. However,
in the next section I shall show that regarding designers’ intentions as sufficient
for establishing proper function causes other difficulties for the theory of artifact
function – difficulties that are far more serious than the prototype problem their
sufficiency so neatly resolves.

Why Designers’ Intentions are not Sufficient

Regarding designers’ intentions as a sufficient condition for function privileges

the intentions of the originating agent in relation to the intentions of subsequent
users. Millikan says specifically that ‘artifacts have as derived proper functions
the functions intended for them by their makers... not their users’ (1999, p.205, my
emphasis). It is important to realize that in this context ‘maker’ primarily means
‘designer.’ The maker and the designer may be the same individual, of course, in
which case the point is moot. Or the maker may be a different individual but have the
same intentions towards the artifact as the designer – the maker may be following
detailed and informative instructions from the designer, for instance. But the problem
is that making – constructing, strictly speaking – can be a mechanical or uninformed
process where the maker has no relevant intentions towards the artifact. The maker
22 Doing Things with Things
may not know what the thing is designed to do, or she may be in a distracted state
of mind – think of someone working on a production line, for example. On the other
hand, a designer can hardly fail to have relevant intentions towards the artifacts she
designs. So what is primarily at stake here is the relationship between the intentions
of designers and the intentions of users. But what is supposed to underwrite or justify
this distinction between the intentions of designers and the intentions of users, and
the privileging of the former with regard to the establishment of artifact functions?
To put the problem in better perspective, consider the role of users in instituting
system functions of artifacts. I may use a can opener as a paperweight, or a table knife
as a screwdriver, or a highjacked airliner as an incendiary bomb. I certainly intend
these artifacts to have the functions mentioned. So why are these functions not the
proper functions of these artifacts? Why are the user’s intentions not efficacious in
the same way the designer’s intentions would be if he were to design a paperweight
or a screwdriver or an incendiary bomb from scratch? Millikan rules this out, but she
does not give any reason for her stipulation. It might seem that she can sidestep this
issue because she defined derived proper function in terms of production relations in
the first place. I must have a desire to open cans by producing (as opposed to utilizing)
a machine to do the job for that machine to have the derived proper function of
opening cans. But this just shifts the problem rather than removing it. A definitional
stipulation should not be arbitrary or ad hoc with regard to the phenomena under
investigation.
One obvious possibility is that the intentions of designers might have some special
cognitive structure or characteristic that sets them apart from the intentions of users.
In his action theoretic account of artifacts Randall Dipert (1993) focuses on the
cognitive processes employed by designers of artifacts. On Dipert’s view, there is a
characteristic and distinctive process from which the artifact derives its function and
other features. Specifically, there is a ‘deliberative history’ in the course of which the
creating agent contemplates the overall function of the artifact as her goal, considers
alternative means for achieving it, and forms a complex set of intentions in the form
of a construction plan. When executed, this plan imposes artifactual features on the
resulting object, chief among them the artifact’s purpose or function (1993, pp.150-
53). The artifact may fail to fulfill this function (1993, pp. 143–50). But the artifact
has the intended function; it is only failing to live up to it. Taking ‘screwdriver’ as
a functional predicate, for instance, you say of a broken or defective screwdriver
that it doesn’t work, or is a bad screwdriver, not that it is not a screwdriver. In short,
the intentions of the designer (or designer/maker – Dipert uses the ambiguous term
‘creator’ throughout) are sufficient to establish the proper function of an artifact in
virtue of the plan construction and execution the designer undertakes.
The problem is that system functional uses of existing artifacts are often the
outcome of exactly the same sort of planful deliberation. You want to pry a large
rock out of your lawn; you don’t have a crowbar; you consider available artifacts like
steel fence posts and wooden rake handles until you find the one that best suits your
purposes. You may even modify it to better suit your purposes – pounding the end
of the steel fence post flat with a sledgehammer so as to get it under the rock more
 The Case of the Recalcitrant Prototype 23
easily, for instance. But if designers’ intentions in the form of a deliberative history
are sufficient for establishing function, and if users must employ relevantly similar
deliberations in order to use artifacts in a system functional way, then why are the
intentions of users not also sufficient for establishing a proper function? It seems
that cognitive characteristics cannot be the criterion that distinguishes designers’
intentions from users’ intentions and underwrites the privileging of the former over
the latter in the establishment of proper function.
Another obvious possibility is that production is creative in a way that use is not.
Designing and making brings something new into the world, whereas use merely
operates with what is already there. But here we encounter a problem analogous
to the one we noticed in Dipert’s case. On the one hand, use is in fact often very
creative. It endows existing artifacts with new, if transient, system functions by using
them in novel ways. And in some cases a system function instituted in this way by
users ultimately becomes a new proper function. For example, the original proper
function of aspirin tablets was to lower fever and alleviate pain, but they are now
so commonly prescribed to prevent blood clots in cardiac patients as well that they
are most reasonably regarded as having acquired the additional proper function of
preventing heart attacks. You might be tempted to say that creative system functional
uses are cases of ‘virtual’ design in order to distinguish them from routine, standard
use. But on the one hand, this only concedes the creativity of use and institutes an ad
hoc move to assimilate creative use to design. And on the other hand, it prompts the
corresponding observation that production is often not very creative, especially in
contexts where what is being designed and produced is a variation on a standard type.
Much design work only changes the form of an artifact, for instance, with little or no
effect on function. Consider the constant redesign of soft drink bottles, for instance.
And much design work is oriented towards the enhancement of an existing function
rather than the institution of a new one. Consider the streamlining of automobiles,
for example. In short, both design and use have a creative dimension, and both
also have a pedestrian, routine dimension. So creativity cannot be the criterion that
distinguishes design from use either.
A third possibility is that design and production involve intentional modification
of materials in forming the artifact, whereas use does not. But in fact many common
use activities involve or result in modification of artifacts, e.g., maintenance
(polishing, washing, sharpening) and repair (mending, patching, rebuilding). These
modifications can be quite extensive, and are typically intended. More importantly,
many artifacts are intentionally and necessarily modified in the very process of use,
for example, perfume and other cosmetics, food artifacts, solvents and other cleaning
agents – in general, anything that gets used up, as opposed to just used. On the other
hand, there are cases of design and production that involve little or no modification
because prototypes can be produced by utilizing existing artifacts as components.
For instance, Les Paul is said to have put together his first electric guitar by using
an acoustic guitar, a telephone mouthpiece, and a radio as an amplifier (Bacon
and Day 1993, p.8). This case does involve some modification in the sense that
the telephone mouthpiece had to be attached to the guitar and wired into the radio.
24 Doing Things with Things
But the modifications were very minor. The basic form and function of the guitar,
the mouthpiece and the radio were not changed, just connected up in a novel way.
So neither the fact of intentional modification nor the extent of the modification
provides a plausible criterion for distinguishing cases of design and production from
cases of use.
The intentions of users do not seem to be significantly different from the intentions
of designers and producers in any relevant way, then. So if the intentions of designers
are sufficient for establishing a proper function, it seems the intentions of users must
be sufficient as well. This conclusion reflects the fact, noted already at the beginning
of this paper, that the roles of designer, maker, and user, while distinguishable, are
not rigorously distinct. Moreover, we can now see that if you want to hold both that
there is a distinction to be had between proper function and system function, and that
the intentions of designers are sufficient for the establishment of proper function, you
also must hold that design and use are rigorously distinct. If not, system functions
are rapidly and exhaustively assimilated to proper functions, and that distinction is
effectively lost for theoretical purposes. Similarly, it will not do to say that creative
use is really a kind of design. This moves the line between design and use, so that
the only activities that count as use are pedestrian employments of artifacts for their
unequivocally proper purposes. But then the distinction between proper function
and system function is erased, because the intentions of designers (a group that now
includes creative users) establish proper functions, and use is defined in terms of the
exercise of proper functions as well, so it is proper function all the way down, so to
speak. In other words, this is just another way of conceding that if the intentions of
designers are sufficient for the establishment of proper functions, so are the intentions
of (creative) users, and no matter which way you do it, the consequence is the loss of
the distinction between proper function and system function.
But conceding that users’ intentions are also sufficient and accepting the
consequences has ramifications that must be laid out in full detail. First, all intended
functions of artifacts, whether initiated by designers or by users, are derived proper
functions if we allow the sufficiency of users’ intentions. Thus all functions of
artifacts, no matter how transiently performed, are what the artifact is supposed to
do. For example, suppose I use my safe deposit box key as a screwdriver to tighten
the screws on my garage door lock. This is now a derived proper function of that
key, a function it is supposed to perform, completely on a par with its function
of unlocking my safe deposit box. In short, we no longer have any theoretical
mechanism for marking the distinction between functions an artifact is designed to
perform and can therefore be criticized for not being able to perform, and functions
an artifact may serve on occasion even though it was not designed to do so and thus
cannot be criticized for failing to do so. But this distinction is a salient feature of the
phenomenology of artifacts and their use (Vermaas and Houkes 2003). So it would
be a signal failing in a theory of artifact function if it were not able to capture this
distinction.
Second, not being able to maintain the distinction between proper functions
and system functions allows the proper functions of artifacts to proliferate in an
 The Case of the Recalcitrant Prototype 25
uncontrolled and disorganized way. Typically the proper functions of artifacts are
regarded as the proper functions of types of artifacts. It is this fact that Millikan’s
theory of direct proper function is designed to capture – in virtue of spoons having
been historically selected and reproduced for stirring food, all spoons have this
as a direct proper function. It is not clear whether the derived proper functions of
particular artifacts would have to be regarded as establishing proper functions for
their type or not. But let us suppose that they do not. Then each exemplar of a
type of artifact would have a different, although slightly overlapping, set of proper
functions. One spoon, having been used on one occasion as a musical instrument and
on another as a weapon, would have the corresponding proper functions; whereas
another indistinguishable exemplar of the same type, having been used once or
twice as a screwdriver, would have that proper function instead. In short, the proper
functions of indistinguishable exemplars of the same type of artifact would be
different depending on their own particular history of use. On the other hand, let us
suppose that every exemplar of a type of artifact has every proper function derived
by any of them. Then spoons, for example, would have not only the proper functions
intended by their designers, but all the functions any spoon user ever intended for any
particular spoon, and these are surely legion. This would be especially problematic
if, as is often held, artifact kinds must be defined in whole or in part in terms of
proper function. Uncontrolled proliferation of proper functions would make such
function-based classificatory systems for artifacts unstable, counterintuitive, or
outright unworkable.
So we are faced with a choice. If we follow Millikan and most other theorists
in holding that designers’ intentions are sufficient for the establishment of proper
function, we will be able to ascribe proper functions to prototypes, even to prototypes
that never work as they were intended to work. But we will have to grant that users’
intentions are also sufficient, and then we will lose the distinction between proper
function and system function for artifacts altogether, and suffer with an uncontrolled
proliferation of proper functions of artifacts in consequence. On the other hand, we
can deny that the intentions of either designers or users are sufficient by themselves
to establish the proper functions of artifacts, and we can rely on the theories of
direct proper function and system function, neither of which requires an appeal to
intentions, for our account of how artifacts get their functions. This option preserves
the phenomenologically crucial distinction between proper function and system
function, and the ways of talking about and thinking about artifacts that depend on
it. But we will not be able to ascribe any functions at all to prototypes that do not
work.
In my view, the second option is clearly preferable. It captures more of the
important features of the phenomenology of our dealings with artifacts. It absolves
us of the thankless task of trying to draw an implausibly sharp line between design
and use. And we can blunt the counterintuitive edge of not being able to ascribe
proper functions to prototypes by noting that our talk about prototypes can just as
easily be interpreted as merely attributing a purpose to the designer, who believes,
hopes, or expects her device will work in a certain way. But this human purpose
26 Doing Things with Things
is not necessarily and without further ado the proper function of the artifact. To
suppose otherwise gives an unfortunate, quasi-theological flavor to the philosophical
enterprise of understanding agents, artifacts, and their relationships.

Conclusion

If designers’ intentions are regarded as sufficient conditions for the establishment

of function, it is unreasonable to deny sufficiency to users’ intentions as well. But
this move threatens the theoretical distinction between system function and proper
function, and results in an uncontrolled proliferation of proper functions. An account
of artifact function in which designers’ intentions are not sufficient can be constructed
by combining Millikan’s account of direct proper function with Cummins’ account
of system function, and it does not generate these problems. So it is deserving of
adoption as a viable alternative to the more traditional intentionalist accounts. The
presence of a few functionless prototype artifacts among us seems a small price to
pay.

References

Bacon, T. & Day, P. (1993). The Gibson Les Paul Book. San Francisco: GPI Books
(Miller Freeman, Inc.).
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72, 741–65.
Davies, P. S. (2001). Norms of nature: Naturalism and the nature of function.
Cambridge, MA: The MIT Press.
Dipert, R.R. (1993). Artifacts, art works, and agency. Philadelphia: Temple
University Press.
Godfrey-Smith, P. (1993). Functions: consensus without unity. Pacific Philosophical
Quarterly, 74, 196–208.
Kitcher, P.E. (1993). Function and design. Midwest Studies in Philosophy, 18,
379–397.
McLaughlin, P. (2001). What functions explain: Functional explanation and self-
reproducing systems. Cambridge, UK: Cambridge University Press.
Millikan, R.G. (1999). Wings, spoons, pills, and quills: A pluralist theory of function.
The Journal of Philosophy, 96(4), 191–206.
Millikan, R.G. (1993). White queen psychology and other essays for Alice.
Cambridge, MA and London: The MIT Press.
Millikan, R.G. (1989). An ambiguity in the notion ‘function.’ Biology and Philosophy,
4, 172–76.
Millikan, R.G. (1984). Language, thought, and other biological categories.
Cambridge, MA and London: The MIT Press.
Plato. (1992). Republic. 2nd edition. [translated by G.M.A. Grube, revised by C.D.C.
Reeve] Indianapolis/Cambridge: Hackett Publishing Company.
 The Case of the Recalcitrant Prototype 27
Preston, B. (1998). Why is a wing like a spoon? A pluralist theory of function. The
Journal of Philosophy, 95(5), 215–54.
Vermaas, P.E. & Houkes, W. (2003). Ascribing functions to technical artefacts: A
challenge to etiological accounts of functions. British Journal for the Philosophy
of Science 54, 261–289.
Walsh, D.M. & Ariew, A. (1996). A taxonomy of functions. Canadian Journal of
Philosophy, 26, 493–514.
This page intentionally left blank
 Chapter 2

Use Plans and Artefact Functions:

An Intentionalist Approach to Artefacts
and their Use
Pieter E. Vermaas and Wybo Houkes

Artefacts are the means as well as the products of intentional human action. Users
intentionally manipulate artefacts such as toasters and screwdrivers for attaining
their goals; designers intentionally create artefacts that users can manipulate for
attaining these goals. Many philosophers therefore seek to describe artefacts in
terms of the intentions of their designers and users: what distinguishes a toaster
from a screwdriver has something to do with the intentions with which humans have
designed or are using them. In addition, many philosophers feel that a philosophy of
artefacts primarily ought to understand artefacts as objects with functionality. After
all, when we describe artefacts we routinely do this in terms of their functions. We,
for instance, classify artefacts in functional terms, such as ‘toaster’ and ‘screwdriver’.
Also when users manipulate artefacts in ways that are different to the use for which
they was designed, we express this in terms of functions: if a screwdriver is used to
extract small objects from cracks, we say that the user manipulates the screwdriver
not as a ‘screwdriver’, but as a ‘lever’. The combination of an emphasis on intentions
and on functionality constitutes what we call the ‘standard view’ of artefacts: artefacts
are understood primarily as functional objects, and their functionality is strongly
connected to the intentions of users and, first and foremost, designers.
In our contribution to this volume we sketch this standard view and then present
an alternative view of artefacts. This alternative endorses the intentionalism that
underlies the standard view, but renounces its strong emphasis on functionality. Our
alternative intentionalism primarily addresses artefact use and design and describes
them in terms of plans, i.e., structured sequences of considered actions, rather than
in terms of (only) functions.
Our alternative view is motivated in part by recent criticism of the standard view.
Beth Preston, for instance, argues in her contribution to this volume that intentions of
designers cannot be sufficient to determine the functions of artefacts, and advocates
a non-intentionalist understanding of the functionality of artefacts.1 It is, however,
our position that the intentionalist intuitions behind the standard view are viable.

1 Beth Preston’s theory of functions can be found in Preston (1998).

30 Doing Things with Things
We have developed our alternative view to demonstrate this viability and to defend
intentionalism as a sufficient, if not necessary, approach to understanding artefacts
and their use and design.
We start by describing some relevant parts of the current situation in the philosophy
of artefacts. We introduce the standard view of artefacts, sketch Preston’s criticism
of intentionalism and present two problems we have with the standard view. In the
next section we go on to describe our alternative view by conceptualising artefact
use in terms of plans. We show that this use need not be understood in terms of
the functionality of artefacts, and in this sense our plan theory is a development of
the intentionalist approach that, in comparison to the standard view, de-emphasises
functionality. Then, we discuss the limitations of planning – in phenomena that we
call ‘embodiment’ and ‘situatedness’ – and counters objections to a plan view, put
forward by Lucy Suchman. Next, we shift our attention from use to artefact design
and to artefacts themselves. We describe design on the basis of our plan theory, and
then draw conclusions about how to understand artefacts and their functionality.
And we defend intentionalism against the criticism and problems raised. Finally, we
consider the consequences of our plan view and comment on the view held by James
Gibson and John Searle that the observation of objects is primarily the observation
of the uses of those objects.

The Standard View of Artefacts and their Use

Artefacts are intimately and rather obviously connected to human intentional action.
Their use is often deliberate, as when we choose a tool to achieve a certain task
or wonder how to use a newly bought household appliance. Moreover, it seems
artefacts can be distinguished from ‘natural objects’ because they are the products of
intentional action. Spelling out this connection between artefacts and intentionality
has been a major task of the philosophy of artefacts so far. Straightforwardly defining
artefacts as the products of intentional action is too simple-minded, as this would
grant the status of artefacts to waste products such as soot and radiation, and to other
‘by-products’ of our actions, such as footprints. More refined definitions have been
proposed, which refer to a specific type of intentional action, usually called ‘design’.
The common core of these philosophical efforts can be called ‘intentionalism’:
general conceptualisations of artefacts, artefact use and design should be built on the
notion of intentional action. Nevertheless, this intentionalism usually is incorporated
in those conceptualisations by referring to intentions, a certain type of mental states
of intentional agents, rather than to the actions themselves.
Many authors combine this assumption with an emphasis on functionality, to
produce what we call the ‘standard view’ of artefacts. On this view, design is described
as an activity aimed at providing means to certain ends, artefacts themselves as these
means, and use as the manipulation of these means to an end. That artefacts play
a role as means to an end is expressed by ascribing a function to them – or even
categorising artefacts in ‘functional kinds’: a toaster is an object with the function of
 Use Plans and Artefact Functions 31
toasting bread; a screwdriver is by its very name a means to drive screws into various
materials. If this emphasis on functionality is combined with the above-mentioned
appeal to intentions, the central task for a philosophy of artefacts becomes to explain
artefact functions in terms of the intentions of designers and users.
There is a further phenomenon that seems relevant to a philosophy of artefacts.
Virtually all artefacts can be used for many different purposes (see also Hysse
Forchhammer’s contribution to this volume). Screwdrivers, for example, can be used
to extract small objects from cracks or to remove staples from stacks of paper. All
of these uses employ dispositions2 of the object, such as its rigidity. On the standard
view, the distinction between the function of an artefact and its other dispositions
should be made in terms of intentions, for example by analysing the function as a
‘privileged’ set of dispositions intentionally produced or selected by the designer
of the artefact for attaining a certain purpose. This difference between the function
and a non-privileged disposition of an artefact shows in a number of ways. For one
thing, it leads to a distinction between proper and improper use, as employing the
function of an artefact versus employing its other features respectively; adding the
connection between functionality and intentions, proper use is use in accordance with
the designer’s intentions. This terminology – or the analogous distinction between
‘proper’ and ‘accidental’ functions – is not just a figment of the philosophical
imagination: Many warranties include a clause that excludes damage caused by
‘improper’ use.
Summing up this brief presentation, we have identified the following elements of
what we call the standard view of artefacts:3

(1) Artefacts are best understood as objects with a function.

(2) The function of an artefact is a set of dispositions of the artefact; this set is
intentionally selected or created by a designer.

2 ‘Disposition’ is used by philosophers to indicate anything that an object does when

acted upon in a certain way; this concept is therefore considerably broader than its everyday
counterpart, which primarily applies to people and their mannerisms (e.g., ‘of a sunny
disposition’) and to possibilities for human action (as in the phrase ‘at one’s disposition’).
Examples of ‘philosopher’s dispositions’ are the disposition of solid objects to cast a shadow
in bright daylight and of a thin layer of ice to crack when someone steps on it.
3 The following quotes all illustrate aspects of this standard view: ‘An object o is an
artefact made by an agent Ag only if it satisfies some type-description D included in the
intention IA which brings about the existence of o.’ (Hilpinen 1992, p.61); ‘Roughly, an
artefactual property is an intentionally-added property that is intended to get us to believe
(in certain ways) that the object possesses certain tool properties [i.e., properties that were
intentionally modified or deliberately left alone]’ (Dipert 1995, p. 129); ‘[…] if the desire is
to produce a certain result by means of making, for example, a certain tool […] then a derived
proper function of the tool (token) […] is to produce this result. Thus, it happens that artefacts
have as their derived proper functions the functions intended for them by their makers […]’
(Millikan 1999, p. 205).
32 Doing Things with Things
(3) Alternative, or improper, use of an artefact is best understood as employing
another set of dispositions – the artefact’s ‘accidental functions’; users are
under normative pressure to use an artefact as intended by the designer of an
artefact, that is, to use it for its (proper) function.

One may doubt the claim, inherent to the standard view, that designers select or create
the set of dispositions that makes up the functions of an artefact. In her contribution to
this volume, Beth Preston develops this doubt into an argument against the standard
view. Preston argues that designers’ intentions are not sufficient to determine artefact
functions. Briefly summarising her exposition: if the designers’ intentions were
sufficient, users’ intentions would also be; after all, there is little or no distinction
between creative use on the one hand and design or production on the other. But if
users’ intentions are also sufficient to determine artefact functions, the distinction
between proper and accidental use collapses, and the classification of artefacts in
terms of their proper functions becomes impossible: there would be no reason to
call a screwdriver ‘screwdriver’ instead of ‘staple remover’. Preston concludes this
argument with a call for a ‘non-intentionalist’ theory of technical functions, i.e., a
theory that does not define functions in terms of designer’s intentions, analogous to
the one that has been developed for the functions of biological items.4 In terms of the
elements of the standard view, she denies (2) and (3), while maintaining (1).
As we have said in our introduction, we regard the intentionalist approach that
underlies the standard view as viable and have developed an alternative intentionalist
view. We present this view in the next sections and defend it against Preston’s
and other criticisms. This does, however, not imply that we are not critical to the
standard view ourselves. We wish to point out two problems of this view, which in
part motivate our alternative view. The first arises because, as we have noted, the
standard view appeals to intentions of agents, rather than to their actions. This leads
to a definition of artefact functions in terms of the intentions of their designers: a
screwdriver has the function to drive screws iff its designer intended it as a driver of
screws.5 The problem is that this definition gives rise to a proliferation of functions,
given additional designer intentions: if a designer intended the screwdriver also as
a cheap Christmas present or regards it as a means to earn money, then on the given
definition ‘to be a cheap gift’ and ‘to earn money’ are functions of the screwdriver – on
a par with its function to drive screws into materials. Moreover, it becomes difficult
to isolate the relevant intentions when artefacts are designed by teams of designers.
We aim at avoiding this problem by conceptualising artefacts in terms of intentional
actions. A second problem of the standard view is that it maintains that artefacts are
to be understood primarily as objects with functions. For designers it most probably
is important to understand artefacts in terms of functions: in design methodology,
for instance, designing is often defined as a process that starts with specific required

4 E.g., Neander (1991), Millikan (1984) and Preston (1998).

5 Millikan’s definition of derived proper function quoted in footnote 3, comes close to
this possibility.
 Use Plans and Artefact Functions 33
functions and that ends with a physical description of an artefact that can perform
these functions.6 But for users this understanding may be unnecessarily full. Users
may take artefacts simply as means for attaining desired ends, without making
the additional step of expressing this as that artefacts have functions. This second
problem is addressed in our view by describing artefact use and design primarily
in terms of plans, i.e., structured sequences of considered actions, by which users
can attain goals, and only secondary (and optionally) in terms of functions. Hence,
in terms of the elements of the standard view, we deny (1) while maintaining (2)
and (3).

A Plan View of Artefact Use

Before presenting our alternative view, we give a hypothetical but illustrative

example of the use of an everyday artefact. Then we discuss how our view can
accommodate the difference between proper and improper use (element 3 of the
standard view) but undermines the idea that artefacts are best understood in terms of
functions (element 1). Later, we extend the alternative view to also artefact design
and to the conceptualisation of artefacts themselves, and show how it can counter
the criticism of Beth Preston, and stop the proliferation of functions due to additional
designers’ intentions.
Suppose Anna has a craving for a piece of toast and decides to use her trusty
toaster. She takes the machine from the cupboard and plugs it in. Then she puts
two slices of bread in the toast rack and pushes down the switch. The toaster has a
small dial that determines when it will switch off. According to the manual, users
should determine which setting of the dial (‘1’ to ‘5’) goes with their preferred ‘level
of browning’ of the toast, put the dial at that level and then wait until the toaster
switches off and the toast rack lifts. Yet what Anna usually does is turn the dial to
some higher-than-desired setting (say, ‘5’). Then, she uses either her experience with
this type of bread or frequent checks of the ‘level of browning’ to determine when
she wants the toasting to stop. At that moment, she turns the knob to the lowest
setting, causing the toaster to switch off and the toast rack to lift. In other words,
Anna uses the dial as a glorified on/off-switch.
By the standard view, with its appeal to intentions, this case would be described
as a series of decisions or intention-formations: Anna’s decision to use a toaster,
to push down the switch, to turn the dial to a higher-than-desired setting, etc. With
each decision, a new intention is formed, which is subsequently realised, after which
another decision is taken and a new intention formed on the basis of the current
situation. The individual decisions are rational and the intentions consequences of
valid ‘practical reasoning’ only as long as the alternative with the highest utility
is chosen.7 Yet this description of artefact use turns out to be inadequate on closer

6 See, e.g., Gero (1990) and Roozenburg and Eekels (1995).

7 The processes of decision-making and intention-formation are described in the
voluminous literature on rational choice theory and theories of practical reasoning. Some
34 Doing Things with Things
inspection. Turning the dial of the toaster to a setting indicating a longer-than-desired
toasting time may be irrational in isolation, given the possibility to turn the dial to
the optimal setting. Yet a natural objection in this case is that the rationality of this
action should not be considered in isolation but as part of Anna’s course of action.
And as part of that course, turning the dial to a non-optimal setting appears to be
rational – or so we maintain.8
It appears to be more accurate to describe this case and artefact use in general
as the realisation of a plan. As a first approximation, a plan can be characterised as
a series of considered actions, execution of which is supposed to achieve a certain
goal. We call the process of constructing a plan ‘deliberation’ and the process of
executing it ‘realisation’. Anna’s plan to use her toaster consists of a number of
steps: plugging it in, pushing down the switch, turning the dial clockwise, waiting
and observing, and turning the dial counter-clockwise. In general, when settling on
a course of action, one considers alternative plans, not alternative actions within
plans – although, of course, it is possible to reconsider steps of plans while realising
them.
On this view, artefact use can be schematically represented as a ‘sequence’ of
actions:9

U.1 The user wants to bring about some state of affairs I, and believes it does
not obtain.
U.2 The user either chooses from a set of available alternatives a use plan P for
bringing about I that involves the intentional manipulation of objects O1,
O2, etc., or he constructs a novel use plan P by deliberation.
U.3 The user believes that the physical circumstances support realising P and
that he possesses the necessary skills.10
U.4 The user intends to carry out P and acts accordingly.
U.5 The user observes I′ as the outcome of P and compares I′ with I.

introductions to these fields are Raz (1978), French (1986) and Audi (1989).
8 This is only one example of theories of practical rationality having to conform to
intuitions concerning rational actions. It may be possible to account for the rationality of
Anna’s dial turning by considering a succession of intention formations; we do not refute
this traditional action-theoretical account. Yet a description in terms of courses of actions
and plans seems more natural. A more sophisticated version of this argument can be found in
Pollock (1995), §5.2.
9 This representation of artefact use is a simplified version of the one presented in
Houkes et al. (2002), §1.
10 Two remarks are in order. First, the beliefs mentioned in U.3 are, of course, used to
assess the relative utilities of the available alternatives or to construct a novel use plan in U.2.
These steps only serve to reconstruct them explicitly. Secondly, the beliefs mentioned are
oversimplified. In many cases, plans will be chosen or constructed on the basis of comparative
and considerably less apodictic beliefs, such as ‘I believe I’m better at carrying out P than P′,
although I’m not quite sure under the present circumstances’.
 Use Plans and Artefact Functions 35

U.6 The user believes that I has been brought about or not. In the latter case, he
may decide to repeat the realisation of P or to repeat the entire U-sequence.
If he repeats U he may reconsider his intended state of affairs I, select
another use plan P′, or do both.

This sequence can be illustrated by means of the toaster example. Anna deliberates
on a way to obtain toast, given her desire for toast and lack thereof. As said above,
this deliberation cannot be understood as merely a sequence of intention-formations,
but is rather the adoption of a single plan consisting of a sequence of intentional
actions for attaining the goal. This plan can be chosen from a (small) set of available
alternatives – using a toaster, buying readymade toast, putting slices of bread in an
oven, etc. – on the basis of Anna’s beliefs about her skills and circumstances. In
most circumstances, using a toaster in an established, instead of newly devised, way
probably has the highest utility of the listed alternatives because it combines a high
chance of success with a small amount of time spent and little chance of overtaxing
Anna’s skills. Once Anna has settled on a plan, she executes it, and judges whether
the result is satisfactory. Suppose the toast comes out distastefully and evidently
burnt. Then she may repeat the execution of the plan with a different slice of bread,
change her plan (e.g., turn the knob at an earlier point or put slices of bread in an
oven), or abandon altogether her intention to eat toast.
We call this the ‘plan view’ of artefact use. And although it is as yet a view of
only artefact use – not of design, which is of central importance to the standard view
– two conclusions can already be drawn. The first is that a detailed description of
the use of artefacts need not mention the functions of the artefacts involved. Our
plan view will in the end also include a characterisation of the functions of artefacts,
but this is motivated by our description of design, not that of use. In the design
context it makes sense to introduce functions. But in the context of use of especially
everyday artefacts11 functions may be taken as redundant. In that sense, we deny the
first element (1) of the standard view.
Another conclusion is that our view accommodates the distinction between proper
and improper use, which was introduced in the previous section. As step U.2 in the
‘use sequence’ shows, use plans may be either constructed by the user herself or
drawn from a set of available alternatives. Some of these alternatives have obviously
been constructed by others, often by designers and sometimes by fellow users. Use
plans constructed by designers are expressed in the user manuals of artefacts and
communicated to potential users by various other means (e.g., explicit instructions,
television ads, features of the artefact (so called use cues), product demonstrations).
Use plans constructed by fellow users are usually communicated orally among users
and by explicit hands-on demonstrations. The concept of proper use can now be
explained by maintaining that use plans for artefacts constructed by designers hold a
privileged position in the set of available alternative plans. Proper use of an artefact

11 The concept of function is most probably helpful when the use of professional
equipment such as aeroplanes is described.
36 Doing Things with Things
is then use that follows a plan approximately identical to the use plan developed
by designers, i.e., their goals are identical and there is an approximate similarity
of included actions.12 Improper use, on the other hand, is use that follows a plan
constructed by the user herself. And use that follows plans developed by fellow users
may be seen as filling up a spectrum between proper and improper use.
If we explain proper use in this way, one can draw on the intrinsic normativity of
plans to make a distinction between proper use and what could be called ‘rational’ use. To
illustrate the normativity, suppose someone undertakes the same actions when writing
poems for profit as another who writes poems for fun, such as jotting down some lines
in the weekends, occasionally reading the results to family and friends, and sending
the most successful ones to a magazine once a year. Although this is a rather good plan
for writing poems for fun, it is exceptionally bad as a plan for writing poems for profit
– if there is a good plan for doing so. Plans are subject to many standards. One of the
most straightforward ones is that plans should be likely to lead to the desired state of
affairs, or at least more likely than other plans. This standard of ‘appropriateness’ of a
plan can be formulated in terms of a calculation of utilities of several available plans.13
Some other standards include ‘means-end coherence’, which requires the person who
executes a plan to obtain necessary means (such as slices of bread for making toast)
and construct relevant sub-plans, and ‘belief coherence’, which means that a plan
should be based on beliefs held by the person realising it (such as U.3 in the sequence
presented above).14 These standards on plans allow for a distinction between good
(rational) and bad (irrational) use, which is independent from the distinction between
proper and non-standard use. Users may assume that proper use is always rational use
as well – designers are on the basis of their technical knowledge and skills, supposed
to construct good use plans. And it is probably safe to say that most of such use plans
are indeed rational. But even use plans by designers can be or become irrational, as
is illustrated by the designed use of asbestos plates to make walls fire-resistant. And
improper use plans developed by users themselves are often rational, as was illustrated
by Anna’s use of the toaster. One may now interpret the ‘normative pressure’ put on

12 The identity conditions for plans are hard to spell out in detail. Making toast with a
toaster and making it by putting slices of bread on a hot roof are obviously different plans,
although they are aimed at the same goal; writing poems for fun and writing them for profit
are different plans, although they may include the same actions. And in many cases, two
different sets of actions may count as the same plan. Is driving a car and checking the tyre
pressure before starting the car a different plan than checking the tyre pressure when the car
has been started? Or even different than checking this pressure only when one thinks there
might be something wrong with the car? This leniency in the identity conditions of plans also
affects the notion of proper use. Suppose the car manual says that one has to check the tire
pressure before starting the car. It seems implausible to maintain that this by itself is sufficient
to characterise checking the tire pressure after starting the car as improper use of the car. Yet
this intuition may be based largely on certain connotations of the term ‘improper’, such as
‘potentially harmful’ or ‘damaging to the artefact’.
13 Cf. Pollock (1995), §5.3 for a method of evaluating plans by calculating utilities.
14 These standards are taken from Bratman (1987), §3.2.
 Use Plans and Artefact Functions 37
users (element (3) of the standard view) as a pressure to follow use plans that are
rational. This interpretation does change the purport of the pressure since, as we have
said, rational use plans need not cover all use plans constructed by designers and does
cover many use plans constructed by users themselves. Specifically users are on this
interpretation not pressured to follow irrational plans of their designers, a consequence
one may easily accept.
So the plan view accounts for the phenomenon of (im-)proper use in a rather
sophisticated way, perhaps outperforming the standard view in this respect. Yet this
account still presupposes that one can make a distinction between the designers of
artefacts and their users. Preston’s criticism of the standard view seems to show that
such a distinction cannot be properly made. However, on the basis of our plan view
of the design of artefact we are able to differentiate designers from users of artefacts,
as we show when we present our view of artefact design.

Objections to Planning

Before turning to design and artefact functions, we first consider an obvious objection
to the plan view: does it not overemphasise the role that deliberation actually plays
in our dealings with artefacts? In this section, we argue that there are limitations to
planning, but that these limitations – which we call embodiment and situatedness –
can be incorporated in the plan view of artefact use. A secondary aim of this section
is to counter arguments to the effect that a plan view is fundamentally mistaken.
Arguments to this effect have been put forward by Lucy Suchman, whose primary
purpose is to show that the ‘planning model of agency’ is incapable of describing
interactions between agents and has led to seriously flawed models of human-machine
interaction. She proposes to replace these theories with an appeal to ‘situated action’,
a notion inspired by the phenomenological tradition of Merleau-Ponty and the early
Heidegger, as interpreted by Hubert Dreyfus.15
It may seem impossible to specify the limitations of planning and deliberation
in general. For one thing, some people tend to think a lot before they act, whereas
others prefer to act quickly and think later. Moreover, in some cases, such as kicking
a stone to let off steam, deliberation makes little sense, whereas in others, such as
constructing a nuclear power plant, it seems one can hardly deliberate enough. We
think that, nevertheless, there are two general limitations to planning.
Once again consider the use of a toaster. Above we gave a rudimentary reconstruction
of this example of artefact use in terms of a use plan: Anna gets the toaster, plugs it in,
puts slices of bread into the toast rack, pushes down the switch, etc. Most steps of this

15 Suchman only refers to Heidegger and Merleau-Ponty as cited by Hubert Dreyfus; cf.
Suchman (1987, pp. 53–54). The theme of ‘situated intentionality’ is developed in Dreyfus’
(1991) influential commentary on the First Division of Heidegger’s Being and Time and many
of Dreyfus’ later papers. We do not go into Suchman’s proposed alternative here, besides
noting that her notion of ‘situatedness’ differs from the one presented in this paper – in fact, it
includes both embodiment and situatedness in our sense.
38 Doing Things with Things
plan can themselves be constructed as a number of steps, e.g., the ‘sub-plan’ of getting
slices of bread. Yet for other steps, such a construction would be contrived. Anna does
not, for instance, deliberate on how to push down the switch. In any plan, one necessarily
reaches a point at which deliberation ends, because further deliberation becomes strongly
counterproductive – no one will deliberate for a week on what to eat – or because
further sub-planning is infeasible, e.g., for simple bodily movements. Furthermore, one
does not, in general, plan everything one is supposed to do in advance, even if this
would be possible in principle. To allow for mistakes in deliberation or the beliefs on
which it is based, to accommodate changes in the situation and other eventualities,
plans are usually not complete. In the toaster case, for instance, Anna probably does not
deliberate on where to put the toaster – she takes a decision while realising the plan. The
first limitation – which we call ‘embodiment’ – is based on a principled incompleteness
of planning: it is impossible to specify all included actions in all details. The second
limitation – which we call ‘situatedness’ – is based on incompleteness in practice: plans
seldom specify all actions regarding a situation.
Our plan view can accommodate both phenomena – without reducing them to
aspects of planning. Let us start with the former, the limitation of embodiment.
One way to spell out this embodiment is to acknowledge that in realising plans, we
necessarily draw on mental and motor skills. Even for the simplest plans, containing
only one step, this is the case. Throwing a ball to someone is realised by taking
one action, but this action combines several motor skills. Most cases of artefact use
ultimately appeal to elementary motor skills without explicitly referring to them:
a user manual of a toaster that specifies how the user should hold her hand when
pressing the switch of the toast rack would be odd indeed. And this ‘skill gap’ is not
closed by deliberation on the part of the user either. So some details of artefact use
are not determined by the communicated plan or autonomous deliberation by the
user; skills take over at some point. This goes not just for elementary motor skills,
which are involved in virtually every plan, but also for more sophisticated skills,
such as operating the gears and clutch of a car (which one does not expect to be
explained in the car manual).
The importance of skills for artefact use can be represented in the use sequence
presented in the previous section: the user of an artefact has a belief about his skills
(as part of U.3) and stops deliberating as soon as he thinks he knows ‘how a certain
thing is done’. If he in fact possesses the skills he thinks he has – i.e., has ‘embodied’
these skills – and if these skills are the ones presupposed by the designer of the
artefact, this decision to stop deliberating and start drawing on skills is a rational
one; otherwise, it is not. The rationality of this decision depends both on the executor
of the plan and on the nature of the task. In the case of a toaster, little harm is done
if it turns out one does not in fact know how to stop the toasting process; but if one
uses a parachute, it may be best to double-check whether one knows the ropes before
drawing on possibly non-existent skills.16

16 Skills are related to, but different from what ‘ingrained’ action patterns – which
can also be described on our plan view. In toasting bread or putting on clothes, most users
 Use Plans and Artefact Functions 39
This phenomenon of embodiment is different from that of situatedness, to
which we now turn. Apart from their incompleteness regarding skills, plans are
also incomplete in practice, to make them more readily executable. Sometimes one
is, for instance, not sure about the circumstances in which a plan is realised, or
about the stability of these circumstances. In those cases, it makes little sense to
construct an elaborate plan for using an artefact – even if one is so inept that ample
deliberation is required to realise the plan. Plans may turn out to be worthless in later
situations. What is more, cognitive resources may be wasted on attempts to continue
acting along the lines of the original plan. Hence, it makes more sense to construct
a ‘lightweight plan’, which specifies only the starting actions and a broad route that
leads to the goal state.17
This practical, rather than principled incompleteness makes plans more adaptable
to the situation in which they are realised. Suppose one decides to use a toaster and
includes in one’s use plan a step of turning the knob to the level ‘3’ (out of five
levels), only to find that one’s spouse has bought a new, more sophisticated toaster
with ten levels. Then one has to revise one’s plan before using this toaster. If, on the
other hand, one uses a toaster by turning the dial to the maximum level and turning
it back once the toast is sufficiently browned, there is no need to revise the plan.
Hence, the second, less detailed plan is more flexible and more ‘situated’, i.e., more
responsive to the details of the situation in which it is realised.
This responsiveness can be accommodated in the plan view. As was argued in
the previous section, plans are based on beliefs that the agent has about herself,
the artefact, and the environment. These beliefs enter into the standards for rational
planning; in this way, we plan our actions in a context-sensitive way. A plan to
make fire with a cigarette lighter in the middle of a storm may be purposeful, but
it is nevertheless a bad plan. Moreover, additional information can lead to quick
revisions of a plan, given sufficient sensitivity to the context. Previous planning may
even, through focussing mechanisms, be helpful in acquiring relevant beliefs about
the environment, which can then be used to revise or complete the plan. In this way,
planning and situatedness may support each other.
Suchman’s arguments against planning conclude impotence from limitations.18
She, for instance, considers the example of a canoeist who has to navigate a series of

probably go through a more or less fixed routine. Both the goal and the method of attaining
it with a certain artefact are so familiar that the execution of the corresponding use plan has
become rote. The difference between routine and skilled use shows in the drawbacks of the
former. Routine users can easily make mistakes when confronted with an unfamiliar aspect
of the standard situation (‘breakdown’) or with a slightly different specimen of the artefact.
Furthermore, routine use may be sub-optimal from the start – many users standardly use
artefacts in an inefficient way.
17 A similar point about the practical importance of lightweight planning is made in
Leudar and Costall (1996), who use a ‘dialogical’ notion of plan different from ours. We lack
the space for a critical comparison of these two approaches.
18 It is important to note, moreover, that Suchman uses a different definition of ‘plan’ than
ours. According to Suchman, a plan is a series of considered actions that can be mechanically
40 Doing Things with Things
rapids and concludes that: ‘When it really comes down to the details of responding
to currents and handling a canoe, you effectively abandon the plan and fall back
on whatever embodied skills are available to you’.19 In our plan view, the canoeist
example is an extreme case in situatedness and embodiment – responsiveness to the
particular circumstances, by means of exercising exceptional skills, is of the essence
to realise the goal state. Yet even here, constructing a lightweight plan – a choice for
a point to enter the rapids, or a rough outline of a route – makes sense. Moreover,
most cases of artefact use are far less adventurous; and if they are, they resemble
rock climbing more than wild-water canoeing: the circumstances in which we most
often use artefacts are relatively stable and do not require exceptional skills.
An analogous argument builds on situatedness alone. Suchman notes that ‘[…]
we constrain and direct our actions according to the significance that we assign to
a particular context. How we do that is the outstanding problem. Plans and goals
do not provide the solution for that problem, they simply restate it’.20 Our analysis
of situatedness points out that planning does not take place in a void, but is based
on beliefs. This may not solve the problem, and it does show that planning must be
balanced with other cognitive strategies, but it does more than ‘restate’ the problem: it
indicates the conceptual resources that can be used to describe the context-sensitivity
of human action. Hence, we conclude that Suchman’s arguments fail to show that
planning in the sense described here is not just limited, but ineffective.21
Summing up, in this section, we have offered a sketch of an integration of the
phenomena of embodiment and situatedness within our plan view. Much work
is still to be done to show exactly how an appeal is made to skills in the context
of constructing a plan and how deliberate actions merge seamlessly with situated
actions. Furthermore, situatedness and embodiment are surely far richer phenomena
than described here. But it has at least been shown in what way embodiment and
situatedness can be accounted for in our plan view, namely by imposing limitations
on the practical and principled possibilities of planning. By explicitly stating these
limitations, general arguments against the role of planning in human action lose their
bite.

implemented (1987, p. 52) – all cognitive resources are spent on planning, and none on
execution. Hence, any kind of incompleteness discredits plans as generative mechanisms. But
surely this notion of ‘plan’ is overly strict, although it may be close to that used in cognitive
science and the planning approach in AI; see, for instance, the highly influential Miller et
al. (1960). Our notion is more liberal; for this reason alone, it is already less likely to be
vulnerable to Suchman’s criticisms.
19 Suchman 1987, p. 52; emphasis added.
20 Suchman 1987, pp. 47–48.
21 David de Léon (private communication) has suggested that Suchman’s arguments can
also be taken to show that plans are not the generative mechanism of actions – although
they may play a part in subsequent reconstructions of our actions (1987; p. 39). Yet such
an alternative view should be supported by independent arguments against planning as a
generative mechanism. As indicated, we do not think Suchman supplies valid arguments.
Thus, her alternative lacks an argumentative foundation.
 Use Plans and Artefact Functions 41
A Plan View of Artefact Design and of Artefact Functions

We have now presented our plan view of artefact use and we have developed it
by accommodating the phenomena of embodiment and situatedness. In this section
we continue by presenting our view of artefact design and of artefact functions.
Since our approach is action-oriented, we first extend the plan view to the process
of designing. We take this as a process that aims primarily at the development of use
plans, and, subordinately, at designing objects.
Because we describe the use of artefacts as the realisation of use plans, it seems
natural to describe the design process that leads to those artefacts as the construction
of such plans. The concept of a ‘use plan developed by designers’ was introduced
earlier to account for the distinction between proper and improper use. Focusing
on the role of this plan, designing can be represented by the following sequence of
actions:22

D.1 The designer wants to contribute to a user’s goal of bringing about a state of
affairs I.
D.2 The designer believes that the state of affairs I′ is the closest consistent and
viable approximation of I, and intends to contribute to bringing about I′.
D.3 The designer believes that if the users follow an appropriate use plan P
that involves the manipulation of objects O1, O2, etc., this will contribute to
bringing about I′, and intends to construct this plan P and to communicate
it to the users.
D.4 The designer intends to contribute to producing the objects Oi, Oj, etc., that
do not yet exist by designing23 them, and acts accordingly.
D.5 The designer intends to communicate P to the users, and acts accordingly.
D.6 The designer believes that I′ can or cannot be brought about by the users
to whom P is communicated. This belief is based on the observation that
some users go through a sequence of actions P′ and bring about I′′, and on
a comparison of I′′ with I′.
D.7 The designer decides that his goal to contribute to bringing about I′ has
been achieved or not. In the latter case, he may decide to repeat the entire D
sequence, settle on another plan (return to D.3), redesign at least one of the
objects Oi, Oj, etc., (return to D.4) or re-attempt communication (return to
D.5).

By means of this description we can define ‘artefact’ and ‘artefact function’. By step
D.3, the use plan P that is constructed by the designer involves the manipulation
of objects. Some of these objects, denoted by Oi, Oj, etc., may not exist. Hence,

22 This representation of designing is a simplified version of the one presented in Houkes

et al. (2002), §2.
23 ‘Designing’ is here used in the broad sense of describing the objects in words, pictures,
gestures or some combination of these, along with instructions for producing the objects.
42 Doing Things with Things
they should be produced if users are to carry out the designed plan. In step D.4, the
designer contributes to the production of these objects by describing them and the
way in which they can be manufactured. We now define artefacts as objects that are
designed and manufactured as part of the construction of a use plan. The definition
of a function is somewhat more involved.24 The whole plan P has the purpose of
bringing about a state of affairs I′ and the plan is supposed to generate a transition
from an initial state of affairs to I′. An explanation of this capacity of P singles out
dispositions of the objects O1, O2, etc., used in P, by referring to those dispositions
as the ones that contribute to P’s capacity to bring about I′. These singled-out
dispositions are the functions of those objects.
Let’s get back to the toast to illustrate all this. By our description, the original
designer of the toaster did not start by intending to make an object that can toast
bread. Instead the designer started with the intention to help people with a taste for
toasted bread by constructing a plan with which these people can acquire toast (step
D.1). The designer may adjust or refine this desire, for instance, by requiring that the
size of the toast fall within certain limits (step D.2). After that the designer actually
develops the use plan as a series of actions he or she thinks the prospective users
can carry out. These actions involve using an object that heats bread, and since this
object did not exist in the pre-toaster era, the designer describes this object as part
of the construction of the use plan (D.3 and D.4). Finally, the designer checks the
resulting plan by, among other things, explaining it to some of the considered users
and determining whether they manage to acquire toast with the plan (D.5 and D.6).
We guess the toaster was an immediate success, but if not, the designer would have
returned to some of the steps in the design sequence to improve on the design of
the plan or the toaster (D.7). Based on the definition given above, the toaster that is
designed and manufactured as the result of this plan is an artefact. The disposition
of the toaster to slightly roast the surfaces of slices of bread is its function, because
that disposition (in contrast to other dispositions such as that to generate an awful
smell if cheese falls in) is referred to when it is explained why toasted bread can be
generated by the plan.
On the basis of this plan view of artefact design and the proposed definitions we
can address Beth Preston’s criticism of intentionalism and show how to avoid the
additional problems of the standard view discussed above.
Firstly, we can make a clear distinction between users and designers of artefacts.
Users primarily realise use plans for attaining the goals associated with these plans.
Designers develop use plans that prospective users can carry out for attaining
the associated goals, and designers check whether they have achieved this by
communicating the developed use plans to users and by observing if these users do
attain the associated goals with the plans. By the use sequence (step U.2) given in
above it is acknowledged that users themselves may also construct use plans, but this
does not automatically turn them into designers: users usually do not communicate

24 In Vermaas & Houkes (2003) we discuss existing theories of artefact functions and
present our own, plan-based, theory.
 Use Plans and Artefact Functions 43
the plans they construct to other users and do not check whether these plans are
(still) successful if others carry them out; designers are by the design sequence (steps
D.5 and D.6) required to do so. Some users may actually communicate a use plan
they have constructed to fellow users but this does not undermine the principled
distinction between users and designers. It merely proves that the class of users can
be broken up in subclasses: passive users that carry out existing use plans (users
that drive screws with screwdrivers); creative users that construct their own use
plans (Anna with the toaster), and innovative users that construct their own use
plans and communicate them to other users (Anna if she starts to inform other toast
consumers of her use of the toaster). And if those latter agents also check whether
their newly constructed and communicated plans have the intend results, they stop to
be innovative users and become designers on our plan view (Anna, if she also checks
whether those other toast consumers achieve their required ‘level of browning’ with
Anna’s use plan).
Secondly, we can avoid the proliferation of functions due to additional intentions
of designers. On our view, the functions of an artefact correspond to the set of
dispositions of the artefact that are singled-out or ‘highlighted’ by the explanation
of why the use plan for the artefact leads successfully to its associated goal: the
functions of an artefact correspond to the set of dispositions that contributing to
this success. This determination of the functions of an artefact is independent of the
specific intentions that the designer of the artefact had: the function of the toaster
is still its disposition to slightly roast the surfaces of slices of bread, regardless of
whether the original designer of the toaster, in addition to his or her intention that
the toaster has this disposition, wanted to earn money with it, or become known as a
great benefactor of mankind.
Thirdly and finally, we can defend intentionalism, if not the standard view,
against Preston’s criticism. Our determination of functions is different from the
simple creed that designers’ intentions determine functions. In that sense Preston’s
criticism does not to apply. Moreover, on our plan view we can, as argued above,
make a clear distinction between designers and users. Hence, if we define the proper
functions of an artefact as those dispositions of the artefact that are highlighted by
the designer’s explanation of the success of the use plan of the artefact, we do not fall
prey to the conclusion that users by their (explanations of) improper use of artefacts
also determine proper functions of artefacts. Instead we can define accidental
functions of an artefact as those dispositions of the artefact that are highlighted by
the explanations of the success of use that follows a plan constructed by the users
themselves, and maintain that these functions are in general different to the proper
functions. One can quarrel about whether the dispositions of artefacts highlighted by
explanations of the use plans developed by users and communicated to fellow users
are to be characterised as proper or accidental functions. These plans may earn the
same status as the use plans developed by designers, especially when the can be taken
as rational. On the other hand, if such a use plan is clearly irrational, people may not
be inclined to call it proper use nor the dispositions highlighted proper functions.
So, in terms of the examples, to drive screws is the proper function of a screwdriver
44 Doing Things with Things
and to pierce is an accidental function if a user incidentally uses a screwdriver for
making a hole in plasterboard. And to open tins of paint may be taken as a proper
function even though screwdrivers are not designed for this use.

Artefacts and their Functions

Our definitions have some surprising consequences for the understanding of artefacts.
First, objects produced by intentional agents outside the context of constructing a use
plan are not considered artefacts. Hence, objects made absent-mindedly, such as a
little pile of sand swept together at the beach during a lively conversation, or a fancy
but useless alloy designed just because nobody thought it was possible to make it,
do not count as artefacts. Secondly, artefacts exist relative to a particular designed
use plan. In that sense one can take them as relational, i.e., as being a particular type
of artefact with respect to that plan. This would mean that the toaster is a toaster
relative to the toasting plan, and not relative to any other plan, say to warm one’s
hands with it. Many artefact names reflect this relativity to some extent. We return
to this in a moment. Plan-relativity does not extend to artefactuality as such, which
would be counterintuitive. In a way, designed objects are more intrinsically artefacts
than they are toasters or screwdrivers: once an object is designed and manufactured
as the result of the construction of a use plan, it becomes an artefact independently of
which use plan one is considering, as long as it is clear that it was constructed to play
a role in some use plan. Thirdly, functions of artefacts are also relative to use plans.
A screwdriver, for instance, has the function to drive screws relative to the use plan
to install or remove screws, and the function to lever relative to the plan to remove
little items from cracks.
On our plan view, artefacts can still be taken as objects with functions. But
artefacts are better understood as objects subsidiary to actions included in use plans.
Functions are defined relative to these use plans. In the context of designing it is
helpful to speak about functions because a function of an artefact or, more generally,
of an object, refers to those dispositions of the artefact or object, respectively, that
the designer puts into action to create a use plan. The use plan as a whole is supposed
to bring about some state of affairs. The designer can, roughly speaking, construct
this plan by choosing actions with specific effects for which it holds that when
the actions are carried out sequentially, their effects result in this state of affairs.
The designer then has to devise the individual actions with their specified effects.
One of the means the designer has for doing this is choosing an object or artefact
with certain dispositions that give rise to a specified effect when employed. The
chosen dispositions are then by definition the functions of that object or artefact,
respectively.
In the context of artefact use, one can introduce functions in the same way as
for artefact design. Everyday language reflects this. Names of some artefacts are
functional, such as toaster, screwdriver, shoe polish, and computer. However,
everyday language does not confirm that function is an all-pervading concept.
 Use Plans and Artefact Functions 45
Many artefacts are not named by the functions that can be ascribed to them: bread
is not called ‘hunger-killer,’ paper is not called ‘sign-carrier’, nitro-glycerine is not
‘explosion-fluid’, and compact discs are not called ‘music-carriers’.25 Moreover, if
we discuss in everyday life how artefacts can be used, we need not mention functions.
An inquiry about the use of a strange artefact on display in a shop can be formulated
as ‘what is the function of this thing?’ And the answer can be ‘Its function is to
absorb fluid and smearing it over a surface.’ But it can also be formulated as ‘what is
this thing for?’ and answered by ‘it is for applying paint’. If our plan view of artefact
use is right, users are primarily interested in goals they can achieve with artefacts,
and in the plans they have to realise for doing so. The question ‘what is this thing
for?’ should then be interpreted as the question ‘how and for what goals can one use
this thing?’ And the answer should be taken as ‘you can use it for achieving goal
I by carrying out plan P’. One advantage of this latter answer is that it provides
users with the information they need, whereas the answer in terms of the function
generally does not. Consider a coffee percolator. Telling somebody unfamiliar with
percolators that it has the function to make coffee with does reveal the goal that can
be achieved by this artefact, but does not tell much about how to achieve it. Saying
that the coffee maker is for making coffee by putting a filter in the percolator, filling
the filter with an amount of grained coffee, and finally pouring boiling water on that
grained coffee, is more informative.
Yet functional lingo is of some help in communicating use plans. By naming
the functions of artefacts, designers (or fellow-users) provide some information
to users about how they may use the artefact. This communication works better if
the functional description is unambiguous, i.e., if the artefact is used in use plans
relative to which it has mostly the same function. So, if our plan view holds, then
artefacts that have functional names are mainly artefacts that have one dominant
use. Examples are toasters, corkscrews, and coffee makers. Artefacts that have
different functions relative to different plans are, on other hand, mainly not named
functionally.26 Examples are concrete, boards, flour, and compact discs.
This discussion on functions has some bearing on the view, taken in other
contributions to this volume, that the observation of objects by agents is primarily the
observation of the uses those objects have for the agents, instead of the observation
of the objects’ physical and chemical properties. One finds this view defended in
– otherwise quite different – works by James Gibson and John Searle. According
to Gibson’s ecological approach to visual perception,27 objects and substances

25 In everyday language there are also cases of artefacts with non-functional names that
became functional. An iron, for instance, was most probably named after the material used.
But one now can say that ironing is its function.
26 Some functional names are ill chosen, if artefacts have acquired new functions. The
Dutch fire brigade (to give an example of a social artefact) has, for instance, recently launched
an expensive advertising campaign to inform people that it does more than extinguishing fires.
Changing its name to a non-functional one would probably be even more costly, in many
respects.
27 Gibson (1979), especially chapters 3 and 8.
46 Doing Things with Things
have ‘affordances’. Gibson defines an affordance of an object or substance in the
environment of an animal – this animal may also be a human being – as that ‘what
it offers the animal, what it provides or furnishes, either for good or ill.’28 These
affordances are taken to be properties of the object or substance concerned and can
be perceived directly by the animals. An open environment, for instance, affords
locomotion to pedestrian animals, water affords drinking, bathing and drowning, fire
affords the cooking of food substances and the glazing of clay, and scissors afford
cutting.29 Searle, on his turn, remarks that we more easily learn to perceive things as
functional objects than as material objects stripped of functionality. So we learn to
see full bathtubs as objects with the function to bathe in instead of ‘enamel-covered
iron concavities containing water,’ and money as an object with the function to store
value instead of ‘cellulose fibers with green and gray stains.’30
It is not our intention to evaluate the tenability of the view that agents can (learn
to) perceive immediately the uses of objects, which belongs to the philosophy of
perception rather than that of artefacts. Instead we wish to point to an ambiguity in
the formulation of this view. By our plan view of use, there is a difference between
perceiving the function of an object and perceiving how an object can be used. The
latter means that observers perceive an entire use plan by which the object can be
used; whereas the former means that observers perceive merely the function the
object has relative to a plan. For most of the examples Gibson and Searle give, this
ambiguity does not surface. Perceiving that water has the affordance to be drunk or
the function to quench thirst, leads rather straightforwardly to the plan to quench
thirst by swallowing the water. Perceiving a full bathtub as affording bathing or
as having the function to bathe in leads similarly to the plan to step in the tub and
start washing. But for the fire example by Gibson, the ambiguity starts to play up.
Perceiving that a fire has the affordance to glaze clay, may lead to the conclusion
that fire has ‘glazing clay’ as a function relative to a use plan that has glazed clay
as its goal. But this perception does not yet inform the observer how to use the fire
for glazing clay. In the case of artefacts that can have more than one function, the
ambiguity is also clearly present. By looking at a pile of iron beams one may perceive
that they afford building a duplicate Eiffel tower or that they have this function. But
this certainly does not reveal how to use those beams for actually building such a
tower. This ambiguity can of course be removed by just choosing one of the options
available. If one takes the position that observers of objects perceive whole use
plans as, one could argue, Gibson is doing, the view becomes more daring: when
observing beams, observers indeed perceive the plans by which they can be used,
for instance, to construct towers. And if one takes the position that observers of
objects perceive only their functions as, one could argue, Searle is tending to, the
view becomes more mysterious: how can one perceive that fire has the function to

28 Gibson (1979, p. 127), his emphasis.

29 Gibson (1979, pp. 36–40).
30 Searle (1995, pp. 4–5, 14). Presumably, Searle regards only dollar bills as real
money.
 Use Plans and Artefact Functions 47
glaze clay without having knowledge of the use plans to glaze clay relative to which
fire has this function?

Conclusions

In this chapter, we defended an intentionalist approach in the philosophy of artefacts.

We identified a standard view of artefacts and artefact use and design, which
describes artefacts in terms of their functionality and builds a strong connection
between artefact functions and designers’ intentions. We showed that this standard
view could be criticised, but that its intentionalism can be salvaged by an alternative,
which we called the plan view. We first described artefact use in terms of use plans
and countered criticisms to a plan view by showing that purported limitations to our
theory, which refer to skills and situated action, can be accommodated within it.
Then we went on to extend our view to artefact design and artefacts. Designing was
described as the construction of a use plan and possibly some subsidiary objects. We
refer to these objects as ‘artefacts’ and ascribe functions to them via explanations
of use plans. Function-talk turns out to be useful in the context of designing, but
of limited use in the context of using. In the latter context, it mainly serves to
communicate to users part of the designed use plan of artefacts designed for a single
purpose. In this way, we have described artefacts and their use and design in terms
of plans and thus salvaged the intuitively plausible intentionalist approach in the
philosophy of artefacts.

Acknowledgements

We would like to acknowledge the participants of the two ‘Doing Things with
Things’ meetings for valuable and pleasant discussions and we would like to thank
Beth Preston, David de Léon, Gustavo Ribeiro, and the Editors for their helpful
comments. Research for this chapter was part of the program ‘The Dual Nature
of Technical Artefacts’, which is supported by the Netherlands Organisation of
Scientific Research (NWO).

References

Audi, R. (1989). Practical reasoning. London: Routledge.

Bratman, M. (1987). Intentions, plans and practical reason. Cambridge, MA:
Harvard University Press.
Dipert, R.R. (1995). Some issues in the theory of artefacts: Defining “artefact” and
related notions, The Monist, 78, 119–135.
Dreyfus, Hubert L. (1991). Being-in-the-world. A commentary of Heidegger’s Being
and Time, Division I. Cambridge, MA: MIT Press.
French, S. (1986). Decision theory: An introduction to the mathematics of rationality.
Chicester, UK: Horwood.
48 Doing Things with Things
Gero, J.S. (1990). Design prototypes: A knowledge representation schema for design,
AI Magazine, 11(4), 26–36.
Gibson, J.J. (1979). The ecological approach to visual perception. Boston: Houghton
Mifflin Company.
Hilpinen, R. (1992). On artefacts and works of art, Theoria, 7, 58–82.
Houkes, W., Pieter E. Vermaas, P. E., Dorst, K., & de Vries, M. J. (2002). Design and
use as plans: An action-theoretical account. Design Studies, 23, 303–320.
Leudar, I. & Costall, A. (1996). Situating action IV: Planning as situated action,
Ecological Psychology, 8, 153–170.
Miller, G.A., Galanter , E. & Pribram, K.H. (1960). Plans and the structure of
behavior. New York: Holt, Rinehart, and Winston.
Millikan, R.G. (1999). Wings, spoons, pills, and quills: A pluralist theory of function,
Journal of Philosophy, 96, 191–206.
Neander, K. (1991). The teleological notion of ‘function’, Australasian Journal of
Philosophy, 69, 454–468.
Pollock, J.L. (1995). Cognitive carpentry: A blueprint for how to build a person.
Cambridge, MA: MIT Press.
Preston, B. (1998). ‘Why is a wing like a spoon? A pluralist theory of function’, The
Journal of Philosophy, 95, 215–254.
Raz, J. (Ed.) (1978). Practical reasoning. Oxford: Oxford University Press.
Roozenburg, N.F.M. & Eekels, J. (1995). Product design: Fundamentals and
methods. Chichester: John Wiley & Sons.
Searle, J.R. (1995). The construction of social reality. New York: Free Press.
Suchman, L.A. (1987). Plans and situated action. Cambridge: Cambridge University
Press.
Vermaas, P.E. & Houkes, W. (2003). Ascribing functions to technical artefacts: A
challenge to etiological accounts of functions. British Journal for the Philosophy
of Science 54, 261–289.
Vermaas, P.E. & Houkes, W. (2006). Technical functions: A drawbridge between the
intentional and structural natures of technical artefacts, Studies in the History and
Philosophy of Science, 37, 5–18.
 PART II
Things in the World of the Child
This page intentionally left blank
 Chapter 3

Autism and Object Use: The Mutuality

of the Social and Material in Children’s
Developing Understanding and Use
of Everyday Objects
Emma Williams and Linda Kendell-Scott

When comparing human interaction to interactions amongst other animal species,

social scientists have tended to centre their attention solely on language. Such a
narrow focus results from a tendency to limit conceptions of the ‘social’ to only
person-to-person interactions, as if we lived in a ‘a world of actors devoid of
things’ (Joerges, 1988, p. 220). Recently, a number of researchers (Costall, 1995;
Tomasello, 1999; Preston, 2000; Williams & Costall, 2000) have begun to question
this restriction, contending that it is problematic for a number of reasons: it fails
to recognise the far-reaching impact that commonplace artefacts (such as eating
utensils) can have in shaping everyday human activity, including interactions with
other people; it ignores the fact that we encounter objects in a social context; and
neglects the role that other people play in introducing children to the ‘proper’ use of
objects. Schiffer (1999) takes this argument a step further, proposing that the most
defining attribute of human beings is not in fact language (though this is obviously
important), but rather our daily interactions with innumerable kinds of things:

... the terms ‘interpersonal’ interaction and ‘social’ interaction which permeate publications
in the social sciences, are misnomers, for nearly always artefacts – at the very least
body paints, clothing and ornaments – accompany individuals in every interactional
setting. Humans ... interact not with other humans per se, but with artefacts and humans
compounded with artefacts …What is singular about homo sapiens is the constant intimacy
of people with countless kinds of things – our immersion in the material medium (p. 3-4).

Western Psychology, in particular, has been taken to task for underestimating the
psychological significance of objects (Costall, 1995; Williams & Costall, 2000). Not
only does traditional Psychology in the West neglect our relations to things, but to
the limited extent that it even touches upon them they have been regarded as existing
primarily in a physical, asocial realm, as distinct from the socio-cultural domain of
people. In addition, many psychologists appear to be convinced that the only real
‘object’ of study is, not what people actually do and the material conditions of this
52 Doing Things with Things
action, but the internal structures and processes underlying human agency. Hutchins
(1995) describes this as an over-reaction to behaviourism:

Behaviorism had made the claim that internal mental structure was either irrelevant
or nonexistent – that the study of behavior could be conducted entirely in an objective
characterization of behavior itself. Cognitive science’s reaction was not simply to argue
that the internal mental world was important too; it took as its domain of study the internal
mental environment largely separated from the external world. Interaction with the world
was reduced to read and write operations conducted at either end of extensive processing
activity (p. 371).

In other words, people’s actions in the material world are regarded as important
only to the extent that they provide clues to the ‘hidden’ mental activity that is
going on. The ‘theory of mind’ approach, which has proliferated in developmental
psychology from the 1980’s onwards, exemplifies this emphasis on invisible internal
cognitive structures. It assumes that people’s minds are the sum of their beliefs,
desires, emotions and intentions and that our awareness of the existence of these
internal thoughts and wants enables us to explain and predict what people will do in
specific situations (Astington, 1994). According to exponents of ‘Theory of Mind’,
individuals who are unable to impute mental states to either themselves or other
people, such as those suffering from autism, will experience severe difficulties in
understanding and engaging successfully with other people (Leslie, 1987; Baron-
Cohen, 1995).

Psychological Theories of Autism and the Material-Social Divide

The most influential models of autism have been derived from the ‘Theory of Mind’
approach (e.g., Leslie, 1987; Baron-Cohen, 1995; Hobson, 1993). Baron-Cohen,
Leslie and Frith (1985) proposed that autism derives from an impairment in a specific
cognitive mechanism (a ‘Theory of Mind’ mechanism) required for second-order
representations, which in turn are necessary for understanding the mental states
of other people (for further information relating to this theory see Baron-Cohen,
Tager-Flusberg, Cohen, & Volkmar, 1993). Baron-Cohen (1995) has recently
expanded his model to take into account joint attention deficits in autism, arguing for
the failure of a cognitive mechanism required to engage in shared attention, which
is developmentally more primitive than the ‘Theory of Mind’ module, but which
in turn is necessary for the development of a theory of mind. Both these models
of dysfunction do not explicitly deal with our understanding of objects, as they
focus on the evolution of social understanding, or even more specifically with the
development of cognitive structures inside the head of the individual, which result
in the understanding that other people also have things going on inside their heads.
Nevertheless, their logic would predict that those activities which do not call on the
use of secondary representation, such as the proper use of everyday objects, ought
to be completely intact.
 Autism and Object Use 53
Contrary to this widely accepted conclusion, we propose that, given the
considerable evidence from sociocultural theory and research suggesting that
learning to use objects in normally developing children is an inherently social
process (Leontiev, 1981; Volpert, 1985; Valsiner, 1987), it would seem reasonable
to suppose that children who have trouble in relating to other people would also
have extensive difficulties entering into the appropriate use of cultural tools. Our
aim in this Chapter is to challenge both the notion of a rigid separation between the
material and the social, and the downplaying of the psychological significance of
the material world. In order to do this we first consider an important body of socio-
cultural research and theory which highlights, on the one hand, the role that other
people play in introducing children to objects and, on the other, the part that artefacts
themselves play in structuring human activity in conventional ways. Drawing on this
work, we call into question the characterisation of autism as a disorder that combines
difficulties in relating to other people with ‘intact’ relations to objects. In contrast,
we suggest that, given the evidence of the reciprocal nature of the interactions
between people and objects, impairment in interpersonal relations should itself
entail corresponding disruptions in the use of objects. We present evidence from
a qualitative investigation of everyday object use in young children with autism in
support of this proposition.1 In this Chapter, we will use the term ‘everyday object’
to refer to material items that are regularly used in mealtime and washing routines
and which can be manipulated by the hand, such as a toothbrush and cutlery. Our
findings suggest that, as a consequence of disruptions experienced in relating to
other people, children with autism have more limited opportunities to pick up the
social conventions surrounding the use of everyday objects than children without
this condition. They may also have difficulty in picking up the notion of ‘properness’
itself – an understanding of the fact that most artefacts have a single, definite,
culturally selected, function. The results also suggest that the tendency of children
with autism to relate idiosyncratically to objects may add an additional obstacle to
their ability to acquire the proper use of everyday objects from other people.

The Social Mediation of Object Use

Not all Psychology neglects the role of objects in our social life. Within developmental
psychology there is an important body of ‘socio-cultural’ theory and research, largely
drawing upon the tradition of Soviet psychology, that does challenge any rigid
separation of the social and the material in children’s developing understanding of
their surroundings. Proponents of this approach argue that our dealings with people
and with objects do not occur in mutual isolation (Leontiev, 1981; Volpert, 1985;
Valsiner, 1987). And this is true not only for adults but all the more so for infants
and young children.

1 Full details of the study can be found in Williams, Kendell-Scott and Costall, (2005)
‘Parents’ experiences of introducing everyday object use to children with autism’, Autism 9,
495–514.
54 Doing Things with Things
According to Valsiner (1987), a number of factors dovetail to ensure the cultural
canalisation of infants’ use of everyday artefacts: direct and indirect guidance by
caregivers (and older siblings), the child’s own eagerness to participate in the activities
of the adults around them, and the nature of the object itself. Valsiner carried out a small
number of longitudinal case studies in which mothers and their infants were videotaped
interacting during mealtime sessions. The infants were followed from the age of 7 or
8 months to 2 years, during which time they were visited by the investigators every
2–4 months. Mothers were observed using a number of different techniques aimed at
channelling their child’s attention and skill in using the implements they were given in
culturally appropriate ways. This included structuring the wider setting in such a way
as to promote desired actions, whilst discouraging undesirable ones. For example, in
teaching their child to use a spoon or fork correctly, the parent usually placed the infant
in a high-chair, restricting their freedom of movement, and moved distracting objects
out of the way. Only then was the cutlery presented, at the appropriate orientation, to
the child. Brief and intermittent periods of joint action tended to occur in the early
stages of an infant learning to use a particular implement. Either the mother placed
her hands over those of the child, who was holding the object, and physically guided
their action, or they performed a sub-part of an action, such as loading a spoon or fork,
leaving the child to do the rest. As the children became more adept at using the cutlery,
the mothers were more likely either to verbally prompt the child, for example if they
became distracted from the task in hand, or to use verbal encouragement when the
child was performing the correct action.
Valsiner (1987) proposed that only short and episodic periods of intensive
instruction by the parent may be required, as the child’s interest in what other people
are doing and their willingness to copy them would ensure that the child’s actions
will, eventually, be canalised even without direct intervention: ‘If a child misses
being explicitly taught a specifically relevant skill, this child may acquire it himself
on the basis of observation and trying’ (p.179).
It is important to note that the participants in Valsiner’s study were white middle-
class Americans and, therefore, the observations he made cannot necessarily be
generalised to other cultures. However, there is evidence to suggest that, whilst
there are cultural differences in the extent of adult mediation and in the specific
cultural messages which are conveyed to infants about objects, adults in all cultures
comment and instruct at least occasionally (Bakeman, Adamson, Konner & Barr,
1990; Tomasello, 1999).
Social influence, however, does not simply occur around objects, but through
them. Our world has already been shaped by human activity, and is full of things
designed by people to be used in specific human activities, by people who share a
common body shape, needs & cultural history. Artefacts such as a child’s toothbrush
provide a means by which the child appropriates new action schemes, and in so doing
enters into the shared practices of society (Volpert, 1985). Furthermore, the child’s
eating behaviour is not only shaped through the direct and indirect guidance of the
more culturally experienced people around them, but also through the very form of
the artefacts commonly used at mealtimes (Leontiev, 1981; Valsiner, 1987; Preston,
 Autism and Object Use 55
2000; Williams & Costall, 2000). Their design reflects society’s expectations for
children to become socialised eaters. Preston (2000) argues that appropriate eating
behaviour at the table is organised systematically with regard to the proper functions
of the tableware: ‘Take cutlery in Western culture. Forks, spoons and knives have a
specific set of eating functions which are proper to them, and these functions govern
the behaviour we call table manners’ (p.41). According to Valsiner (1987), the
cultural canalisation of children’s actions starts from ‘the coding of some suggestive
cultural messages into the form and function of objects’ themselves (p.173). Objects
intended specifically for children are not simply scaled down versions of those used
by adults, but are often specifically constructed in such away as to further constrain
and guide the infant’s activity. For example curved cutlery, ‘Anywayup’ beakers,
and ‘Doydy’ cups, whose lip is slanted in such a way as to require less tipping by the
child to get the liquid into the mouth.

Autism and the Conventional Use of Everyday Objects

In the light of the body of ‘socio-cultural’ research and theory that clearly challenges
any rigid separation of the social and the material in children’s developing
understanding of their surroundings, it would seem highly unlikely that the problems
associated with autism, whatever its underlying ‘cause’, would be restricted, in any
clear cut way, to a particular ‘domain’ of the child’s life, such as his or her relations
with other people. If the child’s relations with people are disrupted, then their
understanding and use of objects should also be impaired – and vice versa. However,
in spite of this there is little published evidence concerning the problems, if any,
experienced by parents in training a child with autism to use them. In our review of
the literature we were able to identify only one empirical study (Ungerer & Sigman,
1981) which bears on the subject of everyday object use in children with autism,
with the remaining, limited, information coming from case study reports.
As part of an investigation into the relationship between play and language, Ungerer
& Sigman (1981) assessed the ability of 16 children with autism (Mental age=24.8
months; Chronological age=51.7 months) to use the objects they encountered in
their daily environment appropriately. At the time of the study all the children were
inpatients at a neuropsychiatric institute. Their primary nursing staff were asked
to complete a checklist consisting of 62 objects commonly found in the home and
residential hospital setting, including eating utensils, items used in grooming, books,
telephones and pencils. For each object listed, the staff were asked to record whether
the child could use it spontaneously, whether they could use it if verbally prompted, or
whether, even with verbal prompting they were unable to use the object according to
its designated function. Of the 62 objects surveyed, the children were able to use, on
average, just under 60% appropriately either on their own or with verbal prompting.
A mean of 25% were reported as never having been used appropriately, even with a
verbal cue. The children’s ability to use the remainder could not be assessed, either
because they were unavailable for use by the child or because the primary caretaker
was unable to give any information regarding them. Methodological problems with
56 Doing Things with Things
this study, however, make it difficult to draw precise conclusions with regard to how
proficient children with autism actually are in their use of everyday objects. Firstly, it is
not clear to what extent verbal prompting was required for the child to use the object in
an appropriate manner, as the children’s ability to use the objects on their own or only
with a verbal cue were not analysed separately. Secondly, no comparative data were
collected from a comparison group of typically developing children or children with
learning difficulties of an equivalent developmental age.
Case study reports appear to present a less positive picture of the acquisition of
everyday object use by children with autism than that suggested by Ungerer and
Sigman (1981). Asperger (1944, p. 39) notes the difficulties experienced by parents
in getting their children to eat normally at the meal table. Park (1983) also describes
the difficulties her daughter with autism had with using spoons and cups. Her ability
to use these things was much delayed and she had to be intensively trained in order
to use them appropriately. Although she did spontaneously use a spoon at 15 months,
this occurred only once, after which she refused even to hold a spoon until she
was over two-and-a-half. The use of other everyday objects, such as taps and light
switches, also had to be extensively trained, using the hand-over-hand method, with
periods of retraining required when her daughter encountered an unfamiliar version
of the same object. A striking feature of some self-reports and adult recollections of
highly functioning children with autism is the extreme affective responses provoked
by certain commonplace artefacts (Bemporad, 1984; Volkmar & Cohen, 1985;
Grandin and Scariano, 1986). Whilst some everyday objects appear to be a source
of deep fascination (Bosch, 1970), others give rise to unusual fears (Darr & Worden,
1951). Such evidence indicates that children with autism may experience some
difficulty entering into the conventional uses of objects.

A Comparative Study of Everyday Object use in Children with Autism

We decided to conduct a comparative investigation of the acquisition of everyday

object use in children with autism, children with Down syndrome and typically
developing children2 with two main aims in mind. Firstly, we wanted to carry out
an empirical investigation that examined systematically the notion that object use
is socially mediated, in order to complement the findings obtained from the case
studies conducted by Valsiner (1987). Secondly, we wanted to investigate whether
disturbances in relating to other people, such as those experienced by children
with autism, would have a disruptive effect on their ability to pick up the social
conventions surrounding those objects used in everyday routines. In addition, the

2 In autism research it is standard practice to include 2 comparison groups, consisting

of typical children and learning disabled children, in order to determine what aspect of any
behaviours identified are attributable to the autism and what simply reflects general difficulties
in learning (Burack, 1997).
 Autism and Object Use 57
findings obtained from the little research that has been conducted in this area to date
are difficult to interpret, due to fundamental methodological flaws.
Semi-structured interviews were administered, asking the parents (10 in each
group) to describe their child’s current use of objects traditionally used during
mealtime and washing routines, the process by which their child learnt to use these
objects appropriately, and any problems they encountered in attempting to guide
their child’s actions. The groups of children were individually matched for general
developmental age (23 months) using the second edition of the Bayley Scales of
Infant Development (BSID II; Bayley, 1993). Each child with autism was matched
to one child with Down syndrome and one typically developing child.
Following transcription a detailed examination of the interview transcripts
was undertaken following the guidelines outlined by Giorgi (1985), which
combine phenomenological and content analysis. In this approach, although
some presuppositions of the investigators may serve to structure the findings,
the participants’ own descriptions of their experiences form the primary basis for
understanding the particular phenomenon under investigation. In this study these
were the parents’ own descriptions of their experiences in introducing their children
to the appropriate use of everyday objects, and in monitoring that usage once it was
established (for further applications of this methodology, see Hétu, Riverin, Lalande,
Getty & St-Cyr, 1988 and Yardley, Todd, Lacoudraye-Harter & Ingham, 1992). The
process of analysis consisted of the following three stages:

Stage 1

All statements referring to the following themes were identified and isolated from the
rest of the transcript: problems experienced by parents in introducing their children
to, and maintaining, the appropriate use of an object; specific parental strategies
designed to introduce children to the proper use of an object; and any means identified
by the parents by which they thought their child had come to use a particular object
in a conventional manner. Once identified, the statements were broken down into the
smallest unit that allowed for the sense of the statement to remain intact3 (Yardley et
al., 1992). This process yielded approximately 1,500 statements in total.4

3 The origin of the interview data (i.e. which parent-child pair and which group the
participants were allocated to) was recorded by each statement, using a letter and number
system in order to preserve confidentiality. In addition the context was preserved in square
brackets where this was essential to making sense of a statement.
4 Reliability in terms of isolating statements from the transcripts was checked by having
a second investigator code approximately 30% of the transcripts (3 from each group), yielding
a mean percentage agreement of 85% (range, 77–90%).
58 Doing Things with Things
Stage 2

Progressive categorisation was undertaken to structure the data and a series of

inter-rater comparisons employed to ensure that the category definitions arrived at
reflected the nature of the statements assigned to them as closely as possible. Based
on the themes identified by the list of statements isolated from the transcripts, a first
classification scheme was arrived at through the consultation of the two authors.
Two of the main categories (parental strategies and problems in introducing objects)
were pre-selected as the interview schedule had been designed to elicit information
relating to these topics. The other main categories, non-parental influences and child
role, as well as all the subcategories emerged directly from the statements made
by the parents. Following Yardley et al. (1992), flow-charts were compiled which
outlined the rules used to classify each particular statement into its appropriate
subcategory.

Stage 3

The statements within each subcategory were then represented by ‘summary

statements’, presented in the form of comparative tables, which indicated the number
of parents in each group who had reported them.5
The findings below are presented under the headings ‘the social mediation of
object use’ and ‘cultural canalisation of everyday object use’. This division is an
analytical, rather than conceptual, one. Its purpose is simply to present clearly for
the reader both the influence of other people on object use and the role played by the
object itself in shaping the child’s activity. It should not be taken as implying a rigid
distinction between relations to people and relations to objects, which are viewed as
interdependent.

The Social Mediation of Object Use

Valsiner’s (1987) contention that the social canalisation of the motor functions
of the developing child towards culturally-specific ways of using objects is over-
determined, in the sense that there is more than one route via which a child can come
to use an everyday object in a conventional manner, was clearly supported by the
findings from the study. Almost without exception, the parents in the typical and
Down groups spontaneously reported that their child’s object use was shaped, not
only through conscious instruction by them, but also indirectly through the child’s
exposure to other people’s (particularly family members and peers) conventional use
of particular everyday objects:

5 Only summary statements based on the accounts of at least four parents from any
of the groups, or ten parents overall (i.e. a third of the total sample) were included in the
summary tables.
 Autism and Object Use 59
In the highchair he would sit with just his food on a tray, a plate and a cup, and we would
sit at the table with a mat and stuff. And quite often he would reach from his highchair for
a coaster and bring the coaster onto his highchair, and put his cup on the coaster to show
that he was copying us ... (child 6, typical).

He loves eating and he loves being part of a group and loves to be seen to be the same as
everybody else, and so because everybody else in the nursery and his school are using a
fork he has a good bash (child 10, Down).

According to all these parents, their children showed signs of actively participating
in the pick up of the conventions surrounding object use. This was demonstrated by
their interest in, and copying of, other people’s actions on objects, their desire to use
a particular object when they saw someone else doing so, and their initiation of the
introduction of an object themselves by asking their parent for it.
In stark contrast to the parents in both comparison groups, those who had children
with autism reported more problems in introducing appropriate object use, less
active involvement on the part of their child and, the use of more intensive teaching
methods. They described their children as lacking interest; both in what other people
are doing with objects, and in reproducing these actions: ‘There can be a whole
load of children around the table, and they will be tucking into lovely food. She will
not be motivated by any of them. She will do her own thing’ (child 10, autism); ‘It
seems to me L. [sibling] does things by example. She would just look at what you
are doing, and she will just copy it because that’s the way it should be done, because
an adult does it ... but with B. [child with autism], you have to teach him how to do
something.’ (child 9, autism). In addition, half of the parents of children with autism
reported that their child refused to eat at the table with other family members. This
means that even if the child was motivated to copy other people’s use of various
eating utensils, the opportunity is simply not there:

He wouldn’t eat his yoghurt if anyone was in the room. I had to show him the bowl outside
the room, put the bowl in here, and he would see me put it in here and then we would
have to go ... the whole family would have to stay out of the room until he had finished
(child 5, autism).

In addition, only one parent of a child with autism made a positive statement
indicating that her child watched what other people did with objects, and this was
only after the child had spent several weeks participating in an intensive training
programme, which involved the explicit modelling of actions to the child. Finally,
no parent mentioned that his or her child was motivated to use an object by
seeing someone else do so, or that they ever initiated the introduction of an object
themselves.
Parents of children with autism predominantly reported using physical methods of
instruction, and needing to successively repeat verbal prompts and demonstrations, as
60 Doing Things with Things
well as emphasising the importance of structured intervention by the school.6 These
are all consistent with a failure of more indirect means of social transmission.
Whereas the redundancy typically present in the socialisation of object use means
that the task of shaping children’s utensil use usually takes the form of episodic
instruction by an adult (or older sibling), rather than that of an intensive instruction
session, the absence of such over-determination in the case of children with autism
leaves their caregivers reliant on more structured, ‘intensive’, teaching methods if
they are to achieve success. The greater, conscious, effort needed on the part of
parents who have children with autism to shape the developing motor functions of
their child toward objects in culturally specific ways is made particularly clear in the
comparisons they draw between their child with autism and that child’s non-autistic
sibling. For example:

I don’t remember having to teach him [older brother] particularly you know, there must
have been a little bit there, but not that I remember [but] everything with J. had to be
taught, and reminded to go carefully. And remind him when he was putting it to his mouth
[spoon]…it was literally saying over and over again (child 4, autism).

In contrast, several parents of typically developing children remarked on the ease by

which their child seemed to pick up the conventional use of everyday objects. Some
referred to the progress as a ‘natural’ one: ‘I don’t think it is a strategy [getting the
child to use an object appropriately]. I think it is something that comes naturally’
(child 8, typical); ‘he is generally very good at wanting to go on to the next thing,
as it were, and we just sort of leave it to be quite a natural progression’. (child 6,
typical). Other parents remarked that their child seemed to acquire the ability to use
a particular object ‘spontaneously’: ‘well I never had to show him how to put it to his
mouth, he just understood that it needed to go into his mouth’(child 1, typical); ‘you
start with a spoon, and … then you give him a fork when it is a more appropriate
thing to eat the meal with, and then you just … you know …one day they are just
using a knife…’ (child 10, typical).
An alternative, although not necessarily mutually exclusive, explanation is that
children with autism may fail to enter into the notion of ‘properness’ itself. Loveland
(1991) has argued that autism involves a specific impairment not only in the child’s
ability to detect the significance of other people’s activity (e.g. the direction of gaze,
facial gesture, non-verbal gesture, speech), for their own action, but also a failure to
appreciate that most of the objects they encounter have a single, definite, culturally-
selected or preferred function In other words, they fail not only to discover the
preferred meanings of particular object, but also do not develop a sense that objects
in general normally have a definite function.

6 Whilst parents who had children with Down syndrome also described a heavy use of
physical methods, this was reported to occur as a result of motor difficulties or ‘laziness’ on
the part of the their child. They also reported using a wider range of strategies than did parents
of children with autism.
 Autism and Object Use 61
Cultural Canalisation Through Everyday Objects

The results also suggested that the tendency of children with autism to relate
idiosyncratically to objects might add an additional obstacle to the difficulties
they already have in entering into meaningful relations with other people. When
questioned about their child’s use of everyday objects, only parents of children with
autism reported that their child’s odd use of certain objects, or interest in isolated
aspects of an object, interfered with attempts to get the child to use the object in
an appropriate way. In other words, rather than the physical form of the object
providing an additional source of cultural influence by helping to shape the child’s
action in conventional ways, the child’s atypical reaction to the object actually
becomes a barrier to the introduction and maintenance of appropriate object use by
the caregiver.
All parents who had children with autism mentioned a problem relating to
the object itself compared to only four parents of children with Down syndrome
(when inexpert use and motor problems are excluded), and one parent of a typically
developing child. Seven out of the ten parents in the autism group remarked that
their child’s own particular interests/ obsessions actually interfered with the parent’s
attempts to get the child to use an object, such as a cup, in the appropriate way. For
example in the following statements the child’s interest in pouring or in whatever
was depicted on the cup, led the child to use the object inappropriately:

He was easy to change from a teat to a spout. Not quite so hot going from a cup with a lid
to a cup with a lid off because he had this thing in those days about pouring…so it was
always very tempting for him to want to pour (child 8, autism).

A cup is not like a vessel for drinking out of, a cup is a toy. Whatever is on the cup is
the thing that he will play with. So if it’s Thomas [the tank engine] on the cup it will just
go straight over…drink comes out and he will just look at it whilst lying down (child 9,
autism).

Another parent in this group compared their child’s co-option of a set of coloured
pencils or crayons for their own particular interest in lining things up in coloured
sequences, with the conventional use of the same set of objects by the child’s sister
who did not have autistim:

K. (sibling) wanted to scribble or draw with crayons or pencils because we introduced

them to her … she would copy. With L. (child with autism) we showed her what to do.
No, no, she did not want to. In fact, drawing and painting was something she did not want
to do at all. She wanted to line them all up, all the pencils we had. She wanted to line them
all up, … she had a different interest. They were colourful, so she was motivated by colour
and the fact that they were similar objects, so she wanted to line them all up (child 10,
autism).
62 Doing Things with Things
A further problem identified from the interview statements of those parents who
had children with autism was an unusual use of certain objects, particularly items
of cutlery:

When he first used his spoon he always used it upside down for some strange reason.
He would always put the stuff on upside down and he would, somehow when he got to
his mouth he would quickly ... he would scoop it up the right way, then when he got to
his mouth he would turn it, so that he would always be using it upside down (child 3,
autism).

It’s [the spoon] in his hand and it’s used to scoop, but he doesn’t actually put it in his
mouth, it’s used to pour the yoghurt into his hand, rather than put in his mouth (child 5,
autism).

It is important to note, though, that these findings do not address whether it is unusual
relations to objects or disturbances in joint attention that come first in development.
To investigate the direction of causality, as well as the extent to which earlier forms of
play underpin later ones, longitudinal studies would need to be undertaken. Ideally,
these should be prospective, following children identified at a very young age as
being at risk from autism.7 However, whether or not idiosyncratic relations to objects
precede deficits in joint attention, it is argued that they are likely to have a cascading,
disruptive, effect on other aspects of development, given the evidence suggesting that
play with, and exploration of, objects are primary modes of interpersonal interaction
and learning (Bakeman & Adamson, 1984; McArthur & Adamson, 1996).

Conclusion

In stressing the problems young children with autism have in using objects in a
conventional manner, it has not been the purpose of this chapter to play down the
social dimension of autism in favour of a material one. Rather, the aim has been to
challenge the rigid dualism between the social and the material, which underlies
most current theories of autism. It is suggested that those theories of autism deriving
from the ‘theory of mind’ approach fail to make the link between difficulties in
understanding others and relations to things, as a result of a failure to appreciate
fully the importance of social mediation, including various aspects of joint attention,
in the education of the child’s attention towards the function of things. Challenging
this rigid social-material divide necessarily involves questioning the restricted use
of the term ‘social’ as referring only to relations to other people. In highlighting the
reciprocal nature of the interactions between the child, other people, and objects,

7 The recent development of screening instruments for autism, such as the CHAT and
M-CHAT, make this an increasing possibility, although considerable work still needs to be
done to increase their sensitivity and positive predictive power (Baron-Cohen et al., 1992;
Baird et al, 2000; Filipek et al., 2000; Robins, Fein, Barton & Green, 2001).
 Autism and Object Use 63
the findings from the studies presented here indicate that relations to objects cannot,
in fact, be considered outside of the social context in which they are introduced
and used. Nor can they be examined apart from the human activities they shape
in common ways. Furthermore, this is not simply a theoretical problem, but also a
practical one. If we are to devise successful interventions for individuals with autism
it is essential that we fully appreciate the inter-penetration between different areas of
development, including the child’s dealings with objects and other people.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual

of mental disorders (4th ed., rev.). Washington, DC: American Psychiatric
Association.
Asperger, H. (1991). Autistic psychopathology in childhood. In U. Frith (Ed.), Autism
and Asperger syndrome (pp. 37–92). Cambridge, UK: Cambridge University
Press.
Asperger, H. (1944). Die ‘Autistischen psychopathen’ im kindesalter. Archiv für
Psychiatrie und Nervenkrankheiten, 117, 76–136.
Astington, J.W. (1994). The child’s discovery of the mind. London: Fontana Press.
Bakeman, R. & Adamson, L.B. (1984). Co-ordinating attention to people and object
in mother-infant and peer-infant interaction. Child Development, 55, 1278–1289.
Bakeman, R., Adamson, L.B., Konner, M. & Barr, R.G. (1990). Kung infancy: The
social context of object exploration. Child Development, 61, 794–809.
Baird, G., Charman, T., Baron-Cohen, S. Cox, A. Swettenham, J., Wheelwright, S.
& Drew, A. (2000). A screening instrument for autism at 18 month age: A six-
year follow-up study. Journal of the Academy of Child Adolescent Psychiatry, 39,
694–702.
Baron-Cohen, S. (1987). Autism and symbolic play. British Journal of Developmental
Psychology, 5, 139–148.
Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind.
Cambridge, MA: The MIT Press.
Baron-Cohen, S., Allen, J. & Gillberg, C. (1992). Can autism be detected at 18
months? The needle, the haystack and the CHAT. British Journal of Psychiatry,
161, 839–843.
Baron-Cohen, S., Leslie, A.M. & Frith, U. (1985). Does the autistic child have a
‘Theory of Mind’? Cognition, 21, 37–46.
Baron-Cohen, S., Tager-Flusberg, H., Cohen, D. & Volkmar, F. (Eds.). (1993).
Understanding other minds: Perspectives from autism. Oxford, UK: Oxford
University Press.
Bayley, N. (1993). Bayley Scales of Infant Development, Second Edition. San
Antonio, Harcourt Brace.
Bemporad, J.R. (1984). Recollections of a formerly autistic child. Journal of Autism
and Developmental Disorders, 14, 85–104.
Bosch, G. (1970). Infantile autism: A clinical and phenomenological-anthropological
64 Doing Things with Things
investigation taking language as the guide. New York: Springer.
Bugrimenko, E.A. & Smirnova, E.O. (1992). Paradoxes of children’s play in
Vygotsky’s theory. In G.C. Cupchik & J. Laszlo (Eds.), Emerging visions of the
aesthetic process (pp. 286–299). New York: Cambridge University Press.
Burack, J.A. (1997). The study of atypical and typical populations in developmental
psychopathology: the quest for a common science, in S.S. Luthar, J.A. Burack,
D. Cicchetti, J.R. Weisz, & D.J. Cohen (Eds.), Developmental Psychopathology:
Perspectives on Adjustment, Risk and Disorder (pp. 139–165). Cambridge
University Press.
Costall, A.P. (1995). Socialising affordances. Theory and Psychology, 5,
467–482.
Darr, G.C. & Worden, F.G. (1951). Case report twenty-eight years after an infantile
autistic disorder. American Journal of Orthopsychiatry, 21, 559–570.
Dawson, G. (1991). A psychobiological perspective on the early socioemotional
development of children with autism. In S. Toth & D. Cicchetti (Eds.), Rochester
symposium on developmental psychopathology (Vol. 3, pp. 207–234). Hillsdale,
NJ: Lawrence Erlbaum.
Filipek, P.A., Accardo, P.J., Ashwal, S., Baranek, G.T., Cook, E.H., Dawson,
G., Gordon, B., Gravel, J.S., Johnson, C.P., Kallen, R.J., Levy, S.E.
Minshew, N.J., Ozonoff, S., Prizant, B.M., Rapin, I., Rogers, S.J., Stone, W.L.,
Teplin, S.W., Tuchman, R.F. & Volkmar, F.R. (2000). Practice parameter: Screening
and diagnosis of autism. Report of the Quality Standards Subcommittee of the
American Academy of Neurology and the Child Neurology Society. Neurology,
55, 468–479.
Frith, U. (1989). Autism: explaining the enigma. Oxford, UK: Basil Blackwell.
Giorgi, A. (1985). Sketch of a psychological phenomenological method. In A.Giorgi
(Ed.), Phenomenological and psychological research (pp. 8–22). Pittsburgh, PA:
Duquesne University Press.
Grandin, T. & Scariano, M. M. (1986). Emergence labelled autistic. Tunbridge
Wells, U.K.: Costello.
Hétu, R., Riverin, L., Lalande, N., Getty, L. & St-Cyr, C. (1988). Qualitative analysis
of the handicap associated with hearing loss. British Journal of Audiology, 22,
251–264.
Hobson, R.P. (1993). Autism and the development of mind. Hillsdale, NJ: Lawrence
Erlbaum.
Hutchins, E. (1995). Cognition in the wild. Cambridge, MA: MIT Press.
Joerges, B. (1988). Technology in everyday life: Conceptual queries. Journal for the
Theory of Social Behaviour, 18, 221–237.
Leontiev, A.N. (1981). Problems of the development of mind. Moscow: Progress
Publishers.
Leslie, A. (1987). Pretence and representation in infancy: Origins of ‘theory of
mind’. Psychological Review, 94, 84–106.
McArthur, D. & Adamson, L.B. (1996). Joint attention in preverbal children: Autism
and developmental language disorder. Journal of Autism and Developmental
 Autism and Object Use 65
Disorders, 26, 481–496.
Mundy, P. (1995). Joint attention and socio-emotional approach behaviour in children
with autism. Development and Psychopathology, 7, 63–2.
Park, C.C. (1983). The Siege: The first years of an autistic child with an epilogue,
fifteen years after. London: Hutchinson.
Preston, B. (2000). The functions of things: a philosophical perspective on material
culture. In P.M. Graves-Brown (Ed.), Matter, materiality and modern culture.
(pp. 22–49). London: Routledge.
Robins, D., Fein, D., Barton, M.L. & Green, J.A. (2001). The modified checklist
for autism in toddlers: An initial study investigating the early detection of autism
and pervasive developmental disorders. Journal of Autism and Developmental
Disorders 14, 131–44.
Schiffer, M.B. (1999). The material life of human beings: Artifacts, behaviour, and
communication. London: Routledge.
Szokolszky, A. (2001). Using an object as if it were another. The perception and use
of affordances in pretend object play. In A. Szokolszky & E. Williams (Chair),
Taking ‘things’ more seriously: The role of objects in constraining developing
action structures in typical children and children with autism. Symposium,
The Tenth European Conference on Developmental Psychology, University of
Uppsala, Sweden.
Tomasello, M. (1999). The cultural ecology of young children’s interactions with
objects and artefacts. In E. Winograd, R. Fivush & W. Hirst (Eds.), Ecological
approaches to cognition: Essays in honour of Ulric Neisser (pp. 153–170).
Mahweh, NJ: Lawrence Erlbaum.
Ungerer, J.A. & Sigman, M. (1981). Symbolic play and language comprehension
in autistic children. Journal of the American Academy of Child Psychiatry, 20,
318–337.
Valsiner, J. (1987). Culture and the development of children’s action: a cultural-
historical theory of developmental psychology. New York: John Wiley.
Volkmar, F.R. & Cohen, D.J. (1985). Infantile autism: A first-person account by Tony
W. Journal of Autism and Developmental Disorders, 15, 47–54.
Volpert, W. (1985). Epilogue. In M. Frese and J. Sabini (Eds.), Goal-directed
behaviour: the concept of action in psychology (pp. 357–365). Hillsdale, NJ:
Lawrence Erlbaum.
Williams, E. & Costall, A. (2000). The functions of things: a philosophical perspective
on material culture. In P.M. Graves-Brown (Ed.), Matter, materiality and modern
culture (pp. 97–111). London: Routledge.
Williams, E., Costall, A. & Reddy, V. (1999). Children with autism experience
problems with both objects and people. Journal of Autism and Developmental
Disorders, 29, 367–378.
World Health Organisation. (1990). International Classification of Diseases, 10th
edn. Diagnostic Criteria for Research (draft). Geneva: WHO.
Yardley, L., Todd, A.M., Lacoudraye-Harter, M.M. & Ingham, R. (1992). Psychological
consequences of recurrent vertigo. Psychology and Health, 6, 85–96.
This page intentionally left blank
 Chapter 4

Object Use in Pretend Play:

Symbolic or Functional?
Ágnes Szokolsky

One evening I was playing with my two-and-a-half-year old son in the bedroom.
Suddenly he suggested playing cooking; he grabbed a pillow from the bed, put it on
the floor and said to me: ‘This is the stove and I am making a soup’. He paused as if
looking for something to serve as a pot, then said: ‘I am making bread’ and he started
to knead the pillow. Soon he claimed that the bread was done and he was going to
cut it. He picked up another pillow right away and holding it vertically, imitated
cutting. This kind of pretend play, in which the child knows the object for what it is
but chooses to treat it as if it were a different kind of object, is a universal dimension
in play from the age of two, across various cultures (Haight, Wang, Fung, Williams,
Mintz, 1999).
Psychologists have been traditionally interested in pretend play because it appears
to involve the increasing symbolic capacity to treat objects representationally. Pretend
object play seems to be a sophisticated cognitive activity that implies the ability
(a) to think of one object as two things at once, (b) to think of one object as representing
another, and (c) to represent mental representations. That the manipulation of mental
representations is the primary driving force and explanatory principle behind pretense
is, in fact, generally not argued but taken as axiomatic. The pretend act of using a
block as a horse is self-evidently described as a process in which ‘one employs one’s
own mental representation of the block and applies that representation to the horse’
(Lillard, 1993a, p.373). Neo-Piagetian research focused on how children choose
pretend objects and found evidence for symbolicity increasing with age. More recent
research found new theoretical significance in pretense because it seems to open
a window on the child’s developing theory of mind inasmuch as it involves the
understanding that the mind has internal symbolic contents that represent the world
and the ability to represent the mental state of another person. For this reason, Alan
Leslie (1987, p. 416) has suggested that ‘the emergence of pretense is not seen as
a development in the understanding of objects and events as such, but rather as the
beginnings of a capacity to understand cognition itself’.
In this chapter I argue for an alternative functionalist-ecological view on pretend
object play. Pretense involves culturally mediated, structured and sequenced actions
that are recreated in a highly ‘compressed’ way. Objects play a central role in such
activities. I suggest that in pretend object play the child’s intention to realize a play
68 Doing Things with Things
theme leads her/him to use an available object functionally in place of the real
object. Functional properties of interest are defined by the demand of the pretend
action. However, perceiving a functional property can also make the child modify
his original intention. In the above example, when launching the play theme the
child used the pillow as a stove and had the idea of making soup. Then he changed
his idea from making soup to making bread, probably because baking bread was an
activity in which he had previously participated and also because he realized that the
pillow was perfectly knead-able. As the play theme unfolded, the action of cutting
the bread came into focus – the child needed an object that could be used as a knife.
Without hesitation he grabbed another pillow and used it as if it were a knife.
From an external point of view, a pillow is very good for a dough but not so
good for a knife. However, pretend objects need to support pretend acts, not real
acts; a pretend knife does not need to afford cutting, it only needs to afford pretend
cutting which is a much less definite act than real cutting. The task of the child is to
select an object that is good enough to support the enactment. As long as the object
functionally fits the pretend act well, the action looks very much as it would with the
real object. When the the object is functionally less fit for the pretend act, the child
compensates for the deficiency by making the best possible use of the object (such
as when holding the pillow vertically so that it has an edge and moving it to and fro
over the ‘bread’). In neither case is the pretend object ‘standing for’ the target object
as a detached representation. Rather, the pretend object is integrated into the action
as a partial embodiment of the target object, supporting the pretend act.
The approach taken here suggests that pretend object play is a perceptually
grounded activity that involves perceiving and using functionally relevant
properties of objects. Such functional properties are best defined as affordances. In
the ecological psychology of James and Eleanor Gibson, affordances are directly
meaningful functional object properties that specify potential actions for a given
actor, with a given intention (J. J. Gibson, 1979, E. J. Gibson, 1982). For an actor
intending to grasp an object or walk on a surface, the object must be graspable and
the surface must be walk-on-able. Graspability or walk-on-ability are not functional
properties in the traditional sense. Traditionally, functional properties are taken to be
predetermined properties of objects that define what the object is good for – pillows
are for sleeping, and knives are for cutting. Affordances, on the other hand, are
defined as action potentials in relation to the particular actor at a more basic level
of on-line coordination with the object world. When climbing a stair, reaching for
an object or walking through an aperture, we perceive and use affordances (Mark,
1987; Warren, and Wang, 1987; Carello et al., 1989). Infants detect affordances
(e.g. Walker, 1982, Gibson and Walker, 1984, Adolph, Eppler and Gibson, 1993)
and affordances provide a basis for object categorization and tool use (Smitsman
and Pick, 1982; Smitsman, Loosbroek and Pick, 1987; Pick and Heinrichs, 1989;
Leeuwen et al., 1994).
Any object has an immense number of action possibilities, but these cannot be
known in advance, in separation from the actor and the action (Costall, 1995). In the
course of everyday actions we routinely use objects differently from what they are
 Object Use in Pretend Play 69
meant for; we may rest our legs on the table, use a heavy book to press a wrinkled
piece of paper, and so on. Unconventional object use deviates from culturally
prescribed object use, even though it still involves the use of affordances. Tomasello
suggested that in using artifacts children use the ‘intentional affordance’ of the object
– what the object is culturally made for. In pretense the child directly perceives the
intentional affordance of the pretend object and then, in a second step, ‘decouples’ it
from the object so that the ‘inappropriate’ object can be used playfully, as a symbol.
He also notes, however, that this process is ‘very poorly understood at this point’
(Tomasello, 1999, p. 158, p.160). I agree but I suggest that better understanding will
stem from viewing pretend object play as the proper use of available affordances.
Certainly, the available affordances of objects differ in their appropriateness to
be used as the target object. But children are satisfied with using a good-enough
affordance and fine-tune the action accordingly. Mentally ‘decoupling’ the culturally
preferred affordance does not present a profound problem because knowing what the
object is made for is not essential for the actual use of affordances.

The Symbolic Framework and its Underlying Assumptions

Piaget and Vygotsky presented the background from which modern cognitive views
about pretend play have derived. Piaget’s widely accepted account of pretend play
was an extension of his general theory of play (1962). According to Piaget, play
‘manifests the peculiarity of a primacy of assimilation over accommodation which
permits it to transform reality in its own manner without submitting that transformation
to the criterion of objective fact’ (1971, p. 338). Viewed as a subordination of reality
to the desires of the self, play could not be a significant factor of cognitive growth.
Play is some sort of compensation for the inadequacy of cognitive development;
it exists because thought is not yet adequate to its task. Pretense, however, plays
a significant role in cognitive development, because it is the beginning of the
differentiation between the signified and the signifier. The differentiation rests
initially on an iconic relationship in which signified and signifier are closely linked
and perceptually similar, as when reenacting one’s own activities (pretend sleeping),
or when a toy car is used as a car. With development the differentiation grows into
a more developed form of symbolism, when signified and signifier are related more
arbitrarily, although this relationship never reaches the fully-fledged arbitrariness of
linguistic signs.
Vygtosky’s account of play (1976) contained two somewhat contradictory
lines of argument. On the one hand, he presented an anti-intellectualistic position,
implicitly criticizing Piaget. He warned that theories of play tend to view the child
as a theoretical being: ‘If play is to be understood as symbolic, there is the danger
that it might turn into a kind of activity akin to algebra in action. ... Here we look
upon the child as an unsuccessful algebraist who cannot yet write the symbols on
paper, but depicts them in action’ (p. 540). Purely cognitive accounts neglect the
affective side of play and also the actual circumstances of the child’s activity, he
70 Doing Things with Things
claimed. At the heart of pretend play is not the beginnings of symbols but imaginative
situations and the practicing of social rules. Vygotsky and his followers concentrated
on the function of activity in establishing pretend relationships between object and
word. Based on the experimental work of Bugrimenko and Smirnova (1992), they
concluded among other things that for object substitution the important condition
was that the child was able to act with the object in certain ways. On the other hand,
Vygotsky (1976) argued that in play the child is liberated from situational constraints.
With development in imaginary play ‘thought is separated from the object and action
arises from ideas (rules) rather than from things’ (p.546). Separation of meaning and
object is a gradual process, which needs, in the beginning, a ‘material pivot to keep
the meaning from evaporating’ (p.548). Thus, pretend play is a transitional stage at
which the child cannot yet separate the thought from the object, and so must have
a substitute object. The substitute object is not a symbol, only a reflection of the
process in which the semantic aspect moves from a subordinate to a dominating
position. At the same time, the specific properties of the pivot are still relevant; thus,
for a child a postcard, for example, can never serve as a pretend horse.
Based on Piaget’s legacy, a distinctively cognitive research program took shape
in the 1970’s and 1980’s (Fein, 1975, Lowe, 1975, Nicolich, 1977, Watson and
Fisher, 1977, Jakowitz and Watson, 1980, Kagan, 1981, Pederson, Rook-Green
and Elder, 1981, Ungerer, Zelazo, Kearsley, and O’Leary, 1981, Crum, Thornburg,
Benninga, and Bridge, 1983, Winner, 1988). Experimental studies on solitary
pretend object play provided children with a range of possible substitute objects to
determine children’s capacity and willingness to use one object as if it were another.
In a typical study, such as Elder and Pederson (1978), substitute objects differed
from the target object in similarity of size and shape and also in having a definite
conventional identity or not. Children between 2.5, and 3.5 years of age were asked
to perform pretend actions (1) with similar objects that had no definite identity (a
tinkertoy stick as a hammer); (2) with dissimilar objects that had definite identity
(e.g. a toy apple as a hammer); and, (3) with no objects present (hammering without
an object). The youngest children performed best with similar objects, the three-
year-olds performed equally well with similar and dissimilar objects but less so with
no object present, while the oldest children performed equally well in all conditions.
The authors suggested that in fully developed pretense, therefore, children are able to
use signifiers arbitrarily, as true symbols. Once this is achieved, ‘similarity between
the substitute and the referent does not appear to be important’ (p.503).
Another study with children between 18 and 34 months of age (Ungerer, Zelazo,
Kearsley, and O’Leary, 1981) also defined perceptual support as ‘high physical
similarity between the object used and the object represented’ (p.189). Object
substitutions were categorized as: (1) high physical support with action (e.g. the
child lifts a rectangular block to his ear and says ‘hi’), (2) high physical support
without action (the child picks up a cylindrical yellow block and says ‘carrot’), (3)
low physical support with action (the child pretends combing with a baby bottle),
and (4) low physical support without action (the child hands a table to his mother
and says ‘this house’). Younger children chose as signifiers objects whose perceptual
 Object Use in Pretend Play 71
and functional properties presented ‘the least conflict with those of the signified
objects and, therefore, were the easiest symbolically to transform.’ Older children
were ‘capable of violating an object’s perceptual and functional features and using
as signifiers objects with clear and conflicting functions which were perceptually
different from the signified objects’ (p.194).
In general, these early experimental studies identified three main factors that
made pretense difficult for young children: 1. No substitute object present, because
this condition requires the child to generate symbols internally; 2. Dissimilar
substitute object, because this requires the child mentally to override conflicting
perceptual cues; and 3. Substitute objects with definite conventional identity, because
this requires the child to prohibit established object-associated motor responses and
violate firm semantic boundaries. Design of the studies as well as their interpretation
manifested the following core assumptions: 1. Perceptual (physical) support
consists of simple structural features such as size, shape and color. Reliance on these
perceptual cues is a limitation that needs to be overcome; development means less
and less dependence on such cues. 2. Objects with definite identity (artifacts) are
represented as concepts with firm semantic boundaries. To use a concept differently
goes against the established order of the mental machinery and this ‘violation’
requires substantial mental effort. 3. Actions are guided by representations and are
unproblematic: once the representation is symbolically redefined, the associated
motor response is in place, too. These assumptions have been basically supported by
more recent research that has revealed hitherto underestimated capacities of young
children for interpersonal communication and coordination in collective pretend
play (Stambak and Sinclair, 1993). Research on children’s coordination with objects
has, however, not been pursued any further because the assumption has persisted
that the ‘separation between action and object’ is integral to mental development
(Stambak and Sinclair, 1993, p.ix).
The assumptions underlying the symbolic approach to pretense leads to
paradoxes. If the ability to violate mental order is a criterion for advanced cognition,
then development manifesting itself in pretense is not the development of a normal
function. The paradoxical situation was noted by Leslie (1987, p. 412): ‘The
perceiving, thinking organism ought, as far as possible, to get things right. Yet
pretense flies in the face of this fundamental principle. In pretense we deliberately
distort reality’. Assuming that significant human behaviors have biological proper
functions (Millikan, 1993), we can assume that, whatever proper function cognition
has, it is not the deliberate distortion/violation of reality. The theme of violation
is familiar from research on metaphor. A deep puzzle presented by metaphors to
traditional cognitivist thinking is that they seem to violate semantic order (Ortony,
1993). A solution to the puzzle is that there is no violation. The idea of violation,
is relevant only if a rigid order of mental representations is assumed. Given the
epistemic assumption that organisms are in a constant dynamic flow of coordination
with their surroundings through action, perception and cognition, reasons to assume
rigid representations, along with perplexing forces that violate them, evaporate.
Metaphor can be defined as the perceptually grounded detection of dynamic
72 Doing Things with Things
properties that remain invariant across different objects and events (Dent-Read and
Szokolszky, 1993) and pretend play can be similarly considered as perceptually
grounded action.
The symbolic approach to pretense is paradoxical, on a second ground, because
it cannot present conceptual limitations on ‘violation’. Once pretense has acquired
its mature, truly symbolic form, practically anything can stand for anything else. In
principle, the stronger the violation, the more highly developed it is. The assumption
that pretense is symbolic transformation does not set constraints on arbitrariness, even
if it is assumed that pretense would not reach the fully-fledged arbitrariness of true
symbols. Empirical evidence shows that such limits exist and that younger and older
children as well are sensitive to object appropriateness (e.g. Golomb, 1977, Copple,
Cocking and Matthews, 1984). Vygotsky argued for the role of functional object
properties, but functionality and arbitrariness do not go together, if arbitrariness is
understood as complete freedom from situational constraints, a sort of ‘stepping out
of reality’. Assuming, however, that functional coordination with the object world is
maintained in pretense, pretend play is a matter of ‘staying in’ as well as ‘stepping
out’.
The symbolic approach to pretense is paradoxical, on a third ground, because the
essence of pretense is regarded as essentially mental in nature, whereas pretense is,
in the first place, a matter of doing. One only has to observe children when they are
pretending: they move, they pick up objects and use them certain ways – they act
and it is the action that is focal. Talk also serves the action; an utterance like ‘I am
making soup’ confirms but does not establish what is going on. In the experimental
studies described above, pretend acts were of interest only to the extent they provided
evidence for symbolic use. For example, when a child used a baby bottle to comb the
doll’s hair, the act was taken as evidence for the symbolic use of the bottle; the action
itself – how the child used the bottle – was ignored as an issue for investigation. As
Reed (1996, p.13) has observed, theories that try to account for complex behaviors
by a set of cognitively stored representations and commands ‘have a tendency to
make psychologists skip over the arduous task of describing behaviors accurately’.
For the perspective taken here, actions are interesting in their own right and the
detailed observation and description of pretend actions is a central task.
The previous research on pretense has neglected the possibility that children
perceive and use available affordances of pretend objects. Perceptual support was
defined as limited to the similarity of simple features. Experimental design based on
the similarity notion has, generally, obscured the role of functional fitness in pretense.
The results, as a consequence, may have been methodological artifacts: children’s
dependence on similarity may have reflected nothing more than reliance on context-
free likeness when only such likeness was offered. However, if we consider task-
specific affordances as perceptual support, it turns out that a similar object is not
necessarily more fit to be the substitute than a dissimilar object. For example, an iron
rod can be very similar to a sword but it can be very heavy for the child; but he can
be perfectly happy to use a pencil, or his finger as a sword. A flat piece of wood is
unquestionably more similar to a comb than is a rubber ball (example from Elder and
 Object Use in Pretend Play 73
Pederson, 1978), but if we consider that pretend combing is nothing but repeatedly
moving the object along the head, both the flat piece of wood and the rubber ball
are appropriate because both of them afford this schematic action. Of course, form
and function go together. Yet it does not mean that surface similarity per se is the
perceptual basis for pretense, or when using a dissimilar object there is no perceptual
support to the pretend act. The task-specific functional fitness of a pretend object
depends on the the structure of the intended action, the action capabilities of the
child, and the affordances of the object, and not on simple resemblances.

The Functional Study of Pretense

Functional fitness of a pretend object is a matter of degree; an appropriately sized

broom supports pretend horse riding better than a fishing rod, although it is not
entirely impossible that the latter also could be used as a pretend horse. A few early
studies produced evidence that object suitability is an important factor in pretense
for children. In a study by Golomb (1977) children between 2.8 and 5.8 years of
age faced a variety of substitute objects ranging from suitable to highly unsuitable
objects. As first choices children selected suitable objects. When all suitable objects
were exhausted, children could be induced to make further selections, but when
only incongruous objects were available, selection sharply declined at all age levels.
Other studies have also shown sensitivity to the functional demands of pretense
situations. For instance, children, as young as 12 to 20 months old, selectively
matched appropriate feeding tools to recipients in pretend feeding acts (Fein and
Apfel, 1979). In another study children between 3 and 5 years of age proved to be
sensitive to critical affordances of substitute objects; when they had to choose a
pretend spoon for scooping they chose an eggshell, but when they were to find a
pretend spoon for stirring they chose a stick (Copple, Cocking and Matthews, 1984).
Criticisms of the cognitivist position have noted the tendency to overintellectualize
interpretations of pretend play. Culturally recognizable pretend activities (such as
feeding a doll) are more parsimoniously interpreted as affordance use; when a child
performs the pretend act s/he can be mimicking routines modeled by adults, or, can be
‘following affordances revealed by objects’. The child’s understanding about ‘how
things are related is not arbitrary but emerges as the co-participants continuously
interact with the perceptually available particulars in the situation’ (Zukow, 1984,
p.166, 168).
Based on these theoretical and empirical preliminaries an experimental study
was designed to investigate how children use the affordances of substitute objects in
solitary pretend play. Sixty-four children, half of them between 3 and 4 and the other
half between 4 and 5 years of age, were asked to play with pretend objects and rank
them according to their appropriateness. The pretend objects differed in their degree
of affordance–based functional fitness and in conventional identity. Degree of object
fitness was judged by groups of undergraduate students who saw demonstrations
of the pretend use of candidate objects and rated them. Objects were repeatedly
74 Doing Things with Things
changed and rated by new observers, until the final selection was made. Three sets
of objects were selected, each set containing one realistic target object and three
pretend objects. In each set the three pretend objects were clearly different in their
appropriateness for the pretend task, one being highly fit, one less fit and one least
fit. In each set the highly fit and the least fit objects had definite identities while the
moderately fit object did not (‘blank’ object). Evidence was gained that children
indeed knew the conventional identities of the definite objects and could not tell
what kind of objects the blank objects were.
This procedure differs from natural pretend play in two basic respects: the
intention is not the child’s own – the play theme is suggested and presented as a sort
of task; neither is explicit verbal judgement of how good a substitute object a part
of natural pretend play. The reason for the design was the goal of comparability to
previous experimental studies that have been briefly sketched in the section on the
symbolic framework. As the main purpose of the present study was to challenge the
work motivated by the symbolic approach to pretense, following the methodology
of previous studies appeared to be necessary. Externally pre-establishing the group
of objects that were highly fit, moderately fit and least fit was also a methodological
aspect of previous experimental studies. In fact, in previous studies the degree of the
object’s appropriateness was simply assumed. The present study established object
appropriateness empirically and in a perceptually grounded procedure: judgements
were based on actually watching the pretend objects being used as the target object
(for example, watching ‘the doll taking a nap’ with a soft lamb-as-a-pillow, a car-as-
a-pillow, and other objects).
Children were individually tested and videotaped in 20 minute long sessions.
The first play episode was about a doll taking a nap. The target object was a pillow
and the pretend objects were a stuffed lamb (fit, definite object), a wedge made of
cardboard (less fit, blank object), and a plastic car (least fit, definite object). The
second episode was about ‘cleaning up mess from the kitchen floor’. The target
object was a small broom and the pretend objects were a bunch of plastic flowers (fit,
definite object), a so called tassel (less fit, blank object), and a fork (least fit, definite
object). The third episode was about a doll wanting to take a boat ride. The target
object was a small toy boat and the pretend objects were a wooden shoe (fit, definite
object), a basket-like object (less fit, blank object), and a toy skate board (least fit
definite object; see Table 4.1 and Figure 4.1).
After some warm-up play, the experimenter modeled the play act with the realistic
target object and asked the child to repeat the act with the same object. Then she put
the target object away and said: ‘Now we don’t have the (target object) but we have
these’ (pointing at the substitute objects). ‘Play with all of these things for a while.’
After 2 minutes of free play the experimenter put the substitute objects in front of the
child and asked her ‘to pick the object that would be the best (target object), then the
second best and, finally, the least good object. The children were also asked to show
how these objects could be used as the target object.
The experimental hypotheses were as follows: In picking the best, second best
and least good objects, the children were basically expected to agree with the pre-
 Object Use in Pretend Play 75
Table 4.1 Target and substitute objects

Target Substitute Object Fitness to

Objects Objects Identity the Task*

First Set PILLOW – LAMB – Definite – High

– ‘WEDGE’ – Blank – Moderate
– CAR – Definite – Low

Second Set BROOM – FLOWERS – Definite – High

– ‘TASSEL’ – Blank – Moderate
– FORK – Definite – Low

Third Set BOAT – WOODEN SHOE – Definite – High

– ‘BASKET’ – Blank – Moderate
– SKATEBOARD – Definite – Low

* As judged by groups of undergraduate students.

established judgements. Specifically, both younger and older children were expected
to pick highly ‘fit/definite’ objects as first, and least ‘fit/definite’ objects as last
choices, since in these cases fitness, or the lack of it, was easy to perceive. This was
in contrast to expectations that would have been generated by the symbolic approach.
Namely, on the symbolic approach one would expect that definite object identity will
cause difficulty for younger children in perceiving object fitness. In addition, one
would expect that older children would show a tendency to disregard lack of object
fitness. The main interest was, however, related to pretend actions since previous
studies tended not to look at qualitative aspects of pretend acts. Pretend acts with
functionally fit objects were expected to be qualitatively different from pretend acts
with functionally less fit objects. This was based on the idea that when no ideal fit is
possible, the child would opportunistically exploit the available affordances and this
‘compromise’ would show up in the qualitative character of the act.
As a first approach to the qualitative analysis of pretend actions, three main types
were differentiated. Integrated actions were smooth and adequate, similar to what
those actions would have been with the proper use of the realistic target object. For
example, the child puts the lamb on its side to make it a good pillow. Discrepant
actions were jerky and fragmentary, different from what they would be with the
proper use of the realistic target object. For example, the child puts the doll’s head
on the point of the wedge, the dolls head is hanging down. Bridging actions involved
special effort to make the object fit for the task. For example, the child wraps the
blanket (on which the objects were presented) around the car to make it an acceptable
pillow. Two trained observers segmented the flow of videotaped actions into basic
level units; agreement was 86% on thirty percent of the data. A basic level unit was
defined by accomplishing a meaningful action with one object in focus (cf. Reed et.
76 Doing Things with Things

Set 1. The Lamb, the Wedge and the Car

Set 2. The Flowers, the Tassel and the Fork

Set 3. The Shoe, the Basket and the Skateboard

Figure 4.1 Photographs of the experimental objects

 Object Use in Pretend Play 77
al., 1992). Action units were then categorized as integrated, discrepant, or bridging
actions by two scorers different from those who scored for action units; agreement
was 89% on 30% of the data.
Results regarding the ranking of the objects basically followed fitness as
previously judged, but not as tightly as expected. Results were analyzed by an Age
x Object ANOVA. (See Figure 4.2 for the ranking of the substitute objects.) In

First Set. Judged Fitness of the Lamb, the Wedge and the Car as a Pillow

Second Set. Judged Fitness of the Tassel and the Fork as a Broom

Third Set. Judged Fitness of the Shoe, the Basket

and the Skateboard as a Boat

Figure 4.2 Ranking of the fitness of the substitute objects

78 Doing Things with Things
the first set children judged the lamb and the wedge as equally fit to be the pillow
(F[2,120]=14.04, p< .01) and the car was ranked last. In the second set children
judged the tassel as somewhat better to be the broom than the flowers (F[2,120]=
21,61, p< .01); the fork was judged as the worst. Older children appreciated the
fitness of the flowers for a broom significantly more than younger children (F[1,60]=
8,405, p< .01). In the third set children in both age groups judged the wooden shoe as
significantly better to be a boat than the other two objects (F[2,120]=11.84, p<.01).
The skateboard was judged as worst, but not a significantly worse choice than the
basket.
The high ranking of the ‘fit/definite’ objects showed that children generally
perceived their affordance-based fitness. The boat-potential of the wooden shoe
was readily detected. Knowing its conventional identity did not prevent even young
children from perceiving its functional fitness in the new context. The choice of
the shoe over the basket could not be due to simple similarity, since both were
approximately similar to the toy boat in shape and size. Justifications confirmed that
children paid attention to those affordances of the shoe that made it a good boat and
not to simple similarity of shape.
In the first and the second sets children ranked ‘fit/definite’ and the moderately
‘fit/blank’ objects as approximately equal choices. Either they did not perceive clear
functional superiority in these cases, or else, if they did, other considerations influenced
their ranking. In the case of the lamb-as-a-pillow, for example, children may have
perceived the pillow-potential of the lamb, but concern for lambs as living beings
may have made them reject it as a pillow (as the children sometimes explained). The
functional superiority of the flower-as-a-broom over the tassel was more difficult
to perceive than originally assumed because the bunch of plastic flowers had to be
held upside down in order to realize its broom-potential. This might explain why
older children were much better at perceiving the broom-potential of the flowers
than younger children. Social expectations also might have interfered with ranking:
flowers are not supposed to be used as a broom. In summary, ‘fit/definite’ objects
received a lower than expected ranking when the the affordance was not obvious
when some aspect of the identity of the object was incongruous with the pretend use.
In spite of these confounding factors, the perception of the functional fitness of ‘fit/
definite’ objects was powerful enough to make them at least equal choices to blank
objects.
Furthermore, the least ‘fit/definite’ objects were ranked far behind the
‘fit/definite’ and the ‘blank’ objects not only by the younger but also the older
children, as well. Older children did not tend to disregard the inappropriateness
of these objects more readily than younger children, as the symbolic approach
would have predicted. In fact, older children proved to be more sensitive to the
functional inappropriateness of these objects than younger children. This finding
is in accordance with other results (Golomb, 1977) showing that younger children
were more willing to make incongruous object selections than older children.
The third set was an exception in that the skateboard-as-a-boat was rated as a not
 Object Use in Pretend Play 79
significantly worse substitute than the blank object. This may have been because
children generally found the skateboard to be an attractive object.
The ranking task required children consciously to judge object appriopriateness.
Experimental procedures that require conscious judgment are known to less
accurately reflect the perception of affordances than actions themselves (Heft,
1993, Runeson and Frykholm, 1986). Under natural conditions children do not stop
to make judgements of potential pretend objects. Therefore, a more ecologically
valid approach to the analysis of pretend object play is to look at what children
actually do when they use pretend objects. Action measures came from free play
and from the ranking task, where children were instructed to show how the given
object would be used as the target object. Results are summarized by object types
in order to reveal tendencies that overarch individual sets (see Figure 4.3).
Regarding pretend actions the main finding was that distinct patterns have
emerged depending on degree of object fitness. Although during free play there
was less pretend object use than in the ranking task, the overall pattern of results
was basically in agreement in both situations. The use of the ‘fit/definite’ objects
(the lamb, the flowers, and the shoe) was characterized by a tendency for a high
number of integrated actions, a low number of discrepant actions, and virtually
no bridging. The use of the ‘least fit/definite’ objects (the car, the fork and the
skateboard) was characterized by an insignificant number of integrated actions
(with the exception of the car in the ranking task; in free play the car was ignored
altogether), a high number of discrepant actions, and a relatively high number of
bridging. The use of the ‘moderately fit/blank’ objects (the wedge, the tassel, and
the basket) was generally characterized by a relatively high number of integrated
actions, as well as a relatively high number of discrepant actions, and an increased
amount of bridging. In conclusion, the three object types were associated with
recognizably different action patterns, along with some variation and exception.
Action patterns associated with clearly fit and unfit objects were dominated by
one particular action type in a relatively stable manner, whereas action patterns
associated with blank objects had a more varied character.
Besides object fitness, the actions also depended on children’s ability to
perceive and willingness to realize the goal-specific affordances. An object could
be ideally suited to the pretend task, yet used discrepantly for the pretend action,
and vice versa, depending on the skills, intentions and motivational state of the
child. Clearly, children were not ‘passive users’ of affordances. Occasionally they
changed the available affordance in order to create better support for the action.
Functional fitness constrained actions but intentional actions could, and sometimes
did, change functional fitness in order to optimize performance.
The explanatory framework of the symbolic approach falls short of an adequate
account of the above results. The symbolic approach proposes that children need
to inhibit established motor responses when using definite objects (e.g. Elder and
Pederson, 1978). This means that when children use, for example, a hair dryer as a car,
they would inhibit motor responses associated with the hair dryer and would apply
80 Doing Things with Things

Action Types with ‘fit/definite’ Action Types with ‘fit/definite’

objects in Free Play objects in the Ranking Task

Action Types with ‘fit/definite’ Action Types with ‘fit/definite’

objects in Free Play objects in the Ranking Task

Action with Blank Action with Blank Objects

Objects in Free Play in the Ranking Task

Figure 4.3 Frequency of action types

IA = integrated action; DA = discrepant action; BR = bridging.
 Object Use in Pretend Play 81
motor responses associated with toy cars. However, the fact that children’s actions
were influenced by affordance-based fitness suggests that they did not simply replace
one fixed set of motor responses with another fixed set, but rather, they adapted their
actions to the properties of the substitute objects. In symbolic accounts the focus on
decontextualized mental representations obscures the possibility that in some basic
ways pretend object play involves functional coordination with objects. The above
results suggest that pretend object use involves a flexible and subtle reliance on
perceptual support and not an increasing disregard of it. Pretend object play has
been traditionally framed as an analytical task, but it is more likely, especially under
natural circumstances, that it is essentially a perception-action task that involves
goal-driven, but not necessarily intellectually driven actions.

Pretend Object Play as Pragmatic Tool Use

Pretense can be viewed as the unconventional use of an object in place of another

object in order to achieve a goal. In this regard pretend object use is, in some ways,
analogous to unconventional tool use when, for example, a screwdriver is used to
hammer in a nail. Conventional tools are specific objects that aid us in achieving
goals that would be otherwise difficult to achieve. When, however, in the course of
everyday activity proper tools are not available, we often use another object that will
do the job. We are perfectly aware that an artifact has a proper function but we do not
hesitate to use it quite differently if this is advantageous. This use is ‘improper’ only
when extrinsically viewed, according to its socially assigned proper function. From
the intrisic viewpoint of the action, however, the unconventional use of the object is
proper to the task at hand. In addition, it can be argued that the unconventional use
of an artifact is more parsimoniously explained by assuming affordance selection
than a process involving representational ‘decoupling’. Objects have a multitude
of affordances and intentional actions flexibly select the required affordance (Reed,
1993, Heft, 2001). Decoupling assumes that socially assigned proper functions are
rigidly associated with objects. However, in our everyday actions we routinely use
objects ‘differently’; we tap on desks with pencils, use a chair to prop open the door,
climb up on the table to reach the curtain-rod, and so on. Whereas the symbolic
approach suggests that such acts involve the cognitive act of decontextualization
and mental redefiniton of pencils, chairs and tables, the framework presented here
suggests that they involve the ever present pragmatic recontextualization of object
use.
Pretend object use is also the dynamic recontextualization of reality aimed at
the playful re-enactment of culturally learned actions. To view pretend objects as
unconventional tools may illuminate an important aspect of the developmental
significance of pretend object use. Tools create new opportunities for action by
enhancing, extending, or restoring the action capabilities of the actor. Tools focus
behaviors and promote the fit of actors to their task environments in more visible
ways than actions that do not depend on tools (Shaw, Flascher, and Kadar, 1993).
82 Doing Things with Things
Pretend objects, as unconventional tools, may help focus behavior and this
may explain why pretense without objects is more difficult than pretense with
objects: the difficulty lies not in the lack of external cues that would activate
representation, but in that there is no material basis that would guide and focus
the activity.
Unconventional tool use by adults is, of course, not playful. The adult is not
pretending that the screwdriver is a hammer, but simply uses it, for that purpose.
For an activity to be playful it means that it has no goal beyond the realization
of the activity. Compare real cutting to pretend cutting: in the former the goal
is that something gets cut. In pretend cutting the goal is to do the cutting except
that nothing gets cut. The material constraints for real cutting are tight: there has
to be a fit between the hand, the tool and the thing; if the thing is hard, the hand
is weak or the knife is dull the act will not succeed. In pretense, however, the
joy and satisfaction comes from ‘as if ’ cutting, where such tight constraints do
not exist. As argued above, this does not mean that no constraints are present,
but only that the constraints are much looser than in real acts. The looseness of
constraints contributes to the basic character of pretense, which is its dynamic,
fluid nature.

Pretend Object Play as Fluid Action

The dynamic and fluid nature of cognition is well demonstrated by metaphors,

analogies, jokes and the like. What is more, these forms of cognition are not
a separate ‘fluid realm’ added to the ‘solid realm’ of literal thought; cognition is
inherently fluid. From an ecological standpoint, it is important to emphasize that the
fluidity of thought basically comes from the fluidity of perception and action. Fluidity
means flexibility, adaptability, pliancy, continuousness (Hofstadter, 1996, p.2).
Pretend play is an especially good example of the fluid and dynamically intertwined
presence of perception, action and cognition. Recontextualization in pretense is a
dynamic process of relating the affordances of the object to the demands of the
pretend intention. The interplay between the pretend intention and the affordances
of the object involves the use of affordances that may support the pretend action to
various degrees; it may also involve the changing of the affordances in case they
are not ideally supporting the pretend action. It may also be that, upon perceiving or
conceiving a new possibility for action, the original intention is quickly dropped and
replaced by a new intention. Pretend intention is volatile, prone to flickering change
in the context of ongoing action and perception in play. Pretend actions typically
involve rapid decisions without reflection on the appropriateness of the pretend
object. The temporary new role of the object is constituted in the action, which only
shows, in an exceptionally clear way, that our dealings with the world is not pre-
given, they require a versatile fit.
The methodology used in the study described in this chapter basically replicated
the experimental procedure established within the traditional symbolic framework.
 Object Use in Pretend Play 83
This was necessary precisely to challenge that body of experimental work. However,
to get a better understanding of pretend object play, a methodology better suited to
the assumptions of the ecological-functional framework is needed in the future. In-
depth, qualitative observation of object pretend play under natural circumstances has
to play an important role in exploring how children actually use objects in pretend
play. Pretense is indeed magic, in a way, with lots of room for imaginative, unusual
use of objects. But this, however, is no ground for regarding it as a distortion of
reality. Earlier I cited Leslie in saying that pretense is paradoxical because it flies in
the face of the fundamental principle of ‘getting things right’. I suggest that there is
no paradox, as claimed by Leslie. Children do get things right in pretense. Pretend
play is imaginative, but not by turning away from the real world. Children perceive
and use affordances while they play and this is surely part of the developmental
significance of pretend play.

References

Adolph, K.E., Eppler, M.A. & Gibson, E.J. (1993). Development of perception of
affordances. In C. Rovee-Collier and Lipsitt, L.P. (Eds.), Advances in infancy
research. Vol.8. (pp.51–98). NJ: Ablex Publishing Corp.
Balzano, G.J. & McCabe, V. (1986). An ecological perspective on concepts and
cognition. In G.J. Balzano & V. McCabe (Eds.), Event cognition: An ecological
perspective. Hillsdale, New Jersey: Lawrence Erlbaum.
Bugrimenko, E.A. and Smirnova, E.O. (1992). Paradoxes of children’s play in
Vygotsky’s theory. In G.C. Cupchik and J. Laszlo (Eds.) Emerging views of an
aesthetic process. Cambridge: Cambridge University Press.
Carello, C., Grosofsky, A., Reichel, F.D., Solomon, H.Y. & Turvey, M.T. (1989).
Visually perceiving what is reachable. Ecological Psychology, 1, 27–54.
Copple, C.E., Cocking, R.R. & Matthews, W. (1984). Objects, symbols, and
substitutes: The nature of the cognitive activity during symbolic play. In
T.D. Yawkey & A.D. Pellegrini (Eds.), Child’s play: Developmental and applied.
Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Costall, A. (1995). Socializing affordances. Theory & Psychology, 5, 467–481.
Crum, R.A., Thornburg, K., Benninga, J. & Bridge, C. (1983). Preschool children’s
object substitutions during symbolic play. Perceptual and Motor Skills, 56, 947–955.
Dent-Read, C. & Szokolszky, A. (1993). Where do metaphors come from? Metaphor
and Symbolic Activity, 8, 227–242.
Elder, J.L. & Pederson, D.R. (1978). Preschool children’s use of objects in symbolic
play. Child Development, 49, 500–508.
Fein, G.G. (1975). A transformational analysis of pretending. Developmental
Psychology, 11, 291–296.
Fein, G.G. & Apfel, N. (1979). Some preliminary observations on knowing and
pretending. In N. Smith & M. Franklin (Eds.), Symbolic functioning in childhood.
Hillsdale, NJ: Lawrence Erlbaum Associates.
84 Doing Things with Things
Gibson, E.J. & Walker, A.S. (1984). Development of knowledge of visual-tactual
affordances of substance. Child Development, 55, 453–460.
Gibson, E.J. (1982). The concept of affordances in development: The renascence
of functionalism. In W.A. Collins (Ed.), Minnesota Symposia on Child
Psychology, The concept of development, 15. Hillsdale, N.J: Lawrence Erlbaum
Associates.
Gibson, J. (1979). The ecological approach to visual perception. Boston: Houghton
Mifflin Company.
Golomb, C. (1977). The role of substitutions in pretense and puzzle games. British
Journal of Educational Psychology, 47, 175–186.
Haight, W.L., Wang, X., Fung, H., Williams, K. & Mintz, J. (1999). Universal,
developmental and variable aspects of young children’s play: A cross-cultural
comparison of pretending at home. Child Development, 70, 1477–1488.
Heft, H. (1993). A methodological note on overestimates of reaching distance:
Distinguishing between perceptual and analytical judgments. Ecological
Psychology, 5, 255–271.
Hofstadter, D. (1996) Fluid concepts and creative analogies. New York: Basic
Books.
Jakowitz, E.R. & Watson, M.W. (1980). Development of object transformations in
early pretend play. Developmental Psychology, 16, 543–549.
Kagan, J. (1981). The second year: The emergence of self awareness. Cambridge,
MA: Harvard University Press.
Leeuwen, L. van Smitsman & Leeuwen, C. van (1994). Affordances, perceptual
complexity, and the development of tool use. Journal of Experimental Psychology,
20, 174–191.
Leslie, A.M. (1987). Pretense and representation: The origins of ‘theory of mind’.
Psychological Review, 94, 412–426.
Lillard, A.S. (1993). Young children’s conceptualization of pretense: Action or
mental representational state? Child Development, 64, 372–386.
Lowe, M. (1975). Trends in the development of representational play in infants
from one to three years. An observational study. Journal of Child Psychology and
Psychiatry, 16, 33–47.
Mark, L.S. (1987). Eyeheight – scaled information about affordances: A study of
sitting and stair climbing. Journal of Experimental Psychology: Human Perception
and Performance, 13 (3), 361–370.
Nicolich, L. (1977). Beyond Sensorimotor intelligence: Assessment of symbolic
maturity through analysis of pretend play. Merrill-Palmer Quarterly, 23, 89–99.
Millikan, R.G. (1993) White queen psychology and other essays for Alice, Cambridge,
MA: MIT Press.
Pederson, D.R., Rook-Green, A. & Elder, J.L. (1981). The role of action in the
development of pretend play in young children. Developmental Psychology, 17,
756–759.
Piaget, J. (1962). Play, dreams and imitation in childhood. New York: Norton.
 Object Use in Pretend Play 85
Piaget, J. (1971). Response to Brian Sutton-Smith. In R.E. Herron and Sutton-Smith
(Eds.), Child’s play. (pp. 337–339). New York: Wiley.
Reed, E.S. (1993). The intention to use a specific affordance: A conceptual framework
for psychology In R.H. Wozniak and K. W. Fisher (Eds.) Development in context.
Acting and thinking in specific environments, Hillsdale, N.J.: Lawrence Erlbaum,
45–76.
Reed, E.S. (1996). Encountering the world. Toward an ecological psychology. New
York, Oxford University Press.
Reed, E.S., Montgomery, M., Schwartz, M., Palmer, C. & Pittinger, J.B. (1992).
Visually based descriptions of an everyday action. Ecological Psychology, 4,
129–152.
Runeson, S. & Frykholm, G. (1986). Kinematic specification of gender and gender
expression. In V. McCabe & G.J. Balzano (Eds.), Event Cognition: An ecological
perspective. Hillsdale, NJ: Lawrence Erlbaum Associates.
Shaw, R.E., Flasher, O.M. & Kadar, E.E. (1993). Dimensionless invariants for
intentional systems: Measuring the fit of vehicular activities to environmental
layout. In J.M. Flach, P.A. Hancock, J.K. Caird & K.J. Vicente (Eds.), An
ecological approach to human-machine systems I: A global perspective. Hillsdale,
NJ: Lawrence Erlbaum Associates.
Smitsman, A.W. & Pick, A.D. (1982). Perception of inclusion in collection of
objects. In D. Rogers & J.A. Sloboda (Eds.), The acquisition of symbolic skills.
(pp. 335–342). New York: Plenum Press.
Smitsman, A.W., Loosebroek, E. van & Pick, A.D. (1987). The primacy of affordances
in categorization by children. British Journal of Developmental Psychology, 5,
265–273.
Stambak, M. & Sinclair, H. (1993). Pretend play among 3 year olds. Hillsdale, NJ:
Erlbaum.
Ungerer, J.A., Zelazo, P.R., Kearsley, R.B. & O’Leary, K. (1981). Developmental
changes in the representation of objects in symbolic play from 18 to 34 months of
age. Child Development, 52, 186–195.
Vygotsky, L.S. (1976). Play and its role in the mental development of the child.
In J.S. Bruner, A. Jolly & K. Sylva (Eds.) Play – Its role in development and
evolution (pp. 538–554). New York: Basic Books, Inc.
Walker, A.S. (1982). Intermodal perception of expressive behaviors by human
infants. Journal of Experimental Psychology, 33, 514–535.
Warren, W.H. & Wang, S. (1987). Visual guidance of walking through apertures:
Body-scaled information for affordances. Journal of Experimenal Psychology:
Human Perception and Performance, 13, 371–383.
Watson, M.W. & Fisher, K.W. (1977). A developmental sequence of agent use in late
infancy. Child Development, 48, 828–836.
Winner, E. (1988). The Point of words. Cambridge, MA: Harvard University Press.
Zukow, P.G. (1984). Criteria for the emergence of symbolic conduct: When words
refer and play is symbolic. In L. Feagans, C. Garvey and R. Golinkoff (Eds.), The
origins and growth of communication. New York: Ablex.
This page intentionally left blank
 Chapter 5

Culture, Language and Canonicality:

Differences in the use of Containers
between Zapotec (Mexican Indigenous),
and Danish Children*
Kristine Jensen de López

W.J. McGee, having observed that the Seri Indians of California had no notion of knives,
believed them to be the most primitive human beings (McGee, 1889, cited in Mauss,
1979).

Compared to the Piagetian infant, contemporary psychology credits infants with the
possession of a surprising amount of basic knowledge or what Spelke (1991) calls
‘core’ knowledge and Mandler (1996) refers to as ‘image-schemata’, concerning
the spatio-temporal properties of material objects.1 Tomasello (1999) has critically
defined this new theoretical and methodological perspective within developmental
psychology as philosophical nativism.2 Here are some of the abilities now attributed to
very young children. For example, at the age of 3 or 4 months, infants express surprise
(as measured in terms of how long they look at a scene presenting a violation of what
would be expected in the possible physical world) when presented with objects that
violate the constraint of continuity. Infants know that objects cannot be in two places
at one time. They also express surprise when presented with the violation of the notion
of solidity. So infants also know that solid objects cannot pass through one another.
Furthermore, at the age of 6 months infants appreciate the constraints of gravity and
inertia (Spelke, 1991). One important insight from this neo-nativist paradigm is that
Piaget’s emphasis on the active exploratory manipulation of objects, as the crucial

* The study was supported by Aarhus University, Denmark, the Danish Research
Council for the Humanities and by Inge Lehmann’s Scholarship of 1983.
1 This revisited view concerning infant’s abilities to predict the actions of objects in
the physical world has emerged over the last two decades from the results gained within the
preferential looking paradigm (Spelke et. al. 1992; Mandler, 1996).
2 Tomasello’s motivation for classifying this perspective in terms of a philosophical
nativism is grounded in the fact that similar to other first generation cognitive scientists, such
as Chomsky (1980) and Fodor (1983), the conclusions drawn by Spelke concerning infants’
innate core-knowledge are drawn from logical considerations only.
88 Doing Things with Things
aspect of the sensory-motoric child’s development of spatial conceptualizations, now
seems a less crucial developmental activity. However, core knowledge or spatial
image-schemata in themselves do not embrace notions concerning the intentional,
relational and socio-cultural meanings of objects, nor the canonical way of relating
two objects to each other in a spatial relationship. Hence, in themselves, these innate
abilities do not account for how the developing child becomes a situated participant
of a specific culture. To develop this ability, infants need to be enmeshed in at least
three correlated scenarios; a) they need to understand others as intentional agents
with an interrelated functioning goal, attention, and behavioral strategy, as expressed
in infants’ ability to engage in joint attentional activities emerging at the age of
nine months (Carpenter, Nagell and Tomasello, 1998), b) they need to become an
active part of a cognitive ‘habitus’ (Bourdieu, 1977) in order to be presented to the
intentional and canonical functions of objects and artifacts, and finally c) they need
some basic source of active instruction from adults, for example, scaffolding (Wood,
Bruner and Ross, 1976) or engagements within a zone of proximal development
(Vygotsky, 1986/1994; Tomasello, 1999; Sinha, 1999).
To illustrate the notion of a culturally situated participant, imagine the ongoing
imitative play of a two-year-old Zapotec indigenous girl, named Mita.3 While
squatting in front of a small container filled with water, Mita suddenly picks up a
20 cm long round wooden artifact, carved in such way that it gradually becomes
thicker at one end. One might perhaps anticipate Mita’s following action to consist
of her inserting the artifact into the container, in order to imitate the well-known
stirring activity of her mother preparing dinner in a large clay pot. Despite the fact
that Mita does insert the artifact into the container, she does not engage in a stirring
action, but instead performs the motoric task of balancing the artifact between the
palms of her hands, while quickly rubbing them against the artifact to create a rapid
circular movement. The above anecdote is based on a video observation I made
during an extended period of fieldwork in the Zapotec village where Mita lives. The
particular artifact employed by Mita consisted of a miniature version of what she,
many times prior to her engagement in the described event, had observed women
using with the purpose of whisking up a foaming topping on the hot chocolate,
which is daily served for breakfast in this Zapotec village. Having now added this
important piece of information, it may immediately seem obvious to the reader why
Mita was so eagerly encouraged to activate the whisking affordance rather than the
stirring affordance equally offered by this particular artifact. When following a strict
Gibsonian interpretation of Mita’s object manipulation, contrary to what might be
predicted, there is nothing inherent in the immediate properties of the object which
should license or motivate Mita to prefer this particular action. The whisking action
is in fact motorically4 more demanding than the more simple action of stirring. In

3 This particular activity involved delayed imitation.

4 Note that this motoric action goes against the human reflex of grasping the object,
because it requires that one’s hands are slightly arched outwards in order to fit the round form
of the object.
 Culture, Language and Canonicality 89
fact, when I presented this particular Zapotec artifact to my Danish undergraduate
psychology students, their best guess concerning its specific functionality was in fact
to categorize it as a type of rolling pin. Two-year-old Mita’s cognitive advance, when
compared to Danish undergraduate students, illustrates the fact that what objects
afford is not strictly objective and immediately equally perceptual to all humans in
an identical way. Affordances are culturally and interpersonally mediated (Costall,
1995) and children become aware of them through their interpersonal experiences,
which form part of their early expectation that others are intentional agents (Carpenter,
Nagell & Tomasello, 1998; Sinha & Jensen de López, 2000).
The above argument is not new. It is the cornerstone of the socio-cultural
approach established by Vygotsky (1978) and Luria (1979/1984). In their approach
the given function of an object cannot be accounted for as an inherent property
of the artifact, but rather as mediating, cultural-historically constructed, meanings
(Jensen de López, 2001). A similar attempt in the same direction was launched by
von Uexküll (1949, in Ingold, 1992) in employing the concept of Umwelt to describe
those qualities of objects which are subjectively added on to the neutral object (e.g.
Mita’s choice to use the wooden stick for whisking rather than stirring). However,
there is an important contrast to be drawn between the concept of Umwelt and the
essences of Vygotsky’s approach. For von Uexküll, qualities are (according to Ingold)
subjectively encountered by each individual and enclosed within his own ‘reality
bubble’. However when following Vygtosky’s socio-cultural approach – recently
extended by Tomasello (1999) – the notions of intersubjectivity and intentionality
should equally be viewed as playing crucial roles for children in understanding how
to manipulate objects in a meaningful way.
Returning to the example of the wooden stick used as a whisk, although certain
socio-cultural potentials of the object were already there prior to Mita’s attempt to
manipulate the object, these potentials were not merely part of Mita’s subjective
Umwelt. They are potentials which have been constructed socially as a particular
habitus into which Mita was born and thus passed on to her through the process of what
Tomasello et al. (1993) call imitative learning. This view is in contrast to Gibson’s
notion of human interaction with objects as a solipsistic learning processing, which
does not rely on the individual being able to imitate the way others use objects. The
functions or affordances of an object are, according to Gibson, directly available to
the user (1979). Imitative learning, in contrast, used for learning how to manipulate
tools and artifacts demands that the child place itself in the intentional space of the
user, for, thus discovering what Tomasello (1999) calls ‘intentional affordances’,
i.e. how ‘we’ as a cultural group use this object. The notion of this specific type of
affordances – and which is in accordance with what Cole (1996) calls the ‘ideal
dimension’ of objects – makes it possible to capture and emphasize the important
distinct between an object’s natural affordances (as expressed by Gibson) and its
intentional or canonical affordances , which – at least in part – are social constructions.
The distinction between natural affordances and canonical affordances may also be
seen as parallel to Vygotsky’s distinction between the natural and the cultural lines
in ontogenesis (Vygotsky, 1978; 1986). However, as pointed out by Emma Williams
90 Doing Things with Things
(personal communication) there also exists a problem within Vygotsky’s distinction
between the natural and cultural line, in that it leaves unresolved the dichotomy
between the individual and the environment or the individual and the culture.

Meanings and Actions

According to the developmental psychologist Bremner, and in opposition to

the philosophical nativism view addressed earlier, infants’ engagement in the
world is best described in terms of: ‘infants being busily engaged in working out
what the world offers for action rather than what the world is in physical terms’
(Bremner, 2001: 176, emphasis in the original). With this quote Bremner stresses the
importance of conceiving children as active explorers of their Umwelt rather than
passive observers of the physical properties inherent to objects. Hence, Bremner’s
alternative view emphasizes the notion of action rather than percept, as mentioned
in his formulation of what the world offers. Despite the advance gained in Bremner’s
reformulation concerning infants’ interest in actions rather than in objects, it may
seem to face similar limits to those seen in Gibson’s notion of affordance. The
limitation is seen in the fact that not just anything offered in the world will actively
be attended to by the child. The child needs to rely on some kind of constraint in
order to know which actions and affordances are important to pay attention to and
which not. In order for the child to become culturally enmeshed, her task consists of
going beyond simply working out any, or perhaps all, the functional properties of an
object and instead working out what the world means in terms of acting on specific
socio-historical constructed objects with purposeful and intentional functions. The
case of the autistic child, as presented in the work of Williams and Kendell-Scott
(this volume), demonstrates that the lack of such ‘natural’ constraint may delay
the child in acquiring knowledge about how to use everyday objects. I therefore
suggest that Bremner’s description concerning the main interest of infants’ becomes
more appropriate, if the notion of what the world offers for action is substituted
for the notion of what the world means for action. A similar view is suggested by
Costall, who argues in favour of an ecological realism, in which the social and the
cultural should be viewed as the primary qualities ‘learning affordances does not
simply concern the uses an object happens to afford, but what it is meant to afford’
(Costall, 1995, p. 472). For developmental psychology, this new approach means
that children’s appropriation of symbolic usage emerges within and as a result of
specific situated, culturally embodied desires and needs (Sinha & Jensen de López,
2000). Having stressed the importance of a situated embodied approach to children’s
learning of object usage, I will later extend this same view to children’s acquisition
of other cultural practices, such as linguistic practices. In this sense objects are not
viewed as culturally neutral, but as mediating cultural schemas or habitus, and hence
the notion of embodiment also becomes constituted and exemplified by participating
with objects in an entire matrix of cultural practices, some which are nonlinguistic
and some which are linguistic (Sinha & Jensen de López, 2000, p. 36). First, I briefly
 Culture, Language and Canonicality 91
review some of the relevant literature concerning children’s early concept formation
for understanding the spatial properties of objects.

Infants’ knowledge about what containers are for

Consistent with Spelke’s notion of infants’ core knowledge, researchers assessing

children’s early concept formation have argued that infants possess a conceptualization
of the notion of containment prior to acquiring the concept in language (Pieraut-Le
Bonniec, 1985; Freeman et al. 1980; Caron, Caron and Antell 1988; MacLean and
Schuler, 1989; Aguiar &Baillargeon, 1998). The results from these studies suggest
that the conceptualization of containment is either an innate, or at least a very early
acquired concept, universally present in all infants. However, three major limitations
can be identified concerning these previous empirical studies; a) they all test the
physical properties of objects only, b) most of them do not integrate the role of
meaningful socio-cultural functions of objects, and, c) they were all carried out in
Western cultures, primarily with American middle-class children.
The first and second problems were addressed by Freeman et. al. (1980) and
Sinha (1982). They suggested that infants’ understanding of the function of an
object does not only rest on the infant’s ability to analyze the physical properties
of that object, but requires the ability to know about the canonical functions of
objects. Take for example the notion of containment as presented by container-
shaped objects in an upright orientation, such as cups or open-topped hollow cubes.
How does the infant learn to use a cup or a cube properly. Freeman et. al. (1980)
and Sinha (1982) suggest that due to children’s experiences with cups used as
containers for constraining liquids from falling onto the floor, they become able
to conceptualize cups as canonical container objects. In an imitation task Sinha
illustrated that infants rely heavily on how object manipulation is modeled to them
by adults in constructing their own conceptualizations with respect to finding out
what the ‘natural’ or proper relationships between objects consist of (Sinha, 1982).
For example, if an object with an ambivalent canonical function – such as an open-
topped cube which can equally be used to contain as support another object – is
previously modeled by an adult to offer the functionality of a container, infants will
treat this ‘neutral’ object in terms of affording a containment function, although the
physical properties of the cube could equally afford that of support. This specific
phenomenon or preference, referred to as the ‘canonicality effect’, is described in
terms of ‘…a specific, and privileged instance of a much more general phenomenon
whereby the significance of objects and events is evaluated by children according to
criteria derived from joint praxis’ (Sinha, 1982, p. 151). Children’s reliance on the
canonical function of an object has been tested with 18 to 36 month old infants, who
were required to evaluate an object in a canonical or a non-canonical orientation
against a relational conceptual schema. The tasks were presented as an imitation
task and as a linguistic task, where the children were tested on the acquisition of
the spatial relational terms IN, ON or UNDER. English-acquiring children were
92 Doing Things with Things
requested to ‘put a toy/block IN, ON or UNDER a box/cup’ (Clark 1973; Wilcox and
Palermo, 1975; Freeman, Lloyd & Sinha, 1980; Sinha 1982). The youngest infants
almost exclusively produced IN responses to these containment-affording objects,
independent of whether the objects were oriented up-right or inverted. This behavior
was interpreted as a canonicality effect. Consistent with this, the errors produced by
the children consisted of responding with IN configurations to ON instructions, and
of responding with IN and ON responses to UNDER instructions. Explanations as to
what caused this specific pattern of non-linguistic strategies expressed in children’s
errors indicated that the children were relying on one of the following non-exclusive
strategies:

a) on the perceptually most salient affordance – and hence they were constrained
by the hypothesis, that if the object looks like a container place the second
object IN it, but if it looks like a support, then place the second object ON it
(Clark, 1993; Clark, 2003).
b) on the easier motoric execution – an upright oriented container, thus offers an
IN configuration, as well as the same object in an inverted orientation offers
an ON configuration (Wilcox and Palmero, 1975).
c) on contextual cues, with the contextual sensitivity being based on a highly
functional and adaptive cognitive mode (Sinha, 1982).

Sinha’s suggestion (strategy c) concerning children’s flexibility and sensitivity

in applying contextual cues was in part supported by the fact that only children’s
errors to UNDER instructions were not so strongly influenced by whether the cup
was first presented in an upright or inverted orientation. He argues that IN and
ON configurations share similar properties, in that when the cup is first placed or
topicalized in its correct orientation, placing the second object is straightforward,
while for an UNDER configuration, the block first needs to be in place or topicalized
followed by the cup. Several studies have, in fact, shown that UNDER spatial
relationships are more difficult for children to acquire in cognition and in language
than are the spatial relationships IN and ON. Rohlfing (2002) suggests that children’s
relatively late acquisition of POD (UNDER in Polish) is in part due to the perceptual
relations referred to by the notion of UNDER (e.g. the underside of an object) being
less salient than is the case of IN or ON, which on the other hand rely on the inside
or the surface of an object. The inside and the surface region of an upright container
object are both perceptually salient, whereas the inner region of an inverted container
is not perceptually salient.
Since these studies were all carried out with Western children acquiring
prepositional languages it is not clear whether this preference for perceiving objects
as containers rather than supports and the later acquisition of the notion of UNDER
is a universal developmental sequence. If this claim is viewed in parallel with the
notion of canonicality effects derived from joint praxis and actions with members
of the adult speech and conceptual community, it seems plausible that the container-
bias might be relative to the specific praxis and speech community of the child. I
 Culture, Language and Canonicality 93
have tested this hypothesis in a comparative study of Danish and Zapotec children’s
action imitation and language comprehension of the spatial configurations IN, ON
and UNDER. The results indicate that the degree to which infants’ responses are the
result of an IN-biased canonicality effect is dependent on the child’s cultural and
linguistic experiences and practices concerning the social-historical and meaningful
functions of physical and symbolic objects. In order to appreciate the results from
the study, it is first necessary to introduce briefly the main differences between
canonical object manipulation within the two cultures and in parallelism to the
different semantic structures employed by each of the two languages in referring to
spatial relationships between objects.

Culturally Specific Usage of Objects

Due to the restricted diversity of objects available within this particular Zapotec5
culture, many objects are used with multi-functional purposes. A stool which to the
Western mind may primarily be viewed as affording sitting on, will to the Zapotec
mind be equally defined with the purpose of supporting a dish while eating one’s
meal (similar to a table), or when inverted, as a pot-holder for constraining a large
clay cooking pot just removed from the fire. Thus, the function of a seat is not
taken to be the essence of the chair, and despite its cultural-historical identity being
that of a seat, it affords a large amount of other functions as well. These kinds of
‘intentional affordances’ of objects (Tomasello, 1999), ‘ideal dimensions’ of objects
(Cole, 1996) or ‘canonical functions’ of objects (Sinha, 1982) are naturally explored
for their multi-functional uses by the villagers. Hence, many traditional artifacts are
purposefully constructed to possess a variety of diverse functions. I will exemplify
this through the description of a second object, namely the traditional one-and-a half
to two meters long woven shawl, owned by all Zapotec women and young girls, and
which conventionally affords several purposes. When worn across the shoulders,
the shawl protects the women from cold and breezes or is used as a handy cloth.
However, when the shawl is placed over one’s head it serves as a protection from
the strong mountain sun. A rolled-up shawl placed in a circle position on one’s head
serves as a ‘cushion’ and support for baskets or other artifacts, which are transported
on the head. Yet, a fourth conventional function of the shawl consists of using it to
strap infants onto their mother’s back, or to transport firewood and other types of
large materials. Finally, the shawl also conventionally serves to cover-up and protect
infants from insects, while taking a nap on a mat spread out over the ground.
Gourds are one of the main container objects used by the villagers for drinking
liquids, for serving food, for measuring and scooping up liquids or dry substances.
However, when inverted, gourds can equally be used to protect items of food from

5 All information concerning the particular Zapotec culture and language were collected
by the author during a field work period of 20 months in total, running from 1995 to 2001.
The community is rural, basically very poor, and villagers are semi-self-sufficient farmers.
Zapotec is the language spoken in the village.
94 Doing Things with Things
dust and insects. A second main container object used in this Zapotec community
consists of woven baskets, which come in various sizes with various names. Rigid
baskets without handles serve both the purpose of transportation (e.g. the harvest
is transported in this type of basket, which is carried on people’s backs with a strap
across their foreheads) and that of the storage of crops, the daily baked tortillas, and
even brooding hens. The lack of a handle is in fact an important aspect of this type of
basket, since it permits the basket – when placed in an inverted orientation – to serve
the purpose of a constrainer rather than a container object. The basket is therefore
conventionally used to capture brooding hens or other kinds of small animals, which
would otherwise normally run freely around in the compound. One more example
of multi-functional tool or object use is the case of tortillas. In the village tortillas
not only function as food to be eaten, but also function as eating utensils, and are the
only object used for wrapping-up ‘lunches’ to take on day trips or to take to school.
The human body is employed as an integrated vehicle of transportation and as an
‘implement’ during daily practices. Heads and backs have already been mentioned as
the standard means of transportation. In addition hands are used as conventionalized
instruments for measuring dry materials, for stirring foods during cooking or eating,
and even a fully extended arm serves perfectly to stir large portions of foods, as for
example in the fiestas. To my knowledge, feet are not employed for anything other
than walking and sweeping fodder to the oxen. So, it seems that Zapotec villagers
make use of a whole set of properties within each single object available in the
community, and this is an integrated aspect of their cultural habitus.
The main purpose of the following study was to investigate whether, and if
so, to what extent the non-linguistic and linguistic spatial conceptualizations of
Western middle-class English and Danish infants – when compared to Indigenous
‘primitive’ Zapotec infants – concerning their abilities to imitate and to comprehend
basic spatial relations can account for universal developmental processes. The
general claim underlying the present study rests on the notion that neither object
manipulation nor linguistic conceptualization develop in a socio-cultural vacuum.
Therefore a brief description of the cultural and semantic differences of the two
groups of children forming part of the study is in place. Since Danish is a Western
culture and since the Danish language employs prepositions for referring to spatial
relations, similar to the structure of English, the focus in describing the cultural and
linguistic differences between the two groups of children will mainly be restricted to
descriptions concerning the case of the Zapotec language.

The Relativity of Linguistic Spatial Concepts

Similar to tools and objects language is also a symbolic system, and the proper
meaning of a word is constructed within a particular social usage. Therefore the view
addressed in an earlier section concerning children’s appropriation of the properties or
affordances of physical objects can equally serve to account for children’s acquisition
of linguistic concepts during the process of language acquisition. According to
 Culture, Language and Canonicality 95
Vygotsky (1986), words should be conceived as symbolic and meditated artifacts in
a similar way as material objects are. A similar view of tools and meaning as related
mediators is also presented by Karpatschof (2001).
Likewise in the traditional view on the function of objects, which holds that
objects possess inherent properties that determine their function, researchers within
formal semantics have claimed that linguistic concepts refer to well-defined objects
or categories in the world. However, more recently within the discipline of Cognitive
Semantics by Lakoff (1987), but also anthropological linguistics, such as Duranti
(2000) and Palmer (1996) it is argued that linguistic meanings are not inherent
properties of the word, but, psychologically speaking, are cognitively and culturally
motivated within cultural practices.
Also of parallel importance to these rapidly emerging new paradigms is the
reassessment by Lucy (1992a) of the linguistic relativity hypothesis proposed by
the American linguistic anthropologist Whorf (1956) which has now been tested
empirically by Lucy (1992b) and by researchers at the Max Plank for Psycholinguistics
led by Levinson (1996). These studies demonstrate that languages vary profoundly
concerning how the spatial environment is segmented, and that adults’ performances
on non-linguistic tasks correlate with these specific language differences. Hence,
within the last two decades several empirical studies have been offering support for
the weak version of the Whorf hypothesis, namely that language influences habitual
thought (Whorf, 1956).
One main problem, however, in most contemporary studies re-examining the
view addressed by Whorf is that they rarely consider the impact that the cultural
habitus of the speakers may have on such cross-linguistic differences (Palmer,
1992) (but see Hanks, 1996, for an exception). A second problem is that so far only
one study has approached this issue from an ontogenetic point of departure. The
exception is the work of Bowerman (1996). In joint research with Choi, Bowerman
compared elicited spatial relational utterances produced by English-acquiring
children with those produced by Korean-acquiring children. One main reason why
these specific languages are interesting for comparison is that they structure the
spatial array in different ways. Some of these language-specific semantic structures
were present in the very early utterances of the children. Bowerman and Choi found
that by the age of two years Korean-speaking, but not English-speaking children
had formed distinctive semantic categories for reference to tight-fit versus loose-
fit spatial relationships. In other words, a difference between these two groups
was already apparent at the outset of their language acquisition. To illustrate this
difference, Korean children used different spatial verbs to express the notion of a
peg in a hole (a tight-fit relationship) as opposed to a pencil in a cup (a loose-fit
relationship), while the English-acquiring children simply referred to both spatial
notions as IN-configuring relations, using the preposition IN. Bowerman argues that
many spatial concepts are not fully formed prior to when children start learning their
first language, and that while acquiring the semantic structure of one’s language,
children are actually acquiring language-specific spatial categories for how to
refer to or how to ‘profile’ a spatial relationship between objects for the purpose
96 Doing Things with Things
of communicating with other people. Thus, the way languages vary in how they
structure or ‘partition’ the spatial environment will motivate infants to conceptualize
the world in accordance with these semantic categories. Presenting the case of
Zapotec, the following section introduces yet another variation of how languages
can choose to partition the spatial array, and predicts that children acquiring such
system consequentially will conceptualize spatial relationships different from those
acquiring an Indo-European language.

Zapotec Spatial Words

The particular Zapotec6 language forming part of this study, has not, to my knowledge,
been described in detail previously, and is still far from fully comprehended in detail.
The main linguistic means for referring to basic spatial relational conceptualizations
in Zapotec (e.g. the configuration expressed in English, as the cup is ON the table),
consist of the reliance on a core group of 7 body-part nouns used for locative reference
(Jensen de López, 2002, 1998). The meanings of body-part locatives derive from
the metaphorical extension of the human body in an upright position (MacLaury,
1989). Hence, the term for head, QUIA, is often used in reference to the highest
region, NII (foot) to the lowest region, LÁANI (stomach.region) to the inner region
of an object, LLAAN (butt.region) to the lower region, and DETS (back) to the
region behind an object and so on. For the present study it is sufficient to understand
the general semantics underlying only 2 locative body-part spatial nouns; namely
LLAAN, which literately means butt and LÁANI, which literately means stomach,
since these are the nouns investigated in the present study.
Despite what at first sight may seem a relatively simple and straightforward
system, the meanings of locative body-part terms are in addition often ‘intrinsic’
(see Levinson, 1996) to the object in such a way that when inverting an object so that
what initially was the lower region now becomes the highest region, the spatial gram
employed does not change. So for example, the lower or bottom region of a basket
in an upright orientation is referred to by the term LLAAN (butt.region), and even
when the basket is inverted, offering a surface region, this region is still referred to
with LLAAN (butt.region), and not (as might be predicted from the metaphorical
extension) as QUIA (head). This notion of an intrinsic frame of reference may be
seen as motivated by the canonical functional of the object. Hence, since baskets
are canonically containers the bottom or butt region remains referred to using the
body-part noun LLAAN (butt.region) independent or how the basket is oriented.
The bodypart LÁANI (stomach.region) similarly translates into meanings, which in
English are expressed using the prepositions in, under or through (Jensen de López,
2002) with many of these meanings derived from cultural and socially canonical
practices. Figure 5.1 illustrates the salient semantic distinction concerning the
containment schema configured in Danish and Zapotec.

6 Zapotec is an Otomanguean language spoken in the Southern Mexican state of

Oaxaca. Zapotecan languages are primarily spoken in the central valleys near Oaxaca City.
 Culture, Language and Canonicality 97
As illustrated in figure 5.1, whereas in Danish (and in English) two linguistic
concepts are applied when referring to the depicted spatial configurations, in Zapotec
just one linguistic concept is employed to refer to both spatial configurations. This
phenomenon, in fact, correlates with the cultural characteristic, noted earlier, of this
particular Zapotec community, that objects are conventionally employed with one-
to-many functions. Hence, the Zapotec linguistic terms used to refer to location, too,
can be said to follow a one-to-many referential system.

Zapotec and Danish Children’s Imitative and Linguistically-Guided Object

Manipulation

A total of 33 Zapotec and 71 Danish monolingual children were tested in a cross-

sectional developmental study for their ability to imitate and comprehend spatial
configurations in a naturalistic experiment involving the placement of a piece of
corn-maize either IN, ON or UNDER a small woven basket. The children were 17 to
46 months of age. The procedure of the task consisted of the child seated facing the
adult. It was explained to the child that it was going to engage in a small game with
the baskets. The game involved an action imitation and a language comprehension
task. In the action imitation task two identical baskets and two pieces of corn-maize
were employed; one manipulated by the child and the other manipulated by the
adult, while in the linguistically-guided task it was only the child that manipulated
just one basket and one piece of corn-maize. In both tasks the basket was presented
to the child in two different orientations. In one condition the the basket was initially
presented to the child oriented in an upright and canonical orientation. In the other,
condition the the basket was initially oriented in an inverted and noncanonical
orientation.
As a whole the two groups of children differed in terms of the response patterns
produced to each of the tasks. The response pattern of the Danish toddlers replicated
Danish
I (In) Under (Under)

Zapotec
Laáni (Stomach) Laáni (Stomach)
Figure 5.1 Illustration of the semantic differences expressed in Danish and
Zapotec
98 Doing Things with Things
the pattern of American and British toddlers, reflecting a preference for responding
with IN responses (respectively 39% of their responses), followed by ON responses
(respectively 31% of their responses) and finally UNDER responses (respectively
23% of their responses). However, the Zapotec children responded differently, in that
their responses expressed an almost equal preference for all three spatial response
types (respectively 30%, 30% and 28%) (see table 5.1).
Thus, the Danish children, like the American and British children, were highly
influenced by the IN-biased canonicality effect described by Sinha, whereas the
Zapotec children were equally happy to respond with any of the three target spatial
response types. However, this pattern, as we shall see, was not a result of the Zapotec
children simply responding in an arbitrary fashion.

Results of the Imitation Task

When modeled to imitate the spatial configuration of placing the corn-maize

UNDER the basket, the patterns of errors and hence, the non-linguistic strategies
employed by the two groups of children differed. The patterns of errors produced by
the Danish children seemed, in particular, to be influenced by the initial orientation
of the basket, resulting in the initial-upright and canonical orientation motivating the
children to mainly produce IN-errors (see figure 5.2), while the initial-inverted and
non-canonical orientation of the basket motivating them to mainly produce ON-errors
(see figure 5.3). Hence, they seemed influenced by both the canonical function of the
basket and the perceptually and motorically most salience response as suggested in
the literature. The Zapotec children, on the other hand, when responding erroneously
to the UNDER imitative instruction, produced ON errors in both orientations, but
hardly any IN errors (see figures 5.2 and 5.3).

Table 5.1 The three main response types to the imitative and the linguistically-
guided task by language group

Task IN ON UNDER Other Total

responses responses responses responses

Imitation
Danish 39% 31% 23% 7% 100%
Zapotec 30% 30% 28% 12% 100%

Language
Danish 44% 28% 13% 15% 100%
Zapotec 34% 36% 20% 10% 100%
 Culture, Language and Canonicality 99

Figure 5.2 Imitation of the UNDER spatial relation to the Initial-Canonical

oriented basket

These differences suggest that the two groups of children were relying on
different strategies with the Danish children directed by the canonical orientation
of the basket, thus mainly putting the corn-maize inside the basket, whereas the
Zapotec children did so to a much lesser degree. In addition, for the initial-inverted
and noncanonical orientation trial, only Danish children responded by turning the
basket into its canonical orientation in order to place the maize-corn inside it, while
the Zapotec children never did this. Overall, the Zapotec children almost never
responded with IN-errors to the UNDER imitation trials, while the Danish children
did – especially for the initial canonical trail.
When investigating the responses of the youngest children aged 17–24 months,
the Danish children responded in a way which suggests at least two nonexclusive
strategies; a) they were not ‘able’ motorically to invert the initial-upright oriented
basket in order to place the corn-maize underneath it and thus simply executed the
motoric easiest action consisting of an IN response, or b) they were not able to ‘resist’
100 Doing Things with Things

Figure 5.3 Imitation of the UNDER spatial relation to the Initial-Inverted

oriented basket
the canonical IN-bias demonstrated by the container function of the basket leading
them to place the corn-maize inside it. When these two hypotheses are viewed in
the light of the level of motor abilities displayed by the Danish toddlers during the
remaining parts of the task, the second explanation seems much more plausible.
Thus, the explanation of why the Danish children responded in such fashion can
be accounted for in terms of all three explanations suggested in the literature
concerning American and English children. These are: the children were affected by
the perceptual properties of the basket, by the motorically easiest response, by the
canonical effect of the basket as a container object or by all three in combination.
As for the youngest of the Zapotec children (age 17 to 25 months), exactly half
of them responded with an IN configuration while the remaining half responded
with UNDER configurations, and none responded with ON configurations, although
ON-responses have been argued to be motorically easier than UNDER-responses.
Thus, the results from the Zapotec children cannot fully be accounted for by the
three explanations available in the literature, and hence these children were relying
 Culture, Language and Canonicality 101
on different response strategies than Danish and English speaking children rely on.
These cross-cultural differences demand further explanation.
First, there seem to exist two immediate and nonexclusive explanations for
the Zapotec children’s response strategy, when manipulating the basket. Both
explanations were addressed in the above descriptions concerning the general culture
and habitus of the Zapotec children and in the description of the specific semantic
structure underlying the spatial grams present in the language acquired by these
children. First what needs to be to recalled is the prevalence of multi-functional
object usage, culturally inherent in the habitus of the Zapotec villagers, and which
(from a Western point of view of objects) seems to offer a much broader range of
canonical and intentional functions. In contrast, the everyday UmWelt of the Danish
children consists primarily of objects with just a few specific or canonical functions.
Now keeping this distinction between one-to-many versus many-to-one functions in
mind, we may now appreciate that: a) the response patterns of the Zapotec children
seem to reflect the children’s conceptualization of one-to-many functions offered
by the objects, while b) the response patterns configured by the Danish children,
conversely, may be seen as reflecting the conceptualization of many-to-one functions
in relation to the functions container objects offer for configuring spatial relationships
between two objects.

Results of the language task

The second study set out to investigate whether the two groups of children also
differed in their linguistic conceptualization of spatial relationships, and, in addition,
whether in accordance with Vygotsky (1978), this might be accounted for by an
explanation that is parallel to the one suggested for the differences in the imitation
task. In the linguistic task the children were assessed for their ability to respond
to linguistic instructions, conveying the same three spatial relational configurations
as in the imitative manipulation of objects task. In accordance with the semantic
structure of Zapotec, the Zapotec children were tested on their comprehension of
the body-part terms LÁANI (inner.region) and LLAAN (butt.region), whereas the
Danish children were tested on their comprehension of the Danish prepositions I
(in), PÅ (on) and UNDER (under).
Similar to in the imitative object manipulation task, the two groups of children
responded to the linguistically guided manipulation of the basket in different,
although language- and culture-specific, ways. The pattern of responses to this
task were, however, not identical to those to the imitative task. For the Danish
group, the discrepancy between the children’s overall responses of the type IN
and UNDER were larger than in the imitative task, whereas although the Zapotec
children responded with larger variation than in the imitative task, their responses
were still more evenly distributed as in the imitation task (see table 5.1). In fact,
when compared to the results for the imitation task, the Danish children responded
with three times as many IN as UNDER responses, whereas the Zapotec children,
on the other hand, responded with slightly more ON responses than IN responses.
102 Doing Things with Things
When instructed to place the corn-maize PÅ (on) and LLAAN (butt.region region)
the basket, the responses of the two groups of children differed. The Danish
children responded in a way which was consistent with the canonical function of
the basket as a container object, thus when initially presented to the basket in an
upright and canonical orientation, 59% of the Danish children’s responses consisted
of erroneous placing the corn-maize inside the basket. Contrary to this response
strategy, only 46% of the Zapotec children’s responses to this item consisted of
the children inverting the basket in order to correctly place the corn-maize ON the
surface region of the basket. This suggests that the Zapotec children managed, to
a larger degree, to disregard the perceptual-based canonical property of the basket
as a container object. When presented to the two language prompts that referred
to the notion of a support relationship only and with the basket initially presented
in an inverted-initial orientation, the Danish children produced relatively the same
amount of IN errors as the Zapotec children produced UNDER errors. Again this
suggests that the Danish children were influenced to a larger degree by the basket
as a container object, than was the case of the Zapotec children. Hence, there was
a clear discrepancy in the nonlinguistic strategies employed by the two groups of
children mirrored by the different types of error strategies produced by each language
group. Also to the instruction LÁANI (inner region) the Zapotec children produced
relatively and significantly more UNDER responses than the Danish children did to
the instruction I (in). Even when presented with the initial-inverted orientation of the
basket, which was expected to be the orientation least motivating an IN-response,
the Danish children still mainly turned the basket upright placing the maize inside
it, while they hardly ever placed the corn-maize under it. In contrast to this strategy,
a similar number of the Zapotec children produced UNDER as IN responses to this
trial suggesting that their overall understanding of the meaning of the term LÁANI
was less dependent on the canonical function of the basket as a container object (see
figure 5.4).
As a whole, the linguistic guidance of the children’s object manipulation proved
to have a larger impact than the imitative instructions on the two groups of children
responding in ‘opposite’ directions when manipulating the container object. This
difference demands further examination, and may at a first glance seem directly
accountable for by the above illustrated differences concerning the semantic
partitioning of the spatial array within each of the respective languages. The
explanation would then follow that, since Danish, at the lexical level, offers a clear
distinction for identifying the spatial relationships of IN versus UNDER this may
serve the Danish children as a ‘hint’ about which type of response might be expected
of them. In the case of the Zapotec children, however, the term LÁANI does not
distinguish semantically between an IN versus an UNDER spatial configuration and
thus only indicates that the object should be placed in an inner region, with no attention
paid to the orientation of the basket. This prediction was in part confirmed.
Thus, when only viewing the between group differences at the linguistic level,
the semantic structuring seems to explain the differences entirely. However, such
an explanation in terms of linguistic relativity tends to leave out the important
 Culture, Language and Canonicality 103

Figure 5.4 Responses by language group to the I/LÁANI configuring

linguistically-guided object manipulation for the Initial-Inverted basket

role of the cultural differences present in the distinction between the one-to-many
versus many-to-one functionality as part of the socio-cultural profiling available to
each of the groups of children. If this cultural difference in object manipulation
is viewed as an influencing factor, then it might instead suggest that children’s
acquisition of language, parallel to their acquisition of meaningful and socio-cultural
manipulations of objects, are best acknowledged as mutually influencing each other
during ontogenesis (Sinha & Jensen de López, 2000). Thus, the different response
strategies produced by the two groups of children in the imitation task and in the
language task were mutually due to the ecological and environmental constitution of
how objects are used in each culture, and also by how the languages as a tool serves
to mediate cognition in partitioning spatial relationships between objects. Hence
it may not be the language per se that is responsible for the differences, but the
semantics as part of the children’s socio-cultural environment that seems to motivate
the two language groups in different directions.
104 Doing Things with Things
Linguistic Relativism

The conclusion from this study indicates that children’s early non-linguistic and
linguistically guided object manipulation for basic spatial relationships between two
objects is determined by their culture-specific cognitive habitus and by language-
specific cognitive strategies. As mentioned briefly, there exist several other possible
interpretations of the results, which are in part similar to my view. First of all, in
accounting for the differences in the linguistically guided manipulation task only,
the view of linguistic relativism serves as one means of interpretation. The American
anthropological linguist Whorf, benefiting from Sapir and indirectly from Boas,
launched the notion of linguistic relativism. Whorf argued that the language we
speak influences our thought. However, according to Whorf it was not all thought
as an absolute and determining property that should be expected to be influenced by
language. Instead he hypothesized a specific type of thought to differ in accordance
with one’s language structure, namely what he called habitual everyday type of
thought. Lucy describes habitual thought in terms of ‘everyday, routine ways of
attending to objects and events, categorizing them, remembering them, and perhaps
even reflecting upon them’ (Lucy, 1992, p. 7).
From observations of exotic languages – a practice Whorf actually argued
necessary in order to surpass our own ethnocentricity – he developed the principle of
linguistic relativism in the sense of ‘ users of markedly different grammars are pointed
by their grammars toward different types of observations and different evaluations
of externally similar acts of observation, and hence are not equivalent as observers,
but must arrive at somewhat different views of the world’ (Whorf, 1956, p. 221).
Whorf even, critically, stressed that his own theory should be seen as also applying
to the premises of science and its ethnocentricity in respect to non-Western modes
of conceptualization. Consequently, he stated that ‘no individual is free to describe
nature with absolute impartiality, but is constrained by the modes of interpretation
even when he thinks himself most free’ (Whorf, 1940, p. 214, cited in Lee, 1996). A
modern and perhaps more dynamic view, which is similar to that of Whorf, is the one
expressed by Slobin (1996) in what he refers to as the active process of ‘thinking for
speaking’, an activity in which the individual is forced to take a specific world view.
Also Tomasello’s (1999) view of language usage as perspective-taking resembles
the idea that language is a tool for thought. Supporting these views, the recent cross-
linguistic studies by Bowerman (1996) and Jensen de López’ (2002) suggest that
children’s linguistic conceptualizations are influenced from the onset of language
acquisition by the specific semantic structure of their native language, hence, even if
children possess ‘true’ concepts concerning the physical properties of objects, these
should be viewed as readily open to language and cultural readjustments during
ontogenesis.
This notion of language as a mediating tool and also the notion of intentional
actions as mediating signs, although widely associated with Vygotsky (1978), was
similarly addressed by Sapir. He drew a parallelism between actions and language
as mediating signs of yet other acts and viewed language per se as an act. He stated
 Culture, Language and Canonicality 105
that ‘a vast number of acts are language…..That is they [acts/language] are not
important to us because of the work they immediately do, but because they serve as
mediating signs of other important acts’ (Sapir, 1949, p. 146, in Duranti, 2000, p.
43). Hence in Sapir’s view language does not only serve an immediate purpose, it
also packages information concerning socio-cultural practices and values. Language,
as a mediated tool, serves children in acquiring such socio-cultural practices and
values, but at the same time these particular practices and values also serve as a tool
for learning one’s language as a semiotic system.

Cultural Embodiment and Intentional Affordances

The notion of linguistic relativity can only partially account for the cross-cultural
differences expressed by the Danish and Zapotec children. Unless, following the
strong and deterministic interpretation of the linguistic hypothesis, the language
differences in themselves do not explain the cross-cultural differences appearing
in the imitative object manipulation task. In order to account for both the cross-
cultural and cross-linguistic differences children’s non-linguistic and linguistically
guided performances must be viewed as culturally embodied within intentional
and purposeful social practices, as argued by Sinha and Jensen de López (2000). A
view similar to that of Sihna and López (although relying on formal rather than on
cognitive semantics) is expressed by Hanks (1990) in his extensive analysis of the
spatial items known as deixis (‘here’, ‘there’ etc) in Yucatan Mayan. Through his
analysis he illustrates that the conceptualization of the human body as a corporeal
field is habitually used by Mayan speakers to interpret speech as situated within
culturally defined living space. In doing so Hank integrates formal semantics with
phenomenological characterizations of the human body as a crucial mediator of our
relation of the world of objects around us (Duranti, 2000). Discourse about space
involves self-awareness or taking the perspectives of others. In describing this,
Hanks relies on what Merleau-Ponty calls a prise de conscience by the speaker in
relation to the listener (Hanks, 1996). Hence to understand an utterance one must be
aware of the perspective of both oneself as well as that of the listener grounded in
the social values, practices and habitus of the individuals. Hanks’ view may also be
extended to account for children’s early manipulation of objects. Thus, the Zapotec
children, in responding to the spatial relational imitation task might be said to be
responding from a multi-functional cognitive habitus or habitual thought concerning
the canonical functions of the basket. This specific kind of cognitive habitus is what
influences these Zapotec children to differ in their response strategies from what
is well reported for Western children’s responding from a uni-functional cognitive
habitus concerning object use.
The child’s development of non-linguistic and linguistically guided object
manipulation, thus becomes incorporated within a whole series of social abilities.
These are constrained by the mediation of cultural practices, different types of
profilings, of which many are socially constructed within the specific environment
106 Doing Things with Things
within which the child is situated (Sinha & Jensen de López, 2000). The parallel
between children’s acquisition of the intentional/canonical affordances of objects
and that of the canonical affordances of linguistic concepts rests on the notion of
children perceived as enmeshed in a specific environmental and cultural surrounding
(Sinha & Jensen de López, 2000). In addition it also relies on the universal ability
of infants to place themselves in the intentional space of the others – discerning
the user’s intentional goal (Tomasello, 1999) – an ability known as joint attention
and which develops between the age of nine to twelve months (Carpenter, Nagell
and Tomasello, 1998). The present study suggests that the embodied culturally
mediated habitus of young children influences both their early object manipulation
and their early linguistic conceptualization of spatial relations in parallelism during
ontogenesis, suggesting a mutual manifestation in an entire matrix of cultural
practices or habitus.
Serving as a closing remark, Wittgenstein’s view concerning the meaning of
words expressed in ‘One cannot guess how a word functions. One has to look at its
use and learn from that’. (Wittgenstein, 1958: 109, § 340, cited in Duranti, 2000),
should in my view thus, be reformulated and extended as one cannot guess how a
word nor an object functions. One has, culturally and embodiedly to experience how
either a word or an object functions in a specific culture.

References

Aguiar and Baillargeon (1998): Eight-and-a-half-month-old infant’s reasoning about

containment events. Child Development, 69, 636–653.
Bremner, J.G. (2001): Infancy. (2nd ed.) Oxford: Blackwell.
Bowerman, M. (1996). The origins of children’s spatial semantic categories: cognitive
vs linguistic determinants. In J. Gumperz and S. Levinson, (Eds.), Rethinking
linguistic relativity. Cambridge: Cambridge University Press.
Brugman, Claudia, (unpublished manuscript). Metaphor in the elaboration of
grammatical categories in Mixtec.
Bourdieu, P. (1997). Outline of a theory of practice. Cambridge: Cambridge
University Press.
Caron, A.J., Caron, R.F. & Antell, S.E. (1988). Infant understanding of containment:
an affordance perceived or a relationship conceived. Developmental Psychology,
24, 620–627.
Carpenter, M. Nagell, K. Tomasello, M. (1998). Social cognition, joint attention,
and communicative competence from 9 to 15 months of age. Monographs of the
Society for Research in Child Development, 63.
Chomsky, N. (1980): Rules and representations. Behavioral and Brain Sciences, 3,
1–61.
Clark, E. (1973). Non-linguistic strategies and the acquisition of word meanings.
Cognition, 3, 161–182.
Clark, E. (1993). The lexicon in acquisition. Cambridge: Cambridge University
Press.
 Culture, Language and Canonicality 107
Cole, M. (1996). Culture psychology: A once and future discipline. Cambridge, MA:
Harvard University Press.
Costall, A. (1995). Socializing affordances. Theory and Psychology, 5, 467–481.
Duranti, A. (2000). Linguistic anthropology. Cambridge: Cambridge Textbooks in
Linguistics.
Fodor, J. (1983). The modularity of mind. Cambridge, MA: MIT Press.
Freeman, N.H., Lloyd, S. & Sinha, C. (1980). Infant search tasks reveal
early concepts of containment and canonical usage of objects. Cognition, 8,
243–263.
Hanks, W.F. (1990). Referential practice: language and lived space among the
Maya. Chicago: University of Chicago Press.
Hanks, W.F. (1996). Language and communicative practices. Chicago: University
of Chicago.
Ingold, T. (1992). Culture and the perception of the environment. In E. Croll and
D. Parkin (eds.), Bush base: forest farm – culture, environment and development.
London: Routledge.
Jensen de López, K. (1998). Organizing space by use of prepositions, body-part
terms and dispositional verbs. Paper presented to the Second Conference on
Conceptual Structure in Discourse, Atlanta, October 1998.
Jensen de López, K. (2001): Sociodramatic play in a Zapotec Indian Community.
Paper presented to the Tenth European Conference on Developmental Psychology,
Uppsala, Sweden, August 22–26.
Jensen de López, K. (2002): Language-specific patterns in lexical acquisition of
Danish and Zapotec spatial grams. Electronic proceeding from the Stanford Child
Language Conference, April 12–14, Stanford Uninversity: 50–59.
Jensen de López, K. (2003). Baskets and Body-Parts: A Cross-cultural and
Cross-linguistic investigation of children’s Development of Spatial Cognition
and Language. Ph.D. Thesis. Institute of Psychology, Aarhus University,
Denmark.
Karpatschof, B. (2002). Activity, tool and meaning. Paper presented at The Second
Conference on Doing Things with Things: The Design and use of objects,
University of Copenhagen, November, 2002.
Lakoff, G. (1987). Women, fire and dangerous things: what categories reveal about
the mind. Chicago: Chicago University Press.
Lee, P. (1996): The Whorf theory complex. Amsterdam: John Benjamins.
Levinson, S. (1996): Frames of reference and Molyneux’s question: Cross-linguistic
evidence. In P. Bloom, M. Peterson, L. Nadel & M.F. Garret (eds.), Language and
Space (pp. 109–169). Cambridge, MA: MIT Press.
Lucy, J. (1992a). Language diversity and thought: A reformation of the linguistic
relativity hypothesis. Cambridge: Cambridge University Press.
Lucy, J. (1992b). Grammatical categories and cognition: a case study of the linguistic
relativity hypothesis. Cambridge: Cambridge University Press.
Luria, A.R. (1979/1984). Sprog of bevidsthed. Translated from Jazyk i soznanie. Nyt
Nordisk Forlag.
108 Doing Things with Things
Mandler, J. (1996). Preverbal representation and language. In P. Bloom. M. Petersen,
L. Nadel & M.F. Garret, (Eds.), Language and space. Cambridge, MA: MIT
Press.
MacLaury, R. (1989). Zapotec body-part locatives: Prototypes and Metaphoric
extensions. American Linguistics, 55, 119–153.
MacLean, D.J. & Schuler, M. (1989). Conceptual development in infancy: The
understanding of containment. Child Development, 60, 1126–1137.
Mauss, M. (1979). Sociology and psychology: Essays. London: Routledge & Kegan
Paul.
Pieraut-Le Bonniec, G. (1985). From visual-motor anticipation to conceptualization:
reaction to solid and hollow objects and knowledge of the function of containment.
Infant Behavior and Development, 8, 413–424.
Palmer, G. (1996). Towards a theory of cultural linguistics. Austin, TX: University
of Texas Press.
Rofhling, K. (2002). UNDERstanding: How Infants acquire the Meaning of UNDER
and Other Spatial Relational Terms. Doctoral Dissertation Thesis, Faculty of
Linguistics and Literature, University of Bielefeld, Germany.
Slobin, D. (1996). From ‘thought and language’ to ‘thinking for speaking’. In J.
Gumperz and S. Levinson, ed., Rethinking linguistic relativity. Cambridge:
Cambridge University Press.
Sinha, C. (1982). Representational development and the structure of action. In
Butterworth, G. and P. Light (eds.), Social cognition: Studies in the development
of understanding (pp. 137–162). Brighton: Harvester Press,.
Sinha, C. (1999). Situated selves: learning to be a learner. In Bliss, J., Säljö, R. and
Light, P. (Eds.). Learning sites: Social and technological resources for learning
(pp. 32–46). Oxford: Pergamon.
Sinha, C., & Jensen de López, K. (2000). Language, culture and the embodiment of
spatial cognition. Cognitive Linguistics, 11, 17–41.
Spelke, E. (1991). Physical knowledge in infancy: Reflection on Piaget’s theory.
In S. Carey and R. Gelman (eds.) The epigenesis of mind: Essay on biology and
cognition (pp. 133–169). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.
Spelke, E., Breinlinger, K., Macomber, J. & Jacobson, K. (1992). Origins of
knowledge. Psychological Review, 99, 605–632.
Tomasello, M., Kruger, A.C. & Ratner, A. C. (1993). Cultural learning. Behavioral
and Brain Sciences, 16, 495–552.
Tomasello, M. (1999). The cultural origins of human cognition. Cambridge, MA:
Harvard University Press.
Ulvund, S.E. (1983). The canonicality effect in search for the hidden object.
Scandinavian Journal of Psychology, 24, 149–151.
Vygotsky, L. (1978). Mind in society: The development of higher psychological
processes. (M. Cole, editor) Cambridge, MA: Harvard University Press.
Vygotsky, L. (1986/1994). Thought and language. (A. Kozulin, editor). Cambridge,
MA: MIT Press.
Whorf, B.L. (1956). Language, thought and reality. Cambridge, MA: MIT Press.
 Culture, Language and Canonicality 109
Wilcox, S. & Palmero, D.S. (1974). ‘In’, ‘on’, and ‘under’ revisited. Cognition, 3,
245–254.
Wood, D., Bruner, J. & Ross, G. (1976). The role of tutoring in problem solving.
Journal of Child Psychology and Psychiatry, 17, 89–100.
This page intentionally left blank
 PART III
Transformation and Things
This page intentionally left blank
 Chapter 6

The Cognitive Biographies of Things

David de Léon

We are currently seeing an increasing awareness in the cognitive and mind sciences
of the importance of physical structure for cognition. The old view of cognition as
something that takes place only in the head has been replaced (or at least tempered)
by a view that recognises the roles played by physical and social structures. The
environment, it turns out, is not just an arena for action – a playground for problem-
solving and plan constructing minds – but is intimately implied in many, if not all,
cognitive processes.
There are several different ways of advancing this claim, but most would agree
that physical structures in the world can, at least, act: as an extension of memory
(Hutchins, 1995b; Beach, 1988; Norman, 1988), to simplify choice, perception and
internal computation (Kirsh, 1995; Clark, 1997), to constrain and even determine
cognitive behaviour (Zhang & Norman, 1994), and to transform tasks in ways that
better harmonise with our cognitive competencies (Hutchins, 1990, 1995a; Norman
1991).
Once this basic idea has been accepted it should be natural to ask how these kinds
of physical structures come about. After all, if physical structure can be an intimate
and integral part of cognition, more so than previously recognised, then asking
questions about the growth, development and appropriation of these structures
should be as natural as asking age-old questions about learning and development. In
fact, both learning and development need to be understood afresh in the light of these
emerging insights. The genesis, evolution and adjustment to cognitively significant
structure ought to be viewed as an essential aspect of most, if not all, of our cognitive
achievements.
Most of the authors cited above do acknowledge that there are interesting
processes responsible for the build-up of cognitively significant physical structure,
but these processes then figure to a negligible extent in their accounts.
Hutchins (1995a), for instance, gives truncated histories of the astrolabe and the
compass rose (both ancient navigational instruments that significantly transformed
the cognitive task of ship navigation), but then simply concludes that practice can be
‘crystallised’ into things, without discussing the process of crystallisation itself.
Kirsh (1995) acknowledges that the interaction of agent and environment can be
studied along different time scales, and does an admirable job of looking at medium
and short-term mechanisms of how people set up their workplaces for particular tasks.
However, the issue of how the workplaces themselves evolve is not addressed.
114 Doing Things with Things
Bærentsen’s (1989) work on the evolution of the rifle is an inspirational exception
which explicitly deals with the influence artefacts have on the cognitive demands of
task performance, and the influence of cognition on artefact development. Although
a bold and innovative attempt, Bærentsen’s analysis relies on an unanalysed notion
of cognitive processes being built into things (see de Léon, 1999).
Activity theory (the tradition in which Bærentsen’s paper is written) places
great emphasis on the historical and cultural foundation of thought and artefacts
and would therefore seem an ideal place to find the kind of analysis sought for. The
activity theoretical concepts of externalisation and historicity also seem to capture
the concerns discussed. However, as Engeström (1999) has noted, there seems to be
a general paucity of work in activity theory on these very topics.
A part of the reason for this neglect is no doubt connected with the difficulty of
reconstructing what are primarily historical processes. Unless we limit our interest in
the ways in which artefacts and practices co-evolve to very short time scales (for a nice
study in this vein see Agre & Shrager, 1990) we have to choose between longitudinal
studies and historical reconstructions. Of these two options a longitudinal approach
may, initially, seem preferable, as we avoid much of the speculation and conjecture
required in a reconstruction. However, longitudinal studies demand great effort
without guaranteeing results: we cannot know beforehand that anything of interest
will turn up at a chosen site or during the time-frame selected. In many cases the
sheer scope and intrusiveness of longitudinal projects also make them unfeasible.
Although less controlled and more speculative than a longitudinal study might
be, reconstructions permit us to explore sites where the occurrence of significant
artefactual change has already been established. They also present an opportunity to
investigate real-life events and changes that span durations for which a longitudinal
approach would be impractical. These particular characteristics make reconstructions,
despite the aforementioned methodological concerns, a compelling and intriguing
possibility – a possibility that has, to date, been insufficiently explored.
In the rest of this paper I will take a shot at constructing what I like to call
a cognitive biography, tracing the life-history of a particular artefact and its use
over a period of roughly three decades, and detailing the mutual influences between
cognition, activity and changing physical structures. In contrast to reconstructions
of events taking place over several lifetimes (cf. de Léon, 1999) the present
biographical time scale makes ethnographic methods and structured interviews part
of the available methodological arsenal.
The artefact in question is an unusually large spice shelf that I encountered whilst
conducting a cognitive ethnography (Hutchins, 1995a, 1995b; Lave, Murtaugh & de
la Rocha, 1984) of people cooking at home in their kitchens (see de Léon, 2003).
Each of the participants of the cooking study were video filmed whilst preparing
dinner and later interviewed about the organisation of their kitchens and about the
origins of their tools and cooking practices. One of the participants of the study was
Robert, a man in his mid-fifties. It is in his home that the shelf described in this paper
resides. Based on the interview conducted at the time of the study I have attempted a
reconstruction of the genesis and evolution of the shelf and pieced together a credible
 The Cognitive Biographies of Things 115
story of the underlying factors behind the various changes to the shelf, as well as
their probable cognitive consequences. The reconstruction was continuously checked
against the tape I had of Robert cooking and a number of supplementary interviews.
The shelf did not always appear as it does now. The collection of spices has been
stored in a manner of different ways and the actual shelf makes its appearance fairly
late in the story. Although I speak of the evolution of the shelf it is really the history
of a constellation of artefacts and practices.
This paper is thus an experiment in reconstructive cognitive biography. The
result is an unusual hybrid: on some occasions I use data from the case study to make
particular claims, at other times I introduce extraneous theories and observations to
bear on the case in question and to explain my observations. I hope that the attempt
might give some indication of what a cognitive biography might look like, and what
sorts of things we might learn by constructing them.

A Cognitive Biography of a Shelf of Spices

First I will give a brief description of the shelf and then outline some ways in which
its current structure and organisation supports Robert’s cognitive activity whilst
cooking. This is followed by a reconstruction of the shelf’s history.

A brief description of the shelf

Most of the spices in Robert’s kitchen are kept on a tall shelves fixed to a wall, a few
steps from the stove and workbench where the main activities of cooking take place
(see Figure 6.1). The shelf consists of two prefabricated units bought at IKEA (a
ubiquitous Swedish furniture store) that have been placed one above the other.
Each shelf is just deep enough to accommodate one spice jar and wide enough
to accommodate a row of about ten jars. All jars have been labelled with embossed
plastic strips and are neatly aligned along the shelving. Almost all are of identical
size, row upon row of yellow plastic lids divulging their origins as reused Coleman’s
mustard jars. Dispersed among these are a couple of tins, a few brand name spice jars
and two pepper mills of disparate design. Through the clear glass of the jars various
dry powders, seeds, flakes and roots can be seen, their colours ranging from beige
to brown.
In the narrow space between the two units that make up the shelf – too narrow
to accommodate an additional row of jars – lies a bulldog clip, some rubber bands, a
small pile of paper twist ties, and a paper packet of black pepper. Along one side of
the shelf hangs a decimated garlic braid.
In all, almost exactly one hundred jars and containers are kept here. This
constitutes an unusually large and impressive collection.
116 Doing Things with Things

Figure 6.1 The shelf (to the left of the door), sink and plate rack (far left of
picture) and corner of fridge (far right side of picture). Part of the dining room
can be seen through the kitchen door

The organisation of the shelf

Robert’s cooking encompasses Western as well as more exotic cuisines. About a

third of the spices found on the shelves (the lower three shelves) are endemic to
French and Italian cooking, the remaining shelves being devoted to spices used in
Indian, Middle Eastern, Chinese, Japanese, Indonesian and Thai dishes.
The spices are organised into categories according to a number of different
principles (see figure 6.2 for the basic categorisation of the shelf).1 The top shelf, for
instance, contains various hot spices, such as chilli and cayenne. Here it is taste and
function that determines placement. Another section contains different kinds of dry
leaves used in South and East Asian cooking, and is thus loosely organised by the
form in which the spices are found. A less obvious category is the group of spices
placed together because of their modest application, being dubbed ‘by-the-pinch
spices’ by Robert; the principle governing this particular categorisation being the
manner in which they are applied. There are also several sections in which the spices

1 The categories in figure 6.2, and in the text that follows, are Robert’s own and
were taken from a picture of the shelf which he had annotated. In the interview Robert also
explicated some of the uses to which he put the spices.
 The Cognitive Biographies of Things 117
are grouped together because they are used together in a particular style of cooking.
There are prominent sections with spices used in, for instance, Indian and Chinese
cooking.

Figure 6.2 Basic categorisation of the shelf2

2 The fastidious reader might note a slight discrepancy between the labels in the picture
and the description of the categories given in the text. The figure is based on a picture taken a
year after the original study and in the interim the organisation of the shelf changed somewhat
(more on this below).
118 Doing Things with Things
In those cases in which a particular spice is found in more than one form (e.g.
whole and ground) these have been placed adjacent to one another.
The bottom three shelves are home to more familiar spices used in Western
cooking. This is the largest regional section and has been arrayed alphabetically. The
spices kept here are shared by Robert and his wife. Since his wife lacks his penchant
for spicy and exotic food this is the only part of the shelf that she ever uses. As she
is also substantially shorter than her husband the placement of the spices, in this
mutually accessible region, is particularly felicitous. The reason Robert gives for the
alphabetical ordering of these shelves is that it was the only obvious categorisation to
present itself that would serve two people. It is instructive to note that the only region
of the shelf that is used by more than one person relies on a culturally conventional
system of categorisation. Alphabetisation ensures a clear, mutually intelligible, and
maintainable order.
There is also a small section of spices that are ‘on their way out’. According to
Robert these are spices that will be sacrificed as space is required. Some of these are
spices that Robert once bought out of curiosity, or that he has finished experimenting
with, or for which he has recently found better alternatives (for instance, access to
fresh spices that were only previously available in desiccated form).

Some ways in which the shelf currently supports cognition

Since each shelf is only deep enough for one jar, almost all the jars can be seen at
any given time; of the roughly one hundred jars that stand here only a handful are
blocked, or partially obscured, from sight. The jars are labelled and their contents
can be clearly seen through the glass, both label and content contributing to the
ease of identification of a particular spice. We might think of the shelf as a kind of
conceptual model, a physical structure embodying the basic spice combinations of
some of the most common, Asian and Western dishes, as well as certain aspects of
Robert’s cooking practice and his personal way of conceptualising cooking.
We should be cautious, however, lest we view the shelf simply as a reflection,
projection, or externalisation of Robert’s inner representations of spices and cooking.
For one, the genesis of the shelf belies such an interpretation. As we shall see in the
following, there are several determinants of the shelf’s structure and organisation
that are incidental, rather than intentional. Moreover, to whatever extent (and in
whatever form) the structure and content of the shelf is actually represented by
Robert, this will have been repeatedly shaped by the presence of the shelf itself.
The visible thematic spice groups arguably function to support Robert’s memory
in a number of different ways. Assuming that he knows what spice he is looking
for, a problem then facing him is locating the appropriate jar on the shelf. Instead of
having to scan the entire shelf for the spice in question, the thematic groups focus
his search to a particular area (obviating the need for knowing the exact location of
every spice). For example, just knowing that a spice is, or is not, used in Western
cooking removes a large part of the shelf from consideration. That the organisation of
 The Cognitive Biographies of Things 119
the shelf is consonant with his own idiosyncratic categorisations, conceptualisations
and habits of cooking also makes the regions more easy to locate.
Even the relationship between the form of his body and the position and
organisation of the shelf comes into play here. The spices that are most frequently
used (the Indian spices, according to Robert) are within easy reach, whereas slightly
less frequently used spices (such as the Middle Eastern spices) require him to stretch
upwards, or (in the case of Western cooking) to bend down. The placement of the
various groups of spices is not only a question of physical effort and comfort, but
ensures that the customary position of Robert’s body when facing the shelf presents
him with the most regularly used sections of the shelf. The same holds true for
Robert’s wife, whose height constrains which shelves she has easy access to. This
is a very straightforward and clear example of the role of embodiment for cognition
(for a critical review of the concept of embodiment see Clark, 1999).
These, then, are some ways in which the shelf can support the task of locating
a particular spice. When cooking a dish from memory an occasional problem is
actually remembering which spices to look for in the first place. In such a case,
all Robert needs to know is what kind of spices he is searching for. Looking at the
appropriate area of the shelf he need then only recognise the spices required – a far
easier task than recalling them (for a primer on recognition and recall see Baddeley,
1997).
The co-location of spices that are commonly used together also serves as a
reminder, throughout the cooking session, of spices still to be applied. Anytime a
spice jar is replaced on its shelf, neighbouring jars can jog Robert’s memory.
There is also an interesting structural feature of the shelf that supports the
replacement of spice jars. Since the spices are stored on the shelf one jar deep, and
each plane is completely full, the removal of any jar leaves a clearly visible gap. This
gap can then function as a perceptual place marker. Again, this saves Robert from
having to remember the exact locations of where particular spices are stored: he has
only to look for a break in the array to know where to replace a jar. Of course, should
Robert pick out several jars together, he would be left with the problem of pairing
the correct jars with the appropriate gaps when time comes to replace them. This is
not an insurmountable problem and would still be easier than having to remember
the exact locations. However, as seen on the video of him cooking, Robert only
picks out and uses one jar at a time, replacing it before picking out the next one,
thus ensuring that there is no confusion as to which jars belong in which gaps. There
are several ways in which the shelf could potentially be used, but significantly,
Robert has settled on a strategy that permits the physical structure to simplify the
cognitive demands of correctly replacing used jars. It is well to point out that it is
the combination of techniques, procedures and/or habits, with particular artefacts
and task environments, in relation to specific tasks, that determines the cognitive
congeniality3 of an activity.

3 Kirsh (1996) calls the measure of how cognitively hospitable an environment is its
‘cognitive congeniality.’ A cognitively congenial environment is one that reduces ‘the number
120 Doing Things with Things
The History of the Shelf

If we want to understand how the shelf came to have its present structure and use, we
must go back in time to the late sixties, to the spice shelf in one of Robert’s previous
homes, and then trace the development of the shelf back to the present. And this is
what I will attempt next, reconstructing the genealogy of the shelf with Robert’s
help.4
In the first kitchen that Robert remembers having (in a flat in Stockholm, Sweden)
the spices were kept in no determinate order, and on a single shelf next to a stove (see
step 1 of figure 6.3). At that time his collection was considerably smaller, consisting
of only a few conventional Western spices, and could fit snugly on a single shelf.
Robert had yet to develop the intense, and broad, interest in cooking that he has
today, and exotic food was still something that was only occasionally sampled in one
of Stockholm’s few Chinese restaurants (in the late sixties Stockholm had only two,
or possibly three, of these).
Although there was no intentional order to the arrangement of the jars at that
time we can speculate that there may have been some incidental grouping resulting
from the handling of the jars. In the kind of arrangement described, spices that are
often used together will tend to gravitate towards each other. There are several ways
in which they may do this. If all, or some, of the spices used in a particular dish are
taken off the shelf and put back at the end of a session of cooking, spices that often
occur together will tend to end up in close proximity. Or, if spices are placed at the
front of the shelf immediately after use, more commonly used spices will be found
near the front, whilst spices in less demand will be gradually pushed towards the
back wall. Needless to say, this kind of grouping would have aided the location of
common, and even uncommon, spices. With just a few spices in the collection there
would have been little incentive to organise them further.
These kinds of processes, in which the repeated performance of an activity
shapes the environment in which that activity takes place, are an important source of
supportive structure.5 The spatial redistribution of artefacts is one basic mechanism
which we have already encountered. In the same way that spice jars can come to
be functionally grouped through their use, so may other artefacts employed in the
kitchen. During my study of cooking, from which this case is taken (de Léon, 2003),
I noticed several of the participants making frequent trips to the plate rack in order
to fetch common objects. Implements that are routinely required are likely to have
been recently used and cleaned.The plate rack is therefore the most probable place to

and cost of mental operations needed for task success,’ ‘reduces cognitive load on working
memory’ and increases ‘the speed, accuracy or robustness of performance’ (Kirsh, 1996).
The use of the term is extended here (for reasons outlined in the above paragraph) to cover
activities as well as environments.
4 I have taken some liberties in truncating the history of the shelf and excising portions
that add little, from the point of view of the reader.
5 Barker (1968) calls this kind of relation between behaviour and environment
‘behaviour-milieu synomorphs’.
 The Cognitive Biographies of Things 121

Figure 6.3 A schematic history of the shelf

find them. In addition, the rack is also conveniently placed in most kitchens and its
contents visible. The processes of using and cleaning kitchen implements, therefore,
sorts out, and makes readily available the most frequently used implements.
Note that this particular use of the rack also conserves effort, the effort of
emptying the rack and replacing the things kept there. The rack is cleared almost
as a by-product of the activity of cooking.6 Much of human activity is like this,
with actions having multiple reasons and serving multiple simultaneous purposes
(cf. Wertsch, 1998).
Basic wear and tear is another mechanism that can generate supportive structure.
For example, think of a footpath kept clear through use (Barker, 1968), or the growth
of meaningful pathways between buildings (see Ribeiro, 1996). Or, consider how
repeatedly opening the phone book to the same section (for instance, the pages
containing your local pizzerias) can weaken the spine of the book making it easier

6 Reading an earlier draft of this paper one of my colleagues pointed out an important
exception to this way of using the plate rack. Although she recognised the use of the rack
described here, she mentioned that she will sometimes leave the rack untouched for a period
of time after washing up so that the full extent of her domestic efforts might be recognised,
and appreciated, by her partner.
122 Doing Things with Things
to locate those numbers in the future.7 In a similar manner, tools and implements
that are stored in designated locations (for instance, around the walls of a workshop)
can discolour, or otherwise mark, those places, thus facilitating their correct
replacement.
In these examples activity results in some cognitively congenial change to the
environment. The natural response to such change is compliance; the net effect is
beneficial and we probably do not pay it much attention. However, it is more common
for activity to have the opposite effect, creating clutter and disorder, rather than
order. One possible way to respond to encroaching entropy is to actively counteract
it, a strategy which Hammond (1990) calls enforcement.8 For example, in the video
of Robert cooking there is a slight lull in the session which he spends meticulously
straightening his spice jars, making sure that all ‘misaligned’ labels face to the front.
As the shelf is subject to rather heavy use (in the session filmed a total of 18 different
spices were taken down and replaced) this kind of upkeep becomes a natural part of
activity, ensuring that the shelf can continue to function as it does. Of course, it could
also be argued that the episode simply reflects an aesthetic preference or ideal, rather
than maintenance of cognitively significant structure (although neither interpretation
invalidates the other). The physical properties of the shelf and jars also help to keep
entropy at bay. The width of the shelf, for instance, greatly limits the ways in which
the jars can shift about on the shelf, and the gaps left by jars that have been removed
facilitate their correct replacement.9

The collection grows

Towards the beginning of the seventies Robert starts to experiment with Chinese
food in an attempt to recreate some of his favourite restaurant meals. There are, as
yet, no Chinese cook books available in Swedish (and books in English are still hard
to come by). However, the Swedish-Chinese Association publishes a small pamphlet
that Robert procures. Through an acquaintance (a supplier to some of Stockholm’s
delicatessens) Robert buys exotic spices in small yellow stackable tins. Throughout

7 This is why your avant-garde books always seem to fall open at the raunchy episodes
in the hands of any guest browsing your bookshelves.
8 Hammond (1990) and Hammond, Converse and Grass (1995) take enforcement to be
an active strategy of order imposed on the environment. It is interesting to note how, in the
present case, enforcement is sometimes incidental (recall how clearing the plate rack can be a
by-product of other activities).
9 It would be interesting to analyse and compare the various artefactual and procedural
means that have been devised to resist entropy. What technical solutions are there? For what
kinds of task is artefactual stability (taken to mean something other than plain robustness) a
desiderata? Does stability stand in the way of other functional aspects? If so, what kinds of
trade-offs are there (cf. Bleed (1986) on the trade-offs between reliability and maintainability
in hunting weapons)? An obvious place to start would be to look at collections and catalogues
and how these are used, and how they have evolved. Another would be to study how surgical
instruments are handled and cared for.
 The Cognitive Biographies of Things 123
the seventies Robert’s spice collection grows, in concert with his steadily increasing
knowledge and interest in cooking. Robert and his family move a couple of times and
at some point Indonesian dishes are added to the repertoire. The small yellow tins are
gradually abandoned as spices become more readily available from other sources.
The spices, that are now bought by weight, are transferred to recycled Coleman’s
mustard jars (these are the jars that can be seen in figures 6.1 and 6.2).

Spices in a box

In 1979 Robert and his family move from Sweden to England. An extended period
followed in which they lived in a succession of more temporary settings: a van, a
trailer, a couple of rented apartments. During this transitional period Robert kept his
spices in a low box (see step 2 of figure 6.3).
Looking down into the open box only the lids of the spice jars could now be
seen, the many identical lids exacerbating the difficulty of distinguishing one spice
jar from another. To find a particular spice, Robert had to rely on his memory and/or
make an educated guess. To determine whether or not the jar selected was the one
actually sought for he would have to lift it above its neighbours to be certain of its
identity. Some of the incidental ordering on the shelf may have survived the transfer
to the box, but may also have been broken up. Assuming that at least some of the
ordering of the spices made it into the box an incorrect guess might still give a clue
to whether he was searching in the right area of the box. Again, just using the box
may have bought some gradual ordering to the spices.
The temporary solution, as so often is the case, turned out to be less temporary
than initially expected. Eventually tiring of the impracticalities of the arrangement
Robert decided to order the spices into thematic groups (see step 3 of figure 6.3). To
locate a jar he would still have to rely on memory. Nevertheless, the organised box
was an improvement over the earlier, mostly haphazard, distribution.
This is the first full and deliberate ordering of the spices undertaken. If there was
already some order in place, as I have suggested, we can speculate that that order
may have influenced the subsequent intentional organisation, perhaps serving as a
rough guide. Any ordering that is conserved through this kind of process corresponds
nicely with the way that the spices are used.

From box to shelf

In England Robert discovers Indian cooking. For obvious reasons to do with the
country’s colonial past, Indian cook books, restaurants and spices were all readily
available. More spices were brought and Robert’s collection started to spill over into
various drawers in his kitchen.
In 1982 the family buy a house and Robert purchases a small shelf for the new
kitchen. The jars in the box were transferred to the shelf and placed in the thematic
categories that had crystallised over the years (see step 4 of figure 6.3). Although
this particular shelf is not the same shelf as the one described earlier in the paper
124 Doing Things with Things
Table 6.1 Summary of main cognitive differences between using the box and
using the shelf

Remembering which spices to use

A cursory scan of the shelf can trigger memory of the spices included in a particular
dish. When taking or replacing a jar on the shelf during cooking, adjacent jars may
serve as reminders of spices still to be applied. In contrast, the content and labels of
the jars kept in the box are not visible.

Finding a sought for spice jar

A guessed at location of a particular spice is easier to confirm using the shelf, since
feedback is instantaneous and category boundaries more distinct. Erroneous guesses
are more costly when using the box.

Correct replacement of spice jars

Gaps are easier to spot and fill on the shelf and neighbouring jars also help to establish
correct replacement. In the box there is a greater risk that jars will shift about, thus
breaking up thematic groups.

(that one still being many years, and many meals, away) it can be assumed that it
supported Robert’s cooking in ways similar to the present day shelf.
In the transition from box to shelf there is a noteworthy qualitative shift that
occurs. Whereas the chief function of the categories in the box was to aid Robert in
locating specific spice jars (remember, only the top of the lids could be seen when
the jars were in the box), the visibility afforded the jars when placed on the shelf
gave rise to new, unplanned for, and unanticipated functionality, in addition to a
general improvement of the previous functionality.10
Let me briefly outline some of the consequences of combining the categories of
the box with the structural features of the shelf (the main differences between using
the box and the shelf are also summarised in Table 6.1).
Finding a sought for spice on the shelf has some similarity to finding it in the box:
in both cases Robert is required to know the relevant category to which the spice
belongs as well as the rough whereabouts of that category. One of the things that
differentiates the two cases, however, is the ease with which the supposed location
of the spice is then confirmed. On the shelf, feedback is almost instantaneous, the
jars are stored one jar deep and can be easily scanned (compare this with the shelf
in his first kitchen which was just a single plane). In the box, a jar has to be lifted
before its identity can be confirmed. Not only is it easier to locate a particular jar on
the shelf, but the cost of a faulty guess is much less compared with the extra effort

10 Those with a fondness for evolutionary metaphors of artefact development (see e.g.
Basalla, 1988; Ziman, 2000) might like to think of this event as a case of exaptation (Gould &
Vrba, 1982). That is, as a feature which currently enhances fitness (i.e. cognitive congeniality),
but which was not originally built for the role it now plays.
 The Cognitive Biographies of Things 125
incurred when picking out the wrong spice from the box. The categories in the box
also have looser boundaries and are harder to pinpoint than the more rigid categories
on the shelf. From just looking at the box the various thematic groups are not readily
apparent and there is the constant risk that the location of particular jars will shift
about during prolonged use of the box.
The removal of a jar from either box or shelf leaves a gap that can later aid in the
replacement of the jar. On the shelf the gap is easy to spot and fill, and the visible
identity of adjacent jars can confirm a correct replacement. In addition, adjacent
spices can serve as reminders of spices still to be applied. In the box the identity of
neighbouring jars can only be established by lifting them up above the level of the
box.
Earlier in the paper it was noted that the vertical positioning of the shelf (in
relation to Robert’s body) also contributes to the ease with which particular spice
jars are found and retrieved.

Reaping the benefits

The ways in which the shelf supports cognition is a, mostly, unanticipated result
of the combination of the categorisation of the spices contained in the box with the
structural properties of the shelf. The claim made here is that the improved cognitive
congeniality of the shelf is partly accidental, and historically contingent. However,
the new ways of working afforded by the shelf are not automatically achieved.
Although there are cases where a change in the material means of an activity entail
a concomitant change in procedure, in this instance some adjustment had to be
made before the benefits of the new set-up could be reaped. There may be many
ways in which the shelf could potentially be used, but only some of these are an
improvement over the previous use of the box. It is by using the shelf in particular
kinds of ways that it is able to scaffold cognition. An example given earlier in the
paper is a good illustration of this. You might recall how Robert’s strategy (or habit)
of taking down and replacing spice jars one at a time permitted the shelf to simplify
the cognitive demands of the task. If several jars were instead taken down together,
then Robert would be faced with the additional chore of pairing each of the jars with
the appropriate gap.
The transition from spices kept in a box to having them arrayed on a shelf, the
adjustments to, and appropriation of the resulting structure by Robert, is an interesting
case in which an artefact (or artefactual complex) grows and develops in cognitively
congenial ways. Needless to say, all artefactual change does not lead to improved
functionality or to cognitive congeniality. Nevertheless, the process described may
have greater generality than this single case.
Similar mechanisms can be found in, for instance, the gradual co-evolution of
the form of books and bookcases (Petroski, 1999). Before the advent of the printing
press, books were rare and expensive luxuries, either locked up or chained to their
bookcases. As they became more numerous, vertical partitions were introduced to
the then standard bookcase design in order to prevent the shelves from sagging.
126 Doing Things with Things
Although the motivation for these partitions was originally structural the partitions
later came to play an important role in locating books. Catalogues, usually posted
on the end of a bookcase, grouped the books in accordance with the partitions that
contained them. Even as late as 1749 catalogues were still not alphabetical, but based
on these tables of contents.11

Concluding the story

In 1988 Robert and his family moved back to Sweden. During the summer of their
return they lived in a caravan and a selection of the spices were again back in a box.
Later, when Robert and his wife moved into a flat, Robert put in an order for a new
shelf. In 1990 they finally bought a house and the two IKEA shelves, described at the
beginning of the paper, were purchased.
At present almost all spices have been transferred to recycled Coleman’s mustard
jars. One could argue, from a cognitive standpoint, that a mix of jars, of varying
appearance, would have been better (providing redundant cues as to identity), but
here Robert prefers to let æsthetic concerns take precedence. Cognitive congeniality
is, after all, but one factor that governs the shaping of our environment.
Today Robert has set a self-imposed limit on the continued growth of the shelf.
He confesses to having been ‘a bit of a collector’ in the past, buying spices in order
to learn about them. Now he knows more about his needs and there is also a greater
pressure on available space with Japanese and Thai cooking having been recently
added to his repertoire.
Since the time of the original study, and the last and most recent interview with
Robert, the shelf has undergone further change and is still in flux. Since the initial
study was conducted Robert’s wife has converted to using organically grown spices.
As a consequence the bottommost shelf has been cleared for that use (as seen in
figure 6.2). The two shelves above it now house Western spices used by Robert alone.
However, a short while after Robert’s wife converted to organically grown spices
many shops in Sweden ceased stocking them and they have become increasingly
difficult to buy. Robert’s wife confesses to now using the ordinary spices on the
shelves above hers to ‘top up’ her own jars.
Will the organisation of the bottommost shelf persist, as a vestige of an ephemeral
fad for organic spices, or will the organisation of the shelf eventually return to the
one described in the text? One thing is certain, the present shape and organisation
of the shelf is unlikely to end here and will undoubtedly continue to change, in
concert with Robert’s unabated interest in cooking and in response to ever changing
circumstances.

11 For some other interesting types of interactions and exchanges between co-located
artefacts see de Wit et al. (2002).
 The Cognitive Biographies of Things 127
Discussion

The story told here is a reconstruction of events taking place during a period of
roughly thirty years. As we have seen, the evolution of the shelf is intimately tied
to the changing circumstances of Robert and his wife, Robert’s intensified interest
and growing knowledge of cooking, and even to changes at a societal scale in
eating habits. The biography of the shelf and its use has allowed us to glimpse some
interesting things, a few of which I think bear repeating.
One important insight is the realisation that much of the structure that supports
cognitive activity may have partially non-cognitive origins. At least some structure
seems to be the result of chance, circumstance, compromise, surrounding agents,
and the shaping force of repetition. Perhaps the most significant moment in the
present case study is the emergence of new and unanticipated functionality from the
combination of previously unrelated structures.
I believe that insufficient attention has been paid to these sorts of processes and
their impact on tasks and task environments. We need to continue to explore and
expand our catalogue of these phenomena, but more pressing, perhaps, is the work
of disentangling and understanding their interplay with other, more purposeful and
intentional, processes. Some changes to an artefact or task environment can impact
the cognitive ease with which a task is performed without any major changes in
procedures or techniques, whereas other changes are accompanied by a concomitant
transformation in the way a task is carried out. How people adapt to a new structure,
and appropriate and incorporate it into more congenial forms of a task, is a key part
that needs to be properly worked out.
A better understanding of these phenomena may also serve as an important
corrective to theories of design, production and artefact functionality, that are
excessively intentional.
Another important and related point is the significance of use. One thing that has
been demonstrated in this paper is that the cognitive congeniality of an environment
is as much a function of an agent’s particular use of that environment as it is a
function of the environment itself. It is the particular ways in which things are used
that permits them to contribute in cognitively beneficial ways. Cognitive congeniality
is a relational property, and cognitive biographies must include both the changing
forms and shifting uses of things.
As was noted in the introduction, the reconstructive nature of cognitive
biographies may be the cause of some concern. The way that the present biography
was created, for instance, was through repeated interviews coupled with study of
the contemporary form of the artefact and associated activities. The biography is,
by necessity, constructed after the fact; consequently there are aspects of it that are
based on conjecture. Though cognitive biographies have some distinct advantages
they also open the doors to speculation and presupposition.
This problem, however, is not unique to the present endeavour. There are a
number of related enterprises that have learned to deal with similar issues. There is
much that we may learn, for instance, from the history of technology (for reviews
128 Doing Things with Things
see Staudenmaier, 1984, 1990), the anthropology of technology (e.g. Lemonnier,
1986, 1992; Pfaffenberger, 1988, 1992), social construction of technology studies
and actor-network theory (for a review of both see Bijker, 1994), as well as the
diverse and numerous branches of archaeology. None of these disciplines (and
there are more than those listed here) are specifically focused on cognition, but
all are concerned with the processes behind changes in material culture. These
areas may provide us with supporting evidence, complementary perspectives and
methodological innovations and insights.
Recent focus on cultural biographies of objects (e.g. Appadurai, 1986; Kopytoff,
1986; Gosden & Marshall, 1999) is an interesting example, not only for the partial
neologistic parallelism, but because of shared methodological issues. Both kinds of
biographies seek to retrace sequences of relations between people and things. In the
case of cultural biographies of objects it is a sequence of shifting meanings that is the
elusive quarry, in the present case it is a sequence of uses and cognitions that needs
to be reconstructed. Each quarry is as intangible and ephemeral as the other and we
might find that there are methodological solutions to be shared.
Although interpretative science is difficult there are some potential rewards to
be had. Cognitive biographies allow us, for instance, to explore real life events and
changes spanning long time periods. And they allow us to concentrate our efforts on
sites where significant change has already been established. But there is a further,
fundamental reason for constructing cognitive biographies of things.
Tracing the history of a thing and its use can help us understand the present use
of that thing. A cognitive biography allows us to better discern the cognitive roles
currently being played by an artefact. Against the backdrop of earlier incarnations of
an activity, and previous forms of an artefact, the cognitive functions of a thing are
able to stand out in relief. For example: a feature of an artefact may be the result of a
response or adjustment to problems inherent in previous versions of a task. Knowing
about these earlier phases, enables us to discern (or, at least, to explore) the current
roles being played by this feature. Overlaying succeeding phases of an activity with
preceding ones can often point us to possible areas of cognitive significance.
A disregard for the developmental trajectories of environments, tasks and
people, will therefore lack some of the essential ingredients necessary for a genuine
understanding of the cognition of task performance.

Acknowledgements

I would like to thank the participants and organisers of the Doing Things with
Things workshop in Copenhagen for their reactions to the ideas that eventually
became this paper, and the participants of the Lund University Cognitive Science
seminar for their comments on an earlier draft. I am also grateful to Björn Nilsson
for numerous suggestions, and Pieter Vermaas for some particularly cogent points
of criticism. Thanks are also due my informant Robert for his help, as well as for
 The Cognitive Biographies of Things 129
his delicious cooking. This work was supported by the Swedish Foundation for
Strategic Research.

References

Agre, P. & Shrager, J. (1990). Routine evolution as the microgenetic basis of skill
acquisition. Proceedings of the 12th Annual Conference of the Cognitive Science
Society. Hillsdale, NJ: Lawrence Erlbaum.
Appadurai, A. (1986). Introduction: commodities and the politics of value. In A. Appadurai
(Ed.), The social life of things (pp. 3–63). Cambridge: Cambridge University Press.
Baddeley, A. (1997). Human memory: Theory and practice, 2nd ed, Hove, UK:
Psychology Press.
Barker, R. (1968). Ecological psychology: Concepts and methods for studying the
environment of human behaviour. Stanford: Stanford University Press.
Basalla, G. (1988). The evolution of technology, Cambridge: Cambridge University
Press.
Beach, K. (1988). The role of external mnemonic symbols in acquiring an occupation.
In M.M. Gruneberg, P.E. Morris and R.N. Sykes (Eds.), Practical aspects of memory:
Current research and issues, Vol. 1: Memory in everyday life. Chichester, UK: Wiley.
Bijker, W.E. (1994). Sociohistorical technology studies. In S. Jasanoff, G.E. Markle,
J.C. Petersen and T. Pinch (Eds.), Handbook of science and technology studies.
Thousand Oaks: Sage publications.
Bleed, P. (1986). The optimal design of hunting weapons: Maintainability or
reliability, American Antiquity, 51(4), 737–747.
Bærentsen, K.B. (1989). Mennesker og maskiner. In M. Hedegaard, V.R. Hansen &
S. Thyssen (Eds.), Et virksomt liv: Udforskning af virksomhedsteoriens praksis
(pp. 142–187). Aarhus: Aarhus Universitetsforlag.
Clark, A. (1997). Being there: putting brain, body, and world together again,
Cambridge, MA: MIT Press.
Clark, A. (1999). An embodied cognitive science? Trends in Cognitive Sciences,
3(9), pp 345–351.
de Léon, D. (1999). Building thought into things, Proceedings of the 3rd European
Conference on Cognitive Science, pp. 37–47.
de Léon, D. (2002). Cognitive task transformations, Cognitive Systems Research,
3(3), 449–459.
de Léon, D. (2003). Actions, artefacts and cognition: An ethnography of cooking.
Lund University Studies, 104.
de Wit, O., van den Ende, J., Schot, J. & van Oost, E. (2002). Innovation junctions: Office
technologies in the Netherlands, 1880–1980, Technology and Culture, 43, 50–72.
Engeström, Y. (1999). Activity theory and individual and social transformation. In
Y. Engeström, R. Miettinen and R.-L. Punamaki (Eds.), Perspectives in activity
theory (pp. 19–38). New York: Cambridge Univeristy Press.
Gosden, C. & Marshall, Y. (1999). The cultural biography of objects, World
Archaeology, 31(2), 169–178.
Gould, S.J. & Vrba, E. (1982). Exaptation–a missing term in the science of form,
Paleobiology, 8(1), 4–15.
130 Doing Things with Things
Hammond, K.J. (1990). Learning and enforcement: Stabilizing environments to
facilitate activity. In B.W. Porter and R.J. Mooney (Eds.), Proceedings of the
seventh international conference on machine learning (pp. 204–210). San Mateo,
California: Morgan Kaufman.
Hammond, K.J., Converse, T.M. & Grass, J.W. (1995). The stabilization of
environments, Artificial Intelligence, 72, 305–327.
Hutchins, E. (1990). The technology of team navigation. In J. Galegher, R. Kraut &
C. Egido (Eds.), Intellectual teamwork: Social and technical bases of collaborative
work. Hillsdale, NJ: Lawrence Erlbaum Associates.
Hutchins, E. (1995a). Cognition in the wild, Cambridge, MA: MIT Press.
Hutchins, E. (1995b). How a cockpit remembers its speed, Cognitive Science, 19,
265–288.
Kirsh, D. (1995). The intelligent use of space, Artificial Intelligence, 73, 31–68.
Kirsh, D. (1996). Adapting the environment instead of oneself, Adaptive Behavior,
4(3/4), 415–452.
Kopytoff, I. (1986). The cultural biography of things: commoditization as process. In
A. Appadurai (Ed.), The social life of things (pp. 64–91), Cambridge: Cambridge
University Press.
Lave, J., Murtaugh, M. & de la Rocha, O. (1984). The dialectic of arithmetic in
grocery shopping. In B. Rogoff and J. Lave (Eds.), Everyday cognition: its
development in social context (pp. 67–94). Harvard: Harvard University Press.
Lemonnier, P. (1986). The study of material culture today: toward an anthropology
of technical systems. Journal of Anthropological Archaeology, 5, 147–186.
Lemonnier, P. (1992). Elements for an anthropology of technology. Anthropological
Papers, 88. Ann Arbor, Michigan: Museum of Anthropology, University of Michigan.
Norman, D.A. (1988). The design of everyday things. New York: Doubleday Currency.
Norman, D.A. (1991). Cognitive artifacts. In J.M. Carroll (Ed.), Designing interaction:
Psychology at the human–computer interface. Cambridge: Cambridge University Press.
Petroski, H. (1999). The book on the bookshelf, New York: Alfred A. Knopf, Inc.
Pfaffenberger, B. (1988). Fetished objects and humanized nature: Towards an
anthropology of technology. Man, New Series, 23(2), 236–252.
Pfaffenberger, B. (1992). Social anthropology of technology, Annual Review of
Anthropology, 21, 491–516.
Ribeiro, G. (1996). Situating action III: Acting, dwelling, and squatting: An ecological
approach to the relation between person and urban environment, Ecological
Psychology, 8(2), 131–151.
Staudenmaier, J.M. (1984). What SHOT hath wrought and what SHOT hath not:
Reflections on twenty-five years of the history of technology, Technology and
Culture, 25(4), pp. 707–730.
Staudenmaier, J.M. (1990). Recent trends in the history of technology, American
Historical Review, 95(3), 715–725.
Wertsch, J.V. (1998). Mind as action, Oxford: Oxford University Press.
Zhang, J. & Norman, D.A. (1994). Representations in distributed cognitive tasks,
Cognitive Science, 18, 87–122.
Ziman, J. (Ed.), (2000). Technological innovation as an evolutionary process,
Cambridge: Cambridge University Press.
 Chapter 7

The Woman who used her Walking

Stick as a Telephone: The Use of
Utilities in Praxis1
Hysse Birgitte Forchhammer

Many resources are spent on developing, issuing and adjusting utilities2 in order to
support and help handicapped people overcome their disabilities in every day life
activities. But in practice these technologies do not always fulfill their purpose of
improving the everyday life of the user. Utilities sometimes get in the way of the
user, and they can prove impractical or even useless! A utility might be designed
in an impractical way, so that the user literally bumps into it, falls over it or lacks
the mental or physical capacities to us it. At other times, utilities are used in new,
unplanned and creative ways to deal with an unexpected problem, as when a stick
is used as a ‘telephone’ when an old woman wants to contact her neighbors by
knocking on the ceiling. The use of utilities cannot fully be understood as a function
of design. Utilities – and other things – are used in complex social practices:
complex arrangements of people and material surroundings, imbedded in larger
contexts of other societal practices and discourses. And, furthermore, the use of
things is an ongoing practice. The things and the people involved, as well as their
interrelationships, change, develop and deteriorate over time.
The aim of this paper is to take a closer look at how utilities are used in everyday
practice by people with different degrees of physical and cognitive handicap. I will
argue that the focus of analysis must be the interrelations between utilities, the people
and situations involved, defined by the actual activities pursued.

What are utilities? – Theoretical references

In professional practice, utilities are defined as those objects that are offered to
patients in order to ease their daily lives by compensating for disturbances in certain
functions. Thus, the concept is operationally defined with reference to the intended
technological use. This definition, however, is restricted to only those objects that

1 With apologies to Oliver Sacks (1987).

2 The term ‘utility’ is used instead of ‘disability-aids’. ‘Utility’ refers to all kinds of
things used as aids, and not only objects that are deliberately produced for that purpose.
132 Doing Things with Things
are made in the technological industries and designed with a certain purpose. Since
my aim is to study how utilities are used in practice, the concept of utilities must be
widened to also include things that patients themselves create or use as if they where
utilities. Furthermore it will be necessary to clarify how the concepts of practice
and technology are interrelated. I will briefly present how these concepts and their
interrelationship are defined in the following analyses.
Consistent with the approach of critical psychology (Holzkamp, 1983, Dreier,
1993, Forchhammer and Nissen, 1994) the following reflections will be based upon
an analytical definition in which utilities are regarded as: technologies – strategies
or objects– that are (intentionally or unintentionally) used in practice by different
parties in order to alleviate a problem that has arisen in connection with the patient’s
illness.
In this way I will be using ‘technology’ as an inclusive term for different degrees
of materialized strategies and procedures,3 objects and artifacts. A walking stick and
a bed lift have a tangible material form, whereas a strategy for putting on socks with
only one hand or mnemonics may be considered as objectified but not materialized
in the same sense. A cognitive strategy such as calculation is an example of a socio-
historically developed objectified strategy we are able to relate to almost ‘as if’ it
was given to us in materialized form. In the following analysis, I will primarily focus
on utilities in actual materialized form.
Technology is thus a socially produced object or strategy. Utilities can be
used for many things and assigned many different meanings. Their function can
be transformed into new and different functions that were not thought of before.
Technologies are, nonetheless, often characterized as being designed in a certain
context, by someone, with certain goals in mind. But it is a theoretical challenge
to grasp the dialectics of predetermination and open-endedness in the use of
technologies. I try to handle this problem by using the concepts of objectification
and subjectification. Objectification is a concept, rooted in Marxist inspired thought
(Leontjev, 1981, Heller, 1981), and is used to describe not only the specific condition
in which technology is socially produced, but also how technology in one way or
another represents specific meanings and functions. Objectification may be an
expression used to describe how objects – possibly over thousands of years – are
designed, developed, and adapted by and for people in order to fulfill specific
purposes in the socio-historically organized (re-)production of living conditions. For
example: a knife as a cutting tool or multiplication as a tool for calculating. We
are in our practice as human beings usually able to grasp that meaning, though not
necessarily in a merely reproductive way. We grasp the meaning by a process of
subjectification, in which the common meaning becomes concrete and subjectively

3 A ‘strategy’ is the verbalized and consciously mediated way of solving a problem by

implementing a specific way of doing it. ‘Procedure’ refers to the actual way of performing
this specific task in practice. This way of using the term ‘procedure’ resembles the term of
‘procedural memory’ in cognitive neuropsychology (Gazzaniga, 1999). The distinction
between ‘strategy’ and ‘procedure’ will be explained further in one of my later examples.
 The Woman who used her Walking Stick as a Telephone 133
relevant. In this process, the common meaning may be realized, but may, instead,
be transformed into a new meaning. In this way, the dual character of the process is
also revealed. The process of using technologies is both a process of subjectivation,
making sense in practice, and also of objectivation, producing (new) meaning that
potentially can transform the concrete situation and be related to by others. The
processes of subjectivation/objectivation come into play when an object, that may
have had a different function in another context, is used as a utility and is assigned
meaning as a utility by the user(s) in a different context, as when two phone books
are used to raise the seat of a wheelchair.
Objectification/subjectivation is the process through which the socially produced
character of technology is mediated through the use and ascription of meaning.
Subjectivation is hence the process in which an object is used with a specific purpose
in mind, being something that I/we can use for my/our particular goal, e.g. when a
rock that happens to be lying on a beach is used as a seat on which I can rest. But
this process of ascribing new use does confront limits. Not only do the physical
characteristics of the object thus place constraints on how the object may be used, but
also societal thought forms, discourses, ideologies and power-relations are manifest
in the concrete use of technologies. The different discourses of how, where, and by
whom, utilities should be used, are a concrete condition in the use of utilities.
Another concept, that of everyday practice, also mainly derives from critical
psychology (Holzkamp, 1995). The concept of the subject as participating in the
societal praxis is fundamental to being human. The subject is seen as a mediating
authority who transforms the objective sociohistorical action conditions into
subjective meaning and concrete action. The social life is organized into different
action contexts and the participants move between several of these contexts. The
contexts are limited by different relatively well-defined goals and means.
Most of the participants in my research live their lives in relatively limited
geographical frames. Most of them seldom leave their apartments. Movements
between different contexts nonetheless, characterize their lives. They move between
their home and different institutions, e.g. hospitals, doctors, rehabilitation center,
etc. New action contexts are continuously created as the ‘domestic carer visits the
participant to help with personal hygiene’ or when ‘a family member comes to visit
and drink coffee’, etc.
Life is thus structured in and around action contexts that are relatively delimited
units, but life is also a continuation through these different contexts. Consequently,
daily life is both homogeneous and heterogeneous, i.e. both continuous in terms of a
series of more or less reflective, goal-oriented, actions and at the same time split up into
smaller units of different routines, relaxation and ‘doing nothing’. A standard reading
of Holzkamp would lead us to think that only activities that contributed directly to
‘the widening of societal production and widening of societal action potency’ would
count as ‘real human activities’ (Dreier, 2002, Huniche, 2002). According to such an
interpretation, the theoretical understanding of everyday life presented by Holzkamp
(1995) would give rise to an almost normative distinction between real goal-oriented
activities, on the one hand, and less important, daily routine hurly-burly, on the other.
134 Doing Things with Things
While looking at human practice as organized in activity contexts has its value,
this theoretical perspective fails to grasp the entire complexity of human practice.
Everyday life is also characterized by non-goal oriented activities. Concepts like
procedural and routinized activities encompass other aspect of activities. But also
concepts of playing (Keller, 2005), relaxing and doing nothing obviously must be
considered as important to understanding this non-goal oriented forms of activity
that form the complex whole of everyday practice.
Finally, I will argue that human everyday practice is social. As will be shown later
in the empirical examples, the everyday use of utilities is socially distributed. Even
though utilities are often designed for individuals, they are often used by, or involve,
a number of other people or institutions. Human practice is also socially organized
from a genealogical perspective (Foucault, 1987): every practice has a history and is
embedded in, and related to, other layers of socio-historically developed practices.
Although such a genealogical analysis will not be central to this paper, I will in
some of the examples refer to historically developed thought forms or discourses on
handicap and rehabilitation with reference to other genealogically based analyses of
the field.

The empirical study

The following examples are based on interviews and observations of a group of 25

stroke victims focusing on their use of handicap-utilities. Twenty-five participants
were selected from the larger group of 150 patients who participated in a study
of stroke and functional ability (Andersen, 2000). According to the standardized
medical rating-scales, the 25 participants were all suffering from mild to severe
symptoms following their cerebral stroke. They had all been hospitalized for 12–24
months before the interview. At the time of discharge from the hospital, all received
one or more specially designed handicap utilities. In most cases, technologies to
facilitate walking, such as a stick, a walking-frame or a wheelchair, or grips and
handles had been installed in their apartment.
As all the participants had undergone a standardized medical and neuro-
psychological test both at the time of their discharge from hospital and 6 month later,
it was possible to select a sample that varied in respect to age, gender, social-status,
functional and medical status and level of cognitive disabilities.
The participants gave their written consent to participate in the interview.
Furthermore, it was suggested that a close friend, a care giver or some other person
involved in the everyday life of the patient should also take part in the interview. In
this way, the socially distributed aspect of use could be included in the study. But,
in addition, the companion would also be able to assist in the interviews themselves
given that some of the participants, due to their aphasia, had problems communicating
(Forchhammer, 2001). With few exceptions, the interviews took place in the homes
of the participants. All interviews were taped and transcribed. Furthermore field-
 The Woman who used her Walking Stick as a Telephone 135
notes were taken at all visits. The interview sessions varied from 30 minutes to 3
hours.

Analysis of daily life with utilities – when utilities are not used in the
intended way

Although, utilities are supplied in terms of one or more functions, sometimes these
functions are not fulfilled. In the participant group, I found several examples of
utilities that were not used but rather were hidden away or seen as disturbing to
the daily life activities of the patient. In a large ‘functionalist’ apartment complex
in Copenhagen, an 82-year-old woman lives with her husband. She had suffered
a stroke. Ever since the apartment was built in the 1930s, the couple had lived in
the same apartment with a balcony. When their two children lived at home, the
apartment seemed too small, but now the husband and wife live alone, and they are
comfortable with the space in the apartment. The woman has a right-side paralysis
and global aphasia, i.e. she has extensive difficulties with speech, writing, reading,
and the comprehension of language and symbols. She can only move around by
using a wheel chair and she needs assistance when she has to move from her wheel
chair. Her husband takes care of almost all the domestic work, and so the need for
home care is minimized. The woman had been supplied with a wheel chair when
she was discharged from the hospital, but now the wheel chair had been left in the
corridor taking up space. The problem was that new doors had been installed in
the apartment shortly after the woman came home from the hospital, and now the
wheel chair could no longer pass through the doors.
The patient’s daily life situation had changed in such a way that the utility could
no longer be used as planned. Small changes in the home, such as a new balcony
door, changed flooring, or new furniture, can disable the functionality of the utility.
In some instances the errors are made from the outset: if a handle has been placed
in the wrong place or in an impractical place, then it may not used at all. Officially,
it is the responsibility of the local authority, as represented by the domestic help
or nurse, to report about the lack of use of utilities to the primary sector. In praxis,
this system seldom works. The professionals who work in the home usually have
other primary tasks to perform, e.g. cleaning, helping with personal hygiene, or
nursing wounds. In praxis, the helper adapts to the impractical placement of a
handle by supporting the patient and developing different compensatory strategies
to make up for the lack of a practical utility. The formal goal of helping the patient
wash herself is thus fulfilled. However, the rehabilitation element of helping the
patient to be able to cope with the task herself has been placed in the background.
Maybe the helper fails to appreciate the double function of the handle. Maybe the
helper is unaware of the possibilities of changing the position of the handle or
of other products on the market. It can also be a question of the frequent use of
substitutes and the turnover of staff so that no domestic helper is ever in the same
136 Doing Things with Things
home over a sufficient period of time to be able to obtain extensive knowledge of
the participants’ special rehabilitation needs.
We can, thus, identify a certain type of problems with the use of utilities, that
is presumably common among most professionals, and which, in most cases,
could be solved with relatively limited resources: through better evaluation of the
utility in its concrete setting and also by the introduction of a more systematic and
frequent re-evaluation procedure in which the actual functionality of the utility is
evaluated and adapted.
In the example of the door handle, the utility had lost its practical functionality
because of physical changes in the apartment. In other cases, the changes that
occur concern the patient. It could for instance be that the functional aspect of the
user’s ability to act has changed, such as a physical deterioration in functioning so
that they are no longer able to use the utility.
For example, a 62-year-old woman had developed a considerable spasticity
in her paralysed hand and could no longer grab the handle on her walking tool.
Consequently, she stopped using the walking support and instead now supports
herself by holding onto different objects in her living room, thus increasing the
risk of her falling.

When use is context dependent

Technologies can be more or less connected to a specific location or dependent on

certain practical conditions, e.g. an electric toothbrush has to have electricity and
is designed to be used in the bathroom, or somewhere at least where there is access
to water and a drain. Other utilities are designed to be moved from place to place
and to be used in different contexts, such as a walking frame.
When a utility is not used as planned, it may be a matter of changes in the
physical environment, or else remission or changes in the participant’s physical
abilities. But it may also be because the particular utility has never been tried in, or
adapted to, the specific environment in which the utility is to be used. For example,
one woman had an emergency button installed, but this was placed on a table in the
living room, thus preventing her from reaching it when she fell in the hall.
It can also be the case that training in the use of a tool is given in one context
but the tool is actually going to be used in a different context. This kind of transfer
can be especially problematic if the patient due to her brain damage has acquired
specific problems of abstract thinking, memory or other specific cognitive
disturbances. A participant learnt, for instance, how to put on a support sock whilst
sitting on a hospital bed with a tall edge against which she could support herself.
At home, however, she has a bed without tall edges, and so she had to re-learn
the entire procedure. In this case, the patient had cognitive problems concerning
abstraction and breadth of view and problems in visuo-spatial orientation, which
made it difficult for her to apply the learning from one context to another context.
 The Woman who used her Walking Stick as a Telephone 137
When the utility is used in new and untraditional ways

In a different case, the physical level of functioning of the participant had improved,
so that she no longer used her utility in the traditional sense. A younger participant,
who suffered a stroke at the age of 45, had to give up her former job and now
receives disablement pension. She spends most of her time in her apartment and
has taught herself to work with word processing and spreadsheets at the computer.
Moreover, she is active in several associations where she writes summaries and
balances the budget. Despite her activities, the days are still long. She watches
TV and cooks her own breakfast and lunch. When she first arrived home from
the hospital, she felt ‘bound to the wheel chair’. Now she moves around in her
apartment by supporting herself against different items of furniture, such as doors
and window frames. She explains how she was told that she would never be able to
walk again. However, she proved them wrong. She has always done more than she
was ‘permitted’ by the personnel concerned with her treatment and rehabilitation.
She claims that her tendency to cross the lines, placed by the professionals, is the
reason that she has come as far as she has. It has, however, also been the reason
that she has fallen a few times, and her friend claims that the participant often
overestimates her own strength. She still has her wheelchair, but she no longer
uses it as a wheel chair. She uses it, instead, as an ‘office chair’. When she works at
the computer, she is comfortable using the wheel chair as an office chair, because
it enables her to drive around and change her positions in a way that she could not
with ordinary office chairs. Furthermore, she uses the wheel chair as a ‘launching
pad’ when she has to stand up. By having a wheel chair, on which she can lock
the wheels, she trusts that the chair will not disappear under her when she has to
stand up.
A couple of times a week, another participant gets help from the upstairs
neighbor. He comes down to help with the shopping and to get a cup of coffee.
They have arranged that the participant will use her walking stick to knock on the
ceiling, whenever she is ready for a visit.
In the examples I have been discussing so far, such as the wheel chair and
walking stick, the utility has continued to be used for its intended function, but
also used for an ‘improvised’ function. In other cases, the technology of utilities
far exceeds being a consciously used utility and is instead an ordinary household
item. Things and objects that had originally been used to compensate for a function
disorder have now become a part of the every day inventory, as when a specially
designed chair for use in the shower is used as a flower-bench.
In other cases, the process is reversed. Ordinary pieces of furniture, or other
items that the participant has always possessed, change their function and are used
as utilities: furniture is used for support and take-off, books are used to raise and
support, bath stools are used as steps or plant stands. Several participants reported
that they have changed their procedures. They had, for instance, found a different
way to put on their clothes than the procedure used in the hospital. The participants
138 Doing Things with Things
were, however, seldom consciously aware of the fact that they had created new
utilities. That is, the participants had made small changes and adaptations – actions
– probably without taking any notice of these changes. The participants did not
mention these changes spontaneously, but the changes became clear the minute
I asked the participants to show me how they used a particular item or how they
followed a certain procedure.

When we encounter limitations upon transformation and meaning ascription

Things change meaning when they are used in new ways. There are, however,
limitations on meaning ascription and transformation of use. Although it is difficult
to define abstract limitations on the meanings that can be ascribed to the technology,
the participants encounter such limitations in praxis every day. A person might well
assign human traits to the walking stick, such as calling it a ‘friend’ or an ‘enemy,’
but when communicating with another person there has to be some agreement
about whether or not the words are meant ironically, concretely or metaphorically
with regards to the utility in question. An aphasic might find a situation such as
this very difficult as a result of their problems with contextualization and framing
of meaning ascription. The aphasic participant may appear childish or unclear
when she identifies her walking-stick as the ‘stupid, stupid, you know, that one’.
The limitations of meaning ascription are thus not embedded in the technology
itself but rather defined by the existing societal discourse and hence situationally
defined.
The material characteristics of the technology might also set limits on what
the object may be used for. The object may fall over, break, roll away, etc. At
the same time, the specific shape of the object objectifies a more or less open
potential of usage. The more advanced a technology, the more firmly defined is
this potential. Transforming this potential of usage demands a certain competence
on the part of the participant. Some technologies require far more competence
than others do. An electronic bed, for instance, requires a certain amount of
learning and conscious appreciation of the specific possibilities and limitations
of the technology. But neither the competence nor the objectification of the usage
potential is independent of the situation in which the technology is used. When a
95-year-old demented stroke participant is about to fall, he automatically reaches
out toward his walking frame – and not the door handle. He cannot consciously
explain this action. The act is rather an expression of a bodily learnt competence,
a procedurally acquired, pre-reflexive, and automatic action sequence. In this
situation, the usage potential of his utility is objectified as meaningful to him. The
utility user’s situational embeddedness helps ‘redeem’ different usage potentials
of the particular technology but this never occurs independently of the subjective
competence.
 The Woman who used her Walking Stick as a Telephone 139
When the socially distributed usage of utilities is overlooked

Utilities are not only used by the people who receive them. Some utilities, such as
a lift, are designed to be a tool for the helpers, whereas other utilities are designed
for the patients to use, as in the case of a support sock.
In praxis, utilities are used in various and frequently unpredictable ways, and
often also in both social and distributed ways. The daughter of an 80-year-old
demented woman has bought a calendar, which has been placed on the table next
to the phone in the apartment. In order to make daily life easier for the mother,
the domestic care bought yet another calendar, that the participant could carry
around with her. The calendar by the phone has a page for each separate day. In
this calendar the program of the day is recorded. In the carry-around calendar,
birthdays and arrangements during the week etc. are written down. Although the
patient is able to read and write in the calendar whenever she is encouraged to do
so, she does not, however, use it spontaneously. Her daughter and the domestic care
explain that it is a recurrent problem that they have to encourage the participant to
use the calendar. The daughter worries that her mother is not oriented in time and
place, and she views the calendar as a way of linking her mother to reality, so that,
for example, she ‘at least knows what day it is’.
Despite the fact that the mother never uses her calendar, it has become a utility
that is frequently and collectively used: it has become a journal that carries messages
between the different parties who come to visit or take care of the mother. Hence,
the calendar fulfills several functions in her daily life of the participant. However,
although it is used as a utility by all the people helping the participant in her daily
life, it does not help the participant with her memory problem.
The old woman has a type of cognitive problem that makes it hard for her to
use the utility in her daily life. When the utility is used in standardized conditions,
for instance in the hospital guided by the occupational therapists, the patient has
no problems using the calendar. At home, however, her lack of initiative prevents
her from using her calendar. A ‘hidden’ disorder of the functional aspects of the
activity hinders usage of the utility.
This example also illustrates that the utility objectifies a particular socially
produced standard. The calendar is a product of a particular historical period, in
which structuring daily life into units of time is a central part of the life conditions,
and is also seen as central to the discourse on ‘the good life’. The calendar is
designed to be used by someone working an 8 a.m. to 5 p.m. day, as a tool for
maintaining and recreating the daily life. Now, most people do not experience
daily life as a homogeneous row of repetitions. Rather, they experience changes,
unexpected events, etc. A calendar thus becomes a central element for instance for
anyone in the process of ‘returning back to daily life’ after a disabling illness. The
calendar is designed in such a way that appointments and special occasions can be
noted, thus implying the routine of daily life and important changes and ‘breaks’
in daily life. Routine usage of a calendar provides the necessary background for
making special occasions stand out. Notes in the calendar are supposed to be used
140 Doing Things with Things
as cues to perform certain actions and procedures; e.g. ‘call my daughter to wish
her happy birthday’ or ‘make myself ready to go to the day-care center’, etc. These
are sequences of action that go beyond daily life in the apartment. They presuppose
and involve participation in other contexts.

Conflict, ideology and power

In praxis, a (latent) conflict may exist concerning what kind of usage is the right usage.
As we saw in the calendar-example, the daughter and the domestic help may have
confused means and goal: the use of the calendar became a goal in itself. The goal of
the participant may have been quite different, for instance to remember what she had
forgotten. In this respect, the calendar was of no use to her. She not only forgot to use
it but also the verbal cueing in the calendar did not help kick-start the expected action
sequences. Because of her brain injury, there is only a weak connection between
conscious, verbal planning and the actual execution of more implicit procedures/
sequences of action. This latent conflict, concerning the mother’s lack of use of
the utility, embodies an element of unawareness of the character of the mother’s
cognitive problems. It is also an objectification of a certain societal discourse on
memory as a social competence. Now, although having a bad memory, as such, is
generally acceptable and something we can all relate to, it is, nevertheless, important
to take care to respect the acceptable and available mnemonics of the culture. It is not
normally acceptable to forget appointments, special occasions or social events. Yet,
if seen as a sign of illness, decay, or senility, it is no longer regarded as unacceptable
or even rude. However, framing the mother’s behavior in this way emphasizes her
deterioration and her loss of personal integrity and autonomy. Hence, one could
claim that some utilities, in specific social contexts, objectify a higher status than
others do. Or, in other words, technology often potentially objectifies a latent social
conflict.

The neglected proficiency of ‘disabled’ action

An important perspective is still missing in the analysis of the calendar-example:

the participant’s own experience of – and method of dealing with – her memory
problem. Although the participant currently speaks of ‘my memory problems’,
one has to wonder whether or not the participant would – independently of neuro-
psychologists, domestic helpers, etc. – have ever referred to it as a memory problem?
When asked about her daily life after leaving the hospital, she primarily focuses on
her problems in maintaining her balance and not being able to move around freely.
She is often in pain and it is difficult for her to get anywhere. She also explains that
she has a tendency to forget things. When I asked her to expand on this, she found
it difficult to say anything more specific. She suggests that I ask her daughter. Later
on, she gets up and walks with a great deal of trouble out of the living room. I can
hear that she turns around in the kitchen and comes back into the living room, only
to turn back and go into the kitchen again. This happens repeatedly until she finally
 The Woman who used her Walking Stick as a Telephone 141
comes out with a bowl of sugar, which she places on the table. These acts illustrate
that the participant – without being consciously aware of the fact – brings herself
back to the situation in which the urge to act originated. Hence, she ‘is reminded’ of
what she wanted to do. As a memory aid, she reconstructs the situation and it helps
her to remember. This strategy seems to relieve her problem, but it does not address
the fact that the people around her still regard her real problem as lacking a memory
for time and place. The people in her immediate surroundings failed to notice the
strategy that the participant herself had developed. Thus the participant’s own
definition of the problem and her own strategies of conduct have been overlooked
and overshadowed by interpretations made of others in praxis. This has led to the
choice of an inappropriate utility, i.e. the calendar.

Concepts of illness, health, and the good life

In an interview, a 65-year-old man spoke about his problems in getting outside his
apartment. He can move around inside his apartment, although with some difficulty,
but he cannot walk outside alone. He has a daughter who comes to help him almost
every day. The patient can walk with a little support, but because of his left-sided
neglect, he keeps bumping into the person supporting him. The father and his
daughter have borrowed a wheelchair a couple of times. This has increased the
action radius of the patient considerably. Now, three years after two serious strokes,
doctors and physiotherapists have given up on further rehabilitation. They tell me
that the patient, however, does not want a wheelchair, and this decision is partly
supported by his daughter. Accepting a wheelchair would be the same as resolving
to give up on rehabilitation. For this participant, not using utilities is a part of the
political struggle for rehabilitation.
The act of not using utilities received from the hospital may be experienced as
an indication of improvement: an experience of being less disabled than before. To
this participant in his daily life, the utility is a symbol of illness, an externalization
of the inner, organic dysfunction. The utility symbolizes a need for help and support.
This is seen as negative in this example, not because of a general aversion toward
receiving help, but because the existence of the utility is a sign of ‘somebody else’
giving up on his rehabilitation. He has already had experience of the professionals
making incorrect judgments, when they insisted to him that he would never walk
again. Using utilities perpetuates the patient in a dependent relation to the judgments
of the professionals. It objectifies a common assumption that functional disorders
can only be rehabilitated within the first year; thereafter the disabilities are seen as
permanent. By accepting the utilities, the participant feels he would surrender to this
defeatist standpoint.
Paradoxically, the participant himself nevertheless reproduces this view of his
illness as a temporary condition: something that has temporarily ‘entered’ into the
body but will leave again when he gets well. Getting well is – in this understanding
– the same as getting totally rid of the illness and returning to the normal condition
(i.e. before the illness) without any injuries or signs of the illness.
142 Doing Things with Things
To some extent, the participant shares this idealistic view of illness as an abnormal,
deviating condition, which contrasts with the normal condition. But at the same time
he states that he knows that he will need rehabilitation for his paralyzed leg and
that his leg will probably never be the same again. These two very different views
of the illness do not seem to represent a conflict for the participant. The idealistic
view, i.e. that he will get completely well again, exists as a goal for the participant,
even though he describes the paralysis of his leg as probably permanent. It seems,
however, that the professionals, who have been in touch with the participant,
have interpreted these apparently conflicting views of the illness as an expression
of lack of realism and inadequate recognition of his own situation. According to
the participant, the doctors and the therapists at the hospital tried to give him a
more realistic view on his situation. But the professional discourse on illness is as
contradictory and incoherent as the participant’s. The concepts used in treatment and
rehabilitation contexts, however, are just as different and contrary: illness can be
seen as either chronic and life long, or else as a process or as an event (Stiker, 1999).
The discourse can concurrently include defeatist, biological deterministic, optimistic
or even unrealistic views of how the power of will and the specific abilities of the
individual can cope with the illness (Rose, 1999).
The discourse of illness and rehabilitation is still being negotiated and further
developed. What may seem a meaningless discussion of words may be important
from a participant’s perspective because it may be a matter of what rehabilitation
offers or whether or not a participant gives up or feels renounced by others. Different
parties participate in different contexts in which important decisions are made
concerning further rehabilitation. Utilitarian considerations concerning the patient’s
possible gain from a particular rehabilitation program compared to other patients’
gains are part of the professional perspective but do not make sense for the patient.
In short, although many patients and relatives have struggled long and hard
to gain access to rehabilitation, the utilities themselves, that may sometimes be
essential to the process of rehabilitation, may nevertheless come to symbolize quite
the reverse. Thus, in praxis, the ascription of negative meaning to the utility may
limit the action possibilities for the participant, even though it may feel like a victory
to be able to put it aside.

Utilities as ‘stigmata’

In my analysis, I have emphasized how the use of utility technology is situated and
how meaning and function is transformed and changed in the social praxis in which
the participant acts. But my interviews and observations also describe another side of
the utility user, that I have not yet discussed in this article. Some of the participants
saw the utility as more static and constraining, a ‘stigma’, a negative sign of disability,
helplessness and exposure (Goffman, 1969). For some of the participants this is not
a temporary meaning ascription but rather an interpretation that is maintained across
situations and different action contexts within which the participants move.
 The Woman who used her Walking Stick as a Telephone 143
A 67-year-old woman, who now lives alone in her apartment in an old apartment
complex, talked about her decision not to use utilities. Her late husband had a small
business and, for most of her life, she had been part housewife, part helpmate.
Following her stroke she has been able to manage almost without any help. In the
past few years, however, she has experienced a growing problem with her balance.
She fears that she might have had yet another small stroke. Her doctor has tried to
comfort her and has started her up in treatment with Prozac. A domestic help, who
comes every fortnight, arranged for the woman to be offered the chance of having
a banister put up in the stairway. But the patient did not want a banister. During the
course of the interview, the woman reported that she had not told any of the people
living on the same stairway as herself about her illness. She did not want to ‘expose’
herself and she did not want strangers to take pity on her. The participant tried in
several ways to conceal and cover up her disability. She views her functional disorder
as a serious problem when meeting strangers in the street or meeting others living on
the same stairway. Her unsteady gait – and its externalization in the form of an extra
banister – becomes, in Goffman’s terms, a socially constructed stigma. Once again,
borrowing a term from Goffman, she tries to ‘pass’ by hiding her disability and thus
turns down the use of utilities, such as the extra banister.
The concept of socially constructed stigma tends to imply that the freedom of
action of the participant is limited by an independent and presupposed meaning
ascription that she merely reproduces across the different contexts in which she
participates. But such meaning ascription is not static nor independent of the particular
situation. The participant can for instance live with – or, as she herself puts it, ‘has
learnt to live with’ – the fact that the domestic help comes once every fortnight.
One can interpret this situation in many ways. Her attitude might be attributed to
a depressive state of mind, or as an attempt to hold on to a former definition of
quality of life. Such interpretations would not, however, be appropriate to make on
the basis of the interviews and observations that I have made. This example itself
demonstrates that for some participants there is an on-going process of reification,
in which the technology is objectified and ascribed relatively independent meanings
of stigmatization. The participant tries to escape from being stigmatized, but
paradoxically, the process of stigmatization is maintained and reinforced by the
participant’s attempts to cover up, hide and ‘pass’.
As a contrast to the above example, another participant illustrates entirely
different attitudes and action strategies: she uses a wheelchair and has had several
different ramps installed so that she can get out of her first floor apartment with help
from up to two people. When she has to get out of the apartment, she yells out of the
window to contact her two neighbors working in a garage across the street. If they
have time, they come over and carry her out of the apartment. Her contact with the
world around her has been extended ever since she fell ill. The domestic help comes
daily, her daughter calls frequently to make sure that her mother has not fallen or
become ill, and the two neighbors carry her out of her apartment – all these contacts
have been established or intensified as a result of her stroke. Her disability and the
144 Doing Things with Things
visible signs or stigmata have in some way been her gateway to a more extensive
social life.

Summing up

I have based this article on an examination and discussion of concrete examples of

how stroke-patients use utilities in daily life. The research reveals that the utilities
issued to such patients are not always used as planned. Such wrong, different or lack
of usage is caused by different conditions. Often the utility is insufficiently designed
for the specific context in which it is to be used. In other instances, certain (hidden)
cognitive difficulties have been overlooked in the patient. Lack of information and
training for the patients and helpers can also be a problem. Furthermore, the follow-
up and re-evaluation of needs and tools does not always function optimally. In these
latter cases the problems with the use of utilities are caused by poor evaluation of the
patients’ needs and changes in levels of function and/or changes in the surrounding
environment.
I have also pointed to the fact that utilities are often used with entirely different
purposes than those originally planned. This is not necessarily problematic. On the
contrary, this new transformed usage can be more meaningful for the user. However,
new transformed uses of a utility can be inconvenient for the patient or the helper.
Moreover, it can be expensive and inappropriate if advanced and/or specially
designed utilities are used as ordinary household items belonging to the user e.g.
using the wheel chair as an office chair or the bath stool as a flower bench.
This analysis has shown how the use of utilities is socially distributed. Often
many parties use the utilities. If this aspect is overlooked, the potential of this
distributed use may also be forgotten. It can also be difficult to understand and
solve the conflicts that can arise as a consequence of the collective or distributed
use of the utilities. Different parties may have different – and not always explicitly
defined – understandings of the participant’s problems and need of utilities. If these
differences are not acknowledged, conflicts can arise. Moreover, these differences in
usage may lead to an overshadowing of the participants’ own strategies of solution.
Utilities become, for some participants, a symbol of illness, and in some cases
the termination of using the utilities is a sign of improvement and getting well. This
act is seen as part of the battle to attain rehabilitation– in order to be acknowledged
as a person with a potential to regain lost functions, and not just seen as a ‘hopeless
case’. In a few cases, the utilities are seen as stigmatizing, and the user may feel
constrained and choose not to use the utilities.
The people involved in this research belong primarily to the group of elderly
people with stroke. They are all retired and with a relatively low level of education
and economic status. It is noticeable that the type of utilities offered to this group
of people is a relatively simple, low technological tool. Apparently, this reflects
the limited goal of ensuring a minimum level of functioning in their own home;
it does not, however, ensure the patients’ participation in leisure activities, and the
resumption or development of new interests. In this article, however, I have not
 The Woman who used her Walking Stick as a Telephone 145
dealt with what wishes and/or needs this group of people formulate themselves for
the development of the technology of utilities. Instead, I have focused on how the
existing utilities are used and what problems these uses bring about.
In order to solve some of the problems discussed and take advantage of the
potentials of distributed use of utilities, it is necessary to look at the concrete and
situated use of utilities. It is hence necessary to examine the practical and individual
barriers systematically, e.g. cognitive difficulties. Moreover, it seems necessary to
develop a more systematic administration system that ensures evaluation and follow-
up procedures regarding the utilities that have been issued. The administrative system
must make sure that utilities are returned if the specific utility is no longer needed,
or if an old utility is replaced by a new one. Thus, utilities are to be re-evaluated on
a regular basis to ensure that the most appropriate utilities are issued and used. This
evaluation process must be based on an analysis of the practical usage, i.e. analysis
of how the utilities are used, who uses them, where they are used and what they are
used for. As part of this analysis, considerations of the character of the problem,
observations of users and their own solutions should be included.
In the development of new utilities, it is essential to take seriously the
considerations of the users themselves regarding their needs, and the functionality
and aesthetics of the utilties. But the analysis also points to the fact that some users
do not like using the utilities because they symbolize disability and passivity and
the abandonment of the very possibility of rehabilitation. It would therefore seem
obvious to try to develop any utility so that it embodies an element of rehabilitation
within itself.

References

Andersen, H.E., Schultz-Larsen, K., Kreiner, S., Forchhammer, B.H., Eriksen, K.

& Brown, A. (2000). Can readmission after stroke be prevented? Stroke, 31,
1038–45.
Bech-Jørgensen, B. (1994). Når hver dag bliver hverdag. Copenhagen: Akademisk
Forlag.
Dreier, O. (1993). Psykosocial behandling. En teori om et praksisområde.
Copenhagen: Dansk Psykologisk forlag.
Dreier, O. (1997). Subject and social practice. Center of Health, Humanity and
Culture, Department of Philosophy, University of Århus.
Dreier, O. (2002). Introduktion til Holzkamps arbejde om livsførelse. Nordiske
Udkast, No. 2, 3–5.
Foucault, M. (1987). Sexualitetens historie I. Viljen til viden. Humlebæk, Denmark:
Rhodos.
Forchhammer, H.B. (1994). Psykologiske sunhedsbegreber i subjektvidenskabeligt
perspektiv. In U.J. Jensen & P.F. Andersen (Eds.) Sundhedsbegreber – filosofi og
praksis (pp. 139–179). Århus: Forlaget Philosophia.
Forchhammer, H.B. (2001). Interviewet som handlesammenhæng. Nordiske Udkast,
2001, no.1, 23–32
146 Doing Things with Things
Gazzaniga, M.S. (1999). The new cognitive neurosciences. (2nd. ed.) Cambridge,
MA: MIT Press.
Goffman, E. (1961). Asylums. London: Penguin Books.
Heller, A. (1981). Das Alltagsleben.Versuch eine Erklärung der individuellen
Reproduktion. Frankfurt: Suhrkamp.
Holzkamp, K. (1983). Grundlegung der Psychologie. Frankfurt/M: Campus.
Holzkamp, K. (1995). Alltägliche Lebendfürung als subjectwissenschaftliches
Grundconzept, Das Argument, 37(212), 817–46.
Huniche, L. (2002). Huntington’s disease in everyday life. Knowledge, ignorance
and generic risk. Institute for Psykology, Copenhagen University.
Keller, K.D. (2005). The corporeal order of things: The Speil of usability. Human
Studies, 28, 173–204.
Leontyev, A.N. (1981). Problems of the development of mind. Moscow: Progress.
Sacks, O. (1987). The man who mistook his wife for a hat. Copenhagen: Borgens
forlag.
Rose, N. (1999). Powers of freedom. Cambridge: Cambridge University Press.
Stiker, H.-J. (2000). A history of disability. (trans. W. Sayers) Ann Arbor: University
of Michigan Press.
 Chapter 8

Politics of Things: Interplay of

Design and Practice in a Design
Workshop with Children1
Estrid Sørensen

In her book Plans and Situated Actions, Suchman (1987) shows how plans and other
things become involved in practices as resources for the formation of actions. Plans
are not prescriptive for action, she emphasizes. The same can be said about any
thing, about any design. Designs engage in practice, they do not impose their order
on practice. But what then, can we say about the interplay of design and practice?
Does a fixed and standardised design result in a rigid practice, and does an open
design bring along flexible practices? These are the central questions of the following
chapter.
With inspiration from later developments within actor-network theory (ANT)
(Mol & Law 1994; Law 1994; Law 2002) I use the term ordering to discuss two
cases from a design workshop with fourth to sixth grade primary school children,
who worked with an open-ended design of a 3D virtual environment. The ways
in which children, design, computers, researchers etc. came to relate to each other
formed different orderings. Talking about how human activity is ordered is in other
words talking about politics. I argue that things have politics.
After first introducing the after-school project Femtedim I describe the design
involved in this workshop. I describe how different orderings were inscribed (Akrich
1992) in the design, defining an open-ended activity. Secondly, I present descriptions
from my fieldwork to illustrate how this open-endedness was continued into practice.
I show that different orderings were at play and I conclude by discussing this more
procedural way of understanding design compared to seeing design as a finished
product, and compared to defining design as linked to designers’ intentions.

1 I would like to thank the editors of this volume and several authors, who have been
very helpful with comments on this chapter, as well as Jeannette Pols, John Law and Lucy
Suchman, who have commented on earlier versions of the chapter.
148 Doing Things with Things
The design workshop

Participants and aim of the research project

The design workshop called the Femtedim project was established in a 5th Dimension2
after-school activity situated in an urban school in Copenhagen. Children from nine
to twelve years of age attended the design workshop which ran once a week from
September 2000 to June 2001. The children enrolled by giving their informed consent
but participation was voluntary and although some of the children attended every
time, others did so only irregularly. About 40 children passed through the Femtedim
project of which 15-17 constituted the main group of regular attendants.
The aim of the design project was to create an open-ended design that kept
the interaction between children, design and other resources evolving. The reason
for experimenting with such an open-ended design was rooted in the general
dissatisfaction with overly standardised educational practices that are not able to
challenge and support children’s individual potentials (Lee 2001). An open-ended
design should not impose any a priori definition of children’s needs, but be flexible
enough to develop challenges and support along the way. Similarly, with the ambition
to make the project take a point of departure in the children’s own activities no
learning goals were set. In this chapter, I will not go further into discussing the
extent to which this aim was reached, but focus on how design can contribute to
establishing such open-endedness.

The design concept

In the Femtedim project, design is understood in a quite broad sense. It concerns

partly the forming of an interface of a graphic 3D virtual environment program
called Active Worlds,3 partly forming the practices in the computer lab with and
around the program. Design comprises more than the interface designers usually
describe in their definition of design. Akrich (1992), Grinth & Woolgar (1997)
among others have shown that even if these ideas are rarely accounted for in
designers’ descriptions of design, designers do indeed have ideas about users and
users’ activities with the technologies they are developing and these ideas guide the

2 5th Dimension is a concept for a joint computer based after-school activity for
children and a research field for researchers studying child-computer interaction developed
originally by Cole and Griffith in the mid-1980s (Cole 1996). It is traditionally based on
an activity theory framework with a number of designed artefacts to support learning and
development. During the past 20 years the 5th Dimension concept has spread to first various
places in the USA, but later also to South America, Australia and Europe. At each place, the 5th
Dimensions are locally adapted, meaning that they may involve different artefacts, different
user groups, different research aims, theories etc. The cases described in this paper is a result
of the establishment of a new 5th Dimension site in Copenhagen in 2000, partly financed by
the EU 5th Frame program. For more information, see http://www.5d.org.
3 http://www.activeworlds.com
 Politics of Things 149
design process. It is simply impossible to design a technology without imagining
something about the practice it is going to be part of. As Winner (1995) has explained
the countless ways in which machines, instruments, and structures of common use
were designed in the 70s – buses, buildings, pavements, plumbing fixtures, and so
forth – made it impossible for many disabled persons to move about freely. These
technologies were not designed strategically to exclude certain members of society,
but, as Winner puts it, were instead the result of ‘long-standing neglect’ (p.32) in
taking these people into account when designing infrastructures. Put differently,
design processes implicitly referred to practices that involved ‘able’ citizens with
full control over limbs and senses. Design is never just about making a technology
but always also about designing practices, whether intentionally or not. This, I argue
below, is a political endeavour.
My definition of design also implies more than what teachers and pedagogical
professionals usually have in mind when talking about planning their lessons. Even
though teachers do not generally talk about designing a lesson, central aspects of
design such as organising and planning are indeed part of what teachers do prior to
each lesson. What significantly distinguishes how teachers conceive of their job and
how interface designers do is that the former mainly understand it as forming the
practice of the class, whereas interface designers generally see their job as creating
technologies leaving alone the specific practice in question. From this perspective,
the teachers’ designing activity can be said to be at the opposite end to the interface
designers’ on a continuum between designing technology and designing practice.
In contrast, I argue that just as designers implicitly design practices as well as
technologies, so do teachers implicitly design technologies when planning the
practice of the class.

Design of the interface and practice in the design workshop

In our project, design was partly comparable to interface design, partly comparable
to design of pedagogical practice. Let me be more specific. Designing the interface
meant in this project designing the graphic appearance of the Active Worlds
browser’s 3D4 virtual environment. Active Worlds is an Internet-based application
which means that any computer user with access to the Internet can enter this virtual
environment. There are various worlds in an Active Worlds universe, each having
a different landscape: the terra-formed landscape of the ‘Babel world’ alluding to
pre-historic sites found in the United Kingdom, the dimly lit theatre auditorium of
the ‘ENZO world’, a gathering of historic villages in the ‘Journeys world’ etc. Based
on Active Worlds technology, the Eduverse5 universe is reserved for professional

4 Technically, Active Worlds is what is called 2½D. It is not based on 3D graphics,

but to the naïve eye, it looks like a spatial 3D environment with height, length and depth,
compared to flat 2D graphics like a html page.
5 Eduverse is one browser – one universe. Eduverse can be downloaded from http://
www.activeworlds.com/edu/awedu_download.htm. The program is executed from a personal
150 Doing Things with Things
educators and researchers of education. As educational researchers we had the
opportunity to achieve a virtual world free of charge which my research team6 did.
We were now ‘world owners’ and we named our virtual world Femtedim.
It is up to the owner of the virtual world to design the landscape of the world.7
The Active Worlds platform provides about 2000 pre-given building blocks which
can be used to build up the houses, forests, rooms or whatever is wanted in the world,
just as it is possible to design new objects. Designing a virtual environment is limited
to designing the world, its graphic landscapes, architecture, navigation paths, etc.8
The Active Worlds browser stays unchanged. The browser can be described as a
frame you use to browse between different virtual worlds. It is set up with menus
for defining the settings of the worlds, the perspective of the avatar9, the range of
sight etc.
Apart from designing the interface to the extent the Active Worlds platform
and the time and skills of the world owners allowed, designing in the Femtedim

computer with Internet access.

6 Comprising Michael Aagaard, Nina Armand, Agnete Husted-Andersen and Tine
Jensen.
7 The world can be designed with a flat ground surface or without such, in the latter case
the experience will be like being in cosmic space. Furthermore, a horizon with for instance
mountains, a cityscape, walls, stars, curtains, other things or just a colour can be installed,
which will constitute the ever-present background of the world.
8 The virtual world can be described as a ‘stage’ for the activities taking place in the
virtual world. Just as a theatre, a historic village etc. invite certain practices to take place it
is possible to create signs with text, pictures, animations and video clips, which enable the
designer through representations to instruct or inspire the users’ activities. This can also be
done by creating hyperlinks from objects in the virtual world to web pages on the Internet.
This has the effect of opening a browser with a web page when clicking on or bumping into
a hyperlinked object. This hyperlink functionality can also work to make an object appear or
disappear when clicking on or bumping into another object, and thereby let doors open, make
illusions of moving objects etc. just as it is possible to make hyperlinks that ‘teleport’ the
avatar to another place. Put into sequences such design solutions allow the designer to make
interactive environments that do not simply frame the practice, but that act on and react to the
users conduct in the virtual world.
9 An avatar is a small figure that the user moves around in the virtual environment using
the arrow keys or mouse. Having chosen a name and a look for her avatar, the user enters
the virtual environment, which means she sees the avatar in the centre of the screen in the
landscape of the virtual world. She can choose an isometric view, looking at her avatar from
a bird perspective behind the avatar. The most used view, however, is the point-of-view. What
the user sees on the computer screen in the point-of-view mode is the landscape of the virtual
world as viewed ‘through the eyes’ of her avatar. To understand the feeling of navigating
in the virtual environment, the closest analogy is hence our everyday bodily experience of
being in a space, able to turn around 360 degrees in the environment we are in, and move in
all directions. You can do the same in Active Worlds’ graphic 3D virtual environment. Only,
of course, your view is limited to the size and graphics quality of the program and computer
monitor, and your movements are restricted by the rigidity of the input devices.
 Politics of Things 151
project also involved designing pedagogical practice. As is already obvious from
my description of designing the interface design of practice cannot be sharply
distinguished from designing the interface. There was more to designing practice,
however, than designing the interface of our virtual world. We also created a frame
story. It was about an avatar called Avafar who had lived all his life in the 3D virtual
world Femtedim. We told the children that Avafar was getting old and with age he
was losing his sight. Mysteriously, the story went, the Femtedim virtual world was
disappearing at the pace of Avafar’s sight failing. Avafar realized that he could not
sustain the world anymore and he asked the children to take over. Melancholically
but with faith in the children’s ability to make the best out of Femtedim, he handed
over the responsibility of rebuilding and maintaining the virtual world to the children.
Avafar met the children in the virtual world, giving them assignments, helping them,
discussing problems, engaging in small-talk, etc. Avafar was ‘played’ by our research
assistant, who logged on to the Femtedim world every week at the same time as the
children, but from a remote location.
Having introduced the design workshop, and having examined the concept of
design and the two aspects of designing – interface and practice – I turn to analyse
some specific details of the interplay of design and practice within the Femtedim
project. First, however, we need a discussion of the metaphors that will be in play in
the analyses of the empirical material.

Analytic concepts

The central analytic concept in use in this chapter is that of ordering. Related to
ordering is the issue of the politics of things.

Ordering

In discussing the interplay of design and practice, I will with inspiration from Law
(1994) use the term ordering. With this verbal mode of the noun ‘order’, which
has long been of central focus in the social sciences, Law emphasises the study of
ongoing processes rather than envisioning a fixed and essential structure. Whenever
something is done, said, thought etc. elements of the world are placed in relations
to each other, in patterns, and some of these patterns are repeated more than others
forming stable orderings, while others are ephemeral and passing.
I have chosen the term ordering for my study of interplay of design and practice,
because it leaves open for empirical investigation how the relations of design and
practice may turn out. This implies not seeing humans and things or design and
practice as separate actors ‘doing things’ to each other, but staying open to less
clear-cut boundaries between these entities. Law and Mol describe rules for what
constitutes objects and relations, the boundaries or transitions between them, etc.
(Mol & Law 1994; Law 2002). Different orderings imply different rules.
152 Doing Things with Things
Even though the term ‘ordering’ hints at somebody actively making an order, it
is crucial for my analyses that this is not implied by the term. Note, how Pickering
describes agency:

One can start with the idea that the world is filled, not in the first instance, with facts and
observations, but with agency. The world, I want to say, is continually doing things, things
that bear upon us not as observation statements upon disembodied intellects but as forces
upon material beings. Think of the weather. Winds, storms, droughts, floods, heat and
cold – all of these engage with our bodies as well as our minds, often in life-threatening
ways. The parts of the world that I know best are ones where one could not survive for
any length of time without responding in a very direct way to such material agency – even
in an English summer (never mind a Midwestern winter), one would die quite quickly
of exposure to the elements in the absence of clothing, buildings, heating, and whatever.
(Pickering, 1995, p. 6, emphasis in original.)

In my understanding of the term ‘ordering’ the study of the relationships of material

forces of winds, storms, clothing, buildings and heating etc. is a study of orderings
of weather. No matter how an ordering comes to be the way it is – as a result of
strategic planning, of contingencies, of ‘natural’ occurrences or otherwise – it is an
ordering.
In the following analyses I describe different orderings to characterise the
different patterns of relations that emerge with the design solutions. I use metaphors
suggested by ANT scholars.

Politics of things

I am not simply interested in mapping orderings describing different patterns of

relations between design and practice, but also in studying how a design may take
part10 in and form practices. In terms of the politics of artefacts, Winner (1999)
discusses how the design of machines, instruments and structures set a certain order,
which excluded disabled persons. Apart from how technical arrangements result in
a form of social order, Winner describes a second way in which technologies may
be political. This is not in terms of consequences of a technology, but in terms of a
specific social order as prerequisites for the working of a technology. Because of the
lethal properties of the atom bomb, Winner maintains, this technology demands a
centralised agency controlling it, and ‘a rigid hierarchical chain of command closed
to all influences that might make its workings unpredictable. The internal social
system of the bomb must be authoritarian’ (op. cit. p. 34).

10 The formulation of ‘taking part’ is inspired by Dreier (1993; 1999), who emphasises
that practitioners participate in practices, in the sense that they always only contribute to and
have access to parts of practices, while other practitioners take up and contribute to other
parts. I however extend the term to also apply to non-human participants of a practice. This
move is inspired by studies of Science, Technology and Society (STS) in general, and by
Haraway (1991) specifically, who talks about being partial, which I take to be the adjective of
taking part or participating.
 Politics of Things 153
While Winner’s discussion of whether artefacts have politics focuses on larger
scale technological systems and their entanglements, Mehan (1993) discusses the
‘politics of representations’ on a more situational basis in terms of ‘the competition
over the meaning of ambiguous events, people and objects’ (p. 241). Through the
analysis of a meeting of an ‘Eligibility and Placement’ Committee discussing the
possible placement of a boy into a program for the learning disabled, Mehan shows
how negotiations and decisions about which terms to identify the boy results in the
formation of a hierarchy (an ordering) in which one mode of representing the world
gains primacy over others and eventually has decisive consequences for the boy’s
life.
Mehan does not talk about politics of technologies or things but of how
representations may have political effects in terms of concepts assigned to events,
people and objects. However, I will borrow Mehan’s understanding of politics as
the interplay taking place in micro-settings and merge it with Winner’s ideas of
politics of artefacts suggesting that it is not simply about humans assigning their own
agendas to technologies, but that these technologies in their entanglements11 indeed
have their own politics.
In the following empirical analyses I describe orderings of design as politics
put forward in practice. First, I will describe orderings inscribed in the Femtedim
design involving both the design of the interface and of practices as they were at the
beginning of the project, and secondly, I look at how the politics of design worked in
two practices in which they were put forward.

The Femtedim design

As I have already mentioned, the aim of the project was to create an open-ended
design for the informal after-school educational setting of 5th Dimension. This already
suggests a specific politics of the design, namely that it was supposed to contribute
to establishing an ordering which allowed children to – or even constituted children
as – acting in a variety of different ways.

The fluidity of the design

The interface design of the Femtedim virtual world was minimal. Corresponding to
the frame story, the virtual world should be quite empty due to its current process
of vanishing. We built five platforms in Femtedim: An entrance platform, where the
characters involved, the avatars, would arrive when entering the world, and four
other platforms for the different groups of children to build villages on. Apart from
the platforms, the 3D virtual world Femtedim was graphically empty.
Together with the frame story this graphic design provided a minimum of resources
for the practice. Given this relatively blank visual and narrative background, the task

11 For a discussion of the term entanglement, see the introductory chapter of Callon
(1998).
154 Doing Things with Things
set for the children was to build up the Femtedim world. A very open space for acting
was thus designed for the children and hence it was they themselves – rather than
the design – who were expected to provide the resources in terms of ideas, skills,
creativity etc. to build up the villages of Femtedim. Hetherington and Lee (2000)
have drawn attention to what they call the ‘blank space’. They give the example of
the Solitaire board game. This game consists of a board with 33 holes. 33 sticks are
ordered in the holes in one sets of three rows of seven sticks crossing another set
of three rows of seven sticks in an arrangement that makes the nine central sticks
be part of two crossing lines. Before starting the game the central stick has to be
removed leaving an empty hole in the board. The game proceeds by moving one
stick over another and into the central empty hole. The stick that has been jumped
over can be removed leaving another empty hole. The next move allows the player
to move another stick into any of the two empty holes. While social scientists usually
focus on the actors and how ‘filled’ spaces (like the sticks) move, they systematically
neglect blank spaces, Hetherington and Lee draw attention to the agency of the
empty holes in the Solitaire game, to their decisive contribution to the proceeding
of the game. This is an interesting observation, which in describing Femtedim lets
us focus on the graphically and narratively empty space. Indeed, the children, the
researchers and Avafar also contributed to setting the Femtedim activity in motion,
but this should not draw our attention away from contribution of the blankness to the
forming the design.
There are however different forms of blankness. The exercise books, which I saw
repeatedly during my classroom studies (Sørensen 2006a) often had the structure
of blank lines for children to fill in with numbers as elements of pre-written sums,
with verb endings in pre-written sentences etc. These blank lines were however only
apparently blank, because the nature of their supposed ‘filling’ was already defined,
and hence rather absent than blank. The blankness of the Femtedim design did not
refer to absent solutions, but to non-existing elements to be produced.
Akrich (1992) suggests the notion of inscription to designate the hypotheses
implicit in a design about the environment into which the design is supposed to be
inserted. Adapted to the vocabulary of this chapter, I will talk about inscriptions
in terms of the orderings the design presupposes and puts forward in the practice
it is going to be part of and take part in. The ordering of the Femtedim design as
described in this section was that of a relatively empty space, which children were
supposed to fill in. Filling in the blankness of the graphics and the frame story meant
building up the villages using the same building objects and functions as available
to the world owners12 described above. Compared to the design of exercise books
which set a fixed standard for how to fill in the lines, the norms for how to fill in the
blankness of the Femtedim design were fluid.
Among their topological metaphors, Law and Mol define fluidity as an ordering of
invariant gradual transformation constituted of optional and exchangeable entities
(Mol & Law 1994; Law & Mol 2001; Laet & Mol 2000). This description fits well

12 World owners distribute and restrict rights to built in the virtual environment.
 Politics of Things 155
with the ordering inscribed in the Femtedim design requiring of the children to fill
out the blanks of the Femtedim design in non-predefined ways, involving varying
entities in terms of building objects, technological solutions, ways of collaborating
etc.
Here we see part of the politics of the Femtedim design. It presents an ordering of
the way the interplay of the children and the design (among other things) was going
to take place. Not in a fixed, standardised way, but varying and transforming. The
blankness provided a demand and a space to be filled in somehow, and the children
should provide ideas, initiative, creativity and skills, together establishing a fluid
ordering transforming with exchangeable entities. This corresponds to the second of
the different aspects of the politics of artefacts identified by Winner, in which design
suggests a certain social order as a prerequisite for the working of the design. If the
children for instance were not to comply with the hypothesis inscribed in the design,
but instead met the design with passive blankness looking dully into the computer
screen, the design simply would not work.

The regionality of the Femtedim design

As I have explained, the graphic layout defined an entrance platform and four
villages, where the children were supposed to build up the Femtedim world. Each
platform was given a different name: ‘Bahilah’, ‘Nisub’, ‘Itnom’ and ‘Akul’. It was
hence inscribed into the design that children were to be divided into four groups.
Privileges were assigned to each group. Only members of the group were allowed to
build on the platform. The inhabitants of each platform were identified by names and
privileges, and thus constituted a homogeneous population. While homogeneity ruled
within the platforms, differences lay between them. Just as much as each inhabitant
was similar to the inhabitants of his own platform, they were different from those of
the other platforms in terms of identity defined by name and privileges.
Law and Mol (Law & Mol 2001; Mol & Law 1994; Law 2002) characterise this
kind of arrangement as regional. It describes a container in which objects are located.
Regions are homogeneous fields divided by boundaries within each of which one
norm and one ontology rules. The extension of a region is limited by its boundaries.
This politics of design is clearly different from the fluid described above. It sets
a standard and fixed structure that is quite different from a transforming ordering
with exchangeable entities described above. The politics of the graphic layout of the
design was that of fixing children and platforms in four separate regions.
The description of Femtedim as regional is clearly different from the description of
Femtedim as fluid. However, it is impossible to decide which of the two descriptions
is more adequate. Just as much as the design is fluid is it regional: Femtedim put
forward a fluid politics inviting children to be active and fill out the blanks with the
effect that elements change along the way, while Femtedim at the same time put
forward a regional politics that set a homogeneous stage with boundaries and a clear-
cut arrangement of separate divisions of elements. There is no time involved in the
156 Doing Things with Things
regional Femtedim – it stays the same – whereas time is indeed a defining element
of the fluid Femtedim.13
Such a double (or multiple) ontology or double (or multiple) politics of things is
implicit in the logic of Law & Mol’s symmetric descriptions of humans and things.
While it is widely accepted that people may exercise different politics with one and
the same thing, a thing itself is generally understood as singular, as only one. Turning
to a symmetric approach that describes humans and things in the same terms makes
it just as possible that things have a multiple politics as it is possible people do (Mol
2000; Law 2002). So yes, the politics of Femtedim was fluid and it was regional.

The Femtedim Design in Practice

This section discusses two instances of how the design worked in practice in the
design workshop, focusing on the ways in which the politics of design put forward
by the blankness and the graphic layout of the design were continued in the interplay
of design and practice.

Continuance and cohabitation of fluid and regional politics of design

The field note summary below starts at the first day of the 5th Dimension project.
Avafar had given the children assignments to go to other virtual worlds and describe
for him what they saw:

After the children had completed their assignments Avafar gave each child a citizenship
of the world, and each group a platform, where they were supposed to build their villages.
The four groups discussed separately how they wanted the layout of their villages. The
Akul group wished to build a main road and construct the village around it. They discussed
vividly whether the road should be of tarmac, stone or other, how the houses should adjoin
etc. In the following months, they built up the platform village with houses next to each
other on both sides of the road. A friend of the two girls in the group joined Akul after
a few months, and assisted by ongoing negotiations a small neighbourhood was erected
following a uniform style of glass walls and single colour furniture.

The Akul group was originally constituted by the graphics of the platform, assignment
of privileges, and its separation from the other groups. Subsequently discussions of
the layout of the village, the establishing of a plan and the building of a uniform
style neighbourhood worked to consolidate their fixed and homogeneous identity
as a region. Apart from continuing the regional ordering, however, the interplay of
design and practice also contributed to continuing the fluid ordering put forward by
the design, filling out the blankness of the Femtedim design, and indeed doing this in
a non-prescribed way, involving new objects as well as children’s ideas, initiatives,
and skills.

13 For further discussion on the interrelation of time and materiality, see Sørensen
(2006b).
 Politics of Things 157
Collaborating closely and sharing the codes of the building objects each of the
Akul girls built up her own house with similar objects differing only in colours and
combination. The procedure of building in Active Worlds involves first finding the
object you would like to build in a printed list or in another world, secondly looking
up the file name of this object, which I call the ‘code’ using the same term as the
children, and thirdly writing the code into the ‘properties’ dialog box of an object in
your virtual world. Subsequently, the colour of the object can be defined. Objects in
Active Worlds are strictly related to individual users, which enables certain ways of
collaborating. One way of collaborating is, as these girls did, to build with the same
objects – even if they were all of different colours – since this allowed children to
share part of the building procedure, namely that of finding the objects with which
to build. While collaborating the two girls also worked in parallel, one for instance
making herself busy in adding animated pictures and pictures taken of her own
and the friend’s avatar in the virtual environment, while the other was busy finding
furniture as well as pictures from the Internet to decorate the walls of the houses.
Soon after a new piece of furniture, animation or other was put into the house of
one girl, it was copied to the other girl’s house, which resulted in a long chain of
interplay of new building objects, possibilities set by the structure of the Active
Worlds building function, and the girls’ managing and ideas about those possibilities
and objects. Step by step, each newly developing building solution was adapted to
the existing ones or the existing were transformed to make them fit to the new.
What was ‘new’ was sometimes an object that had to fit in terms of size, style
and function. But it could also be an idea, like that of having a warehouse on the
Akul platform, which quickly changed the existing logic, style and functions and
changed the set up to fit in with the new idea. The same was the case for the third
girl who entered the little team after a few months. She was quickly acquainted with
the procedure and fit into the already existing buildings and procedures just as these
were adjusted to make her fit into the platform. Soon a third house in similar style but
of another colour was erected at the Akul platform and the procedure went on.
In this way, the politics of the design putting forward a fluid ordering of
transforming with exchangeable entities through the blankness of the design continued
in the interplay of design and practice. It continued not only as an interplay of the
girls and blankness but also by involving other materialities such as the structure
of the building function, animations and other objects, procedures, ideas etc. As an
ongoing interplay of girls, design and building solutions etc. the fluid ordering was
continued as a co-construction of design and practice.
This fluid ordering, however, in no way compromised the regional ordering put
forward by the graphic layout. The regionality started the fluid process linking the
core elements involved in the constitution of the fluid ordering, girls and building
procedures. And apart from staying within the fixed region of the platform the fluid
ordering also contributed to sustaining the homogeneity of the platform through the
creation of houses of similar style with a homogeneous and stable identity. On the one
hand the two orderings were each the result of the interplay of design and practice:
the politics of design were put forward and continued through the interplay of design
158 Doing Things with Things
and practice as described, which in this case resulted in unchanged continuance of
the politics of design. And on the other, the two orderings took part in a mutual
interplay sustaining each other’s maintenance, constituting harmonious cohabitation
of fluid and regional orderings.

Discontinuance and conflict of fluid and regional politics of design

The coherent story of the Akuls suggests that the fluid ordering was limited by the
regional ordering, keeping it from flooding beyond the boundaries of the platform,
while this did not compromise the fluid functioning within the region. However,
looking at what happened at the Itnom platform shows us a quite different picture
of interplay of design and practice. Summarizing the fieldnotes about the Itnom
platform creates the following description:

One of the boys at the Itnom platform stayed in Femtedim while the other two scanned
other virtual worlds in Eduverse to find nice building objects. When one of the ‘scouts’
found a new object they shouted the code across the computer lab to the boy who stayed
back in Femtedim and who would subsequently build the object while the others went
on searching for more building object codes. The boys quickly, eagerly and continuously
built new things on their platform. But after only about a month it happened more and
more often that they came to the design workshop and started a game of Warcraft or
Quake on the computers instead of logging into Femtedim to continue their project. When
asked about this change, one of the boys explained: ‘we have all the coolest stuff, so what
more is there to do’?

At the Itnom platform the interplay of the blankness of the design and the practice
of the Itnom group and platform did indeed continue into an activity of ‘filling in’ in
a non-predefined way. The division of labour of the Itnom boys divided the steps of
the building procedure given by the structure of the Active Worlds building function
among the involved actors as was done at Akul. Two of the Itnoms took care of the
time consuming task of finding objects, and the last built the found objects at the
Itnom platform. A crucial difference between the Itnom and the Akul arrangements
was that the Akul resulted in individual houses within the platform, while the product
resulting from the interplay of design and practice in the case of the Itnoms was one
undivided layout of the platform.
This was clearly due to the division of labour, or rather to, what Law (1994)
calls ‘labour of division’. Compared to the Akul arrangement, the arrangement at
the Itnom platform made clear divisions between the roles of each boy, making each
entirely responsible and in charge of any decisions to be taken within his sub task,
while there was no clear division between roles or tasks at the Akul platform. The
girls performed the same roles, and shared the outcome. In the latter arrangement
the girls worked in parallel on two products which required interplay and adjustment
involving transformation with exchangeable entities and exchange between the two
parallel ‘lines of production’ keeping up a uniform identity of the platform. Due to
the sequential process of production no mutual interplay was necessary at the Itnom
 Politics of Things 159
platform, but only a one-way delivery. Neither adjustment nor transformation with
exchangeable entities relevant as long as each boy filled out his role in the sequence
satisfactorily.
Thus, an important element of the fluid ordering was missing in the interplay of
the design and practice in the case of the Itnom platform, namely that of ongoing
transformation with exchangeable entities. The ordering of the interplay of design
and practice at the Itnom platform is more reminiscent of what Law and Mol (Mol &
Law 1994; Law 2002) drawing on Latour (Latour 1990; 1987) describe as a network
ordering,14 characterised by drawing and keeping things together and functioning
as a running machinery, in which every part – technology as well as humans – play
their specific role.
Out of the politics of design putting forward a fluid ordering came a network
ordering. From transforming with exchangeable entities through blankness the design
turned into an ordering of keeping things together, dynamic but unchanging. This
transformation happened through the interplay of design, the platform, new building
objects, the structure of the Active Worlds building function, and the allocation of
each boy into one step in this sequential production process.
What about the politics of design that put forward a regional ordering through
the graphics of the Femtedim design? This ordering was indeed continued in the
interplay of design and practice: through the determined work directed towards
the end of achieving the ‘coolest stuff’ the Itnom platform was consolidated as a
region with its own identity. However, compared to the regional ordering of the Akul
platform, which contained the ongoing development of the homogeneous identity of
the platform within its boundaries, the regional ordering of the Itnom platform came
in another version, very much dependent on relations across the boundaries of the
platform, between the Itnom platform and the other platforms. I base this conclusion
on the boy’s comments that they had the ‘coolest stuff’, which set a comparison
between the Itnom platform and other platforms, thereby creating a ‘them-and-us’
relationship, where the identity of ‘us’ is defined in relations to ‘them’.15 This ‘them-
and-us’ relation involved the ‘others’ as specific dynamic partners in a competitive
game. This regional ordering did not have the self-containing character of the Akul
platform independent of what went on outside of the platform, but was on the
contrary dependent on an interplay and contrast to the other platforms.
No other platforms, however, entered into a competition of getting the ‘coolest
stuff’, but preserved the more self-containing version of the regional ordering,
consolidating a stable identity. The continuance of a regional ordering at the Itnom

14 The network metaphor of actor-network theory is often criticised for neglecting the
failing, temporary, and unstable (e.g. Lee and Brown 1994; Star 1991). Law and Mol (Law
& Mol 2001; Mol & Law 1994) argue for the need for differentiated topological metaphors
in order to be sensitive to the variety of orderings. In this step, they have narrowed the
traditional and broader notion of network, generating more sensitivity to other orderings, and
to distinguishing between orderings. The description of network in this chapter follows Law
and Mol’s notion, not Latour and Callon’s.
15 Similar to the in-group out-group dynamics described by Sherif et al (1961).
160 Doing Things with Things
platform with a stable and homogeneous identity was hence established once and for
all at the moment the Itnom had achieved the ‘coolest stuff’. As the boy explained,
there was ‘no more to do’ in order to establish the regional ordering. Contrary to
the Akul, where the achievement of a regional ordering was a process due to the
fluid ordering intertwining with it, the regional ordering at the Itnom platform was
constituted as a product quickly achieved once and for all. In this logic, there would,
as implicit in the boy’s explanation, only be ‘more to do’ if another platform would
challenge their established regional ordering of having the ‘coolest stuff’.
In this case, the politics of design continued in the interplay of design and practice.
The regional ordering put forward through the graphic layout was continued in the
design workshop. Not through a direct reproduction, however, but by linking the
regional politics of design to the network ordering which gave rise to the ‘coolest
stuff’ as a regional end product. The regional ordering was maintained even without
the boys’ active contribution. As long as no other platforms had ‘cooler stuff’ the
regional ordering would be upheld at the Itnom platform. Contrary to this, the Akul
girls’ active attendance was required continuously in order to maintain their regional
ordering, according to the logic of the intertwining of the fluid ordering and the self-
contained version of the regional ordering.

Conclusion

I have described two politics of the Femtedim design: the blankness putting forward
a fluid ordering, and the graphic layout putting forward a regional ordering. And I
have shown through two examples how these politics were continued, changed and
unchanged in the interplay of design and practice in a design workshop for children.
By describing the design in terms of politics of design put forward in practice, it
has become clear that the fluid and regional orderings were not simply there as
already settled, stabilised products when the blankness and graphic layout were
set. They were not effortlessly inserted into practice. The design had to be realised,
maintained, and continued in practice. The orderings inscribed in the design had to
change in practice in order to continue as ‘the same’. The continuity of the design
was dependent on involving other elements such as new objects, children’s ideas,
discussions, divisions of labour etc. Hence, rather than describing design as a fixed
structure, my analyses have revealed it as a process of politics of design put forward,
picked up and continued, changed or unchanged.
When a teacher designs a lesson, the artefacts involved are usually understood
as tools that teachers simply use without changing the design. However, by the
vocabulary put forward in this chapter, I argue that a teacher does indeed design
artefacts. He or she involves artefacts in specific practices, which as a result of the
interplay of design and practice constitute specific orderings. Thereby, it is not simply
a teacher inserting a finished design into practice. Among many possible directions
she is contributing to continuing, and thereby constituting, the design in one specific
way (with help, of course, from the children and the other participants involved).
 Politics of Things 161
The aim of creating an open-ended design for our specific research project could
have been realised in many different ways. I have spent more space on describing
design and practice than on the aims and intentions of designers in for instance
pedagogical terms. This is because the way design and practice is realised concretely
proves to be more subtle and complex than the designers’ intentions. A narrow focus
on designers’ intentions may lead us to overlook aspects of concrete materiality and
practice, and hence of politics of design that are not formulated as explicit pedagogical
intentions. Had I focused on designers’ intentions to create open-ended design,
the fluidity of the design would have been highlighted, while a standard and fixed
regional ordering like that of the graphics might have been overlooked. Referring to
the designers’ intentions, we could see the regional politics of the design as a failure,
as an inability to live up to the pedagogical ideas. I prefer, however, to see it as a
practical solution, which turned out to not even be contrary to the ideas of fluidity
implied by the intentions of creating an open-ended design. While the blankness
of the Femtedim contributed to establishing a fluid ordering, the regionality of the
graphics contributed in the Akul case to keeping the activity gathered as one. Mulcahy
(1998) distinguishes between prescriptions and framework. The latter involves multi-
strategic work allowing for both homogeneity and heterogeneity, while prescriptions
tend to silence heterogeneity. Instead of seeing these two strategies as opposites
we can understand the graphics as securing the smooth working of the Femtedim
activity, which allows the fluid activity to flow heterogeneously with the blankness
of the design. In order to sustain the fluid ordering by avoiding it flowing in all
directions and thereby undermining itself, regional standardising and fixing forces
had to be involved. We could even say that the regionality inscribed in the design
turned out to be insufficiently fixed or bounded to maintain fluidity. In the Itnom case,
it flooded the boundaries of the region involving ‘other platforms’ as a constitutive
element of its ordering, transforming the ever changing and adjusting fluid ordering
into a functioning machinery of a network ordering, keeping the elements in place.
But only until the ‘coolest stuff’ was achieved and regionality produced rendering
the boys’ involvement redundant.
Returning to my initial question about whether a fixed and standardised design
results in rigid practice and whether open design brings about flexible activities, I have
to conclude that there is no golden rule. A fixed and standardised design may result in
flexible practices like the regionality sustained fluidity, whereas an open design may
give rise to rigid practice, as when the blankness gave rise to an inflexible network
ordering just focussed on achieving the ‘coolest stuff’. By approaching design in
terms of politics of design putting forward orderings in practice, we realize that
design is an achievement of the interplay between inscriptions and practice. It also
means that when we discuss design apart from practice then we miss the procedural
aspect of design and we come to depict only a brief stage of the procedural life of
design.
162 Doing Things with Things
References

Akrich, M. (1992). The De-Scription of technical objects’, in Wiebe Bijker & John
Law (Eds.), Shaping technology/building society: Studies in sociotechnical
change (pp. 205–224). Cambridge MA: MIT Press.
Callon, M. (1998). Introduction: the embeddedness of economic markets in
economics. In Michel Callon (Ed.), The laws of the markets (pp. 1–57). Oxford:
Blackwell Publishers.
Cole, M. (1996). Cultural psychology: A once and future discipline. Cambridge MA:
Belknap Press,
Dreier, O. (1993). Psykosocial Behandling, en teori om et praksisområde [Psycho-
social treatment – a theory of a field of practice]. Copenhagen: Dansk Psykologisk
Forlag.
Dreier, O. (1999). Personal trajectories of participation across contexts of social
practice. Outlines: Critical Social Studies, 1, 5–32.
Grint, K. & Woolgar, S. (1997). The machine at work. Oxford: Blackwell.
Haraway, D. (1991). Situated knowledges: The science question in feminism and the
privilege of partial perspective. In Donna Haraway, Simians, cyborgs, and women
(pp. 183–202). New York: Routledge.
Hetherington, J. & Lee, N. (2000). Social order and the blank figure. Environment
and Planning D: Society and Space, 18(2), 169–184.
Laet, M. de & Mol, A. (2000). The Zimbabwe bush pump: Mechanics of a fluid
technology. Social Studies of Science, 30, 225-263.
Latour, B. (1987). Science in action. Cambridge, MA: Harvard University Press.
Latour, B. (1990). Drawing things together. In M. Lynch & S. Woolgar (Eds.),
Representation in scientific practice (pp. 19–68). Cambridge, MA: MIT Press,
Law, J. (1994). Organizing modernity. Oxford: Blackwell Publishers.
Law, J. (2002). Objects and spaces. Theory, Culture and Society, 19(5/6), 91–105.
Law, J. & Mol, A. (2001). Situating technoscience: An inquiry into spatiality.
Environment and Planning D: Society and Space, 19(5), 609–621.
Lee, N. (2001). Childhood and society. Buckingham, UK: Open University Press.
Lee, N. & Brown, S.D. (1994). Otherness and actor-network: The undiscovered
continent. American Behavioral Scientist, 37(6), 772–790.
Mehan, H. (1993). Beneath the skin and between the ears: A case study in the politics
of representation. In S. Chaiklin. & J. Lave (Eds.), Understanding practice:
Perspectives on activity and context (pp. 241–268). Cambridge: Cambridge
University Press.
Mol, A. & Law, J. (1994). Regions, networks and fluids: Anaemia and social
topology. Social Studies of Science, 24, 641–671.
Mulcahy, M.D. (1998). Designing the user/Using the design: The shifting relations
of a curriculum technology change. Social Studies of Science, 28, 5–37.
Pickering, A. (1995). The Mangle of practice: Time, agency, and science. Chicago:
University of Chicago Press.
 Politics of Things 163
Sherif, M., Harvey, O.J., White, B.J., Hood, W.R. & Sherif, C.W. (1961). Intergroup
conflict and cooperation: The robbers cave experiment. Norman, OK: University
of Oaklahoma Book Exchange.
Sørensen, E. (2006a). STS goes to school: Spatial imaginaries of technology,
knowledge and presence. Ph.d. Dissertation, University of Copenhagen
Sørensen, E. (2006b). The time of materiality. Forum Qualitative Social Research,
7(4).
Star, S.L. (1991). Power, technologies and the phenomenology of standards:
On being allergic to onions. In J. Law (Ed.), A sociology of monsters? Power,
technology and the modern world (pp. 27–57). Sociological Review Monograph,
no. 38. Oxford: Basil Blackwell.
Suchman, L. (1987). Plans and situated actions. Cambridge: Cambridge University
Press.
Winner, L. (1995). Do artifacts have politics? In D. MacKenzie & J. Wajcman (Eds.),
The social shaping of technology (pp. 28–40). Buckingham: Open University
Press.
This page intentionally left blank
 PART IV
Organisation and Things
This page intentionally left blank
 Chapter 9

Working with Material Things: From

Essentialism to Material–Semiotic
Analysis of Sociotechnical Practice
Finn Olesen and Randi Markussen

One can start from the idea that the world is filled, not in the first instance, with facts and
observations, but with agency. The world, I want to say, is continually doing things, things
that bear upon us not as observation statements upon disembodied intellectuals but as
forces upon material beings. – Andrew Pickering, The Mangle of Practice

In the last few decades, a fruitful body of boundary-crossing research has developed
in cracks in the walls between traditional science and technology institutions.
Representatives from various disciplines, like anthropology, history, sociology,
philosophy and psychology, feminist studies, cultural studies, information studies, etc.
now investigate science and technology as cultural expressions rather than timeless,
rational activity. There are several reasons why this turn to Science and Technology
Studies, STS, has come about, as has been narrated in various publications (e.g.
Pickering 1992; Callebaut 1993; Biagioli 1999).
In this chapter, however, we will only emphasise one aspect of the STS approach.
It has to do with the way we study things around us as part of learning about human
practice. One may study things as if they are somehow pre-given objects with
definitely known properties, or they may be studied as if they are dynamic elements
in a continuous flow of activities. We will follow the latter route and investigate the
relationship between things we (can) say about material things, and the stance from
where we are able to say such things. Essential to our argument is a field study we
conducted at a medical ward at a large Danish hospital in the fall of 2000. During a
five-month period we studied the incorporation of an electronic medication module
(EMM) into daily work at the ward. Our field study was centred on how medicine,
staff and writing were translated and transformed in order for the EMM to become
a usable part of the ward’s work routines (Markussen & Olesen 2002, 2003). In this
chapter we will emphasise only one aspect of our study: The situated practices of
dealing with medications after the introduction of an electronic medication system.
Prior to entering the details of the study, we wish to ponder on some philosophical
and theoretical issues regarding material things and their doings in order not to take
too much for granted in our understanding of medication as a situated practice. After
168 Doing Things with Things
a brief, overall introduction to our field study, we will discuss some philosophical
peculiarities tied to the concept of material things, to show that certain influential
ideas abut things depend on contingent presumptions rather than timeless facts about
things; then, based on the writings of the STS scholar, Donna Haraway, we will
challenge a number of conventional notions of materiality to establish a malleable
concept of material things; finally we will tie this understanding to some core findings
from our field study. The basic claim throughout the paper is that if the researcher
only takes a confined approach to concrete reality, the actual workings of material
things will remain hidden.

Doing Electronic Medication

In the autumn of 2000, the two authors concluded a field study at a plastic surgery
ward at a large Danish Hospital. Through a number of weekly visits we studied
the successful incorporation of an EMM into daily work routines in the ward. The
module is part of an electronic patient record, Medicare, already used by various
professionals, individually and in groups, at the ward. From the very beginning
the intended purpose of the EMM was to simplify a number of basic procedures
related to medication. A sketch of the context of medication at the ward will help to
understand the intentions.
Before October 2nd, 2000, the doctors at the ward prescribed medicine using a
dictaphone, and typists wrote the prescription into the electronic patient record, e.g.
‘Rp T Cipramil 10 mg x 2 daily’. The transcription was done on the first floor of the
tower block hospital, and the resulting prescription text was sent to the 10th and 12th
floor, to the patient unit, either on a loose sheet of paper for the physical version of
the journal, or as electronic text. In the patient unit the nurses would either read from
one of the two available computer monitors, or from the paper-based prescription
sheet, what drugs the particular patient is to have, and in which dose.
The nurses on duty would copy the prescriptions for the particular patients
assigned to them with a ball pen1 to the medication form. This was then inserted in
the cardex, which is the nurses’ version of a patient record. They also wrote down
the same details on a small paper label to be placed at the bottom of the medicine
tray, and on a control form to be attached to the wall in the medicine room. Often
they would write on yet another sheet of paper, the nurses’ report, to be placed in
the cardex after duty. All in all the nurses and other staff members had to copy the
prescriptions to different kinds of paper forms to be placed at four or five locations.

1 The ink of the ball pen was significant in itself, as the colour signals on what shift the
text was written. During the day shift the nurses would use blue or black ink, in the evening
they would use green ink and the night shift nurses would use red. This material manifestation
of daily rhythms appears to be a widespread convention, as we have learned from comments
by nurses and other health professionals in various countries.
 Working with Material Things 169
It is no wonder if they occasionally felt uncertain whether the right medicine was
administered to the right patient in the right dose in the right form at the right time.
In the early months of 2000 four nurses and a consulting surgeon, making up the
prescription group at the ward, decided to implement the EMM from October 2nd.
– At this point one should note that before this date, the nurses clearly had a right
to write down medicine as an integrated part of their work. Seen from the outside
they gave away this right as they began to incorporate the EMM in their work.
Meanwhile, the doctors, who did not do much writing apart from making notes for
their own use, took over a right they had never asked for. We will elaborate further
on this transformation in a later section. For now it is sufficient to note that in terms
of doing things with things, some basic daily work practices were about to change.
It was the doctor, who previously made the (oral) prescription, who now had
instead to write down the prescription in the EMM. At the same time the nurses were
to refrain from repeating medication and instead use the doctor’s prescription on
printed lists as the basis for their drug administering. Furthermore, the writing space
had been altered. Writing medicine was no longer about a nurse using a ball pen or
a pencil to write down the name, dose etc. of a drug on a piece of paper. Instead, it
was about a medical doctor using a computer keyboard to type into a module in the
electronic patient record what drugs she has prescribed to a particular patient.
From a modernist perspective one may be tempted to claim that the hand and
mouth of the doctor has been harmonized in an input activity, while the nurses and
other professional groups have been aligned to mediate the output to the patient.
The new situation would then appear to underpin the modern, dominant idea of
technologically driven progress and autonomous, rational subjects. In our opinion,
however, writing down and printing out medicine is far more complicated than can
be expressed through this unilinear perspective on change. Writing and printing
involves a number of complex coordinations of existing competencies and agencies,
which defy the outcome of any such perspective (Markussen & Olesen 2003).
To support our argument it is necessary to investigate, and possibly subvert,
certain images of social interaction and materiality. As we have just outlined, the
staff use pens, keyboards, forms, etc. to engage in a delicate, but necessary work
practice. From a modernist perspective these are straight-forward examples of
how to do things with things, i.e. how rational human agents master work-tools.
From our alternative perspective within STS, however, there is more to the case
than autonomous subjects operating standardised tools in intentional actions. Both
things and people are transformed during processes of change. It is precisely the
dynamic, interdependent evolving of things and humans which makes it possible to
stabilise particular relations between them over time. It raises the question, how such
restrained changes can be studied? By what means can the co-evolving entities be
dragged out from behind modern dichotomies of free human beings and determined
things?
In the above example, writing is at stake as a specific mode of health care
practice, that has to be transformed in order for medication to become safer and
more efficient. In STS terms this raises the question: What is writing if it is not just
170 Doing Things with Things
a simple tool-using human practice? Let us try to hint at an answer as a first step to
describing our approach to the study of material things and the merging of humans
and things.

Doing writing

First a strategic claim. Writing at the ward, or in similar work situations, is not
a neutral, instrumental skill. On the contrary, it is part of a constant microsocial
stabilising of numerous sociotechnical relations, which in effect will transgress
paper, databases, pens, keyboards, graphic interfaces and words. Why do we say
that?
Historically, writing has always been a powerful means of social control and
principle of organization (Ong 1982, Heim 1987, Martin 1988). That goes not least
for our own time, in which communication has become a cultural beacon carried
forward by the effects of internet use and other social interactions involving micro-
processors. Also historically speaking, one can point out strong idealist traditions
in Western cultures to separate language from its material contexts and machines,
whether those are slate pencils, a printing press, or some other tools for lettering.
Materiality represents simple work, while thoughts and language represent our
privileged human form as animal rationale (Masten et al., 1997).
It follows that we have inherited a somewhat contradictory relationship with
writing. At a modern hi-tech hospital it is evident that writing (as well as reading) is
an extremely important mode of ensuring stability in daily work routines. Here one
also finds the ambiguous relationship to writing where, on the one hand, materialized
textuality and documentation, e.g. test results, measurement values, codes, are sine
qua non, but, on the other hand, many agents aim at a paper-free reorganization of
work processes supported by the computer. If decision makers are not aware of the
materiality of the textual production, or the textuality of the material tools, they will
often be tempted to aim at technological simplifications of delicate work processes,
without considering the situated microprocesses that are involved. Consequently,
intended improvements may be distorted by unintended changes.
The above reflections embed some fundamental presumptions about things and
materiality which to a high degree determine the convictions and choices made,
or not made, regarding our dealings with things. Some of them are philosophical
questions, but not in a highbrow sense of intricate, scholarly debate. Rather, they
involve a more mundane kind of philosophy. They are basic questions we could
ask, but seldom do, with regard to our everyday practices. Most of the time we do
not need to reflect on our work, we just do it!2 But sometimes the situation involves
a breakdown, or an alteration, and we may then begin to reflect on what is going
on. On a few occasions the reflections will raise doubt about foundational elements

2 See for instance Garfinkel (1967) for a classic portrayal of ethnomethodology as one
important line of approach to studying non-reflexive, everyday routines.
 Working with Material Things 171
of the everyday situation. And this, indeed, is a risky kind of endeavour, as it may
literally undermine one’s everyday conceptions of ... things.
Nonetheless, we claim that such philosophical assumptions make up a vigorous
part of everyday life, and we might as well pay attention to this domain. In the
following sections we will do exactly that. We begin the investigation into mundane
philosophy of things by a conceptual analysis of this most ordinary word ‘thing’. To
be economical we will only look at how material things may be understood.

Parts and Properties

The term ‘thing’ is arguably one of the most indispensable, linguistic devices in any
language, but what does it mean, do and represent? One line of initial questioning
could be: Is there more to a thing than the sum of its parts or properties? Does it
have an essence? One reason to assume this is that we apparently use one kind of
concepts to describe the parts and properties of things, and another kind of concepts
to describe, or refer to a thing in itself. It seems that we individuate things all the
time in various ways, so what are we actually able to say about them? Two issues
come to the fore.
First, is a thing, say a car, the sum of its parts, and if so, when will it lose its
unitary status? Is the car a wheeled vehicle? Is it still a car when the wheels are
removed, and are the wheels not now new things in their own right? Hence, before
there was the original unified thing, the car, but now there are five independent
things with their own, unique properties. The problem turns into a puzzle: When do
we talk about a thing, and when do we talk about a context? Does, for example, the
teapot act as context for the teapot lid? Is the tea tray context for various comestible
things. – The puzzle has obviously to do with classifications, and one may wonder
if there is an infinite number of things in the world, so that no classification scheme
whatsoever may be able to exhaust the essence of a thing (Mary Douglas, 1994,
Hacking, 1991).
Second, if we disregard the puzzles about the possibility of an infinite number
of things in the world – and all those questions following from that, regarding
classification as such and the risk of making arbitrary separations – we may return to
the speculations about a thing as the sum of its parts. This, however, raises another
and just as important, philosophical question about the relations between a thing,
and the properties we assign to it. A car may be identical with the totality of its parts,
but that does not account for the relation between the car and its properties, such
as colour, shape, weight, and size. At the manufacturing plant the individual car is
built by combining parts in a certain, planned fashion, but it is not possible to build
a car in that material sense by combining its properties. So what is the ontological
difference between the properties and the parts of a thing? (Quinton, 1973). Is it
perhaps the context that yields properties to the thing? The situatedness?
Seen in this light there seems to be good reason to claim that ‘there is more
to a thing than its properties’, simply because properties are not immanent to a
172 Doing Things with Things
thing, rather they transcend it. The same properties can apply to indefinitely many
individual things, because to ascribe some properties to a thing is to classify it and
compare it with other things that resemble it in certain ways due to some shared
properties. While the car has the property ‘red’ in common with a lot of other things,
it has the property ‘car-shaped’ in common with a number of things, which are only
partly compatible with red things. Also, the colour and weight properties of a car
will relate it to a lot of other things, as there will be a number of additional overlaps
between other properties of these things. The actual properties of a car will limit the
number of properties it has in common with other things, because there are fewer
things in the world that have both red colour, car shape, car weight and car size in
common, than things, which have only some of these properties in common.
It follows, then, that it can only be contingently true that the sum of properties of a
thing is sufficient to individuate this thing from other things. There will be numerous
other things that also have these properties as well. It will thus be extremely difficult,
if not impossible, to produce a complete specification of the properties of a thing.
Neither is it feasible to claim that any particular list of specifications is final. So
we are back at the opening questions: How do we individuate things, so that we
may claim that this thing is what it is because it has these properties, and that these
properties are uniquely belonging to it? It is in this entanglement that the problem
of individuation lies: The total amount of properties ascribed to a particular thing
does not guarantee that this thing necessarily has been individualised. If the total
amount of properties belonging to a thing is indefinite, it follows that the presumed
qualitative identity between two things is inductive, and – as we have learned from
David Hume long ago – empirically accumulated knowledge about the world may be
refuted in the light of tomorrow’s findings. In short, the unique, individual core of a
thing must be something else than its properties. But what could it be, then?
We have so far been speaking about parts and properties of things, but what about
functions? Not least, material things seem to be characterised by having functions
for use, so what about them? Roughly formulated, a functional definition would
assert that a thing is what it is due to the function(s) it has in a system or a structure.
A car may thus be defined as a road-based carrier of people and things, that is its
function! As such it is comparable with other means of transport, and one can start
to list its advantages and disadvantages, like it is motorised, it is fast, it is reliable,
it makes noise, it pollutes, it is expensive, etc. But functionality is both too broad
and too narrow as a means to identify things. First, functional stances tend to be
too liberal. Functionalists generally consider themselves uncommitted to ontology,
hence it is legitimate to overlook the materiality of things. Whether a carburettor is
made from e.g. clay, wood or steel it may serve exactly the same function in a petrol-
driven engine – but intuitively these differences are significant to our understanding
of the thing. Second, functional analysis seems to presume a privileged design
purpose built into things, that warrants a particular kind of result from using them.
Functionalism, thus, advocates a teleology in things, but that reminds us of the point
made above, that it is difficult, if not impossible, to establish what are the immanent
properties of a thing. The proper purpose of a thing does not appear to be any more
 Working with Material Things 173
immanent than properties. Hence, the assumed functional economy of a thing is only
contingently true (and not guaranteed by its designers, as will be evident below). If
nothing else, this suggests that functions somehow resemble the general category of
properties, at least with regard to material things.
It was hinted earlier on that the researcher can benefit from studying the context
of things. ‘Context’ is a somewhat over-worked term which in recent years has been
applied by numerous groups of researchers, not least to characterise the joint social
fabric. It does seem safe, though, to make a minimalistic claim that certain things
occasionally serve as context for other things. But talk about a context implies talk
about people, too, in the sense that ‘context’ suggests linguistic relationships between
various entities. ‘Context’ literally means ‘with text’ or ‘text-based relationship’.
Relationships are established by people as an ongoing practice of linking things
and people in this or that fashion to bring order into their life. This points to another
feature mentioned above, that classifications do much work to constitute things and
to place them in relations to each other and to people.3 Not least is classification an
ever present imperative at a hospital, and it may seem obvious that an EMM, or in
broader terms, an electronic patient record, EPR, constitutes an especially efficient
classification machine that can ensure order in many corners of the health systems.
But is the EPR a thing after all, and if so, what can we say about it?

Is the electronic patient record a thing?

Let us begin with the puzzle about a thing as more than its parts. In recent years,
there has been a struggle in Denmark between various players for the right to define
the EPR. The National Board of Health, politicians, computer companies, several
counties, singular hospitals and wards, appear to have overlapping, but different
ideas about what is required for something to be an EPR.4 How is that possible? One
reason is that, unlike cars, most candidates to be an EPR do not (yet) exist in a stable,
globally recognised clinical setting. That is, there may be national visions about
EPRs, locally applied EPR-like systems, Beta-test versions of EPRs, etc. But there is
no single classification scheme available to unify these candidates, to exhaust what
is an EPR, and, as we saw regarding things as the sum of their parts, such a scheme
might not be possible. It is not unreasonable to suggest that the struggle for definition
rights is often politically motivated – ‘political’ being understood in a broad sense.
Winning the right to define what an EPR is, is also winning the right to classify what
is the right context for its use, and what parts are necessary for its proper working.
So far, the struggle goes on.
What about EPR candidates and their properties, then? It looks like it can only
be contingently true that the sum of properties of a thing is sufficient to individuate
it from other things. Other things, too, have these properties. It will thus be next to

3 See Star & Bowker 1999 for more on the doings of classification systems. See also
Olesen 1995 regarding the foundation of modern, technoscientific order.
4 See EPJ-observatoriets reports, e.g. 2002, 2003.
174 Doing Things with Things
impossible to produce a complete specification of the properties of a thing. Neither
is it meaningful to deem any particular list of specifications final. To speak about
the ‘right’ or the ‘complete’ EPR is to individuate it based on a particular list of
specifications of properties. Whether we talk about a unique EPR, or a class of
identical EPRs, the problem with the contingencies of property lists will arise. If
we accept this analysis, the struggle to define the Danish EPR appears to be an
ascriptionist war, in which properties like ‘effective’, ‘rational’, ‘economical’,
‘safer’, ‘easier’ are ascribed or not ascribed to various EPR candidates. No system
has these properties in the ontological sense mentioned before. It appears that e.g.
‘efficiency’ is not an immanent part of any EPR, and cannot belong to it in the same
sense that a printer port or a database belongs. Again, political and other motivations
to tie certian properties to specific EPRs must be considered alongside concerns for
design and contexts of application.
So the EPR is difficult to make out as a thing on its own. Rather we must look at
certain properties tied to the thing in a context. This is where writing suggests itself
as a promising point of departure. Proponents of EPRs and EMMs assure us that
work will be more orderly for the health professionals, and treatments safer for the
patients, if these computerised things are applied. As we have already pointed out,
writing is basic to tightly organised ordering practices in a hospital. In accordance
with our approach it is not just a physical representation of mental operations to
write things down. Instead, it is a situated practice that includes enfolded textual
and material variations and shifts. Hence, we need to develop different conceptions
of writing to understand it as a practice. To follow up on this lead, the next sections
will investigate some consequential ideas about the relationship between people and
things. The focus will be on the material fabric of the world and things. On the face
of it, it would seem that the material world is the most stable point of departure if one
wishes to learn about things and their contexts. As will be evident, however, we may
well have to develop new ideas about the materiality of things in order to talk about
their practical, but slippery nature. In this paper we choose to accompany the much
debated science student and feminist scholar Donna Haraway. The main reason is
that she has introduced an exciting new vocabulary to talk about people and things
which avoids unproductive separations and dichotomies. Hence, Haraway invites us
to take a stance, from where we can say something different about things and our
relations with them.

Situating the Material World

In the beginning of her book, Modest_Witness@Second_Millennium. FemaleMan©_

Meets_OncoMouse™ – Feminism and Technoscience, Donna Haraway claims the
following: ‘Challenging the material-semiotic practices of technoscience is in the
interest of a deeper, broader, and more open scientific literacy, which this book
will call situated knowledge’ (Haraway, 1997: 11). Below we wish to pursue her
challenge to traditional conceptions of science and technology in order to get a fresh
 Working with Material Things 175
perspective on how to investigate the doings of people and things. It is fruitful to
zoom in on her important concept of a material-semiotic systems (or fields, worlds,
practices, bodies, knots, etc.). As will be evident in later sections on our ethnographic
findings, this provocative coalition of terms is very productive, if one wishes to
confront hidden philosophical assumptions in everyday practices. ‘To write’ a
prescription, say, is not just about mindful agents putting premeditated words to the
paper in deliberate actions, but about performing and being performed in a situated,
interdependent network of entities.
To do the elaboration, we move somewhat deeper into the details of some
influential ideas about the concepts involved, i.e. ‘material’ and ‘semiotic’. That will
make it possible to relate to the kinds of understanding of being and knowledge
production Haraway wishes to transcend. It will be evident that her understanding of
material-semiotic things entails a challenge to established philosophical ideas which
have been built into well-working, modern scientific narratives about ‘objective’
reality.

Taking a materialist stance

Let us continue by trying to answer a straight-forward, solid question: What is

materiality and materialism? First materiality. – The concept is rooted in the latin
materia, meaning ‘matter or substance’. The latin word is derived from the Greek
hyle, meaning ‘wood’, but not wood in general, though, only that kind of wood used
by the carpenter to build things. It follows, that matter is related to form from the
very beginning; it is the carved wood that is implied in the concept of hyle. Matter
and form are tied together in our conceptions and our experience (Flussel, 2000).
If one conceives of matter as the stuff things are made from, it follows that things
are assigned certain properties, e.g. spatial and temporal extension, mass, mobility,
divisibility, measurability. Often these are labelled ‘quantitative properties’, because
they have to do with size and weight. Aristotle, who lived some 2300 years ago, saw
raw matter as form-less or undetermined, while the form of a thing is that which
actualises its potentiality and make it real. To him, properties were to be found in
the individual thing. If we take a leap in time to the natural philosophers of the
17th century, we can see how matter begins to be conferred many of the basic
meanings familiar to us moderns, not least through the development of experimental
science. Most claims produced by natural science prove immensely powerful, which
has to do with the unique possibilities of scientific institutions to make controlled
measurements of material processes. Their results can be circulated and compared
with data from other scientific institutions to develop stable determinations of natural
behaviour, i.e. natural laws.5 As will be evident below, later conceptions of reality

5 See Cantor (1989) for some interesting reflections on the growing needs to circulate
objective knowledge following the development of new scientific institutions in the 17th and
18th century.
176 Doing Things with Things
have criticized this understanding for being fallacious, not providing a fair picture of
the complexity of the world.
What about materialism, then? If by ‘materialism’ one points to a stance from
where the world is seen as the sum of physical matter, we are immediately led
back to an understanding developed by 17th century empiricism and taken over by
enlightenment philosophers. The basics of this understanding had been expressed in
the theory of atoms by the Greek philosopher, Democritus, some 2400 years ago:
Materia is what is given in knowledge about nature; there is no basic power or spirit
intervening in the processes of nature. What exists is the extended, material reality,
and it can be measured. Knowledge about the natural world is limited to what can be
experienced through the senses – without interference from speculative phenomena
and concepts. But even if there is no intention in the spiritless, material nature, things
do not happen by accident. On the contrary, everything follows the clear laws of
nature.
Also, mental processes and states can be reduced to material reality due to the
atomic structure and material fabric of all things. Already Democritus had talked
about the soul as made out of especially smooth and round ‘soul atoms’. In the
mechanical philosophy, introduced by the British empiricists in the 17th century
and further developed in the 18th century by materialist philosophers, the interior of
the soul was precisely investigated on experimental-material grounds. In the same
period, inventors, such as Baron von Kempelen and Vaucanson, produced mechanical
dolls that were not just expressions of mechanical enthusiasm and skills. These
automatons also symbolised a basic mechanical-material conception of reality: We
do not need the Cartesian ‘double-ontology’ that not only contains calculable matter,
but also slippery soul substance of a very speculative nature! In this firm rejection
of body-mind dualism one finds, however, an important recognition of the soul
stuff or mentality as the distinct other in contrast to the material reality. As we well
know today, this rigid contrast between mind and matter did not stay uncontested in
philosophical discourse.
In the middle of the 19th century, materialism got a different meaning, very
much in conflict with the mechanical materialism. The ‘material’ was interpreted
dialectically by Karl Marx and Friedrich Engels, who sought the developmental
principles behind the empirical, sensual world. It is no coincidence that the young
Marx wrote his academic dissertation on Democritus and Epicurus, arguably the
two foremost materialists of antiquity. As Witt-Hansen (1973) has been pointed,
Marx never offered a straightforward answer to the question: What is your kind of
a philosophical materialism? He would talk about materialism and idealism through
aphorisms or in peripheral remarks in his social analyses. He does say, in opposition
to Hegel, that the ideal is nothing but a mental transformation and translation of
the material. This act of transformation and translation is what Marx calls spiritual
production. By emphasizing human productivity Marx distances himself from
Feuerbach and other materialists. In the first thesis on Feuerbach he criticizes him for
ignoring the role played by human sensuous activity in practice, i.e. in history, politics
and economy. Instead, Feuerbach stays satisfied with a contemplative materialism
 Working with Material Things 177
that yields a passive, observing role to human beings. The distinctive mark of Marx’s
materialism is that, while he buys into the claim made by contemporary materialists
that people are products of their material conditions, he adds that because we are
productive beings we also actively change our material and spiritual environments.
His philosophical anthropology is thus much more generous to our participation in
the world.
The materialism advanced by Marx and Engels, historical materialism, would
reject mechanical materialism, or ‘vulgar materialism’, of the kind described above,
because of its basic claim that only physical bodies in motion exist as the basic
building blocks in the universe. For Marx and Engels, on the contrary, it is necessary
to investigate reality on separate ontological levels, where each level has its own
reality in accordance with dialectic laws of historic development. This assumption
makes it very sensible to talk about historical materialism. While historical
materialism does not presume reality to be just extended matter, it presupposes that
reality is developing, or built, on top of a material foundation. But in the same way
that a building is not reducable to its foundation, so is reality not just extended
matter. The world also comprises social, political, cultural, aesthetic phenomena.
The concepts of ‘basis’ and ‘superstructure’ made it possible for Marx and Engels
to point out dialectic tensions between the material and the socio-cultural levels of
any society. They thereby offered a different, much more dynamic, conception of
materiality and materialism.6
The strong political dimension, central to historical materialism, has set its mark
on a large number of subsequent positions, trying to come to grips with our practical,
shared existence, either as endorsements, or as attempted refutations of this dynamic
mode of materialism. Marx and Engels’ materialistic, political arguments and
philosophical anthropology will consequently also be found in those positions. This
observation allows us to move several steps further to Haraway’s stance on material
reality. As will be evident, she attempts to dissolve a number of strong tensions
between philosophical anthropology and contemporary, sociotechnical reality in
which not everything is what it seems to be.

Material-semiotic stuff

Donna Haraway has acknowledged that she is ‘... something of an unreconstructed

and dogged Marxist’, who remains ‘... very interested in how social relationships get
congealed into and taken for decontextualized things’ (Haraway 1997:8). Like Marx,

6 It is true that higher levels of reality depend on lower levels, i.e. the social superstructure
depends on the basic material and economical conditions. At the same time one should not
aim at reducing the moral, religious, or philosophical institutions to some basic material
structures. Furthermore, there are constant interchanges between the various levels. When
reality is thus parted into several ontological levels, each with its own kind of existence, the
physical reduction presupposed by mechanical materialism is totally misleading. Reality is
much too complex for such an understanding to be final.
178 Doing Things with Things
she rejects simple, scientific attempts to reduce social reality to ‘congealed’ objects
in a mechanical-material grid. She rejects any straightforward, Cartesian contrast
between subject and object reality. Rather, she prefers to study sociotechnical
alliances and relations, which she takes to be preconditions for scientists and
philosophers, who mistakenly substitute frozen images for social life.
Haraway willingly subscribes to Marx’s determination to demonstrate how objects
and nature are full of social process and work – ‘... even if many current science
scholars have forgotten his priority here’ (Haraway 1997: 43). At the same time she
disagrees with his basic assumption about the given, human nature and its privileged
role, or rather, she disagrees with seeing work and practice as exclusive human
characteristics. Marx has expressed such views in relation to the anthropocentric
interpretation of technical change through history. To him, the technico-scientific
development represents a progressive, dialectic process in which human beings have
emancipated themselves from nature’s constraints in attempts to realise their own
nature.
The history of science and technology has in other words served as an argument
to demonstrate the human desire for emancipation, i.e. the progressive humanisation.
This makes it sensible to claim that Marx’s thinking was based on a positive view
of technological development for the good of humanity where technoscience would
eventually contribute to the eventual emancipation of man (Siggaard & Skovsmose
1986). Similar anthropocentric conceptions are found in influential Marxist
readings from the 20th century. It is evident in Adorno and Horkheimer’s historico-
philosophical social research and in Habermas’s subsequent social philosophy,
not least expressed through the well-known term ‘technical knowledge interests’.
Other Marx-inspired traditions like activity theory and critical psychology seem to
presuppose this principal, ontological separation between human agency and non-
human things and phenomena.
This is where Haraway distances herself from the Marxist tradition, or perhaps
better, tries to carry over the tradition to let it become a post-humanist social theory
in accordance with contemporary social life:

But unlike Marx, and allied with a few prominent and deliberately crazy scholars in science
studies, with armies of very powerful and paradigmatically sane scientists and engineers,
and with a motley band of off-the-wall ecofeminists and science-fiction enthusiasts, I insist
that social relationships include nonhumans as well as humans as socially (or, what is the
same thing for this odd congeries, sociotechnically) active partners. All that is unhuman is
not un-kind, outside kinship, outside the orders of signification, excluded from trading in
signs and wonders (Haraway 1997: 8).

It follows that, while Haraway buys in on his broad conception of material reality,
she asks us to bend Marx in two significant ways. First, historical materialism fails to
consider non-humans as actors besides humans; and second, as indicated at the end
of the quote, their status as actors must include the handling of signs and other non-
 Working with Material Things 179
material work. That leads to the weaving of the material and semiotic dimensions of
Haraway’s conception of reality.7

Figuring things out

As we argued in the section on writing, Western cultures seem to act from a basic
idealist conception of words as something hovering over the paper. Just think of
Plato’s philosophy of the real world of ideas, or the magic role of the word in the
Judeo-Christian tradition. In these cultures, there is a gulf between the practices of
putting letters, words, and sentences to paper or on hard disk, and the metaphysical
understanding of words as transcending their material context. Gradually, with, for
example, the replacement of hand-written books by mass-produced, printed type
volumes in the renaissance, and with the hermeneutic-humanist interest of modernity
in trying to grasp the meanings hidden in our texts, there has come to be a culturally
accepted separation between the materiality of the written word and its meaning. A
separation between things and signs.
Words are often taken to be representations of the essence of human being, that
is, reason; and language has in paradoxical ways gained status as our connection
to eternity, not least through religious and cultural practices. Whether language
represents God or Nature, it points beyond the specific, social and material context
in which it was first expressed. That seems to be a special property of language:
the ability to free itself from its original context, from its own history. It is largely
because of this cultural imagery that we tend to dismiss the possibility of significant
bonds between words and their contexts of production. Western cultures have, more
precisely, gotten into the habit of making deep, meticulous analysis of the numerous,
symbolic representations of our unique, mental essence, and of the emancipating
and/or alienating effects of materiality on our life. Typically, we presuppose a
boundary between these two reality domains, whether it is a Cartesian sign-thing
ontology, or it is a more advanced understanding of reality, as in Marx and Engels.
In a sociotechnical culture it may not be enough, however, to see nuances through
the lenses of dialectic materialism. They have not, so to speak, been ground to show
the continuous transitions from the material to the semiotic dimensions of reality
and back again. We need to find a different stance from where to investigate the
relationship between humans and things. Not a radically different one, as should be
obvious by now, but one that is able to overcome the culturally-born dichotomies
between humans and things, and between signs and matter.
Donna Haraway, together with the French sociologist Bruno Latour (1988, 1993)
and other STS scholars, has demonstrated the merits of trying to dissolve absolute
ontological boundaries between material and symbolic dimensions of sociotechnical
practices. She has identified figures and figurations as explicit expressions of ‘...
the tropic quality of all material-semiotic processes, especially in technoscience’

7 In Haraway’s writings ‘semiotic’ should be taken in a broad sense to express the

textual and symbolic dimensions of the world as they are woven into fields of practice.
180 Doing Things with Things
(Haraway 1997: 11). By that she wishes to point out that a figuration is not just a
figurative ornamentation of literal speech. Nor are figurative utterances ‘just’ images
or symbolic expressions of literal meaning. Rather, we live with and through such
figurations. Figurations are performative images that one may dwell in. The famous
cyborg-figure is arguably a special sign of a sociotechnical culture’s transgression
of any subject-object dichotomy, of any fixed material-mental boundary-work
(Haraway 1991). But there are, Haraway insists, many other such figurations (Lykke,
Markussen & Olesen 2000).8
In Haraway’s use of images, literal and figurative modes are always intertwined
(Bartsch et al. 2001), and this entails that rhetorical practices are also in effect
political practices. This goes for technoscientific practices too. To talk about the
conquering of extended, lifeless nature in the 17th century, or historically developing,
material world divided into manifest levels of reality in the late 19th century, are two
examples of political practices. Both bring to life a number of figurations that are
embodied and thus able to perform sociotechnical work as such on top of human,
verbal practice.
This brings us back to the initial Haraway quote: ‘Challenging the material-
semiotic practices of technoscience is in the interest of a deeper, broader, and
more open scientific literacy, which this book will call situated knowledge.’ With
her concept of material-semiotic reality Haraway makes it possible to challenge
a number of ‘hard’ facts about nature’s eternal, invariant atomic structures. Her
conception also invites us to point out those rhetorical devices by which material
and semiotic reality apparently is ordered in mutually fixed, static arrangements, not
just within scientific institutions, but also, for instance, in health care contexts. Let
us just mention two such figurations.
‘Modest Witness’ is a figure that was formulated in the 17th century by the
famous British natural philosopher Robert Boyle in an environment of experimental
physics, still in its infancy. His intention was to create a neutral, direct connection
between the scientific matter of fact and Nature without intervention of human
intermediaries (Shapin & Shaffer 1985, Olesen 1995). The experimental scientist is
precisely a modest, almost invisible figure, and due to this arrangement the scientist
also becomes a political figuration. First, the modest witness is more or less equal to
a British gentleman with special moral obligations and standings, i.e. women or poor
people cannot be in a position to become modest witnesses to the basic processes of

8 Haraway assumes that all language is figurative, and even if they are not mentioned
specifically in Modest Witness, she is aligned with the two writers on metaphor, Georg Lakoff
and Mark Johnson, on this view of language. In a number of books, the latter describe how we
practically live through metaphors (Johnson & Lakoff, 1980; Lakoff,1987, Johnson, 1987).
Our social life is characterised by projections from basic, experiential knowledge into abstract
conceptions of the world. For instance, our personal, physical experiences of learning how to
stay in an upright position may be projected into abstract ideas about how difficult it is to live
a balanced life in society. A discussion may be seen as a fight, or the memory as a container,
etc.
 Working with Material Things 181
nature. Second, Boyle and other experimentalists succeeded in establishing a unique
moral order, which is still considered exemplary to scientific and non-scientific
social practice. Basically, this order demands of its members that they are morally
committed to endorse the outcome of any experiment they have witnessed – of their
own free will – in the laboratory. Third, nature and human society are arranged
in a strong dichotomy, and natural scientists are, in a sense, commissioned as
administrators to determine possible relations. Finally, the modest witness stands as
the figure who deals with context-free, non-situated knowledge – not least due to his
unique moral position. For all these reasons, the modest witness is a political figure,
who influences the way we live. Not least, Haraway wants to show how the idea
of the scientist as a modest, neutral observer of natural processes is a translation of
ideas about masculine identity, which excludes the feminine gender from culturally
significant interaction with nature. She is able to demonstrate such political issues by
dissolving rhetorically founded, and very solid walls put up by modernists between
material and semiotic reality.
If we take Haraway’s insight to our own case, it becomes possible to look at
‘the electronic patient record’ as another political figuration in a Danish context,
stretched out in material-semiotic systems of reality. As a figuration its presence
is evident in campaigns to rationalise and improve the Danish health care systems,
not least in press coverages, professional journals and political debates.9 As such
‘the electronic patient record’ (or just the EPR) is made up of fictions, facts and
situated materiality. If the EPR was just a physical machine to be used by rational
professionals in health care work, its alleged advantages would not be discussed,
negotiated and defined as much as is the case. At the moment the political debates
are still undecided about its effects and identity, but certain features are discernible:
The EPR acts as an agitator for the responsible management of patients. It warrants
better political and economic control with health care spendings. Not unlike the
modest witness, the EPR stands as a figure who deals with health care knowledge in
reliable, unbiased ways. Furthermore, improved social order is spoken through the
figuration in the sense that it gives all citizens equal access to their own health data,
which, by the way, will not be forgotten somewhere by a tired doctor. Hence, the
EPR is a dedicated champion of democratic rights and orderly behaviour. Another
figuration, the ‘electronic medication module’, or the EMM, has begun to stand up
as a close aide de camp to the EPR. Competent, safe handling of medicine will save
lives and counteract fears of human mistakes, and the EMM is there to take on that
role. At the moment the EMM seems to work very well as a political figuration to
promote the national Danish EPR precisely due to its lifesaving abilities.
Now, this reading of certain material things may seem exorbitant to some readers,
but it does make it possible to disclose some of the rhetorical devices by which
material and semiotic reality is ordered in mutually fixed, static arrangements, in
science and in health politics alike. The modernist image of intentional humans and

9 It has e.g. been named the ‘catalyst that is changing the Danish Health Care system’
(Rapport fra Elektronisk PatientJournal, Sygehus Fyn, maj 1998).
182 Doing Things with Things
their technological devices may thus be opened up for discussion. By appointing
figurations, like the modest witness or the electronic patient record, as representatives
of a category or a species, one creates a setting in which to disclose some features
that may demonstrate, in a post-Marxian fashion, the political dimensions
behind prominent human practices. Also, such a setting shows the socio-material
embeddedness of the figure and what it represents. The material aspects of reality
are never ‘just material’ as has been argued. Materiality, too, embeds fictional
constructions.
At this point we leave the discourse about the embedded, situated material-
semiotic arrangements of knowledge and things with a sharpened awareness of the
dynamic mingling of things and humans, signs and materiality. Let us return to some
of the empirical findings from the plastic surgical ward to discuss them in the light
of these ideas.

Working with Medication Lists

By focusing on the ward as a dynamic and situated, material-semiotic space of

multiple practices we attempt to demonstrate how the successful incorporation of
the EMM can be seen as an effect of numerous work processes entailing many
reconfigurations of humans, things, and meanings. We will also look at the EMM
as a material-semiotic figuration that appears to play a steady role in the successful
improvement of handling medicine, much like modernist conceptions of technology.
This focus will hopefully demonstrate the price to pay for rationalising medication
in terms of transformed material-semiotic agencies.
Let us take a closer look at the most obvious result of the new writing-reading
practice at the ward after the introduction of the EMM: The printed paper list with
the doctor’s prescriptions bearing the headline: ‘Medicine list – current’. This list
makes up a very important stage in the dynamic process from verbal prescription by
a doctor to physical absorption by a patient.
On the (modernist) surface, it is a piece of paper partly covered with black laser-
printed text. When we begin to scratch the surface, however, it becomes clear why
perceiving the list as a material-semiotic system is productive, because it reveals
the intertwined materiality and linguistics of the ‘Medicine list current’. Here
we will linger on a particular set of properties of the system, which concerns the
reconfigurations of doctors, nurses and medicine lists.10 As will be evident, the core
of the new reconfigurations is to be found in 1) the creation of a stable column in the
medicine list, and 2) the politics of daily rhythms. To justify these odd bedfellows,
it will be beneficial to look at an example of a ‘Medicine list – current’, slightly
modified to fit into this paper (Figure 9.1). It was created on the first day of the
EMM’s introduction into the ward, and printed for use on that same day. In other

10 We will touch only on a few points of the subject in this paper, but will make it the
main theme of a forthcoming book on health care informatics and STS.
 Working with Material Things 183

Figure 9.1 A reproduction of one of the very first Medicine lists of the EMM to
be used for measuring out medicine at the ward

words, the list shows a still unsettled material-semiotic system, trembling on the
borderline between the banal and the exotic.
In this reproduction of the list, originally written in Danish, we have attempted
to retain the headline, fonts, and font sizes along with the horizontal lines, in order
to give a fairly precise idea of what a list looks like when it is printed to be used by a
nurse for measuring out medicine. The reader should imagine that the list covers the
top half of a standard (A4) sheet of paper. At the bottom of the sheet is a line of text,
stating the date and time and the initials of the person who logged in to Medicare.11
At the time of production the fields in the prescription module were filled out
partly based on the doctor’s entering data, on system designers and programmers
ideas with regard to adequate set-ups for data, and on the software’s suitability for
creating such fields. At this early stage of entering data into the EMM, the doctor
primarily filled out the blanks based on her requirements and ideas of meaningful
information as a medical doctor. For instance, the doctors consider ‘1 pill at bedtime’
to be a clear, unambiguous message. It is a message that is communicated daily
between health care professionals in the verbal form preferred by most doctors as
basic mode of communication. But this conception of unambiguity does not take

11 The initials of the person who logged onto Medicare in the morning would almost
always be put at the bottom of that day’s subsequent medicine lists. The individual health
professional at the ward rarely felt any need to log off Medicare after use. The local
working collectives, especially those including nurses, had in other words no need to act out
differentiated identities in regard of password formalism.
184 Doing Things with Things
into consideration the column as a material tool for the nurses to ensure greater
precision in the task of measuring medicine at the correct time. One doctor stated
directly to us that he found the 1+1+1+1 code ‘idiotic and a nuisance’. To him there
was no connection between this notation and the daily rhythm of individual patients.
Most people take their medicine during the main meals, i.e. 3 times a day, hence, the
four times daily built into in the formalism do not fit: ‘You could say that the patient
is forced to adjust to the hospital’s rhythm, as defined by the four times’, he stated.
It is necessary to go into this subject more deeply in order to understand why the
nurses’ and Medicare’s need for a new material-semiotic configuration very much
depends on a shift in the doctors’ writing practice.
Previously, it was the nurses themselves who filled out pre-printed lists
distinguished by grids, rows and columns, and designed for the tasks involved in
measuring out medicine. But the prescription module is not (yet) designed for such
tasks!12 When the nurse, responsible for the patient in question, has printed the
medicine list, she puts it in a plastic folder at the front of the cardex made for the
individual patient. The cardex is a thin, dark blue dossier containing various forms
in a specified order, and placed along with the other patients’ cardex on a shelf in the
ward office. When the next normal time to administer medicine is approaching, i.e. 8
a.m., 12 noon, 6 or 10 p.m., the medicine list is retrieved from the cardex and taken
together with the other medicine lists into the medicine room next door. The lists are
laid beside the medicine closet, and the nurse then begins to measure medicine into
small plastic cups based on the information she extracts from reading the lists, one
by one. During our field study we gradually became aware how closely the process
of measuring out medicine was connected to the material-semiotic conditions of
writing.
The nurse will measure out medicine by following a column with her eyes and
perhaps a finger. But it is evident in Figure 9.1, that there is a total lack of vertical
lines for support in the ‘Medicine list – current’! The table does fulfil its function in
the light of the physicians’ and programmers’ requirements. On the other hand, the
health professionals, who wanted the new list to prevent misreadings, have lost the
stabilizing effect of vertical lines on the up-and-down movements of the eyes, when
compared with their pre-electronic charts and forms. The column ‘Name/Dosage’
is meant to show whether the patient should be given medicine at this time, e.g. in
the evening. Even though it says ‘1 pill at bedtime’, that is not in accordance with
the concrete needs in the medicine room, when the nurse measures out the products.
Does it mean ‘6 p.m.’, ‘in the course of the evening’, or ‘when the light goes out’?

12 It is important to stress that, to us, the effective design process is not caused by
‘designers’. Like ‘scientists’ and ‘technologists’, designers only become ‘freestanding’ agents
with causal powers as a result of delegations at the end of a number of heterogeneous social
processes. The continued design of the EMM is evolving thanks to situated, micro-social
processes of translation and transformation. See also Latour (1987) for more on this stance to
design and technology.
 Working with Material Things 185
Previously, the text in the columns of the medicine list was well assimilated into
the process of measuring out medicine, thus helping to maintain a high degree of
certitude in this procedure. But the computer-assisted ‘Medicine list – current’ is
not compatible with the former process, as we now can see. The new situation was
therefore awkward for the nurses. It may well be that they no longer need to write
down the medical products themselves, a task now left to the prescribing doctor
and an electronic database system. At the same time, however, a shift in the balance
between text and materiality had happened in this crucial phase of the medicine’s
transformation from verbal prescription to physical consumption. The very concrete
measurement phase was, so to speak, marked by a weak, vertical materiality, and
something had to be done to the unstable things at hand.
An unplanned, new kind of practice started to evolve around the writing of
medicine prescriptions. Hence, some of the nurses gradually took on the task of
pushing the doctors to write correctly and on time – if not with enthusiasm, at least
straightforwardly. A nurse expressed the sentiment:

We can’t be bothered to be nursemaids for the doctors. They must remember to write the
prescriptions themselves. Of course, they differ; some just write them, while others have
to be reminded. It’s probably a question of the generation gap, where we younger nurses
refuse to do the doctors’ remembering for them. … So it’s a conflict that, on the one hand,
we leave it to the doctors, and on the other hand, we are dependent on being able to print
out the medicine list!

In other words, a new kind of dependency to be handled in the complex, collective

arrangements at the ward had arisen. The material-semiotic and the sociotechnical
dimensions clearly intersect. At the same time the new situation illustrates the
ambiguity of the work needed to incorporate the EMM. A clearcut initial sharing
of work tasks in connection with the incorporation is not simply transmitted to the
relevant participants, as a modernist diffusion model of technological circulation
would have it (Latour 1987). Instead, the incorporation prompts a number of
pragmatic transformations of roles, needs and agencies. One senior intern did see
herself as complying to the new demands – but in accordance with her medical
logic. At one point she entered a prescription and wrote down the dosage as ‘0+0+1’.
When asked, she had no doubt that any nurse would make out from the text that the
patient is having this product in the evening, but not in the morning or at midday.
On the other hand, she did not find it necessary to write the 4th digit to tell that the
patient is not having this product at night time: ‘The nurses know that!’ she stated
confidently. This is an example of a partial connection between some of the involved
actors, where shared knowledge, daily rhythms, and ink-based columns on a sheet of
paper are colliding in the attempts to make a new formalism work.13
In the local context, the medicine list must provide some kind of columns, which
the nurses can use for measuring out medicine. This can be illustrated by another

13 See Strathern (1991) for a detailed unfolding of the concept.

186 Doing Things with Things

Figure 9.2 A medicine list in harmony with the situated needs of the nurses

medicine list, produced about a month later, and correctly completed – in the nurses’
eyes.
We have added the vertical lines to the above list to illustrate the nurses’
requirements. Note, that it is a question of training the eye and finger to check
whether this patient is to have a specific product at this time. For instance, the patient
is to have two 500-mg Panodil pills, but no Diural. If readings from the list were to
be relatively stable, the doctors had to learn to write according to the 1+1+1+1 code,
and not ‘four times daily’. It is a compromise, where the nurses will add vertical
lines – with ink, graphite or thoughts – if doctors, the computers and the printer
will only live up to a disciplined writing practice, where morning, noon, evening
and night are translated into digits. We see shifts across well-established borders
between professional groups, humans and non-humans, and between textualities and
materialities. The involved doctors, nurses, the EMM, the medicine lists, all had to
perform and be performed differently, for the new system to be ‘more reliable’.
The material-semiotic transformations developed further according to the
needs arising in the course of the Fall 2000. Even though it was banned, several
nurses began writing on the printed medicine lists, because they need to write as
part of their medicine-related practice. It appeared that, for some, the writing acted
as a marker signaling uncertainty as to the correctness of the medicine list. In the
sociotechnical arrangement of the EMM there had been no allowance made for this
alliance with paper as a mnemotechnical tool. The reconfigured writing practice was
expressed in various ways, from pencilled question marks by the name of a product,
corrections written over the dosage text, to pencilled vertical lines, or the crossing
out of discontinued medicine. Often the nurses reflected on medications in concrete
work situations. For instance, a nurse looked at an item in a medicine list that stated
that a patient was to be given morphine for severe pain. She knew from experience
that morphine often causes patients to be constipated. Consequently this patient,
too, might well need something for that condition. She therefore chose to write her
thoughts down on the medicine list to discuss the issue with others later.
 Working with Material Things 187
‘Medicine list – current’ contains a column headed ‘Change/Disc’. It is intended to
show when a product may have changed dosage or been discontinued, i.e. treatment
has been completed or interrupted for a period. Some nurses had mentioned the
discrepancy between the title’s claim to be a ‘current’ medicine list and the reference
to products not currently being given. A nurse described how she had measured out
medicine for a patient who, from his bed told her that he was no longer to have this
product according to the doctor. She had not noticed the discrepancy during the
measuring in the medicine room, where her eyes were literally fixed on the Name/
Dosage column. Again, we see that an ideal simplification demands a good deal of
work by the actors involved, in this case even a patient. The nurse further stated that
she had begun crossing the discontinued medicine out with a pencil before going to
the medicine room, so as not to measure it out by mistake. Another nurse showed us
a list she had written on because she was not sure of the dosage shown for a certain
product. Her eyes had stopped at a surprising dosage in the text, making it necessary
for her to put words on paper. That way she materialized her uncertainty, so that it
was not forgotten but could be followed up by herself or other nurses. The uncertainty
was kept alive and passed on through the paper. One begins to see the dawning need
to reclaim the materiality of the medicine list. The rule against writing is in effect
capable of running counter to the aimed-for simplicity in handling medicine, so it
must be betrayed with a layer of graphite added by hand to the machine-printed layer
of ink.
Our last example illustrates well the often sublime transition between the
material and semiotic dimensions of medicine. It proved impossible to prescribe
children’s vitamins through the EMM, because the product was not found in the
module’s database. In consultation with a doctor, a nurse therefore wrote ‘Multitabs
Junior 1+0+0+0’ with a ballpoint pen at the bottom of a ‘Medicine list – current’,
so that her child patient could be given the necessary vitamins. When the patient’s
other medicine was altered and new medicine lists printed out, the handwritten
prescription was cut off the original medicine list, put in a plastic pocket, and laid
with subsequent lists. This mix of text and paper was in no way planned, but proved
to be a necessary and practical local counter-program to the formal program of the
medication module, which did not satisfy situated needs and interests.
The incorporation of the medicine list arose from rules that were intended to
ensure a simpler, more reliable procedure, but it meant that the nurses lost a writing
tool that was significant for their collective work form. In other words, for the first
couple of months after October 2nd, 2000, the EMM was exclusively seen as a set
of rules, and not as tools for the nurses and other health professionals.14 This pair
of concepts expresses quite well what was put at risk on the ward. There was a
continual transformation of the rules laid down, so that they were not obstructing
gradually recognized needs by (transformed) nurses for tools with which to make
note of their deliberations. Halfway through the planned period of incorporating
the module, the prescription group allowed the nurses to write on the ‘Medicine list

14 This play on words occurs in Engeström (1990, pp. 171f).

188 Doing Things with Things
– current’ (but only in the right margin!). With this, both map and landscape, module
and medicine handling, were significantly altered in ways that were not planned
prior to the inauguration of the system. For a time, at least, the EMM had become a
stable actor in the hypercomplex, sociotechnical arrangements at the ward.15

Conclusion

In this contribution we have explored some relationships between the things we

(can) say about things and some theoretical stances from where we can say specific
things about things. It has also been emphasized, that the way we study things around
us is inherent to what we may learn about human beings in a situated practice. In
other words, the philosophical conceptions of the materiality of things we have
presented here, entail a view of humans and things as intertwined in heterogeneous
sociotechnical arrangements. One may study things as if they are somehow pre-
given objects with certain known qualities; but they can also be studied as if they are
dynamic elements in a continuous flow of activities.
A conceptual analysis of the term ‘thing’ showed that detached, essentialist notions
of parts and properties are rather futile. Following this, we argued that attempts to
define a national electronic patient record based on a fixed list of properties cannot
rely on singular classifications. There is no single classification scheme to exhaust
what an EPR is, and it may never come about. To speak about the ‘right’ or the
‘complete’ EPR is to individuate it based on a particular list of specifications of
properties. Whether we talk about a unique EPR, or a class of identical EPRs, the
problem with the contingencies of a property list will arise. We suggested that
political motives will be a prominent part of the arguments to favour one list over
others. Giving up the attempts at essentialist definitions of the EPR, we went on
to investigate things in meaningful contexts of material practice. Two established
notions of materiality and human practice – mechanical materialism and historical
materialism – were discussed. We ended up in support of the material-semiotic, post-
Marxian stance proposed by Donna Haraway, not least because she questions the
basic philosophical anthropologies of the former stances. Her notion of material-
semiotic figurations offers a different, politically motivated understanding of social
practice, allowing for a rather broad understanding of human-things relationships.
This allowed us to see the EPR and the EMM as figurations with moral and political
bearings, and thus opening up the human-thing dichotomy for closer inspection.

15 About a year after the introduction, we learned that the EMM did not fulfil its roles
so well anymore. The consulting surgeon had left the ward. Hence, the doctors were not
encouraged by an eager colleague to fulfil their roles in the local incorporation of the system,
and they were left with struggles between medical and collaborative logics. Several nurses
had begun to write too extensively on the ‘Medicine list – current’, ‘so we will probably have
to do something to straighten up the discipline on the use of the list’ as a member of the group
said. In any case, the situated sociotechnical practices are clearly dynamic and new stabilities
will (have to) replace older ones.
 Working with Material Things 189
In the last section we tried out this understanding on findings from our own
field study. The aim was to demonstrate that a dynamic sense of the material-textual
reality will disclose – often in surprising ways – the effectual doings of things.
By looking at the ‘medicine list – current’ as a representative of a category or a
domain, it became possible to exhibit some situated, micro-political entanglements
of sociotechnical practices, in which humans and things are intricately woven into
one another. Material things are never ‘just material’, as we have argued. Materiality
also comprises fictional constructions.
Based on our analysis and study we find it misleading to motivate the introduction
of a new sociotechnical system by pointing to its successful use in other sociotechnical
arrangements, or by claiming that the system carries certain inherent qualities
guaranteeing some specific ends through its (correct) use. Instead, we suggest that
a successful incorporation of technological devices in a work practice can be seen
as an effect of multiple undertakings, performed within a gradually expanded chain
of situated, partially connected processes.16 Both actors and processes will undergo
some alterations due to the work.
As we have tried to make evident, the field of STS, and Haraway’s work in
particular, can be used to perform playful, politically motivated excavation of the
field of dynamic tensions between what is material and figurative in things. But
the tensions will remain hidden for those who do not look for the semiotic work in
material reality, and who overlook the material agency of semiotic figurations.

References

Bartsch, I., DiPalma, C. & Sells, L. (2001). Witnessing the postmodern Jeremiad:
(Mis)understanding Donna Haraway’s method of inquiry. Configurations, 9(1),
127–164.
Biagioli, M. (Ed.) (1999). The science studies reader. New York: Routledge.
Callebaut, W. (Ed.) (1993). Taking the naturalistic turn – Or how real philosophy of
science is done. Chicago: University of Chicago Press.
Cantor, G. (1989). The rhetoric of experiment. In D. Gooding, T. Pinch, &
S. Schaffer (Eds.), The uses of experiment: Studies in the natural sciences
(pp. 159–80). Cambridge: Cambridge University Press.
Douglas, M. (1994). The genuine article. In S. H. Riggins (Ed.), The socialness of
things – Essays on the socio-semiotics of objects. Berlin: Mouton de Gruyter.
Engeström, Y. (1990). Learning, working and imagining -Twelve studies in activity
theory. Helsinki: Orienta-Konsultit Oy.
Flusser, V. (1999). The shape of things: A philosophy of design. London: Reaktion
Books.
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, New York:
Prentice-Hall.

16 See Strathern (1991) for a detailed unfolding of the concept of partial connection.
190 Doing Things with Things
Hacking, I. (1991). A tradition of natural kinds. Philosophical Studies, 61,
109–126.
Haraway, D. (1991). Simians, cyborgs, and women – The reinvention of nature. New
York: Routledge.
Haraway, D. (1997). Modest_Witness@Second_Millennium. Female Man©_Meets_
OncoMouse™: Feminism and techno science. New York: Routledge.
Heim, M. (1987). Electric language – A philosophical study of word processing.
New Haven: Yale University Press.
Johnson, M. & Lakoff, G. (1980). Metaphors we live by. Chicago: The University
of Chicago Press.
Johnson, M. (1987). The body in the mind – The bodily basis of meaning, imagination,
and reason. Chicago: The University of Chicago Press.
Latour, B. (1987). Science in action. Cambridge, MA: Harvard University Press.
Latour, B. (1988). The Pasteurization of France (Part one: ‘War and peace of
microbes’; Part two: ‘Irreductions’). Cambridge, MA: Harvard University
Press.
Latour, B. (1993). We have never been modern. New York: Harvester-Wheatsheaf.
Lakoff, G. (1987). Women, fire, and dangerous things – What categories reveal
about the mind. Chicago: The University of Chicago Press.
Lykke, N., Markussen, R. & Olesen, F. (2000). ‘There are always more things going
on than you thought!’ – Methodologies as thinking technologies. Interview with
Donna Haraway, part two’. Kvinder, Køn og Forskning, 9(4), 52–60.
Lykke, N., Markussen, R. & Olesen, F. (2003). ‘Interview with Donna Haraway.’ In
Don Ihde and Evan Selinger (Eds.), Chasing technoscience: Matrix for materiality.
Indiana: Indiana University Press.
Markussen, R. & Olesen, F. (2001). Information technology and politics of
incorporation – The electronic trading zone as coordination of beliefs and actions.
Outline, 3, October, 35–47.
Markussen, R. & Olesen, F. (2002). Kyborger på arbejde – Elektronisk
medicinordination. NIKK Magasin, 2( 2), 20–22.
Markussen, R. & Olesen, F. (2003). Reconfigured medication – Writing medicine in
a sociotechnical practice.’ Configurations, 11, 351–381.
Martin, H.-J. (1988). The history and power of writing. Chicago: The University of
Chicago Press.
Masten, J., Stallybrass, P. & Vickers, N. (Eds.) (1997). Language machines –
Technologies of literary and cultural production. London: Routledge.
Olesen, F. (1995). Kendsgerningernes verden og orden. Philosophia, 24(1–2),
43–70.
Ong, W. (1982). Orality and literacy. London: Routledge.
Pickering, A. (Ed.). (1992). Science as practice and culture. Chicago: University of
Chicago Press.
Quinton, A. (1973). The nature of things. London: Routledge and Kegan Paul.
Shapin, S. & Shaffer, S. (1985). Leviathan and the air-pump – Hobbes, Boyle, and
the experimental life. Princeton: Princeton University Press.
 Working with Material Things 191
Siggaard Jensen, H. & Skovsmose, O. (1986). Teknologikritik. Copenhagen:
Systime.
Star, S.L. & Bowker, G. (1999). Sorting things out – Classification and its
consequences. Boston, MA: MIT Press.
Strathern, M. (1991). Partial connections. Savage, Maryland: Rowman & Littlefield
Publishers.
Witt-Hansen, J. (1973). Historisk materialisme [Historical Materialism]. Copenhagen:
Berlingske Leksikonbibliotek.
This page intentionally left blank
 Chapter 10

Words and Things: Discursive and

Non-Discursive Ordering in a
Networked Organisation
Steven D. Brown and David Middleton

We wish to consider what is done with a kind of artefact that has become ubiquitous
in the practices of any organisation – archives of email messages. The ordering and
use of these archives bring into the focus the relationships between words and things,
and sociality and things within unfolding organisational practices. In particular, the
ways in which archives of messages are configured and used in organisational settings
collapses any distinction between sociality and materiality. We aim to demonstrate,
using an analytic approach derived from discursive psychology, how archives
configure interdependencies between persons and things within organisational
practices. In so doing, we challenge the conventional assumptions concerning the
role of computer-mediated communication in organisational settings where ‘the
social’ is taken as the gold standard against which communicative effectiveness is
judged. Rather, we need to understand how archival practices of computer-mediated
communications are devices that order the socialities of organisational practice.
Computer mediated communication (CMC) clearly allows organisational
members to realise collectively relevant plans and goals. It might also be taken as
making a workgroup or organisation more effective by either democratising relations
between colleagues or by empowering particular people. Alternatively, CMC by its
very nature makes users orient to emergent group norms when communicating with
fellow users. These kinds of issues are at the heart of classic studies of CMC carried
out within a social psychological framework (e.g. Hiltz & Turoff, 1978; Sproull &
Kiesler, 1998). If these diverse studies have something in common, it is that they all
take ‘the social’ as the standard against which communication is to be judged. Or to be
more precise, that specific instances of interactions between users can be understood
with reference to a version of an ideal form of sociality. A given CMC practice can
then be evaluated in terms of how far it realises this ideal form or approximates to
it in its actual application. In such a view, people are taken as offering information
to one another in face-to-face settings with all manner of additional information
gleaned from observing non-verbal stimuli and other features of the immediate
setting (Kiesler et al, 1984; Sproull & Kiesler, 1986). In other words, people are
194 Doing Things with Things
treated as isolated information processors, who exercise permanent vigilance in an
attempt to glean whatever information they can from the immediate moment.
Now that is an interesting model of the person, one which seems perfectly
reasonable for understanding certain sorts of interactions. But not perhaps a model
that can serve as a default description of everyday social life. For example, how
can we understand the emergence of shared meanings in a given interaction when
such meanings are clearly not a property of the information exchanged? At a more
fundamental level how does meaning arise at all from information processing? That
is, how does the ‘signal’ encode not only the information to be exchanged but the
grounds upon which it should be understood, contextualised and potentially enacted?
(Dreyfus & Dreyfus, 1986; Winograd & Flores, 1986). Models of the social based on
information processing are clearly problematic.
So maybe what we need is a better model of the social? Perhaps one like that
contained in Hutchins’ distributed cognition approach? This does at first appear to
move us forward. Hutchins & Klausen’s (1996) description of the communication of
representational states in a simulated airline cockpit, for example, is able to empirically
identify instances of shared tacit knowledge amongst flight personnel who have just
met. But whilst at an empirical level this is to be welcomed, another set of problems
is brought into focus. How is this tacit knowledge derived (see for example Leudar,
1991)? How can we account for its use? In order to do so, Hutchins and Klausen
claim that we must posit a shared ‘intersubjectivity’ at work. Intersubjectivity refers
to a shared understanding of the matter at hand which underpins a given interaction.
It is this prior intersubjectivity alone which, they claim, enables us to make sense of
the following:

0254 OAK24L NASA nine hundred … roger contact Oakland center one thirty
two point eight.
[F/O pulls his hand back from the altitude alert knob when ATC
says ‘contact Oakland center.’ 2.5 seconds after the end of ATC
transmission, F/O looks at Capt.]
[Capt looks at F/O.]
0300 F/O Thirty two eight.
Capt Thirty two eight?
F/O Yeah.
Capt OK
(Hutchins & Klausen, 1996, p. 16)

This extract is a small piece of interaction recorded with a commercial airline crew
training in a flight simulator. There is an awful lot of detail contained in even such a
small extract. The extract opens with the crew being given an unexpected instruction
by the local air traffic control centre. The crew are at a moment in their flight where
they expect to receive confirmation to move to a higher altitude. Control of the flight
is instead passed to a new centre (Oakland). What fascinates Hutchins & Klausen,
however, is the significance of the two looks exchanged between the First Officer (F/
O) and the Captain (Capt.) and the short verbal exchange which follows. Immediately
 Words and Things 195
after the radio message, the Captain is responsible for tuning the flight radio into the
new frequency used by the Oakland control centre. The Captain conspicuously fails
to do so, and this failure occasions the First Officer’s look. The Captain returns
a blank look, displaying to the First Officer that he is aware of the nature of the
problem which has occasioned the look. The First Officer then volunteers the radio
frequency of the Oakland control centre, to which the flight is being passed. The
Captain then seeks confirmation of the frequency from the Flight Officer before the
interaction terminates.
Now the analysis of the interaction supplied by Hutchins and Klausen is certainly
detailed, and it allows us to see that the kind of social interaction which goes on
around the complex technological routines of commercial air flights is subtle and
nuanced. However, we would want ask to what extent we must invoke a notion
of ‘intersubjectivity’ in order to understand the extract. Drawing on discursive
psychology – an approach to talk-in-interaction developed in social psychology
(see Edwards & Potter, 1992; Edwards, 1997) and the application of conversation
analysis to work place studies (e.g., Drew and Heritage, 1992) – we can in principle
analyse this kind of interaction by looking at the discursive formulation of each
turn in the sequence. The Flight Officer’s look, for example, constitutes a fairly
obvious prompt to the Captain, who replies in kind. We do not need the concept of
intersubjectivity to explain this, nor do we need any special model of the social. It
is enough to say that the kind of things the Flight Officer and Captain engage in are
fairly routine question-answer sequences, where each turn – whether verbal or non-
verbal – affords the next in a more or less direct manner.
So our first substantive point is that it is not necessary to have a model of the
social in place before analysing interaction. Indeed, if a model of the social is at
work, then we would treat this as an analytic matter. That is, we would want to
show how speakers display an orientation to a particular model of the social and
the way in which such an orientation is worked up in the interaction itself. And
this leads to the second substantive point. If models of the social are things that
participants use and thus things to be analysed and explained, rather than things we
draw upon to make sense of interaction, it follows that the role and use of CMC in
organisational settings are up for re-negotiation. All of which brings us to our third
and crucial point. If we assume that speakers in CMC interactions have a problem
with ‘things’, that managing joint attention to objects or artefacts is a problem, then
such an assumption implies that ordinarily in face-to-face interaction we do not have
this problem. Again, we would want to treat this as an empirical matter rather than
a theoretical assumption. This is precisely what we will be doing in this chapter.
The approach to things in discursive psychology and the analysis of work place
communication looks not at the artefacts or objects themselves, but rather to an
analysis of how they are formulated and described by speakers. For example, Charles
and Marjorie Goodwin’s (1996) discussion of seeing as a situated activity provides
numerous examples of how objects, (in this case aircraft) are ‘an ongoing contingent
accomplishment within a particular community of practice’ (op cit p. 87). Things,
then, can be viewed in terms of the kinds of descriptive practices which speakers
196 Doing Things with Things
bring to bear upon them. The mutual orientation of speakers to things is a discursive
accomplishment, which can be studied as such. Go back for a moment to the flight
crew in Hutchins and Klausen’s work. They clearly formulate a variety of objects in
their utterances to one another. The significance of the Oakland control centre, for
example, is worked up in the interaction between the First Office and Captain. But
Hutchins and Klausen make an interesting further point. They claim that in order
to understand the interaction, one must recognise that there is something about the
nature of the technology involved in commercial airflight which has an effect on the
interaction. They describe this as a matter of the durability of a medium for retaining
a given representational state.
Now that presents an interesting challenge to discursive psychology, and we
would put it like this: if there seems to be a kind of stability or a form of restriction
placed on a discursive interaction, how is this to be accounted for? In order to address
this issue we will introduce the idea of the ‘non-discursive’. What we mean by this
is the production of order through a means which is not primarily linguistic. In using
this term, we will develop Hutchins & Klausen’s point that the technology has a
kind of grounding effect on interaction and places limits on the kinds of descriptions
and explanations speakers can trade with one another. But, as we hope to show, this
does not mean that one has to assume a given object-world in advance of a piece of
interaction. Quite the contrary, the object world is as put together, as constructed,
and hence as amenable to empirical study. So what we are ultimately trying to do
is to show how the discursive and the non-discursive serve as two interdependent
means of producing order and regularity.

Case Study: Oilsite and CMC

We will make our argument by discussing some case study materials. These are
drawn from a project which examined groupware supported activities in two multi-
site organisations (see Brown & Lightfoot, 2002). We will be drawing upon materials
from one organisation in particular, which we have given the pseudonym ‘Oilsite’.
Oilsite designs and builds infrastructure for the oil industry. They put together
petroleum refineries across the world. Wherever there is crude oil to be pumped out
of the ground, Oilsite have an interest. The company is owned through a holding
operation based in the US and Japan. In the foyer of their main UK office, there are
three large clocks, showing the time in Dallas, Yokohama and London. Oilsite has
a great many problems trying to co-ordinate these three centres. It sends personnel
round the globe from one site to another. Co-ordination is important not only because
of geographical distance. Schedule is the sovereign measure at Oilsite. They order
activities in time. Plants must be up and running by exact deadlines. Failure to follow
and meet schedules involves a complex and costly series of financial penalties.
Schedule is then back-flushed through all of Oilsite’s activities, as are penalties.
Suppliers who fail to deliver to Oilsite on time, or who supply materials which fail
 Words and Things 197
to meet quality requirements, are contractually obliged to cede to a shifting scale of
‘backcharges’.
Here’s the first problem then: how to co-ordinate a mass of personnel, materials
and activities which spreads around the globe? Schedule is one answer. However
schedules are more than the adherence to a particular set of deadlines. Deadlines
set up sequences of complex practices across time and space (Abbott, 2001; Dreier,
1999). They make everyone, including suppliers, irrespective of where they are or
what time it is, adhere to the same strict programme of events. It is also necessary
to put into place technologies which ensure that all sites and departments can
communicate continuously with one another: phones, faxes, video conferencing
facilities. Furthermore there is significant use of email.
Scheduling sequencing is almost but not quite a ‘virtual’ technology. It exists
initially as a series of inscriptions – tables and charts. These inscriptions are
distributed via a cascade of meetings. But in the course of this, they are transformed,
giving rise to different kinds of sub-scheduling sequences and deadlines. Scheduling
sequences become more virtual, less localised in a core set of materials, as they
proliferate throughout the organisation. Which means that the problem then shifts to
ensuring that this mass of inscriptions in which scheduled sequences are embedded
are themselves co-ordinated and interlocked.
This task is tackled in a novel way by Oilsite. The kind of massive refineries in
which they specialise are built in modular form. They are designed as separate bits
with standardised connections that ensure that the bits all eventually fit together into
the finished product. Oilsite has translated this modular format to organise its own
systems of reporting. Modularisation of the organisation and associated artefacts
is also an answer to the problems arising from a multi-site company producing a
complex range of petro-chemical products. There is a core group who deal with the
day to day running of the company (e.g. finance, IT, personnel) and then a number
of distinct project groups. Project Definition, for example, is separated from Project
Engineering. The different parts of the company, then, purportedly fit together in
much the same way that the different modules of the refinery do – thus linking
differences in space and time. Which means that the key task of scheduling sequences
is to ensure that the work of each modular part of the organisation comes together at
the right moment. Conversely, it means that the failure of a given modularised part of
the organisation to reach its own particular deadline can be detached from the overall
schedule. A replacement can be attached in its place.
There are many examples of this process of detaching and reattaching modules.
A gas conversion unit falls off the barge used to transport it. Equipment from a
particular supplier fails to pass quality checks. Relationships between personnel in
two different departments become strained. In each case, Oilsite responds in a similar
way: it sends out messengers. Troubleshooters are ‘parachuted’ into field sites or
seconded to recalcitrant departments. They open out new lines of communication
between modules, make decisions on whether the module itself can be salvaged
or else dispensed with. Scheduling a sequence is therefore time critical but in an
interesting way. Modularisation ensures that the overall sequencing of production
198 Doing Things with Things
is not tied to a linear arrangement of modules. Each module can set up its own
scheduling sequence in relation to local constraints such as the availability of
material and staff. Overall scheduling of sequences consists in the detachment and
reattachment of modular events.
Within Oilsite, then, a version of sociality, of how social relations might best be
structured within the organisation, is formulated in terms of engineering practice
– from the materials themselves. Modularity becomes the form around which the
‘human side’ of the business is managed. It is also the means by which scheduling
sequences are driven down through the narrowest capillaries of the organisation. But
this particular version of sociality also requires rapid and generalised communication.
Materials must flow. People must co-ordinate their actions both in time and space
in terms of scheduling sequences in relation to the detachment and reattachment of
modules. This is why information communication technologies gain overarching
importance for Oilsite. Modularity ensures that the schedule is not threatened by the
failure of a component part. But it also creates hidden spaces in which potentially
disruptive practices and activities may be developing. It is therefore not enough
just to ensure that the modules are all connected one to another. There must also be
methods put in place for producing transparency, for guaranteeing visibility between
modules and the ongoing accomplishment of scheduling sequences.
Computer-mediated communication, in particular email, is the major tool for
engineering this visibility. All employees routinely use email, contribute to electronic
discussion lists, refer to information held on the intranet pages and so on. Moving
upwards, the diaries of senior managers are often accessible through the electronic
network, as are strategy-related documents. The residue of this computer mediated
communication remains available in a variety of ways. The storing of things like
emails, systematic or otherwise is a focus of concerted effort by many members of
the organisation. Elsewhere (Brown, Middleton & Lightfoot, 2001) we describe this
at some length. However, here we can see in the following brief example drawn from
a corpus of interviews with Oilsite managers that archiving the products of CMC can
be a key concern and resource in organisational engagement and accountabilities.
The following extract opens with a question about the storage of email messages:

Interviewer: Do you take them and store them?

OilSite Manager Engineering Computation: Far, far too many. But I put them into
categories and then I actually try and archive them all onto CD, as I have access to a CD
writer, so I archive them again. I do sometimes go back to them but I do store. Putting
them into the categories of the folders is what takes the time and that is why there is
always a bit of a back log. But I do try and do that as you never actually know, it is
amazing how often I do have to go back and look up one of these things.
I: What sort of things do you go back to then?
M: Well, it is if somebody comes up and says or somebody sends a message and says that
I want to do so and so. And you think that we have already dealt with that. So we go back
and it has been dealt with and perhaps that person will find it from that sort of thing. But
there isn’t a lot of protectionism; generally speaking we do not write the sort of memos
which absolve, where they absolve ourselves for more blame for any future reference. But
 Words and Things 199
there is not a lot of that that goes on. So I do not spend a lot of time going back there and
looking for get out clauses. But quite often things come up again and you think that I know
I’ve looked at that and I don’t want to look at that again. Or maybe just to find out who
was interested before so you can pass information on and that sort of thing.

In the first part of the extract it becomes clear that not only does the manager store the
emails he receives but also that this practice of storing is attended to as a problematic
issue (‘far, far too many’). The manager then goes on to point to a tension that exists
between the laborious routine of organising and storing electronic archives against
the potential benefits of doing so when it proves necessary to recall something from
the archive (‘it’s amazing’). The occasions which necessitate such activity are then
presented as instances of colleagues making requests that the manager would like to
deflect through offering up archived material as evidence that the issues have already
occurred and been addressed in the past (‘and you think that we’ve already dealt with
that’). In this way the manager attends to the past as both an issue that is continually
re-invoked in the daily activities of the organisation, and as a practical resource
which can be mobilised to perform rhetorical work of establishing and maintaining
order in the present of ongoing and complex processes across time and places.
So what occurs here is a discursive deployment of the past in the course of
communicative action. This deployment is quite complex. As the extract progresses,
the manager begins to attend to a dilemma raised by his earlier comments, namely
that he may be seen to be producing ‘interested’ or tailored versions of the past in
order to simply avoid having to comply with other requests. His way of managing
this dilemma is to emphasise that what is stored in the archives is not itself partial.
He is not seeking ‘get out clauses’, nor is there any tendency to ‘write memos which
absolve’. By describing the contents of the archive as pure information, which may
be usefully passed on occasionally to others, the manager attempts to frame his
own remembering as an activity which simply serves to benefit the organisation
by recalling its own history to itself in a way which strategically informs its future
directions. Indeed this is often the way that electronic archives are described by their
advocates. They present information communication technologies as the means of
actively storing up the past in a way that bears dividends in the future (Khoshafian
& Buckiewicz, 1995). The past is routinely equated with information, which is
presumed a value free commodity. Information cannot be ‘interested’ or ‘biased’, it
can only be partial or corrupted in some way. These shortcomings can be addressed
by ‘an intelligent organization’ (op cit p.42).
Now in order to describe an electronic archive in this way, the archive itself
must be organised in such a manner as to afford this kind of logic. That is, a work
of ordering needs to be performed on the constituent pieces of equipment whose
relations come to define what the archive is. We can see this already in the first
extract, where the manager describes the electronic archive in terms of a system of
folders which are duplicated across a number of different materials (e.g. a central
system and then CD’s). This work of ordering the archive, which is driven by
the current projects of the organisation, goes on alongside the discursive work of
200 Doing Things with Things
remembering, and acts as an essential support and buttress to it. Without the non-
discursive labour of forging relations between components in the archive there is no
possibility for playing out the evidencing strategies which dominate the manager’s
rhetorical performance of the past.
So the visibility which CMC produces within Oilsite does not come about through
unlimited access to all, or through a liberating of information. It is not the fact of
connection itself, or the circulation of new kinds of informational objects which
makes modules visible to one another. It is instead through the creation of a highly
public space for the performance of accountability. When everyone is connected,
this places an obligation on each employee to display their own attention to that
connection. Such obligations are about visibly making sure that all members and
parts of the organisation attend to connection as part of the evolving ordering of
production processes. This can take the form of replying to requests delivered over
the email such that others can see one’s activities. Or ensuring that one contributes
to debates on discussion lists, even if this is confined to ‘me too’ types of responses.
To be judged to be ‘out of the loop’ is a serious threat to personal accountability
within Oilsite.
Two problems then emerge. First of all, there is an awful lot of email traffic
within Oilsite which is considered to be wasteful, non-productive. These messages
are written in a mode of sociability or ‘phatic’ mode. Their sole purpose is to remind
others of the fact of the connection of the sender – ‘I’m here’. Second, there is
a tendency for copying messages to multiple addresses. This is a proliferation of
the phatic. Ensuring that one’s presence is noted by as many people as possible.
But it also reflects something else ‘at work’. Something which is cutting across the
modularity impressed upon sociality by the material design of Oilsite.
Looking at the kinds of extracts like the above, we may be tempted to conclude
that people at Oilsite spend all their time in conversation, passing the time, doing
work, exchanging gossip, fixing a schedule. And we might then go on to conclude
that it is the interactional structure of such conversation that holds them all together,
that binds them as co-workers. But that is not what we are suggesting here. Already,
it seems to us, that there is something to be explained in the way that managers at
Oilsite spend so much time worrying about the archiving of email messages. This is
not, we argue, a purely discursive matter. This displays an orientation to something
else, something that cannot be explained by looking at the discursive formulation
of objects. Email archives are a peculiar species of objects. They are powerful,
certainly, but they gain this power through their ability to lodge themselves in
everyday interaction. They are objects inasmuch as one can point to them, describe
them, evaluate them and so on. But because they are so important for interaction,
they become in effect a kind of subject. That is, something to be negotiated and
taken into account in a given interaction. Objects are therefore not external to social
relations, but in them, both being parts of unfolding activities and practices. We can
illustrate this point if we examine the multiple ways in which email archives are at
Oilsite. Computer mediated communication readily affords the creation of multiple
email archives. Electronic archives are far more than the storing up of the past, either
 Words and Things 201
actively or passively of an organisational record. Electronic archives are subject
to a continual ordering and reordering. Non-discursive work of ordering must be
performed on the chards of electronic traffic. Such ordering affords relations that
define what the archive is and make possible further forms of communicative action
that also orders sociality and equipmental relation in organisational practice. The
archive is ordered as the third party in CMC, as a quasi object (Serres, 1980; 1995).
Managers in Oilsite complained about the number of emails they received and
the practical problems this created in terms of the time spent archiving. We may
read this complaint in two ways. On the one hand it is a complaint about the literal
problem of spending time ensuring that messages can be archived in the right folder,
and that this archiving maintains the integrity and utility of the archive as a whole
(i.e. archiving by project may prove less successful than archiving by month). That
is, the effort involved in non-discursive ordering. At the same time we should also
understand this non-discursive labour as about tying in senders to projects, recruiting
alliances, and control at a distance (see Brown, Middleton & Lightfoot, 2001). Too
many messages simply undermine the ability of the managers to achieve this kind
of order effectively. Viewed as a discursive problematic, time spent on archiving
exposes managers to claims on their own accountability – namely that their time
would be better spent elsewhere. We may then hear in formulations of the problems
of archiving a dual struggle with establishing order amongst both discursive and
non-discursive elements. In other words, the archives are introduced in the ongoing
activity to establish and maintain order in complex ongoing activities. In addition,
this very means of doing so introduces new relations and dynamics into that activity
which resolves some problems of maintaining order in ongoing complex activities
but creates others.

Distributive Archives

How is this dual problematic addressed? One common method is to tackle the
problem by introducing novel elements into the archive. In the following extract the
manager in question describes how she informally enrols members of her section as
de facto archivists by ensuring that they have access to messages which she is aware
that they will find ‘interesting’ and feel compelled to keep:

Interviewer: The ones that you do tend to keep – what kind of general category are those
things?
Oilsite Personnel Manager: Erm, things that are still ongoing. I design the on-the spot-
bonuses and rather than everytime I get one deal with it and then there it goes into a file
and so that gets kept. I run the management employee communication team which is the
consultative care-team, it works with council and I keep the back notes to that. So it is
things that I might read in the future, like if I don’t really think I am going to or I think that
somebody else is keeping them. Like I say filing is not my forte. Having done my building
I am not wanting to evaluate it at all. You can guarantee that those people absolutely drive
me potty. (Laughter) Have you got it now?
202 Doing Things with Things
I: They are quite useful to have around?
M: Yes they are very useful to have around. They keep the filing for you. You want those
people to be doing the filing for you. You don’t want to be doing it yourself.

The extract begins with the manager pointing to different messages that she would
tend to store in electronic archives, but then an interesting hedge is introduced.
Messages that she does not ‘really think I am going to have to’ keep or those that she
suspects may be held by others can be safely discounted. The clear implication here
is that as a manager, section members can be mobilised as informal archivists. One
can establish the likelihood that an archive may be maintained by others which can
be accessed at some later date without having the trouble of maintaining a complete
archive oneself. This is a non-discursive solution to the problem. In other words,
perceived inadequacies in the upkeep of the archive are resolved by adding new and
novel elements into what is to be counted as the archive – that is, colleagues and
subordinates. This is in effect a means of binding external others into the archive by
informally mobilising more proximal others as archivists. It is as though she were
building a pathway towards the senders who are tied into the archive by using fellow
workers as material extensions of the archive itself (‘You want those people to be doing
the filing for you’). The archive is the third party which makes the communication
possible and to which Oilsite members appeal as a guarantor of deeds and words.
It is an object through which organisational players establish identities in terms of
their relationship to the archive. It is an object defined by transformational orderings
and the effects these produce and have on those who engage and are engaged with
it. Clearly this is a delicate issue to discursively negotiate, and hence some of the
peculiarities of the manager’s replies (‘Have you got it now?’). This is more than a
matter of the archive as a distributed network of the discursive and non discursive
that can be used in settling issues of past accountability, what might or might not have
been done in the past. As the manager points out, it is also about ‘things that are still
ongoing’. The distributed archive is crucial in maintaining the order, or re-ordering
of complex activity across times and places in the scheduling of sequences.
Another way of negotiating the same issue is for a manager to claim that they are
similarly mobilised by their fellow managers:

Interviewer: Would you ever resend stuff from archives or as the basis of one of those
conversations?
Oilsite Manager Design Engineering: Oh yes, often managers who do not archive will
ring me and say do you remember back in May and have you still got (laughter).

Managers come to rely upon distributed archival practices within their own
organisation to buttress their own archives as a feature of maintaining order in their
ongoing activities. But at the same time they also attend to the possibility that these
archives may be used against them, as in the following extract. The extract begins
with a question about how users write emails which attempt to promote a favourable
version of themselves to senior managers:
 Words and Things 203
Interviewer: How would that kind of thing, say that people ... the way in which people
say it is almost like they’re being helpful but they are sticking the knife in at the same
time? How would that type of thing be done?
Oilsite Personnel Manager: I don’t know of any good examples I have had recently. It
is the sort of thing that you know someone is dealing with something and then you write
a memo copying their four senior people. Written sort of you are the one that has picked
this up and what a good person you are to have done this and to have found it out and it
does actually then ... say that the person that you are also including on the mail message
that you are actually already doing something about it. It’s sort of a subtle ‘I am a good
boy,’ but you are only giving them half of the story.
I: I thought that I would just let you know?
M: Actually I can remember a good one, I did have a good one it was something to do
with the wiring not being done and we had asked to, I would ask ISB to do the plan for
the building and then a month later they sent us a message saying further to our letter of a
month ago. It came into us saying why have we not done the wiring yet on 1E and copying
their boss and turning it by saying that we now need to do this because we have been
asked to have it done by Friday or the following Monday or something, and fortunately
both myself and the facilities manager have got a copy of the message asking them for
the plan. Which in spite of the fact that all the other messages were on this email it was
forwarded to the boss the last one wasn’t and boy was he furious when I forwarded him
the final message that he had not been copied on. They had shaved off the top lesson, they
had just come back to a point before my message was actually on it and said look that
Facilities had dropped the ball again and we had already agreed on what we were going
to do was to do.

In the first part of the extract the manager describes how personal accountability
is often at stake in email interactions and how users can strategically exploit this
through practices of copying in superiors (see Brown & Lightfoot, 2002). In the
second part, this is described by recounting an episode where her own section was
involved in a dispute with another. At the heart of this problem was the doctoring of
an email communication by the other section, since email interactions are routinely
copied by the system into the body of successive messages. What this means is that
a sender can tailor how the complete interaction is recorded in a current email simply
by editing out chunks of the words that have been copied over to a new mail in the
interaction. Luckily the manager describes how she was able to expose this tailoring
of the past by forwarding a complete copy of the interaction extracted from her
own archives. As a technology of ordering the archives are introduced to deal with
conflicts between parts of the company (customers, local authorities and others) over
such issues. Nonetheless, precisely because these conflicts do exist, the archives too
can become subjected as technologies for those struggles. Again, the archives do not
make conflicts disappear but affects their dynamics and modes of performance by
being a technology that can be enrolled in them.
What this extract points to is a wider problem with distributed archiving. It is
notable that such kinds of disputes were typically settled in face-to-face meeting (the
irony here of course is that in Oilsite the introduction of CMC was meant to eliminate
the need for precisely these kinds of meetings). Settlements are typically arrived
204 Doing Things with Things
at when speakers can display another form of evidence – such as a written copy
of an email. We would argue that at such moments CMC stops becoming a solely
discursive activity and becomes also a non-discursive activity. That is, a matter of
ordering bits of paper, stored messages, print-outs and so on as well as managing the
demands of immediate interaction.

The Non-discursive and Organisation Structure

So where we have arrived at is the idea that there are always two sets of ongoing
activities: the discursive moves made in interactions and the non-discursive ordering
of things (e.g. emails, archives, materials). So to ask how we talk about things is to
focus on one activity alone. We might equally ask ‘how are things moved around
and ordered in such a way that they become fit objects about which to speak?’ The
email archive, as we have shown, is a particularly intriguing instance of this. It
is something which is worked upon, ordered and continually shaped by ongoing
organisational activities. Yet it is also something which we can display as something
continually oriented to in the talk of Oilsite personnel. In effect, the email archive is
integral in social relations.
Let us try to contextualise this by asking one final question: where does an
organisation actually remember? How are the present-future functions of the archive
deployed as part of the ongoing organisational practice of scheduling sequences. One
thing we can be sure of is that the place where archiving occurs is not the same place
where any demands for accountability and free flowing coordinated production are
initiated. Or more crudely, that archiving occurs in spaces that are entirely distinct
from where the major decisions occur, from where the aura of power and central
executive discretion emanates. We can show this empirically. In the following long
extract a senior manager is questioned about differences in record keeping at various
levels in the organisation:

Interviewer: The impression that we seem to be getting is that a lot of the strategy type
meetings are being done very informally, even since there are not a huge lot of formal
minutes being taken at the meetings?
Oilsite Manager Executive Development: I don’t think that that happens anyway. We
always take formal minutes when you have got a third party there. So if you have got
a meeting with a client, erm, and on project task forces and stuff you will take internal
meetings. That is part of the project discipline. But our side of the project meetings, the
meetings I go to very rarely are they minuted, very rarely. And that is ok as well, to me
it would not be appropriate. We had one meeting a few days ago where we did record it
because it was a project action and I actually just sent around that meeting note today
formally, but that is quite unusual actually for myself. But in the job that I am doing now
it does not come up.
I: We have got the picture especially with the OMG, perhaps they meet ten times a week,
but no one is thinking papers are wrong?
M: That is an internal, that is a fairly informal grouping, yes it is. It is informal but it is
basically four operations managers, you have to bear in mind ... But no there are never ...
 Words and Things 205
no secretary’s there. There are four or five managers working together on issues around
the table progressing some issues. I would tend to get some stuff done to work out because
most of the resources of the company are represented by the people in that meeting. But
no we never record minutes. Absolutely unnecessary. It is not that sort of a meeting and
that is typical of the meetings that go on. There are very few internal meetings of projects
that get recorded formally, in my opinion. Some disciplines might do it differently.
I: It is an intriguing ... as that is entirely where the big decisions are made ... in meetings
... therefore you would record them in the minutes? It seems that, you are OMG and must
be making decisions all the time though, and they are not recorded?
M: Yes. You know you are basically colleagues together. There is not any, you do not
want to leave that meeting having to defend what you have decided at the meeting against
others. If there was something that came up at OMG which was divisive, there was not an
agreement or there was something like that it would go to another level. I have never been
to one of those meetings. It tends to be a consensus pretty well. And very often actions
won’t be decided on the day it will go forward to another day and it will be progressed
over a period of weeks, and that is ok as well as it is the nature of the beast.

What emerges in the course of the extract is that very senior meetings, the places
where big decisions are made, are rarely minuted. This is felt to be ‘absolutely
unnecessary’. The reasons for this as they are mobilised by the manager in question
are interesting. At that level ‘you are basically colleagues together’, which is to
say that matters of routine accountability are on a very different footing. One is
not expected to have to mobilise extensive evidence to defend one’s position. As
such there is no need for the present to be continually consigned to the archive
through a supplementary mechanism in the form of minutes. The joint recall of these
‘colleagues together’ suffices. So it is the introduction of diverse parties, customers,
diverse sections with diverse tasks, responsibilities, interests – and therefore pending
conflict – that minutes are archived. The bosses not taking minutes is not a matter of
their being bosses, he says. If he is right, other meetings of a small group of regularly
meeting equal colleagues would not be archived either – lest they had to take care
that they might be held accountable by others, say, higher up.
What we are presented with here is the space prior to the place of archiving, the
space from which archiving is itself commanded, but which stands outside of the
archive. This is a space that is not bound by the usual rules of audit. Managers here
keep no formal records, save a few notes in their own personal workbooks which
are never entered into the organisation’s wider archives. Why should this be the
case? Because entering such records into the archive would make them subject to
the usual rules of truth. Records of senior meetings would then become part of the
play of correctness and its strategic undermining which constitutes the usual activity
which goes on around electronic archiving. By refusing to allow any supplement to
their own highly fallible human memories senior managers take themselves out of
the economy of correctness.
But in staying out of this economy, managers do not reject truth altogether. In
fact they make available to themselves a more potent form of truth. Occupying an
unauditable space allows senior managers to challenge archiving practices without
206 Doing Things with Things
themselves being challenged in return. Since their own past is not captured in any
archive maintained by the organisation, senior managers are able to claim that these
archives are merely supplementary, that they do not capture the real essence of the
business – that they fail to express the guiding vision or project of the organisation. But
– and this is crucial – such a claim can only be launched if managers simultaneously
have complete access to the organisation’s archives. Because it is only by coming to
terms with the established practices of correctness, with the authorised version of the
past, that they are able to mobilise a radically different version of the organisation’s
past, one which, they are able to claim, is more fundamental and more ‘truthful’ than
any existing version. Since it alone is based on a vital and proximal memory rather
than a supplementary, technical and external practice of archiving.

Conclusion

All of which is another way of saying that our focus on how things are talked about
can be reversed. Things make us talk in particular ways. They make us communicate
in particular ways, they make us manage in particular ways. Now this is not to say
that things are ‘just there’, that they can serve as the a priori grounds upon which
interaction can be understood. To argue that would be just as bad as arguing from an
a priori description of the social. No, things are ordered, put together and made sense
of in precisely the same way that words are. But these are two different activities
which we must draw upon different sets of resources to understand. This we then
take to be the immediate analytic task – to develop a vocabulary for talking about
how things afford talk, for how things are ordered in such a way that they become
fitting topics for talk. And, of course, also how talk gets things done, because clearly
this company is not just in the business of moving messages around, and having
meetings.
Overall then this chapter offers a conceptual re-orientation to how we think
about the relationship of subjects and objects and how ordering and so forth are
part of getting things done. We examined how objects are formulated and disputed
in CMC, paying particular attention to the kinds of discursive resources brought to
bear by users. The role of electronic archives in resourcing disputes arising from
contested formulations shows how the maintenance of electronic archives requires
the enrolling of other subjects, and the sustaining of networks of artefacts. In addition
the ‘non-discursive’ work of archiving and the discursive formulation of artefacts
are accomplished simultaneously. Finally we concluded that attending to the non-
discursive enriches our understanding of how conversation is managed within CMC
settings. Such management of CMC is shown to be more than establishing some
bridge over to the object world that is not to hand by means of communicative
devices that substitute for gesture, gaze (deixis) etc., in the interactional generation
of shared context. The process is better understood as one whereby objects are
brought between speakers. That is, the way objects configure dialogue and at the
same time are ordered as topics in such communicative action. This is not to argue
 Words and Things 207
that things serve as some given but, rather, that things are also ordered in precisely
the same way that words are. Organisational practice emerges in the way the material
resources of the archives order such practices within interdependent orderings of the
discursive and the non-discursive – of words and things.

Aknowledgements

Thanks to Ole Dreier and Alan Costall for their most helpful comments on this
chapter.

References

Abbott, A. (2001). Time matters: On theory and method. Chicago: University of

Chicago Press.
Brown, S.D. & Lightfoot, G.M. (1998). Insistent emplacement: Heidegger on the
technologies of informing. Information Technology & People, 11(4), 290–304.
Brown, S.D. & Lightfoot, G.M. (2002). Presence, absence and acccountability: Email
and the mediation of organisational life. In S. Woolgar (Ed.), Virtual society? Get
real!. Buckingham: Open University Press.
Brown, S.D., Middleton, D. & Lightfoot, G.M. (2001). Performing the past in
electronic archive: Interdependencies in the discursive and non-discursive
organisation of remembering. Culture & Psychology, 7(2), 124–144.
Dreier, Ole (1999). Personal trajectories of participation across contexts of social
Practice. Outlines, 1(1), 5–32.
Dreyfus, H. & Dreyfus, S. (1986). Mind over machine. New York: Macmillan.
Drew, P. & Heritage, J. (1992). Talk at work: Interaction in institutional settings.
Cambridge: Cambridge University Press.
Edwards, D. & Potter, J. (1992). Discursive psychology. London: Sage.
Edwards, D. (1997). Discourse and cognition. London: Sage.
Goodwin, C. & Goodwin, M.H. (1996) Seeing as a situated activity: Formulating
planes. In Y. Engeström & D. Middleton (Eds.), Cognition and communication at
work. Cambridge: Cambridge University Press. (pp. 61–95).
Hiltz, S. R. & Turoff, M. (1978). The network nation: Human communication via
computer. Cambridge, MA: The MIT Press.
Hutchins, E. & Klausen, T. (1996). Distributed cognition in an airline cockpit. In
Y. Engeström & D. Middleton (Eds.), Cognition and communication at work
(pp. 15–34). Cambridge: Cambridge University Press.
Kiesler, S., Siegel, J. & McGuire, T. (1984). Social psychological aspects of
computer-mediated communication. American Psychologist, 38, 1123–1134.
Khoshafian, S. & Buckiewicz, M. (1995). Introduction to groupware, workflow, and
workgroup computing. New York: Wiley.
Leudar, I. (1991) Sociogenesis, coordination and mutualism. Journal for the Theory
of Social Behaviour, 21(2), 197–220.
208 Doing Things with Things
Rice, R. & Love, G. (1987). Electronic emotion: Socioemotional content in a
computer-mediated communication network. Communication Research, 14(1),
85–108.
Serres, M. (1980). The parasite. Baltimore: Johns Hopkins Press.
Serres, M. (1995). Angels: A modern myth. Paris: Flammarion.
Sproull, L. & Kiesler, S. (1986). Reducing social context cues: Electronic mail in
organizational communication. Management science, 32(11), 1492–1513.
Sproull, L. & Kiesler, S. (1998). Connections: New ways of working in the networked
organization. (6th ed.) Cambridge, MA: The MIT Press.
Winograd, T. & Flores, F. (1986). Understanding computers and cognition. Norwood,
NJ, USA: Ablex Publishing Corp.
 Chapter 11

Learning to Do Things with Things:

Apprenticeship in Bakery as Economy
and Social Practice
Klaus Nielsen

Introduction

Today the concept of learning is everywhere. Learning labs, learning centres, courses
on self-reflexive learning, etc., just to mention a few areas in which the concept of
learning has disclosed itself in the last couple of years. The concept of learning has
become a major ‘buzz’ word in academic discourses in Denmark. However, when
listening to the interview tapes with apprentices, as part of analysing qualitative data
in the research project ‘Learning in Modern Danish Vocational Training’, I realized
that they did not address issues of learning as much as notions of economy when
they discussed education. I began to notice this discrepancy and started to wonder
why I, and many other people, with a profound interest in learning never relate to
issues of economy in discussions about learning. Instead, we seem to neglect the
economic influence on learning entirely.
First of all, this article presents a further development of the general problem of
the neglected economic aspects in discourses of learning, starting with a research
project of how apprentices learn their trade. It seems that comprehension of learning
in practice becomes much more transparent when we include aspects of economy,
rather than merely basing our understanding on epistemological perspectives of
learning. In this connection I will define an epistemological perspective as a specific
approach to learning in which matters of knowing the world is central. Below, I will
elaborate on the various definitions used in the article.
The arguments in this article are primarily based on empirical material from a
study of how baker apprentices learn their trade, and subsequently, the methodological
design is described. Examples from the empirical material are laid out and discussed
in two paragraphs. The first section describes how the economic and educational
discourses in the bakery are interwoven. In the second empirical section, examples
are given of how the baker apprentices try to mediate learning the trade of baking,
as well as the economic aspects of this practice. Finally, in the discussion, the
consequences of including the economic aspects of people’s everyday existence are
considered in relation to issues of learning generally.
210 Doing Things with Things
The general neglect of economy

In academic discussions of learning in practice the dualism between learning in

schools and learning in a workplace has been dominant. It has been a discussion
based on the assumption that the main difference between the two settings was
one of knowledge. It became a discussion apparently about the different kinds of
knowledge, which are supposedly embedded in the different social arrangements of
schools and workplaces (see e.g. Nyeri, 1988).
The baker apprentices attended a trade school as part of their training in order
to become bakers. However, according to the baker apprentices the primary
difference between learning at the trade school and learning in the bakery was that
all perspectives related to economy were neglected in the trade school. One of the
apprentices describes the difference between trade school and bakery like this:

[In trade school] we can take as long as we like to make a perfect Danish pastry, the
teacher does not say anything. You can’t take that back to real life, there you need to
produce, and that you only learn by producing, getting better and better, until you figure
out what a Danish should be like, instead of measuring its diameter – it is way too old
fashioned. I believe the school should apply the production method. Here we make rolls
by hand, I mean, where do they do that today? If the machine breaks down, we just get
hold of a new dough machine (Kasper).

According to the apprentice, it is a matter of producing large quantities in the bakery

with the aid of different kinds of machines. At the trade school apprentices work
with small quantities of dough and they do not encounter any hysterical masters who
shout at them because they have made a mistake, or are too slow. Rather, the teacher
does not say anything. Furthermore, there is time pressure in the bakery, which is
not present in the school context; there the apprentices have all the time they need in
order to work on the details. Looking at the physical arrangements in a bakery and
the bakery at the trade school, it is hard to detect any differences. In both settings
there are ovens, large tables, mixers, etc. Even though it is not said directly, it is
clear in the quotation that aspects of economy and production play a crucial role in
the bakery.
Returning to the discussion mentioned above, the baker apprentices shared an
important insight. According to the bakers, the primary difference between learning
at school and at the workplace is not a matter of different kinds of knowledge, but
first and foremost a matter of how schools neglect the economic aspects of practice.
Or using the reverse argument, economic aspects play a crucial role when addressing
issues of learning in the bakery, but not at school. In the following I will explore
these issues more thoroughly. The following is a closer look at how we understand
learning from an epistemological perspective and an introduction to an economic
perspective in relation to learning.
 Learning to Do Things with Things 211
Learning and epistemology

One of the many reasons why the notion of economy is excluded from discussions of
learning has to do with the way we relate learning to epistemology. A paradigmatic
understanding of epistemology is found in Descartes’ thinking. According to Descartes
our knowledge of the world becomes more certain if we follow a systematic method
where we first analyze/divide the problem we are facing into its smallest recognizable
parts. Secondly, we synthesize/reconstruct the parts by following strict logical rules
to a more complex recognition of the subject matter at hand (Descartes, 1996). The
dialectic of the process of analyzing and synthesizing is what Descartes prescribes as
the method, which provides us with a certain knowledge of the world. This systematic
way of knowing the world has a great impact on what we consider to be science, and
consequently, also on how education should be organized. It is the latter, which is
of interest in this article. Descartes’ understanding of epistemology has a bearing on
how educational situations are arranged. Educational curricula are often organized on
the principles of confronting the student with the most simple and basic element of a
subject matter and then moving into the next element (Woolfolk, 2001). This way of
thinking is influenced by Descartes’ conception of knowledge. The same applies to
the way problem solving is represented in textbooks of educational psychology; one
solves a problem by identifying the basic elements and then moving onto the more
complex elements (Woolfolk, 2001). Behind these institutional patterns of arranging
education we find the Cartesian conception of epistemology and knowing the world/
learning about the world. In the epistemological perspective, learning is a question of
learning about the world.
The main problem with this understanding of learning is that the process of
knowing the world divides our recognition of the world into small parts, and thereby
divides the issues we want to know about into discrete unites. Consequently, the
problems we face and analyze are removed from their context in the process of
reducing complexity. From this it follows that matters concerning emotions, desire,
morality, power and economics are excluded from the process of learning/knowing
the world since these issues address complex qualities in our being in the world.
However, it does not make sense to reduce these issues to the smallest recognizable
units. Reading through the index of mainstream textbooks on learning, or in
educational psychology, none of these matters are described to any large extent. In
this article it is especially issues of economy that will be pursued further since, in my
opinion, they have the greatest implications for learning.

Learning and economy.

By addressing economy, I am taking the stand that humans maintain, express and
understand themselves through the use of material goods and services. My argument
in the following is that aspects of economy and learning support each other. Issues of
economy are central and inseparable when addressing issues of learning in practice.
Additionally, I will argue that by paying attention to the interaction between economy
212 Doing Things with Things
and the organization of a practice, a better comprehension of learning can be reached,
than merely by taking an epistemological approach to learning. The perspective on
learning laid out in this article is inspired by theories of learning as socially situated
(see Lave & Wenger, 1991; Lave, 1997; Dreier, 1999). In this perspective, learning
concerns how participants in an ongoing social practice change their participation.
By including this perspective on learning, issues of economy become important in
order to understand changes in ongoing social practices, and in that connection also
issues of learning.
The term economy comes from ancient Greek and is defined as somebody
who manages a household (oikonomia: household manager). Furthermore, it is
a concept relating to the production, distribution and consumption of goods and
services (Merriam Webster Dictionary, 1999, p. 365). In economic anthropology,
this double connotation of the term economy is preserved and differentiated in
terms of community economy and market economy (Gudeman & Rivera, 1990). It
is a helpful distinction for this article so I will shortly define the two concepts. The
notion of community economics describes how smaller groups locally produce in
order to support their independence. Sharing is central to community economics.
Land, material resources, knowledge is what a community shares and the source
of its maintenance. Market economy is based upon competition among buyers and
sellers who exchange goods and services. The goal of participation in the marked is
to secure a money profit, which is private property.
In everyday life both these perspectives on economy exist. However, whether
it is aspects of the dominant community economy in the bakery or aspects of a
market economy make a big difference to issues of learning, so I will maintain the
distinction between the two understandings of economy.
If we merely perceive learning from an epistemological perspective then only
learning, which takes place in institutions where the economic aspects of practice
are not accessible for the students, can be considered real learning. Learning in work
places or in families where issues of economy play a crucial role is a contaminated
kind of learning, a reproductive and second-rate kind of learning. The critique of
apprenticeship has often had this character.
The following is a more concrete elaboration on the relationship between learning
and economy. It is an example of how economy develops the trade and influences
what the apprentices need to learn and how.

The Socially Situated Bakeries

Bakeries have a long tradition in Denmark and Northern Europe in general, going
back at least a thousand years as an organized trade. The baking trade was protected
in the Middle Ages in the market towns by the guilds, and the dominant way of
learning the trade was through apprenticeship. As the names of the cakes and the
bread indicate (e.g. French bread, Sarah Bernhardt cakes, Viennese pastry, brown
sugar pastry, etc.), the baking trade has always been internationally orientated
towards Europe by developing new products, finding new recipes in order to please
 Learning to Do Things with Things 213
the customers. One could argue that aspects of economy have always been part of
developing the trade. The following is a peculiar illustration of this. The famous
Danish pastry was developed in Denmark in 1850 when conflicts about wages in a
strange way led to the introduction and advance of Danish pastry. During a strike
the journeymen in Copenhagen demanded higher wages. However, the masters used
lockout against the journeymen, and invited journeymen from Vienna to work instead.
As the Viennese journeymen could not read the Danish recipes, they introduced
their own cakes instead, where especially the ‘Plundergebäck’, a cake made of yeast
dough became very popular with the customers. After the strike Plundergebäck was
transformed by the Danish journeymen into what we today know as Danish pastry.
A very concrete example of how economy and learning/developing the trade is part
of the same process.
However, today the bakeries are in an economically difficult situation. The
concept of ‘bake off’ is putting the bakeries under economic pressure. ‘Bake off’ is
a concept used by gas stations and kiosks. They have a small oven where they can
‘bake off’ the prefabricated bread and cakes and sell them freshly baked to customers.
Consequently, within the last ten years the number of bakeries has dropped from
1800 in Denmark to about 1300 bakeries today. The confrontation between the
trade embedded in old craft traditions and the economic demands is very visible and
significant in the everyday working life in the bakeries.

Methodological considerations

In the following I will shortly address some of the methodological considerations on

which the present article is based. Learning in practice is a field that has only been
examined to a limited extent. Thus, the purpose of the study is exploratory which
means that interesting statements and considerations that occur in the course of it
generate new hypotheses to be pursued and examined. This requires that the design
of the study has a certain openness and flexibility to make it possible to generate new
hypotheses without reformulating the entire study.
Two methodological approaches were applied, participant observation and semi-
structured interviews. The process of observation is dynamically built around the
interview as the center of rotation. The reason why the interviews are at the center
of rotation is partly that it is through them that the participants in the context can
express themselves and the significance of being a baker becomes visible. Partly
because the material from the interviews is by far the most extensive, comprising
about two hundred and fifty transcribed pages. However, the interviews cannot stand
alone, as they do not represent the context of which the interviewees are a part. To
use several different methodical approaches is to regard the process of research as
a dynamic process, where construction of knowledge is the essence and where a
dynamic interaction between observation and interviews is possible. Social reality is,
indeed, in many ways immense and it requires several complementary perspectives
to be able to form a complete picture of it.
214 Doing Things with Things
Participant observations
By observing what actually happens in a community of practice, it is possible to gain
insights into how it operates. Furthermore, through the observations it was possible
to create a fertile background for the interviews. The method of observation in this
study is seen as a systematic attempt to select, produce and report on those parts of
behavior and surroundings related to individuals’ interaction ‘there and then’, and
which are in accordance with the purpose of the study. In this study, I have worked as
a first-day apprentice in three different bakeries. One bakery was located in a small
supermarket in Aarhus with a master and two journeymen. Another one was located in
Ry (a small town 45 kilometers from Aarhus) where one master and two journeymen
worked. Finally, I worked for a day in a bakery in a larger supermarket, where there
was a master, ten journeymen and two apprentices. The latter supermarket was very
close to being a bread factory.

Semi-structured single interviews and group interviews

Seven individual interviews with six men and one woman were carried out for this
study. Four of the interviewees were baker apprentices; two were journeymen and
one a master. Furthermore, three group interviews were conducted with three groups
of baker apprentices with four apprentices in each group. Six men and six women
were interviewed in these groups. The criteria for participant selection included
willingness to participate in tape-recorded interviews, and the ability to articulate
experiences of learning as working bakers. The participants were volunteers and
were not paid for any part of their involvement. The interviews were designed as
semi-structured interviews centered around particular themes, but with the option of
exploring specific themes in greater depth (Kvale, 1996). The quotations used in this
article are the ones which best illustrate the points of research interest and all names
mentioned are pseudonyms.

Empirical Examples: Economy and Learning

Economy and learning are highly intertwined in the discourse of education in

apprenticeship. This is both a historical fact (Coy, 1989), but also a central facet of
today’s discourse about how apprentices learn in the workplace. In the following I
will argue that the economic aspects play a crucial role both when it comes to the
structure of the education in the bakery, and how it also influences the contents of
what is taught. Furthermore, I will argue that issues of learning are inseparable from
issues of economy when we focus on the bakery. However, I will try to show that
issues of economy are not seen as something that prevents learning from taking
place, rather to gain insights it is important to understand how economy and social
practice in the bakery are related to issues of learning.
 Learning to Do Things with Things 215
The structure of the education

The economic discourse plays a crucial and visible role in the structure of the
apprentice’s education in the bakery. It is certainly a central topic in the bakeries.
One of the interviews with a master provided us with a good example of how the
aspects of learning and economy are interwoven. This particular bakery is relatively
small and employs a master, two journeymen and two apprentices. The bakery is
primarily organized according to principles of a community economy.
After having listed all the things the apprentice needed to learn, the master went
further into the rationality as to why he had chosen this sequence of what was taught
first, second and so on:

Well, I need to teach him so he knows it, and he will know it. It may be an advantage that
it is a small bakery, the more he knows, the more help he will be, right. It will be easier
for us, gradually. The first year with an apprentice, right, it is hard, isn’t it, that is also why
he is sent over to make bread because he is of most use there now. But when he is moved
away from there, he will also get to do all the other things (Ole).

As seen in the master’s description, the rationality of the educational sequence in the
bakery is also embedded in a kind of thinking, which includes economic aspects of
practice. However, as the quotation from the interview with the master shows, the
notion of economic thinking is not something that prevents learning. The master
underlines that the more the apprentice knows the greater the benefit for the bakery.
According to the master the apprentice is also an investment, which costs the master
and the bakery a lot of work in the beginning, that is, the first year or so. However, if
the apprentice is well trained, he will become a profitable investment for the bakery.
The apprentice begins in the bread section of the bakery because this is where most
help is needed, and this is where he can be of most use initially.
The social arrangements surrounding the apprentice are also organized from a
perspective which includes economical thinking. The first year or so, the apprentice
works next to a journeyman who keeps an eye on the apprentice. As stated again
by the master, the rationality behind the arrangement is: there is still someone [the
journeyman] to keep an eye on him [the new apprentice], so he does not make
any major mistakes (Ole). Again the social arrangement with the journeyman who
constantly supervises the apprentice is found in economic thinking because it is
too expensive to let the apprentice make too many or too grave mistakes which
could prevent the production flow. However, as the apprentices express themselves,
working together with a skilled journeyman is the crucial way of learning the trade
because the journeyman functions also as a role model of how to do things in the
bakery, somebody who supports the apprentice, helps him or her in difficult situations
and somebody who sets the standards of the trade. One of the apprentices describes
how the journeyman whom he works right next to supports his learning process:
216 Doing Things with Things
Yes, first he shows you how to do it, and he shows it very fast, it just passes by. You don’t
see much. Then afterwards he shows it slowly and explains a bit why you do so and so. I
believe that is the most practical way.

Interviewer: And then you try it afterwards?

Then I try it afterwards, yes. That is the way it is, and then if you have any questions, he
shows you how you shouldn’t do it. That is fine, I think (Anders).

As this quotation shows, it is often the journeymen who provide the practical
instruction and keep an eye on the apprentice in the beginning of their practical
education in the bakery. However, if the apprentices in the beginning of their practical
education in the bakery are able to make quality products, then supervision from the
journeymen will be less intense, and the apprentices will have the opportunity to
work more independently.
The whole social arrangement of the apprentice’s education is aimed at enabling
the apprentice to work independently of the master and the other journeymen. Again
there are both considerations about learning and economy involved in the reasoning
behind this arrangement. The master nicely described it like this:

Because then he [the apprentice] can do many things when he has become good at it, right,
when he can work on his own, and the rest of us can do something else. Because there are
days when there are only two of us because of holidays or days off…then he will have
to be able to work very independently in the course of another couple of months, when
he will be working on his own with a lot of things. And must be able to do it. And…and
then the biggest help will be that he can make bread. And…that is why we have done it
that way (Ole).

The overall aim of the practical education in the bakery is to provide opportunities for
the apprentices to work independently because that is what the apprentices should be
able to do when they end their education and start working as journeymen. However,
there are more reasons why the aim of the apprenticeship education in the bakery is
to make the apprentice work more independently. Firstly, master and journeymen are
able to attend fully to other assignments in the bakery by not constantly keeping an
eye on the apprentice. Secondly, if the apprentice is able to work on his own, he is
also able to replace staff; he would be able to work during holidays or when the other
staff members are ill, etc. The aim of independence in the apprenticeship education
is not, however, to enhance individual autonomy for its own sake: the emphasis upon
autonomy is embedded in the economical thinking.
However, the apprentices in the bakery appreciate the opportunities to work
independently of the journeymen and master. They feel a sense of responsibility.
One of the apprentices describes the situation where he has just been given the
responsibility for part of the daily production in the bakery, and finds it a significant
learning experience:
 Learning to Do Things with Things 217
Lately while I have been in the section of cakes and confectionary, I have started to get
used to checking what needs to be discarded, what is old, what is dry, where we need to
fill up, doing it independently, filling up the shelves, whipping cream, figuring out what
needs to be done, what we are short of. Sometimes I consult the senior journeyman by
saying: we need this and that. And he says: then you do some of that. And then I will fill
up. Sometimes he shouts, if there is something that needs to be filled up, or that I should
have been more attentive – but otherwise it is my responsibility. You learn a lot that way
– positively (Poul-Erik).

In the apprentice’s description, it is clear that he is no longer just supervised by

a journeyman or merely doing what he is told by a journeyman or master. He is
responsible for part of the production in the bakery, and he has to maintain some
kind of overview and fill up what is lacking, etc. Furthermore, the journeyman is no
longer supervising the apprentice; rather the apprentice consults with the journeyman
when there is something he needs to know. The above quotation is a fine example of
how issues of learning and economy are not necessarily in opposition to each other.
Confronted directly with the notion of economic reasoning, or the academic
notion of exploitation, the master states:

Well, of course you are exploiting him slightly that way, it is cheap labor…. for the 3 years
you have them, 3½ years I believe, the first year you don’t make a lot on them, so to speak,
because they make so many mistakes and don’t manage to do what they are supposed to.
But the last 2½ it is still good, cheap labor. Of course it depends on what you mean by
exploitation, right, because they have to learn it, don’t they, and, really I don’t think that
we exploit them (Ole).

It is clear from the quotation that the organization of the apprentices’ education is
planned from the criteria of what is economically profitable. The apprentice is an
investment, who is expensive in the first year, but becomes profitable in the last
couple of years when he or she works like a journeyman, but is paid like an apprentice.
It should be added that this notion of exploitation is no secret to either the other
journeymen, or the apprentices. When confronted with these issues in the bakery, the
master called out to the apprentice first to find out what his wage was at the moment,
then to the journeyman about what his wage was, and then he subtracted the wages
and came up with how much he saved by having an apprentice. The apprentices find
it important that they are people that are being invested in, and receiving a paycheck
for their efforts every month is of crucial importance to them.
However, it should be noted in connection with the quotation above, that the
master added that he himself did not feel that he exploited the apprentices. He
stressed in the interview that he made sure the apprentices learnt the trade of baking,
that he made sure the apprentices went through all the different aspects of the baking
trade, and that he tried to give the apprentices meaningful assignments right from the
beginning of their time in the bakery.
The point is that the economic aspects are important, and often decisive when
we want to understand issues of learning. However, in the course of planning and
executing the apprentice education, the master, and also the journeymen, mediate
218 Doing Things with Things
aspects of learning and economy without any problems. One could argue that
insights into aspects of economy provides one with a better understanding of the
learning processes in the bakery, rather than consulting learning theories developed
on an epistemological basis.

The customer’s perspective

When we take a close look at a specific instruction situation in the bakery in which
a master instructs an apprentice, we find a complex interwoven pattern of economic
and learning perspectives in that same situation. It is unnecessary for the master to
break the task at hand into discrete elements, as suggested by the epistemological
perspective mentioned above.
When the apprentice learns to make certain products, he or she also needs
to include the perspective of the customer. To learn to see the products from the
perspective of the imaginary customer is an important part of what the apprentices
need to learn. The look of the product has a higher priority than how it tastes. It is
important for the apprentices to learn to make cakes that look delicious.
The master gives a concrete example of how he shows the apprentice to make
cakes:

The first times he does it, I take part in it. First, I do it while he stands next to me and
observes, I may be decorating with chocolate and nuts on top and stuff like that.

Interviewer: Decorating?

Yes, decorating. A cream cake, for example, has in most cases some decoration. It may be
a grape or a piece of tangerine, some chocolate or a chocolate coating. He decorates it and
makes sure the cake looks nice. He will begin with that. He will get the cake, and then he
will have to finish it by decorating it. Sometimes I will say: try to figure out your own way
of decorating it. Then he will have to think and try to decorate it so it looks nice. And we
can talk about it afterwards. After he has done that some times, then you can allow him to
make it from the start. He has to start by making raspberry jam and then put cream on and
the different things that need to go in it and stuff like that (Ole).

As this shows, the aspect of decoration plays a crucial role when it comes to making
cakes. The decoration is the pivotal point in the processes. As stressed in other
interviews, this should be seen in the perspective of how the apprentice needs to
learn that cakes and bread that look delicious sell. Bread or cakes with funny shapes
are thrown out.
Furthermore, in this quotation it is clear that the learning processes start
backwards in relation to the production processes. The overall aim of the learning
process is introduced first to the apprentice and then the master and apprentice
work their way backwards in the instruction process compared to the production
process. The apprentice is introduced to the nearly finished product that only needs
a little decorating. After the apprentice has learned to decorate, he will have the
 Learning to Do Things with Things 219
opportunity to make the cake from the beginning. In this perspective, it is the social
arrangement which is important rather than the method suggested by Descartes and
the epistemological perspective. If we imagined how the process would be arranged
from an epistemological perspective, it would begin with the smallest recognizable
elements. In this case it would be the ingredients for the cake, the raspberry jam, the
cream, the chocolate, and the tangerine. The second step would be to synthesize the
different ingredients into a cake, something that would probably mean a lot of waste,
while the perspective of the costumer would easily would be lost.

The threat of technology

In the previous paragraphs in this article the dominant economic perspective has been
what was initially termed community economics. This was linked to observations
and interviews in smaller bakeries. However, turning to observations and interviews
from a larger bakery with one master, ten journeymen and two apprentices, the
market economic perspective becomes more obvious. It is a bakery, which is very
close to being an industrial bakery, where advanced technology plays a dominant
role in taking over the skilled work of the participants in the bakery.
So far, I have argued that issues of economy are closely related to issues of
learning. However, the examples given have been positive in the sense that they have
shown that taking economic aspects into account does not exclude the apprentices
from finding the bakery a resourceful place to learn. However, in the following I
will argue that the introduction of a specific kind of new technology in the bakeries
can lead to a disqualification of the apprentices. A disqualification in the sense that it
prevents the apprentices from participating in other contextual settings – the bakeries
– in a qualified manner. It becomes obvious how vulnerable the bakeries are to the
development of the market economy.
One of the journeymen described how new technology influences what it means
to be a baker, because it is not necessary for the apprentices to learn to make white
bread, rolls, etc. He said this:

We have two bread maker machines, we don’t have to make bread anymore – it all rolls
out, we just have to put the loaves into moulds. And with rye bread, we throw the dough
into a funnel, and then they come out with the same weight and all, ready to be dipped in
some poppy seeds and then put into a mould too (yes). They [the apprentices] don’t weigh
them themselves, mix them or anything. Our roll machine – they don’t do the morning
rolls themselves or learn to shape them by hand – the roll machine does it all, puts them
on a baking tray – it does everything for them. They just have to turn them over in poppy
seeds and then stuff dough in the other end. They don’t really learn the trade, they don’t,
I don’t think so (Carsten).

The journeyman describes how the technology of making white bread makes many
skills superfluous. Furthermore, a lot of the cakes are prefabricated in a factory, and
all the apprentices and journeymen need to do is to take them out of the refrigerators,
place them on trays and put them in ovens. These are activities which disqualify
220 Doing Things with Things
the apprentices, and which makes it difficult for them to pass the journeyman
examination. The alternative, according to the journeyman, is to let the apprentices
work and learn in smaller bakeries which cannot afford new technology nor to employ
many journeymen. It is in these bakeries that apprentices will become responsible,
and consequently learn from taking part in the everyday practice in the bakeries. It
is worth noticing that several of the apprentices and journeymen stressed that less
technology in the bakeries would provide the apprentices with better opportunities
to learn the trade of baking. Furthermore, in the interviews it was emphasised that
the prefabricated cakes were of a lower quality than the handmade ones. According
to bakers and apprentices, the prefabricated cakes contained more artificial sugar
and were much dryer than handmade cakes because they had been frozen and were
supposed to last for longer periods of time.
The positive aspects of introducing new technology were also outlined. The
amount of work was less, and it was possible for journeymen and apprentices to
have a more flexible work schedule, as they were able to produce more and faster.
Obviously, it is easier to take out cakes from refrigerators and put them on trays,
than actually bake the Danish pastry itself. On the other hand, the reduction in the
workload was only temporary, within the last year five journeymen had been made
redundant in the bakery.
However, it is crucial to stress that to understand the learning processes of
the apprentices, it is important to focus on issues of economy, rather than merely
addressing issues of epistemology. One has to examine closely the relationship
between the bakery (the context) and the market in order to understand how different
skills are developed or extinct, and how this influences the lives of the bakers.
To sum up briefly, I have given three different examples of how the economic
discourses influence the learning process of the apprentices both directly and
indirectly. First, by the way aspects of economy influenced the organisation of the
assignments in which the apprentices take part. Secondly, by working on products
they need to sell, and finally, by focusing on how new technology disqualifies the
skills of the baking trade.
My conclusion for the first section of the article is that the economic aspects are
a crucial part of what the apprentices learn. As mentioned in the beginning of the
article, traditional concepts of learning do not include issues of economy. However,
in the following I will try to introduce another way of addressing issues of learning,
which includes aspects of economy.

Empirical Examples of Different Kinds of Orientation

In the following I will argue that we need to take another stand on how we approach
issues of learning in which aspects of economy are included rather than excluded.
Inspired by a situated perspective to learning, my focus will be on how matters
of identity are a pivotal phenomenon when approaching learning in practice. As a
consequence of how aspects of economy and learning were inseparable, the baker
 Learning to Do Things with Things 221
apprentices developed different kinds of orientation to being-in-the-world-as-bakers.
A central concept in this connection is the concept of orientation. Before outlining
the specific kinds of orientation central to the bakers, I will say a few words about
the concept of orientation.
One way of conceptualizing issues of economy in relation to learning is to focus
on the different kinds of trajectories of participation, which are arranged in the
various bakeries for how the apprentices can participate and, consequently, what is
relevant to learn. Following this line of thinking, it is important to conceptualize how
the personal trajectories in the bakeries which the apprentices themselves arrange,
and the different kinds of orientations develop in relation to the various trajectories
of participation. The concept of orientation tries to capture how apprentices develop
different ways of living with the different contradictions and conflicts in the bakery,
often related to issues of economy. Aspects of economy are not just something which
happens to the apprentices. By using the concept of orientation it is stressed that this
is about how the apprentices live their lives in general, and it also relates to how the
apprentices act in and across other contexts as well (e.g. in their spare time, how they
organize their family life, etc.). A similar concept that could have been used is the
concept of strategies, which, however, often becomes associated with only cognitive
processes. Furthermore, the concept of orientation indicates that the subjects are
aiming at realizing certain goals, though not always completely clear about what
these goals are. Working towards and trying to realize these goals in their everyday
lives is what constitutes the meaning of the ‘here-and-now’ activities in which the
subjects are involved. Finally, the concept of orientation has a social basis as they are
related to the activities in other contextual settings.
This section outlines the three different and most dominant kinds of orientation
found in the material. However, at the present moment the analysis of the material
is not concluded, so other kinds of orientation may be disclosed. The three kinds of
orientation are the craft-dominated orientation; the social/materialistic orientation,
and, finally, the educational orientation (inspired by Wilbrandt, 2002; Tanggaard,
2003). Even though, these orientations are outlined very categorically, they do not
exist in the ‘pure’ fashion developed below. Fortunately, real life is more complex.

The craft dominated orientation

The craft dominated stand is orientated towards the traditional virtues of being a
baker. There are actually different versions of this orientation. The primary orientation
is directed at becoming a master oneself one day. We found this implicitly stated in
different ways in the interviews with the apprentices, however, it was most clearly
formulated by a journeyman. The journeyman had been working in several different
bakeries, one of the reasons being that he wanted to expand his skills as a baker with
the explicit aim of owning a bakery himself. He said:

You become increasingly more skilled everywhere you go (...) and it is because I have this
dream of having my own bakery that I have to see as many things as possible (Thomas).
222 Doing Things with Things
The journeyman had developed a strategy of working in a bakery for a year or so and
would then leave the place in favour of another. In order to advance his skills, and
consequently obtain his own shop, in the end he had to work his way through a lot of
bakeries. The journeyman gave us an example of how he himself actively organized
his learning process in relation to the practice he was part of with regards to a future
participation. This way of learning the skills to become a master in one’s own bakery
is learnt by ‘going waltzing’ (to change learning places), which is the traditional model
for learning a craft. Within this orientation we find a clear scepticism in relation to
learning something at trade schools. According to this orientation, the only way to
learn is by being an apprentice or journeyman in different kinds of bakeries. The craft
dominated stand is broadly orientated toward the trade in general, and those who hold
this view are ready to move to other bakeries in order to learn more.

The social/materialistic orientation

The social/materialistic stand is orientated towards the wage, the need for good
colleagues and good working hours. Paradoxically, the transformation of the bakeries
into larger units of production with significantly more use of technology gives way
to the social/materialistic orientation. The larger industrialized bakeries have more
employees, which offers the possibility of meeting more colleagues. Furthermore,
within the larger and technologically advanced bakeries it is no longer necessary for
bakers and apprentices to start very early in the morning to begin production. Much
of the bread and cakes are prefabricated, and it is only necessary to take them out of
boxes, place them on trays and roll them into ovens, which automatically switch on
in the morning before anybody begins work. This orientation is more dominated by
other considerations than the actual craft. This stand is orientated towards having a
successful private and family life and living a normal social life, where one does not
sleep when other family members are awake and engaged in social activities. The
social/materialistic orientation comes closer to regarding work as industrial work.
However, there is a tendency to accept that maintaining the craft is the duty of the
trade schools, and something that cannot be done in the bakeries, so some resign
when it comes to upholding the standards of the craft in the local environment. Those
who share the social/materialistic stand are orientated towards other bakeries if they
provide better opportunities for a higher wage, better working hours, etc.

The educational orientation

The educational stand is orientated towards pursuing more education. For many
bakery apprentices being a baker is not the ultimate goal. The education as a baker
is merely a step on the way to another kind of education or profession. Several of
the apprentices I spoke to would like to become chefs, however, they could not find
a restaurant that would accept them as apprentices, so they chose the bakeries as the
second best opportunity hoping to become chefs later on. Other apprentices regarded
 Learning to Do Things with Things 223
being a baker apprentice as the first step to becoming an engineer within the food
area. One said about his future career:

Well, yes, what if I don’t want to be a baker, but would like to work in a bread factory or
be a consultant or just not use the training practically, but would rather move on, then it is
nice to know a little bit more about the material’s characteristics and so on (Poul Erik)

As the quotation shows, this apprentice had another perspective on what it meant
to be an apprentice in a bakery. He perceived it as a temporary process, providing
him with interesting insights that could be useful in his own career. The educational
stand is not involved in the local bakery, but has a more distanced perspective on
what happens in a bakery. The trade school and courses are important because they
potentially provide access to other types of education.
As argued so far in the examples from my study on learning in the bakery, the
economic discourse plays a significant role regarding issues of learning. Taking aspects
of economy into consideration when addressing issues of learning it becomes clear
what was possible for the apprentices to learn. In this article I have primarily focused on
what was possible for the apprentices to learn in relation to aspects of economy. Only
one case described how market economy and new technology disqualified learning.

Discussion
I will argue that it is not a coincidence that the economic aspects of human existence
are neglected in discussions of learning. As shown in the discussion of the Cartesian
influence on what we consider as learning, this frame of thinking provides us with
limited space in order to comprehend issues like learning in practice.
The epistemological perspective on learning becomes an idealized version of
human existence in which desires, emotions and interests are left out of the picture.
This also leads to the assumption that learning and education are in themselves
positive things, which nobody should question. To question and discuss learning in
relation to power, economy and various interests is tantamount to blasphemy.
In this article focus has been on apprenticeship because it is so obvious that
economy and learning are closely interwoven. Furthermore, it has been argued that
aspects of economy are not merely something which determines what it is possible
for the apprentices to learn. The apprentices integrate the economic conditions into
their life in general. However, one could argue that learning and economy are closely
interwoven in formal institutions like schools as well. Just to mention one example:
To place twenty-five students in each class with one teacher is not motivated by
learning theory, but by economic considerations. Only by addressing an economic
perspective do we understand why and how such arrangement have been organised
as the central mode for education in modern culture.
Another central feature when introducing issues of economy into discourses on
learning is the notion of contradiction as a central dynamic for processes of learning.
When approaching learning from a traditional perspective, all aspects of conflict seem
to vanish since learning appears to be a process of acquiring the different kinds of
224 Doing Things with Things
knowledge embedded in the various contexts. However, once we introduce issues of
economy, then different social interests become central and we see how contradictions
become a key motivator for learning.

References

Coy, M.W. (1989). Apprenticeship. New York: State University of New York.
Descartes, R. (1996) Om metoden. Copenhagen: Gyldendal.
Dreier, O. (1999). Læring som ændring af personlig deltagelse i sociale kontekster.
In K. Nielsen & S. Kvale (Eds.), Mesterlære – læring som social praksis.
Copenhagen: Hans Reitzels Forlag.
Gudeman, S. & Rivera, A. (1990). Conversations in Colombia: The domestic
economy in life and text. Cambridge: Cambridge University Press.
Kvale, S. (1996). InterViews – An introduction to qualitative research interviewing.
Thousand Oaks, CA: Sage.
Kvale, S. (1997). Research apprenticeship. Nordisk Pedagogik, 17, 186–195.
Lave, J. & Wenger, E. (1991). Situated learning: Legitimate peripheral participation.
Cambridge: Cambridge University Press.
Lave, J. (1997). Learning, apprenticeship, social practice. Nordisk Pedagogik, 17,
140–152.
Nielsen, K. (1999). Musical apprenticeship: Learning at the Academy of Music as
socially situated. Skriftserie for Psykologisk Institut, Århus Universitet.
Nyeri, J.C. (1988). Tradition and practical knowledge. In J.C. Nyeri & B. Smith
(Eds.), Practical knowledge: Outlines of a theory of traditions and skills, London:
Routledge and Kegan Paul.
Schön, D.A. (1987). Educating the reflective practitioner – Toward a new design for
teaching and learning in the professions. London: Jossey Bass.
Tanggard Pedersen, L. (2002). To læringsbaner ind i elektronikken. (In press).
Tanggaard, L. (2003) Læringsbaner igennem elektronikken. In K. Nielsen &
S. Kvale (Eds.), Praktikkens læringslandskab (pp. 161–173). Copenhagen:
Akademisk Forlag.
Webster’s Third New International Dictionary. (1968). Springfield, Massachusetts,
USA.
Wilbrandt, J. (2002.) Vekseluddannelse i håndværksuddannelser. Lærlinges oplæring,
faglighed og identitet mellem skole og virksomhed. Uddannelsesstyrelsens
temahæfteserie, nr. 14.
Woolfolk, A. (2001). Educational psychology. Boston: Allyn and Bacon.
 Chapter 12

Urban Makings and the Formalisation of

Informal Settlements
Gustavo Ribeiro

This chapter looks into patterns of urbanisation that are marginal to formal1
development undertaken by the public sector or by private developers. In doing so it
identifies two types of urban making:

1. An urban making which is implemented through unplanned, collective

intervention in the environment. This is effected by individuals performing
intentional actions, like going to work, to the bus stop, or to the stream,
gathering in public spaces, playing, etc. In the process of performing such
actions, people are collectively contributing to the creation of networks of
footpaths (beaten tracks) and urban spaces (Ribeiro, 1997). This form of
urban making is a continuous event which stretches through months/years
and transcends individual intentions. It is characteristic of shantytowns, which
dodge the control of official agencies and which have basically to rely on
the limited human and economic resources available within each settlement.
But such process rarely occurs in isolation, without the influence of external
actors. Even a shantytown entirely built with the effort of its inhabitants will
be pressured to interrelate with its urban surroundings. This may occur through
the action of authorities or private landowners and may lead to changes in the
settlement and not rarely to evictions.
2. Processes of improving the conditions of shantytowns and of incorporating
them into a broader pattern of urbanisation involve regulating and formalising
such settlements to various degrees. The intervention of public institutions,
foreign donors and Non Governmental Organisations (NGOs) introduces
organisational structures and formal mechanisms for implementing projects,
which affect people’s interaction with their environment and the quality of
urban spaces and of individual dwellings. Urban making through direct action

1 The term formal refers to urban development, which is planned, regulated or financed
through conventional/authorised institutions/channels (such as central or local government,
banks or credit institutions etc.). The term informal refers to developments outside those
channels (see Drakakis-Smith, 1982 for a discussion of the terms formal and informal).
226 Doing Things with Things
by individuals and the implementation of long term patterns of urbanisation
as unplanned events are replaced by mediated urban making through the
intervention of external actors and the introduction of formal planning
procedures.

This chapter is based on a study of low-income settlements in Bangkok and Chiang

Mai (Thailand) which have undergone formalisation, through the implementation of
projects targeted at supporting people and improving their living conditions. Such
projects are carefully designed to promote the participation of local inhabitants
at all levels: from problem formulation, through project design and budgeting, to
implementation.
These participatory mechanisms support processes which involve urban
interventions, but which are primarily aimed at the education and empowerment of
low-income communities. Still, besides visible impacts towards strengthening low-
income groups and besides attempts at promoting community-led social changes,
new urban conditions and patterns of interaction between inhabitants and settlements
are emerging as a result of these processes.
The central argument of this paper is that processes of education and empowerment
facilitated by a public organisation and a NGO cannot be seen in isolation from
changes in the physical environment effected through daily social practices and
through the very implementation of the above-mentioned processes.
Another important argument presented here is that formalisation in processes
of urban production raises issues about the role of design and planning in relation
to mechanisms of public participation. And to the extent that upgraded or resettled
neighbourhoods are presented as improved alternatives to informal settlements, it
becomes particularly relevant to draw comparisons between the two. This paper
looks at unplanned, planned and hybrid forms of urbanism and the processes of
making and design which they engender.

Background

In 1996 the Danish Government, through its environment and international aid
agency, The Danish Co-operation for Environment and Development (DANCED),2
provided the Thai Government with a US$ 1.3 million grant for a project targeted at
the improvement of environmental conditions in urban areas occupied by low-income
communities. The Urban Communities Environmental Activities (UCEA) project

2 DANCED was established in 1984 with the objective of contributing to the greatest
possible extent to the protection of the environment and nature in developing countries. With
the ascension of Venstre (a right wing liberal party) to power in November 2001, international
development aid by Denmark was severely cut down. In addition, the environmental concerns
which figured very high on the political agenda of the Social Democratic Party, were relegated
to a second plan. DANCED was done away with and its projects were placed under the Danish
Development Aid (DANIDA).
 Urban Makings and the Formalisation of Informal Settlements 227
(1996-2002) follows, as the name suggests, DANCED’s overall policy of placing
local communities at the centre of environmental programmes and giving them
responsibilities for decision making, design and implementation. (Boonyabancha,
1999, p.103)
The public institution co-ordinating the implementation of UCEA in Thailand,3
The Community Organisation Development Institute (CODI) has set up, since its
inception in 1992, a programme which aims both at improving the living conditions
of the urban poor communities and at strengthening their organisational capacity,
through the promotion of community savings and credit groups4 (UCDO, 2000,
p.1).
In its role as a support mechanism CODI has stood as a catalyst in, and
arguably as an initiator of, a process of social change. Such a process, which aims
at promoting ‘a large scale community-driven development movement,’ places the
decision making and managing responsibilities largely on community networks and
community groups.5
In line with CODI’s approach, the Urban Communities Environmental Activities
(UCEA) project focuses on a process where environmental improvement is not an end
in itself, but a means for promoting social change (Ribeiro et al, 2000). In addition,
the environment is attributed a highly transitory role for two main reasons: (a) the
focus of UCEA is on activities as tools for people’s education and empowerment;
(b) the low-income communities in question are under the threat of eviction by the
local authorities.
The above outline places the environment of squatter settlements (dwellings,
canals, public spaces) in a condition, which is not simply that of a setting for

3 On the 28th July 2000, the so called Urban Community Development Office (UCDO)
was merged with the Rural Development Fund (RDF) and a new public institution was created,
namely, the Community Organisation Development Institute (CODI).
4 In order to achieve that, CODI relies on a revolving fund which is made available to all
urban poor groups who organise themselves to apply for loans for their development projects.
According to a recent survey, by 2000 ‘over half Thailand’s 2,000 urban poor communities
in 50 provinces… linked together into 103 networks through a broad range of community
development activities, including housing, income generation, environmental improvement,
community enterprise and welfare’ were members of CODI. (UCDO, 2000, p.1)
5 CODI relies on collective instances represented by community groups and
increasingly by community networks to achieve a type of management and decision making
which aim at representing the interests and wishes of socially and economically excluded
groups. In that way, CODI aims at providing alternatives to decision-making initiated by
authorities (local or central), by learned professionals (specialists, academics), by NGOs and
by individuals with illegitimate power within communities (e.g. community leaders [called
Nakleng] involved in criminal activities, such as drug traffic), (see for instance Phongpaichit
and Piriyarangsan, 1994). CODI reports occurrences of misuse of power and funds within the
sphere of its activities (UCDO, 2000, p.12), but has documentation to support the argument that
community-lead processes are self-regulatory and that mismanagement and corrupt practices
will be counteracted in time within community groups and networks. (UCDO Update, 2000)
228 Doing Things with Things
activities neither that of a mere object of intervention. The environment appears
first and foremost as a means for educating and organising people. Making in this
context is not only production (e.g. the rebuilding of a canal embankment) but also
and centrally the constitution of social processes (organisation, empowerment and
education).
The present chapter investigates this multiple role of the environment in UCEA
as means for social change, as object of intervention and as setting. It raises questions
about the process of design involved in the implementation of environmental
improvements in selected low-income settlements in Chiang Mai and about the role
of such environmental improvements in future interactions between people and their
environment.

Squatter Settlements

China is the weed in the human cabbage patch. … The weed is the Nemesis of human
endeavour… Of all the imaginary existences we attribute to plant, beast and star the weed
leads the most satisfactory life of all. True, the weed produces no lilies, no battleships, no
Sermons of the Mount … . Eventually the weed gets the upper hand. Eventually things fall
back into a state of China. This condition is usually referred to by historians as the Dark
Age. Grass is the only way out … . The weed exists only to fill the waste spaces left by
cultivated areas. It grows between, the poppy is maddening – but the weed is rank growth
… : it points a moral. (Henry Miller quoted in Deleuze, 1999, p. 19)

Squatter settlements in Thailand, follow a pattern which is found elsewhere in the

world, and normally occupy publicly owned land that has remained marginal to
formal processes of urbanisation; that is, land with poor accessibility, marshy areas,
flooded sites along canals, etc. The poor occupies such land, through strategies of
survival, with minimal resources and means of construction, which may include
recycled timber, plywood panels, bamboo mats, corrugated tin roofing sheets,
amongst other materials, to create low-cost shelters.
Squatter settlements have an inherent power to change and adapt, for their
investment is minimal in the creation of shelters – a power to fold and unfold in
accordance to changing political conditions. These architectures and urbanisms of
‘necessity’ operate with extremely limited material conditions and challenge the
involvement of professionals such as planners and architects in the provision of
cheap, large scale housing: at the very moment the architect or planner steps into
this process an increase in the costs of shelter production already occurs and their
very involvement may rapidly become an unaffordable commodity!
Of special importance to this paper is the fact that the formation of squatter
settlements involves, simultaneously with the creation of a physical environment,
the constitution of social structures. The act of making the settlement consists not
only of generating products: footpaths, streets and dwellings – as would be the case
in the building of a new urban area. Making is in itself a social process, involving
 Urban Makings and the Formalisation of Informal Settlements 229
the interaction between people and the use of urban spaces. Appropriation of public
spaces is therefore tightly related to their production.
The production of physical structures invites in turn a different approach to that
of consolidated urban areas. Instead of investigating permanence and persistence of
urban structures, we are presented with changes, instead of morphology, patterns of
use.

UCEA

‘UCEA aims at changing interactions and power relationships between organisations,

networks, groups and individual stakeholders in a complex setting,’ through a focus
on environmental management. (DANCED, 1999, p.9)
Public participation in environmental projects aims amongst other things to create
ownership of interventions (sidewalks, bridges, etc.) by the community involved.
By actively contributing to a project, from decision-making, through design, to
implementation, the community will, thus, be in a better position to appropriate the
project as its own and to look after its maintenance. UCEA could be seen as a radical
example of that approach, through its focus on environmental activities instead of
environmental products.
UCEA gives priority to the process of learning and to strengthening the
organisational capacity of communities (Boonyabancha, 1999, p.102, Ribeiro et al,
2000, p.3). An example of that is the project for the restoration of the embankment of
Mae Ka Canal in Chiang Mai where it crosses Kham Phang Ngam settlement.
Mae Ka Canal used to be very narrow at Kham Phang Ngam, causing floods
during the rainy season. Access to and from the community was cut off and children
were prevented from going to school. The community decided therefore to widen
the canal in order to channel the excess of water. The community planned and
executed the restoration of the embankments of Mae Ka canal, but it was destroyed
by new floods. After a second restoration, the embankment was destroyed once more
(Ribeiro et al, 2000, p.16).
The case of the reconstruction of the Mae Ka Canal embankment was taken up in
an interview with the director of CODI and overall responsible for the implementation
of UCEA, Ms. Soomsook Boonyabancha, and her point was that the priority in that
as well as in other UCEA projects was the process of learning by the local people.
The fact that the community identified a problem, provided a solution, implemented
it and then learned from their mistakes was the most important thing. The actual
physical result was of secondary importance. (Ribeiro et al, 2000, p.3)
That approach raises questions about how professionals such as architects may
contribute to a participatory design process. It is clear from the above example that
UCEA gives priority to activities, as heuristic, empowering gestures, over products.
Making in the environment is an event whose main goal is to promote learning and
social change in the long term.
230 Doing Things with Things
Activities can be seen in the context of UCEA as methods for actualising the
community, for the empowerment of people, for promoting participation, social
networking and learning in the community. Such an approach reflects CODI’s
overall policy:

CODI helps people to play a role in the development of their own areas. People see,
people learn and they can believe they can do it. [CODI is interested in] the processes that
will get the essence of the people out (Boonyabancha quoted in Ribeiro et al, 2000, p.4).

UCEA’s focus on process rather than on physical products seems to highlight

the dynamic and transitory nature of the informal settlements in question. Such
settlements as physical structures are not likely to survive as such, rather, they may
quite probably be removed to another site. The process of learning and the building
up of self-confidence is something that the people will be able to take with them6.
In Tung Pattana, a squatter settlement comprising 30 houses built on stilts along
the Mae Ka Canal in Chiang Mai, UCEA has supported through a $2,500 grant,
the implementation of a boardwalk designed and built by the community, without
the assistance of architects or engineers (Asian Coalition for Housing Rights, 1998,
p.29). Before its construction, residents had to negotiate their way to their dwellings
by ‘wading through the water or hopping along bamboo poles strung between the
houses’ (Asian Coalition for Housing Rights, 1998, p.29).
With its ingenious system of concrete posts and movable wooden panels which
can be bolted at different levels according to the level of the water in the Mae Ka
Canal, this boardwalk has become a point of reference for those involved with
community development projects in Thailand as well as internationally. Compared
with Tung Pattana’s boardwalk the walkway in the community of Bang Na, made of
bamboo reinforced-concrete resting on layers of compacted garbage, has been met
with rather less enthusiasm by visitors (Asian Coalition for Housing Rights, 2000,
p.11).
When drawing such a comparison, those supporting these projects draw our
attention to the fact that we should look beyond the products themselves ‘into the
complicated circumstances which yielded these walkways, … [to] find a gold mine
of learning material in both settlements: one devastated by AIDS, lack of jobs and
basic services, and many-times displaced by eviction; the other facing environmental

6 Seen in the context of architectural design, such an approach has led to a conception
of environments or buildings primarily as sites for social activities, as settings for action
and not as finished compositions for aesthetic contemplation. That approach can be found
in the architecture of Joao Filgueiras Lima (Ribeiro G Sarah Network in Arkitekten No 22
September 2001 pp 16-23; Latorraca G (2000) Joao Filgueiras Lima. Lisbon: Blau; Lima JF
(1999) CTRS. Sao Paulo: Sarahletras.), who has developed flexible pre-fabricated complexes
for the Sarah Network of Hospitals in Brazil and in the work of Lina Bo Bardi who based
her architecture on a careful investigation of popular productions (Ribeiro G. Lina Bo Bardi:
Times of Rudeness in Arkitekten No 4 February 2000 pp 2-9; Ferraz M (1993) Lina Bo Bardi.
Milan: Charta).
 Urban Makings and the Formalisation of Informal Settlements 231

Figure 12.1 Tung Pattana Boardwalk (Photo: Jørgen Andreasen)

Figure 12.2 Tung Pattana Boardwalk (Photo: Jørgen Andreasen)

hazards, flooding, eviction threats and hostile officials.’ (Asian Coalition for Housing
Rights, 2000, p.11) It is proposed that both ‘projects are rich in lessons in making
innovative, strategic investments in establishing the right way to stay, mobilising
scant resources, knitting communities together, making decisions democratically,
232 Doing Things with Things
negotiating with hostile city administrations, managing money… doing many, many
things beyond the act of simply upgrading access.’ (Asian Coalition for Housing
Rights, 2000, p.11)
Besides the importance of the process, which has led to the construction of the
raised walkways in both settlements, it is undeniable that such interventions have
radically changed the conditions of accessibility to individual dwellings. In addition,
they have provided a new habitable surface and a social space, where before there
was none. Seen in relation to the process that has led to their production, such
boardwalks stand at the point of their conclusion, as a clearly defined goal and,
one would be inclined to believe, as an element which plays a relevant role in the
motivation of the members of the community in mobilising their efforts.
These boardwalks, as urban elements, as spaces which allow for uses such as
accessibility and social interaction, can hardly be downplayed in their role as physical
products. Their potentials to be maintained by the residents can be measured by the
extent to which they support basic daily actions; actions which will wear them down,
but which contain, at the same time, the key towards their sustainability.
Also relevant here is the fact that the designers and builders are local people with
established social practices that involve the physical environment before and after the
construction of the boardwalks. Their design is therefore invested with such social
practices together with the limits and possibilities offered by physical structures.
If we shift for a moment to the case of a self-built shantytown on firm ground
(like that of Vila Paranoa in Brasilia), the spatial elements which allow for access
and for the performance of social activities are there from the start. The primary
appropriation of the site is effected by creating beaten tracks on the overgrown
field (Ribeiro 1997). Its maintenance as an urban site is effected through the very
performance of activities. People vote with their feet, they walk, meet and play in
certain places and on those spots the ground is cleared from vegetation.
In the case of Tung Pattana and Bang Na the site is not yet appropriated, periods
of flood and draught prevail and impose therefore conditions of access to individual
dwellings as well as the performance of activities between them. The project of
raised walks in those communities can be seen as primary gestures of appropriation
of the site, which perform a role in the transformation of urban spaces.
In that sense, a dissociation between processes and products and a promotion
of the former over the latter may be seen as an a posteriori discourse which risks
misrepresenting the role and performance of the walkways in Tung Pattana and
Bang Na.

Collective Action in the Environment

The aforementioned study of Vila Paranoa investigates how collective action in the
environment led to patterns of urban change and to the generation of urban landscapes
(Ribeiro 1990, Ribeiro 1996, Ribeiro 1997).
 Urban Makings and the Formalisation of Informal Settlements 233
Such collective action occurred in the course of about three decades, from
1960s to early 1990s, in the case of the shantytown Vila Paranoa. It consisted of
people moving between different places of interest (home – bus stop, home – lake,
neighbour – market, etc.). That movement generated a beaten track on the ground
and defined a network of footpaths. Such network of footpaths is at the source of the
layout of urban spaces of Vila Paranoa (Ribeiro 1990, Ribeiro 1997).
Collective action in that case is the result of individual intentional action and will
to perform particular tasks. It is therefore self-motivated and without the interference
of external agents. The will and initiative to act by the members of the community
is integrated with the performance of daily actions, such as going to work, washing
clothes, visiting friends, shopping. It is therefore a self-supporting/sustainable
system, which is at work in the creation and development of these types of squatter
settlements. The intentional action of each individual will be superimposed in the
course of time producing an urban landscape which is not the contribution of a single
individual, but of a community.
The process outlined above has been described by some authors, in my view
misleadingly, as spontaneous (e.g. Chifos & Yabes, 2000 or Epstein, 1973). But
reference to unplanned urban growth as spontaneous masks the fact that such processes
are motivated both at individual and collective levels. People’s comings and goings
and related impacts on the environment, such as the creation of footpaths, can, as
argued elsewhere (Ribeiro 1997), be understood in the context of intentional actions
– not the intention of creating urban spaces, but the intention of reaching the bus stop
to catch the bus or the intention of going to the market to do the daily shopping of
groceries. Merleau-Ponty refers to such intentions as practical intentions:

… when I move towards a world I bury my perceptual and practical intentions in objects
which ultimately appear prior to and external to those intentions, and which nevertheless
exist for me only insofar as they arouse in me thoughts and volitions. (Merleau-Ponty,
1962, p.82)

The creation of urban spaces may be described as the overlapping of individual

actions in time. But it is, in fact, more than the accumulated effect of individual
actions, for each action creates a footprint in the environment and such footprints,
or changes are incorporated in the new environment. Just as collective action can
be understood as a dynamic process which unfolds in time, so can the environment
itself. Such a dynamic conception of the environment has, for instance, been explored
by McCabe and Balzano in their writings on event cognition (McCabe & Balzano,
1986).
The changed environment or the environment under continuous change plays a
dynamic role in subsequent patterns of interactions with users (Ribeiro 1997, Ribeiro
1996). The more people walk along a footpath, the more they wear it down, the more
consolidated it becomes as a track as opposed to the ground surfaces alongside the
footpath.
234 Doing Things with Things
Through that perspective, the concept of collective action is understood as the
succession of individual actions which do not simply appear in isolation but which
become related because they change the environment and because the changing
environment plays a decisive role in the unfolding of future interactions and thus
affect how further physical changes take place.

Participation in Development Projects

When considering people’s participation in development projects, as is the case of

UCEA, we have a different state of affairs. The prime motor for collective action
is not within the community, but outside it, it is namely the agencies that initiate
development projects. This is inevitable as a project is conceived prior to action.
People’s participation in such conditions, already presupposes the setting up of
a priori frameworks which provide an environment for people’s involvement in
projects. Participation as a project event can be contrasted with patterns of usability
(see Ribeiro 1990, and Ribeiro, 1997) which create the layout of squatter settlements
as illustrated earlier.
Participation can be seen as project related, collective action. In that sense,
collective action occurs within the framework of a project. Such a project precedes
collective action, and by creating space for its implementation it is also imposing
limits to how such an action may occur.
A project adopting a participatory approach is still faced with the crucial step
of becoming self-sustainable (sustainability being one of the main goals of UCEA
(Boonyabancha, 1999, p.109, URDI, 1998, p.4, CODI, 1998, p.16). Project co-
ordinators talk about the importance of elaborating exit strategies, where facilitators
ensure that the activities they have promoted are sustained by the communities when
the project formally comes to a conclusion. In order for that to happen, the project
needs to stop being an externally initiated event and turn into a self-motivated
social practice based on the will and resources of a community as well as on the
collaboration of local authorities.
Projects such as UCEA can be seen through the perspective of formalisation of
processes of production and reproduction of urban settlements. The integration of
such levels of formalisation into the daily life of people and into informal practices
is, through this perspective, central for reaching sustainable development within
such communities.
Although it is not the aim of this paper to assess the sustainability of UCEA,7 it
can be said that besides establishing partnerships with local authorities, its potential
to become integrated in the daily practices of the community will be of crucial
importance in the coming years, if the project is to fulfil one of its central aims.8

7 The reader is referred to UCEA evaluation reports (eg URDI, 1998).

8 This chapter discusses sustainability as a parameter introduced by UCEA. It does
not deny the fact that sustainability is in itself a problematic concept, which needs to be
 Urban Makings and the Formalisation of Informal Settlements 235
A focus on the process dimension of UCEA projects denotes a reference to the
duration of the activity of production (from problem formulation, through design, to
making) which may overlook the fact that use and maintenance are just as constitutive
of the intervention as the process of production itself. In the case of self-built
shantytowns, as pointed out earlier, such a distinction between product and process
is challenged to the extent that use involves continuous production and reproduction,
respectively through change and development of existing urban structures on the one
hand, and through their maintenance on the other (Ribeiro, 1990; Ribeiro, 1997). In
its turn, the consolidated environment is constantly promoting new uses.
CODI’s discourse promoting processes over products has two notable
consequences:

1. It downplays the role of environmental conditions (space layouts, building

materials, landscape elements, etc.) in affecting patterns of use and
appropriation. It can be argued that a built element (bridge, walkway,
embankment) will acquire a particular role already to the extent that is has been
constituted as an object of intervention in the context of a project; from problem
formulation by a community (e.g. limited access) to the implementation of a
design proposal (e.g. a walkway). Through collective action, a proposal will
be reflected upon, discussed, designed and built. But as it stands as a realised
product it becomes integrated in the development of future actions. Bourdieu
makes the following comment about this constitutive gesture: ‘in taking a
point of view on the action, withdrawing from it in order to observe it from
above and from as distance, … [the knowing subject] constitutes practical
activity as an object of observation and analysis, a representation. (Bourdieu,
1977, p.2 emphasis on the original). Here, we see a marked contrast to spatial
processes of production and reproduction in self-built shantytowns, which
cuts across such a constitution of discrete objects, and which in my view are
best described as the non-meditated perception of affordances as the basis for
action (Gibson, 1979, Ribeiro, 1996).
2. In addition, CODI’s conception seems to overlook the dimension of
maintenance through use, that is, the incorporation of the physical intervention
into the daily life of the community. It downplays, according to the argument
presented here, an important dimension of the sustainability of a project.

UCEA, in its conception as a forum for learning, resists the accommodation of

qualitative assessments of environmental improvements from a technical and
functional points of view. This is clearly visible in the implementation of the Mae
Ka Canal embankment reconstruction at Kham Phang Ngam.
Continuity and sustainability of environmental projects seems to be formulated
in CODI primarily through a focus on participatory structures, educational/

critically examined. Such a critical discussion is presented by Peter Marcuse (1998) in his
article entitled ‘Sustainability is not enough’.
236 Doing Things with Things
empowerment forums, networks, which come into being and are maintained precisely
through the performance of organisational activities. Changes in organisational
structure, networking and education rely on people’s continuous commitment. In that
way, continuity seems to be defined at an organisational level, overlooking the role
of physical interventions/constructions and their uses. If, for instance, a particularly
engaged community leader moves to another city, the whole community is affected
by the absence of that individual. To be sure, community organisation and leadership
play key roles in environmental sustainability, but so do daily social practices, which
are closely integrated with physical structures, their production and reproduction.

References

Angel, S., Archer, R.W., Tanphiphat, S. & Wegelin, E. A. (1983). Land for housing
the poor. Bangkok: Select Books.
Angel, S. & Amtapunth, P. (1987). The low-cost rental market in Bangkok. In
PADCO/NHA, The land and housing markets in Bangkok: strategies for public-
sector participation. Vol. 3, Technical Report 6.
Asian Coalition for Housing Rights. (2000). Face to Face: notes from the network on
community exchange, January 2000.
Asian Coalition for Housing Rights. (1999). Housing by People in Asia: Newsletter
of the Asian Coalition for Housing Rights, No 12, April 1999.
Asian Coalition for Housing Rights. (1998). Housing by People in Asia: Newsletter
of the Asian Coalition for Housing Rights, No 11, April 1998.
Boonyabancha, S. (1999). The urban communities environmental activities project
and its environmental fund in Thailand. Environment and Urbanization, 11(1),
April 1999.
Bourdieu P (1977). Outline of a theory of practice. Cambridge: Cambridge University
Press.
Chifos, C. & Yabes, R. (2000). Southeast Asian urban environments: Structured and
spontaneous. Arizona: Arizona State University Press.
CODI (1998). Completion report: Urban Communities Environmental Activities
(1996–1998). Unpublished Report.
DANCED (1999). Urban Community Environmental Activities (UCEA) Project,
Thailand, Phase 1. DANCED Ref. No. M123-0057. Draft Evaluation Report.
DANCED (1998). Participation in environmental assistance: DANCED Reference
Note.
Deleuze, G. (1999). A thousand plateaus: capitalism and schizophrenia. London:
Athlone Press.
Epstein, D.G. (1973). Brasilia, plan and reality: a study of planned and spontaneous
urban develpment. Berkeley: University of California Press.
Ferraz, M. (1993). Lina Bo Bardi. Milan: Charta.
Gibson, J.J. (1979). The ecological approach to visual perception. Boston: Houghton-
Mifflin.
 Urban Makings and the Formalisation of Informal Settlements 237
Latorraca, G. (2000). Joao Filgueiras Lima. Lisbon: Blau.
Laursen, J.H. (2001). Miljø fra top til bund. Politiken 1 Section, Sunday 11th February
2001.
Lima, J. F. (1999). CTRS. Sao Paulo: Sarahletras.
McCabe, V. & Balzano, G.J. (Eds.) (1986). Event cognition: An ecological
perspective. Hillsdale, New Jersey: Lawrence Erlbaum.
Marcuse. P. (1998). Sustainability is not enough. Environment and Urbanization,
10(2), October 1998, 103–111.
Merleau-Ponty, M. (1945). Phenomenologie de la perception. Paris: Gallimard.
Merleau-Ponty, M. (1962) Phenomenology of perception. [Colin Smith, trans.]
London: Routledge.
Nørgaard, L. (2001). En mand med overblik. Arkitekten, No 8, April 2001.
Phongpaichit, P. & Piriyarangsan, S. (1994) Corruption and democracy in Thailand.
Bangkok: The Political Economy Centre, Chulalongkorn University.
Ribeiro, G. (2001). G Sarah Network. Arkitekten, No 22 September 2001, 16–23.
Ribeiro, G., Diaw, K. & Andreasen, J. (eds.) Community development in Thailand:
The case of Community Organisation Development Institute (CODI) and the
DANCED supported UCEA project. Report on a visit to Bangkok and Chiang
Mai in November 2000. Under ENRECA Project: Community Initiatives and
Democratisation of Planning (unpublished manuscript).
Ribeiro, G. (2000). Lina Bo Bardi: Times of Rudeness. Arkitekten, No 4 February
2000, 2–9.
Ribeiro, G. (1997). An ecological approach to the study of urban spaces. Journal of
Architectural and Planning Research, 14(4), 289–300.
Ribeiro, G. (1996). Acting, dwelling and squatting: an ecological approach to the
relation between person and urban environment. Ecological Psychology, 8(2),
131–151.
Ribeiro, G. (1990) .Perception and design in the urban environment. PhD Dissertation,
University of York.
UCDO. (2000). UCDO Update: A publication of the Urban Community Development
Office in Thailand, October 2000, no 2.
URDI (1998). Summary report of the evaluation of Urban Community Environmental
Activities Project.
Yap, K.S. (Ed.) (1992). Low-income housing in Bangkok: A review of some housing
sub-markets. Bangkok: Division of Human Settlements Development, Asian
Institute of Technology.

Acronyms

DANCED – Danish Co-operation for Environment and Development

UCEA – Urban Communities Environmental Activities
CODI – Community Organisation Development Institute
ACHR – Asian Coalition for Housing Rights
This page intentionally left blank
 Name Index

Agre, P., 114 Holzkamp, K. 11, 132, 133

Akrich, M. 147, 148, 154 Houkes, W. 2, 3, 21, 24, 34, 41, 42
Ariew, A. 18 Hume, D. 172
Hutchins, E. 5, 52, 113, 114, 194-196
Bærentsen, K. 114
Barker, R., 121 Ingold, T. 89
Baron-Cohen, S. 52, 62
Bourdieu, P. 88, 235 Jensen de Lopez, K. 4-6, 89-91, 96, 103-106
Bowker, G. 7, 173 Joerges, B. 1, 51
Brown, S. 159, 196, 198, 201, 203
Bugrimenko, E. 70 Karpatschof, B. 1, 95
Keller, K., 134
Clark, A., 113, 119 Kendell-Scott, L. 4, 90
Cole, M. 89, 93, 148 Kirsh, D. 113, 119, 120
Costall, A. 4, 39, 51, 54, 68, 89, 90 Kitcher, P. 18
Cummins, R. 17, 26 Kvale, S., 224

Davies, P. 18 Lakoff, G. 95, 180

Descartes, R. 178, 179, 211, 219, 223, 224 Latour, B. 159, 179, 184, 185
Dipert, R. 21-23, 31 Lave, J. 9, 114, 212
Douglas, M. 171 Law, J. 7, 147, 151, 154-159
Dreier, O. 7, 9, 132, 133, 152, 197 van Leeuwen, L. 68
Dreyfus, H. 37, 194 Lemmonier, P. 130
de Léon, D. 5, 40, 114, 120
Engeström, Y. 114, 187 Leontiev, A. 53, 54, 132
Leslie, A. 52, 63, 67, 71, 83
Forchhammer, H. 6, 8, 31, 132, 134 Leudar, I. 39, 194
Foucault, M. 134 Lucy, J. 95, 104
Luria, A. R. 89
Garfinkel, H. 170
Gibson, E. J. 68 Markussen, R. 8, 167, 169, 180
Gibson, J. J. 4, 30, 45, 46, 68, 89, 90, 235 Marx, K. 132, 176-179, 182, 188
Goffman, E., 142 McLaughlin, P. 15, 19, 21
Goodwin, C. 195 Merleau-Ponty, M. 37, 105, 233
Goodwin, M. 195 Middleton, D. 8-10, 198, 201
Millikan, R. 2, 17-22, 26, 31, 32, 71
Hacking, I. 171 Mol, A. 7, 147, 151, 154, 155, 156
Haraway, D. 8, 152, 168, 174, 175, 177-181,
188, 189 Neander, K. 32
Heft, H. 79, 82 Nielsen, K. 9, 10, 224
Hegel, G. W. F. 176 Norman, D. 113
Heidegger, M. 37
240 Doing Things with Things
Olesen, F. 8, 169, 173, 180 Suchman, L. 30, 37, 39, 40, 147
Szokolsky, A. 4, 8, 72
Piaget, J. 3, 67, 69, 70, 87
Petroski, H., 125 Tomasello, M. 51, 54, 69, 87-89, 93, 104, 106
Pickering, A. 152, 167
Preston, B. 2, 18, 19, 29, 30, 32, 33, 37, 42, Ungerer, J. 55, 56, 70
43, 51, 54
Valsiner, J. 53-56, 58, 60
Reed, E. S. 72, 75, 82 Vermaas, P. 2, 3, 21, 24, 42
Ribeiro, G. 10, 11, 121, 225, 229, 230, Vygotsky, L. 3, 12, 69, 70, 72, 88-90, 95,
232-235 101, 104

Sapir, E. 104, 105 Walsh, D. 18

Schatzki, T. 7, 11 Wenger, E. 212
Schiffer, M. B. 51 Wertsch, J. 3, 121
Searle, J. 30, 45, 46 Whorf, B. 95, 104
Sinha, C. 88-93, 98, 103, 105, 106 Williams, E. 4, 51, 54, 90
Smitsman, A. 68 Winner, L. 149, 152, 153, 155
Sørensen, E. 7, 154
Star, S. 7, 173 Zukow, P. 73
 Subject Index

activity theory 114, 178 ecological psychology 4, 45, 67, 68, 82, 90
accountability 203, 205 economy 9, 176, 209-224
actor-network theory (ANT) 2, 127, 147, 159 education 7, 9, 10, 148, 209, 211, 214-218,
affordances 4, 46, 68, 69, 72, 73, 78, 79, 81- 221-228
83, 88-90, 92-94, 105, 106, 235 embodiment 30, 37-41, 68, 90, 105, 119
agency 37, 52, 152, 154, 167, 178, 189 epistemology 209-212, 218-220
apprenticeship 9, 10, 209-224 everyday practice 8, 10, 133, 170, 175, 220
archaeology 1, 127
archiving 8, 9, 193, 198-207 fluid ordering 154
Aristotle on matter and form 175 formal organization 225
artefact function 2, 31-35, 37, 39, 41, 43, 45, function 2, 3, 15, 16-26, 29-47, 53, 55, 60,
47, 127 62, 70, 78, 88-90, 120, 127, 132,
autism 4, 52-63 137, 172

beaten tracks 10, 11, 225, 232, 233 information processing approach 193-194
intentions 2, 15-26, 30-37, 42-44, 47, 68,
canonicality 91-93, 98 88, 104, 127, 161, 181
categorisation 116-118
classification 7, 8, 32, 171, 173, 188 language 51, 87-106, 170, 179
co-evolution of things and practices 169 learning 9, 10, 53, 54, 62, 89, 90, 105, 113,
cognitive biography 114, 115, 128 136, 209-224, 229
cognitive congeniality 6, 119, 124, 125, 127 linguistic relativism 104-105
cognitive ethnography 114
collective action 232-234 maintenance 6, 11, 23, 122, 158, 206, 212,
computer mediated communication (CMC) 232, 235
193 materialism 175-179
computer mediated communication 193, materiality 161, 168, 170, 172, 175, 179,
198, 200 187, 189, 193
conflict, 140, 144, 158, 203, 205, 223 material-semiotic systems 175
context 3, 10, 17, 39, 136, 170, 171, 173 means-end relation 36
conversation analysis 195 multifunctionality 19, 22, 31, 82, 93, 105,
coordination 63, 71, 72, 81, 169 137, 138, 144
craft 221
creativity 23 nature 176, 178, 180
critical psychology 132, 178 non-discursive ordering 8, 9, 196, 201, 202,
customer’s perspective 218 204, 206-7
normativity 36
design 12, 13, 18, 20, 25, 31-36, 40, 47, 51-
54, 57, 157, 161, 171 objectification 133, 139
disability 6, 142, 143, 145 ordering, 151
discursive psychology 8, 193 organizational remembering 204
dualism 1, 62, 176
242 Doing Things with Things
participation 212, 221, 222, 226, 234 signs 69, 104, 105, 179, 182
Piagetian theory 3, 67, 69, 70, 87 situatedness 30, 37, 39, 40, 88, 90, 142, 167,
plans and planning 2, 3, 29, 30, 33-47, 147, 171, 174, 180, 188, 189, 212
193 sociality 8, 193, 198, 200, 201
Plato on the designer and user 15 squatter settlements 228
play 4, 6, 7, 55, 62, 67-83, 134 subject-object dualism 1, 62, 176
politics of things 152 sustainability 11, 232, 234-236
proper function 2, 3-5, 7, 15-26, 31, 32, 43, system function 2, 15-26
55, 81, 82
properties and their relation to things 2, 7, 8, technology 5, 127, 132, 133, 140, 143, 149,
87, 95, 167, 171-175, 188 152, 167, 196, 220, 223
prototype 19-21, 23, 25, 26 teleology 172
purpose 21, 22, 24, 31, 47, 82, 132, 168, 172 theory of mind 3, 4, 52, 62, 67
thing, definition of 171-2
rationality 3, 34, 38, 215
regional ordering156 urban making 225 et seq
rehabilitation 135, 141, 142, 145 use plans 3, 35, 36, 37, 41-47

science and technology studies (STS) writing 8, 167-170, 174

selection 2, 17-21, 81
sequencing 197 Zapotec spatial words 96-97