Bendik Manum

Apartment Layouts and Domestic Life;

The Interior Space and its Usability
A Study of Norwegian Apartments Built in the Period 1930 - 2005
© Bendik Manum, 2006

ISBN: 82-547-0205-5
ISSN: 1502-217X

CON-TEXT
Dissertation #26

A Doctoral dissertation submitted to:

The Oslo School of Architecture and Design

PUBLISHER:
The Oslo School of Architecture and Design

COVER:

PRINTED BY:
Unipub AS / AiT E-dit 2006
Abstract

The background for this study is Norwegian housing where floor plans of
new apartments seem to differ significantly from what previously has been
built as well as from what architects have considered good quality. The study
consists of two empirical surveys. The first is a diachronic analysis
examining the development of apartment layouts since the 1930s, while the
second is a synchronic interview-based survey where different apartments
identified through the diachronic study are examined as dwellings for
contemporary living.

The features of apartments particularly examined in the diachronic analysis

are the sizes of rooms and the spatial configurations of the apartments, two
features that are decisive for the degree of generality concerning functions or
use. These features of space have been analysed in a sample of 150
apartments built in Oslo since 1930. The conclusion of this analysis is the
identification of three generations of apartments. Apartments of the first
generation, which was common until about 1955, were general with respect
to sizes of the rooms as well as to the spatial configuration. Around 1960,
there was a change towards larger apartments and functional specificity. In
the second generation of apartments, those that were typical in the period
from the 1960s until the early 1980s, the individual rooms were highly
differentiated in size and positioned in accordance with their very specific
function. Since then, the number of rooms has decreased and the spatial
layout has become simpler, the kitchens are now usually in the living room
and the bedrooms have become smaller. These apartments, which are the
third generation, are specific with respect to use in that the bedrooms are
rooms for sleeping while the “living and kitchen room” is the place for all
daytime living.

1
The three generations of apartments defined by the diachronic analysis are
not just a theoretical classification of floor plan layouts but also a typology
that captures features relevant for real domestic lives. Sizes of rooms and
configurational aspects of the interior spaces are decisive for what kind of
households that lives in a particular apartment as well as for how they use
their rooms. A conclusion from the interview-based survey is that generality
works; the first generation of apartment, the apartment characterised as
general due to large “second largest rooms” and a spatial layout where all
rooms have access directly from the entrance, is the kind that houses the
largest range of households. This is very different from the apartments being
built now, which are appropriate only for a limited range of households.
Since they rarely have more than one place for daytime living, they are
unsuitable for the many households where daily lives consist in simultaneous
and not easily co-existing activities.

Where theory and methodology are concerned, the field of architectural

research named “space syntax” has been a basis for figuring out the subject to
examine as well as for carrying out the analyses. This study not only
illustrates how space syntax can be useful for identifying patterns across a
sample of dwellings, but also how the configurational features of space
captured by the space syntax methodology are relevant for households
preferences of dwellings and for their daily living.

2
Preface

This thesis comes out of an interest in apartment layouts born while studying
architecture in the late 1980s. When observing the Norwegian dwellings built
at that time, they seemed to differ from what previously had been considered
good architectural quality. This was particularly the case when comparing
them to the tradition of Scandinavian modernism as represented by Bengt
Espen Knutsen, who in his esteemed architectural practice as well as in his
teaching has emphasised to create spaces for living rather than buildings as
esthetical objects. An aim of this thesis has been to develop a knowledge-
based contribution to a discussion about layouts and qualities of dwellings.
Two theoretical and methodological references have been essential in order
to develop a subject of interest into a research question and in order to figure
out relevant empirical inquiries; one is the Scandinavian tradition of survey-
based housing research, the other is the methods for analysing architectural
space developed within the field of architectural research named space
syntax. This positioning of the work does not imply that other aspects of
housing or alternative theoretical and methodological approaches would not
be interesting, only that it would have led to different studies than the one
carried out.

I am grateful to The Oslo School of Architecture and Design for my

education here, for inspiring periods with part-time teaching in the years
thereafter and most of all for giving me the opportunity to carry out the study
that is summarised through this thesis. I will thank Halina Dunin- Woyseth
for important support at crucial stages of the work and for guiding me to
Björn Klarqvist and his PhD-symposiums at Chalmers Institute of
Technology. Klarqvist’s teaching of space syntax provided a basis for this
study. I am also very grateful to Julienne Hanson at the Bartlett, University
Collage London, for her highly skilled guidance in space syntax analyses of
dwellings. Where more local condition are concerned, I must thank Jon Guttu
for his advices about housing research in theory and in practice and
supervisor Bjørn Sandaker for his support and for his reading of drafts that
were rarely on schedule. I must also thank Espen Rusten and Paul Benze for

3
managing to transform my description of a space syntax tool into the very
easily applicable software AGRAPH.

This thesis has an empirical basis that I alone would never have managed to
establish. I am therefore very grateful to Husbanken and to OBOS for their
financial supports that have made it possible to get highly skilled students’
assistance. I would first like to thank Erlend Torkildsen and Cleas Cho Heske
Ekornås for all their hours spent in the archives of the local authorities as
well as for their computer skills applied in digitising and organising the
hundreds of drawings. I must then thank Ida Lenander, Line J. Musæus,
Maja Fjøsne and Håvard Brevik for their excellent work in carrying out the
interviews. I am also very grateful for the hospitality of all those letting us
into their private home and participating in the interviews. When it comes to
the finishing stages of the work, I am grateful to Ann Giæver for her patient
attempts to improve my English and to Jonas Adolfsen for handling the
numerous illustrations that were originally in all kinds and qualities of digital
as well as analogue formats.

Finally, I am thankful for the love and patience of AnneLise, Birk and
Jarand; a support that this study could never have been done without.

Oslo, September 2006

4
CONTENTS

ABSTRACT ...................................................................................................1

PREFACE ......................................................................................................3

CONTENTS...................................................................................................5

1 INTRODUCTION ......................................................................................8
1.1 PROBLEM ................................................................................................8
1.2 BACKGROUND .........................................................................................9
1.2.1 The Layout of Apartments...............................................................9
1.2.2 The Standard of Living and Housing............................................14
1.3 CONTENT AND STRUCTURE OF THE THESIS ............................................19
2 FOCUS AND LIMITATIONS ................................................................20
2.1 A REVIEW OF HOUSING RESEARCH ........................................................21
2.2 VOID AND SOLID ...................................................................................38
2.3 SPATIAL CONFIGURATIONS ...................................................................40
2.4 GENERALITY VERSUS SPECIFICITY ........................................................45
2.5 DWELLINGS AND TIME PERIOD ..............................................................54
2.6 THE STANDARD OF HOUSING .................................................................55
3 METHODOLOGY...................................................................................57
3.1 THE QUANTITATIVE AND THE QUALITATIVE; COMMENTS ON DISCIPLINES
AND THEORIES ............................................................................................58
3.2 THE RESEARCH STRATEGY ....................................................................61
3.3 THE CHOICE OF CASE ............................................................................63
3.4 BASIC SPACE SYNTAX METHODOLOGY ..................................................66
3.5 THE SPACE SYNTAX ANALYSIS OF THIS STUDY ......................................82
3.5.1 The Kind of Space Syntax Modelling............................................82
3.5.2 Identifying the Spatial Units .........................................................83
3.5.3 The Analysis, Applying the Software AGRAPH............................85
3.6 COMMENTS ABOUT ASSUMPTIONS, FINDINGS AND GENERALISATIONS ..89
3.7 SUMMARY .............................................................................................92

5
4 APARTMENT LAYOUTS; A DIACHRONIC INQUIRY...................94
4.1 THE CASE AND THE SAMPLE ..................................................................94
4.2 THE DATA; COLLECTING, EDITING AND ANALYSING FLOOR PLANS ........97
4.3 RESULTS ...............................................................................................99
4.3.1 Sizes of Rooms and Apartments....................................................99
4.3.2 Spatial Configurations................................................................103
4.3.3 Conclusion; Three Generations of Apartments ..........................110
5 SPATIAL LAYOUT AND CONTEMPORARY LIVING; A
SYNCHRONIC SURVEY ........................................................................114
5.1 THE SURVEY .......................................................................................114
5.1.1 The Sample .................................................................................115
5.1.2 The Interviews and the Questionnaires ......................................121
5.1.3 Response.....................................................................................123
5.1.4 Questions and Answers...............................................................124
5.2 THE RESULTS ......................................................................................126
5.2.1 General.......................................................................................126
5.2.2 Spaciousness of Living................................................................127
5.2.3 Kinds of Households...................................................................131
5.2.4 Time Spent in Different Rooms ...................................................139
5.2.5 Rooms and Activities ..................................................................142
5.2.6 Changes Made by the Residents .................................................154
5.2.7 Preferences about Alternative Layouts.......................................160
5.2.8 Some Additional Findings ..........................................................166
5.2.9 Summary of the Synchronic Survey ............................................169
6 THE CONTEMPORARY DWELLINGS ............................................171
6.1 BACKGROUND; OBOS THROUGH 75 YEARS, THE THREE GENERATIONS OF
APARTMENTS ............................................................................................171
6.2 A COMPARATIVE CASE ........................................................................172
6.3 THE NEW LAYOUT ...............................................................................177
7 FINAL REFLECTIONS ........................................................................182
7.1 THE HOUSEHOLDS IN THE NEW DWELLINGS ........................................182
7.2 THE EXTENT OF THE RECENT CHANGES ...............................................183
7.3 MARKETING; THE CONCEPT OF BRANDING ..........................................184
7.4 SOME SCENARIOS ................................................................................186
7.5 DWELLINGS AND GENERALITY ............................................................188
8 FURTHER RESEARCH .......................................................................190

6
APPENDICES

A.3.1 “AGRAPH; Software for Drawing and Calculating Space Syntax

Connectivity Graphs” (appended paper)

This software is free download from http://www.aho.no/ahograph/.

A.4.1 List of the projects examined in the diachronic survey.

A.4.2 Space syntax connectivity graphs, all apartments.
A.4.3 Some room sequences, all apartments.
A.4.4 Results from space syntax calculations, all apartments.

A.5.1 Floor plans and key information about the projects examined in the
synchronic survey.
A.5.2 The interview-forms, an example.
A.5.3 Floor plan drawings with furniture, selected apartments.

NOTE:
The appendices listed above were originally in A4-format. The floor plans in
the printed appendices are therefore not in original scale. For floor plans in
original scales, please se the separate appended volume.

Appendix (as a separate volume):

”OBOS 1930–2005, et snitt gjennom norsk bolighistorie”

This appendix is a “catalogue” presenting the housing projects and the

apartments examined by the diachronic analysis.

7
1 Introduction

1.1 PROBLEM

During the recent 15 years, urban housing has developed into a major section
of Norwegian building construction. This is due to attractive urban sites
being made available after relocation or closing down of industry and
institutions, rising economic wealth and an increasing interest in urban life.
As a result of this development, planning and design of apartment buildings
have recently become a major field of the Norwegian architectural practice;
architects are at the moment designing apartments that will be the future
homes for a huge number of people. These new apartments seem to differ
from earlier ones concerning basic architectural features such as floor plan,
sizes of rooms and daylight conditions. Parallel to the extensive construction
of new urban dwellings and the change in apartment layouts, the households
and their preferences have become more diversified. Two questions can
therefore be raised. The first is about what is going on: if the layout of
apartments has changed significantly, what explicitly are these changes? The
other questions relate to domestic life and to the quality of housing: if the
existing Norwegian apartments are diverse in terms of floor plan layout, how
do the different layouts relate to households and their daily living? By
studying these issues, it might be possible to shed some light on how our
future stock of dwellings, a stock that will include all apartments under
construction and planning now, will correspond to the needs and preferences
of the population.

8
1.2 BACKGROUND

The following two sections elaborate on the background for this study; the
first describes the layout of recently built apartments while the second
comments upon contemporary conditions of living and housing in more
general terms.

1.2.1 The Layout of Apartments

The interest in housing has been changing in both intensity and character
throughout the history of architectural practice. Studies of housing
contributed to the development of the Modern Movement within architecture
between the World Wars. This was a period when architects were deeply
involved in housing politics as well as in housing design. One aim of
architecture was to simplify the trivialities of life and to reflect use was
considered an essence of beauty. Le Corbusier stated that ”A great epoch has
begun. There exists a new spirit. ... we shall arrive at the ”House-Machine”,
the mass-production house, healthy and beautiful in the same way that the
working tools and instruments which accompany our existence are
beautiful.” 1 Even though Le Corbusier’s house-machine has been seriously
questioned as a vision for living, later generations of architects have not
rejected basic ideals of modernism such as comfort, light and usability, from
being essential features of a dwelling.2 However, if we compare recently built
apartments with these ideals of the modernist tradition in housing, the floor
plans now appear to be less determined by concern about actual living.
Architectural features of interior space that in the modern tradition have been
essential and where recently built dwellings seem to differ are indoor daylight
conditions, the size and shape of the rooms and how the different rooms of a
dwelling are positioned in relation to each other.

As far as daylight conditions are concerned, the debate in Norway during the
1930s was whether bathrooms and staircases without windows, and thereby
without daylight and views to the outside, were acceptable. In studies of the
loss of daylight caused by increasing building depths, the range of depths

1
Le Corbusier at the 5th C.I.A.M. congress in Paris in 1937. (Buhl, 1965)
2
This is very different from housing in larger scales in the sense that contemporary ideals about the latter
clearly differ from the early modernist vision of large freestanding buildings in suburban park-like landscapes.

9
then being evaluated was between 7 and 12 metres.3 In new apartments of
today, kitchens without direct views to outside are common; most kitchens
are appendices to the living room, appendices where the only daylight comes
through the living room. As for the depth of the buildings, housing blocks are
now rarely as slender as 12 meters; the housing project shown in figure 1.1
has a building depth of about 17 meters. Regardless of what is attempted
through the design of walls and windows, such a deep building block implies
much floor area where it is hard or impossible to achieve good daylight
conditions.

As regards the size and shape of the individual rooms, at least two aspects are
worth comment. Firstly, living rooms of new apartments often have the
character of “space left over after planning”; living rooms appear as
“leftovers” after minimum versions of the other rooms have been positioned.
This happens in several ways; one is that outside corners of adjacent rooms
appear as boxes into the living room; another is that the positioning of
openings towards adjacent rooms limits the potential use of the living room.
Figure 1.1 and figure 1.2 show two apartments where the “living-and-
kitchen-room” has this character. Secondly, while the diversity of use and
preferences with respect to domestic life is increasing4, the rooms of the
dwelling seem to be moving towards becoming more mono-functional in the
sense that the living room is the only room for daytime living while the other
rooms are tiny and appropriate only for specific and predefined use.

When it comes to the positions of the rooms in new apartments, a common

layout is main bedrooms accessible only through the living room. In order to
access the bathroom from the main bedroom in such an apartment, a mother
would have to pass through both the living room, where her teenage daughter
might be entertaining friends, and the entrance, where dirty winter boots are
likely to be slipped off. Figure 1.2 shows such an apartment. Given the option
of choosing among otherwise equal conditions, these features of a floor plan
layout would hardly be the first choice.

3
This was an important issue for debate within the Norwegian architecture profession in the 1930s, see for
instance Bedre boligtyper – bedre boligbygging” (Øvergaard,1938), Olika husbredder och deras
bggnadskostnader (Markelius, 1935) or other articles in the magazine Plan (1933 and 1935).
4
Preferences about housing and living are strongly influenced by cultural and national background. Due to the
various traditions with respect to the roles within the families and to activities such as cooking and dining, the
preferences about apartment layouts are diverse. The increasing number of persons with a non-western
national background implies an increasing diversity in the preferences of living among the population. Another
phenomenon that makes the preferences more diverse is the increasing variety of households caused by the less
dominant position of households permanently consisting of two parents and the children they have in common.
(See tables 1.3.a and 1.3.b.)

10
Figure 1.1
“Badebakken”
(1999-2002)

Figure 1.2
“Pilestredet Park”
(2002-2006)

Figure 1.3
“Stranden”
(1990)

11
The fact that the features mentioned above are trivial makes their frequent
occurrence in new dwellings remarkable. What is even more remarkable is
the lack of debate within the architectural profession about this issue. Most
remarkable is the fact that architecture containing very strange interior spaces
are not only rarely criticised5, but that such projects are formally described as
excellent by the architectural profession. The project “Stranden” on Oslo’s
best seafront site was not only presented in the magazine “Byggekunst”
without criticism6, it has also been honoured by “Anton Christian Houen
Fund’s Certificate for Outstanding Architecture”, a prize considered to be
Norway’s most prestigious architectural reward.7 Figure 1.3 shows the floor
plan of an apartment in this building. The size and shapes of the area by the
entrance as well as the bedrooms are corridor-like and seem to be accidental
consequences of struggling with the fact that large parts of the floor areas are
distant from windows and direct daylight. It is hard to imagine that this
project would have been acknowledged if the interior space of the apartments
had been the subject of evaluation.

The fact that a housing project like “Stranden” is awarded, does not imply
that critical remarks about recently built dwellings do not exist. In fact,
critical remarks are not hard to find informally. The problem is that the
disadvantages of many new dwellings rarely are pointed out explicitly and in
public. Much of the professional skill and knowledge of practicing architects
has the character of being implicit or tacit.8 When tacit knowledge is no
longer familiar to the majority of a profession, at least two problems arise.
First, it is hard for those not having the skills to be aware of this. Second, for
those who suspect the existence of relevant knowledge that they do not have
themselves, this knowledge is hard to find. Within the profession of
architects, such a lack of previously familiar tacit knowledge now seems to
be the case where housing and dwellings are concerned. Skills of designing
dwellings in general and floor plan layouts in particular, skills taken for
granted by earlier generations of architects, seem to have been neglected or
lost. It might be that previous knowledge and skills, which focused on
improving the standard of housing and were concerned with many trivial
aspects of usability, is now out of date, but, as argued above and in the
following section, this does not seem to be the case.

5
One of the few critical comments is “Acceptera inte” (“Do not accept”) by Klarqvist and Thiberg (2003).
6
The project was published in Byggekunst 1991: 4. Byggekunst is the main Norwegian architectural magazine,
published by NAL, The Norwegian Architects Association
7
“Anton Christian Houens fonds diplom for god arkitektur , …Norges mest prestisjetunge arkitekturpris. ---
De prisbelønte byggverkene utgjør i dag en del av den nasjonale kulturarven. Samlet representerer de det
ypperste som er skapt av arkitektur i Norge i det forrige århundre – til inspirasjon for det som skal bygges i
det neste.”, quote from Grønvold (2000), pointing out the superiority of the architecture honoured by this
reward.
8
For those reading a Scandinavian language, the tacit knowledge of the architectural profession is elaborated
by Ulf Janson (1998) in his thesis on the practice of the Swedish architect Jan Gezelius.

12
Even though housing design is still an important task for architects today, it is
a field where other professions also have a strong influence; housing
developers employ estate agents as consultants and pay as much attention to
them as to the architects concerning basic design questions such as floor plan
layouts. When a particular layout is the best seller, this layout is likely to be
regarded as the most profitable one. This assumption is not necessarily
correct as households that intend to buy an apartment, rarely imagine kinds of
dwellings other than those in existence and being offered for sale. The fact
that a particular kind of apartment is sold most easily, does not prove that it is
as desirable as possible; it just indicates that it is the one most preferred
among the apartments on offer. Apartments of other and unknown layouts
might very well sell better. Due to the influence of other professions, the lack
of excellent new apartments is not solely the responsibility of the architects.
However, given that we as architects aim to improve the situation, we should
rather focus on issues within the field of architecture as long as we can
achieve improvements here. The assumption that this is possible is the
background for writing this thesis.

13
1.2.2 The Standard of Living and Housing

Concerning the contemporary standard of Norwegian dwellings, there are at

least two conceptions that are worth comments. One is about the standard of
housing in general and political interpretations from this, the other is about
the term and the subject “need for housing” in particular.

The contemporary standard of Norwegian housing and its political

implications was a main subject at a seminar at NTNU (The Norwegian
Institute of Technology) in June 2004. It was here argued that the quality of
housing and living conditions in Norway was the best in the world and that
the housing politics of the government had therefore been successful.9 This
argument is not convincing. Firstly, the quality of housing in terms of
technical installations and floor area per person, which are the features
measured in order determine “the standard of housing”, does not necessarily
correspond to the population’s quality of living in more general terms.
Secondly, even if we accept that the Norwegian quality of life is excellent
due to high quality of dwellings with respect to floor areas and technical
standards, this does not indicate any success in recent Norwegian planning
and politics as far as housing is concerned.10 The liberal housing market,
which was the policy argued for, is a new phenomenon in Norway. It has
developed since the early 1980s and it represents ideals for building, buying
and ownership of dwellings very different from those of the preceding post-
war policy. Less than one third of the Norwegian dwellings built after the
Second World War, have been built since 1985. If there is any Norwegian
housing-policy that should be given the credit for a high standard of living, it
is therefore not the recent one but the very different, governmentally planned
housing production that went on up to the 1980s.11

9
This was particularly pointed out by Roger Iversen from “Kommunaldepartementet”, which is the
Governmental section responsible for housing policy.
10
A reason for the high quality of Norwegian housing in terms of technical standard and floor areas per person
is the few dwellings of very low standard. An explicit aim of Norwegian post war politics has been to reduce
and prevent bad housing, an aim that until the early 1980s was carried out by means of legislation and by
norms to be followed in order to achieve cheap governmental loans. These loans were given gy Husbanken,
the Norwegian governmental bank for housing. See Husbanken (1985 and 2000) about their terms and
guidelines for design.
11
Since 1985, 450,000 dwellings have been built in Norway, while 1,150,000 were built between 1945 and
1984. (Numbers of dwellings according to Statistisk Sentralbyrå (Statistics Norway). The conservative
government of Willoch came to power in 1983. Due to the time required for planning and constructing housing
projects, 1984/1985 is here set as the time limit when counting the dwellings built.)

14
A conference in 2003 about Norwegian housing research, illustrates the
perception of needs for housing in Norway.12 The fields pointed out to be of
importance, were sustainability (in terms of pollution, energy consumption
and economy of building materials), the cost of building construction and the
dwelling as representing personification and identity. Several key speakers
explicitly stated that there was no longer any “need” for housing in Norway.
This perception, which is a pleasant one for politicians as well as for planning
authorities, can be questioned. Within a liberal market, “need” is understood
as potential buyers` demands. In a fully liberal housing market, which is now
more or less the case in Norway, a particular household’s need for a better or
a larger dwelling is not noticed unless the household enters the purchasing
field. If we understand the need for housing to be nothing but the demand in
terms of such an interest in buying, and determine the standard of living by
average floor area per person, groups of the population might have serious
personal needs for better housing conditions without its being noticed at all.
If “need for housing” should include the actual living conditions and personal
wishes of real households, then the information relevant for determining the
current standard of living is more than the average floor areas pr. person.

In order to get some information about contemporary households and their

living conditions, a minor survey was carried out at an early stage of this
study. The survey was based on questionnaires and the participants were the
pupils of two classes at the School of Lambertseter in Oslo. Table 1.1 (at the
end of this section 1.2) shows the sizes of the households (by number of
persons) and the sizes of the apartments (by number of rooms) for the pupils
of these two classes at the School of Lambertseter in 2001. Lambertseter is
one of the early suburbs of Oslo; it was under construction from the mid-
1950s and until the 1960s. The population of Lambertseter is today a mixture
of people who have been living there since the 1950s and -60s and many who
have grown up there and moved back in addition to people who have moved
there without living there before, representing a mixture of cultural and
national backgrounds. Even though the data is not statistically representative,
due to the limited sample, they describe the “spaciousness of living” for
households at Lambertseter with schoolchildren in 2001. As shown in table
1.1 the number of persons per room was 1.18 on average, which represents
remarkably cramped living in Norway. By comparison, the average number
of person per room in Norway was 0.6 in 2001.13 The living conditions for

12
This was a conference summing up a Norwegian research program on housing, at Hotel Bristol, Oslo, in
December 2003.
13
According to Statististisk Sentralbyrå / Statistics Norway, http://www.ssb.no/fobbolig/tab-2002-09-23-
01.html, table 1, page 1.

15
households with children at Lambertseter in 2001 was more cramped than the
average for Oslo 50 years ago, the latter being 1.1 person per room in 1958.14

From this brief look at the results of the “Lambertseter-survey” and other
simple statistics of households and living conditions, we can draw at least
two conclusions about standards of living and the “need for housing”. One is
that the complacent attitude about contemporary housing conditions should
be revised. The survey at Lambertseter shows that living conditions for many
households today are similar to what was considered unacceptably cramped
living in Norway 50 years ago; the problem of cramped living has not been
solved, even today; many households need larger dwellings. In the present
fully commercial market, this is not noticed as long as the families living in
cramped conditions cannot afford larger apartments and therefore do not
enter the market as potential buyers. These needs do therefore not influence
the planning or the design of new dwellings. A second conclusion is that
cramped living is far from evenly distributed with respect to national and
cultural backgrounds15; housing conditions have become part of the complex
problem of immigration and physical and social segregation, a problem that
maybe represents the most urgent challenge to western democracies.

When it comes to the discipline of architecture, the situation of contemporary

living and housing described above should influence the subjects focused
upon at all stages from urbanism down to design of particular dwellings. As
far as apartments are concerned, a renewed interest in the layout of the
interior space should be appropriate if we intend to respond to the
contemporary variety of preferences, preferences that include a need for
better housing conditions among all those still living in cramped conditions.
An aim of this study is to contribute to such a renewed interest in the quality
of dwellings with respect to the daily lives necessarily being lived in them.

14
The living conditions in 1958 are described by Brochmann (1961, p. 30).
The same patter is found if comparing with Ås (1971); the percentage of households with children living in
cramped conditions was 44 at Lambertseter in 2001 while it on was 40 in all Norway in 1967, see table 1.2.
15
This fact is not captured by data on average living conditions, but can easily be seen by a look at official
Norwegian statistics that clearly point out the cramped living of non-western immigrants. 66 % of such
households in Oslo that consist of more than two persons live cramped (in the sense having less than one room
per person). Among Western immigrants and non-immigrant households, the percentages are 27 and 18,
respectively. (See http://www.ssb.no/fobinnvbolig/fig-2002-11-12-01.html) The cramped living conditions of
non-western immigrants is not at all surprising when comparing data on households and data on income; the
households of non-western background are larger and have significantly lower incomes than households of
other backgrounds.
(see (http://www.ssb.no/fobinnvbolig/tab-2002-11-12-04.html and http://www.ssb.no/emner/02/01/fobinv/fig-
2002-09-09-01.gif) for the distributions of households (by size of the households) among households with
different background by nationality) and http://www.ssb.no/emner/05/01/inntinnv/tab-2001-08-29-04.html for
incomes by nationalities.)

16
Table 1.1 Spaciousness of living, for the pupils in two classes at Lambertseter School in 2002 (8.class)

Apartm. Persons in the household Rooms Persons Remarks

number Parents C.>16 C. 6-16 C. 0-5 Others >16 Total pr. room I II III
1 2 1 1 4 6 0,67 detatched house x
2 1 2 3 3 1,00 x
3 2 2 4 3 1,33 changed to 4 rooms x
4 2 1 1 4 3 1,33 x
5 2 2 1 5 3 1,67 x
6 1 1 2 4 0,50 x
7 2 5 7 3 2,33 very well being x
8 1 2 3 4 0,75 x
9 2 2 4 4 1,00 x
10 2 1 2 5 5 1,00 x
11 2 1 1 4 4 1,00 x
12 2 2 1 5 4 1,25 x
13 2 1 1 4 4 1,00 x
14 2 2 1 5 3 1,67 x
15 2 2 4 3 1,33 x
16 2 2 4 4 1,00 x
17 1 1 1 3 3 1,00 x
18 1 2 3 3 1,00 x
19 2 2 1 5 3 1,67 x
20 2 1 3 6 5 1,20 detatched house x
21 1 1 1 3 3 1,00 x
22 2 2 4 4 1,00 detatched house x
23 2 1 1 4 2 2,00 x
24 1 1 2 2 1,00 x
25 1 2 1 1 5 5 1,00 (married brother) x
26 1 2 3 3 1,00 x
27 2 1 3 5 0,60 x
28 2 1 3 1 7 4 1,75 x
29 2 2 1 5 3 1,67 x
30 1 1 2 4 0,50 detatched house x
31 1 2 3 4 0,75 x
32 1 2 3 4 0,75 x
33 1 1 2 2 1,00 x
34 1 1 2 2 1,00 x
35 1 3 1 5 3 1,67 x
36 2 1 1 4 3 1,33 x
37 2 3 5 3 1,67 x
38 2 2 4 3 1,33 x
39 2 4 6 3 2,00 x
40 2 4 6 5 1,20 detatched house x
41 2 1 3 6 5 1,20 detatched house x
42 2 3 5 5 1,00 detatched house x
43 2 2 4 5 0,80 x
Persons per room (on average) 1,18 Numbers 19 16 8
Percetages 44 37 19
Numbers 97 56 27
When detached houses not included 1,23 Percetages 54 31 15
Number of rooms includes bedrooms and living rooms but NOT kitchen
Rooms smaller than legal minimum size is not included (apartment nr. 24 og nr.3)

Legend of "spaciousness":
Spaciousness according to Ås (1971, p.19)
I Cramped: number of rooms < number of persons
II Appropriate: number of rooms = number of persons
III Spacious: numberof rooms > number of persons

17
Table 1.2 Spaciousness of living
(according to categories of Ås (1971), see table 1.1)

Spaciousness of living (percentages of households)

Table Households
1.2 All households Households with
with children schoolchildren
Norway 1967 Oslo 1967 Lambertseter 2002
Narrow 40 31 44
Appropiate 22 29 37
Spacious 38 40 19

Reference Ås (1971, p.19) Ås (1971, p.21) Table 1.1

Table 1.3.a Households by kinds of households. 1980, 1990 og 2001. Norway.

Kind of household 1980 1990 2001
Numbers
Total 1 523 508 1 751 363 1 961 548
Singles (without children) 425 725 601 095 739 563
Couples (without children) 306 924 340 634 412 611
Single parent with youngest child < 18 years 54 272 88 953 106 987
Couple with youngest child < 18 years 485 386 442 349 452 950
(Total, households with youngest child < 18 years ) 539 658 531 302 559 937
Several adults (see note) 251 201 278 331 249 437
Percentages
Singles (without children) 27,9 34,3 37,7
Couples (without children) 20,1 19,4 21,0
Single parent with youngest child < 18 years 3,6 5,1 5,5
Couple with youngest child < 18 years 31,9 25,3 23,1
(Total, households with youngest child < 18 years) 35,4 30,3 28,5
Several adults (see note) 16,5 15,9 12,7
Note:
these are households with more than one adult, where at least one is not part of a couple
households with children older than 18 living at home and younger children are not included here

Table 1.3.b Households by kinds of households. 1980, 1990 og 2001. Oslo

Table
Kind of 1.3
households 1980 1990 2001
Numbers
Total 222 291 244 440 266 856
1 Singles 101 776 128 272 138 659
2 Couples 46 841 43 638 48 527
3 Singles with children 9 220 13 265 13 755
4 Couples with children 38 021 34 665 41 308
5 "Several adults" 26 433 24 599 24 607
Percentages
1 Singles 45,8 52,5 52,0
2 Couples 21,1 17,9 18,2
3 Singles with children 4,1 5,4 5,2
4 Couples with children 17,1 14,2 15,5
5 "Several adults" 11,9 10,1 9,2

18
1.3 CONTENT AND STRUCTURE OF THE THESIS

The study described in this thesis has examined the spatial layout of
dwellings through rather extensive empirical surveys. The work has thereby
been different from studies that focus on developing theory and apply
architectural objects as examples shedding light on subjects developed on the
theoretical level. The content of this thesis corresponds to the study carried
out, in that theories and methods are not the main subjects but issues
elaborated in order to describe and position the research question and in order
to explain the surveys and the analyses carried out.

The study consists of two main empirical inquiries: one that has examined
how apartment layouts have developed through time and one that has
examined how different apartment layouts relate to households and their
preferences and well-being. The results of these inquiries are a large amount
of empirical data. In order to prevent the main text from becoming too long,
larger tables and the more extensive figures, such as forms from the
interviews, tables and floor plan drawings, are presented as appendices at the
end of the thesis. The total sample of apartments analysed is presented by
drawings and some key information in a volume separate from the thesis; a
catalogue-like volume that documents a historical development of Norwegian
apartments as well as an important segment of the existing Norwegian stock
of dwellings.

The structure of this thesis is conventional. After this introductory chapter 1

explaining the background for the study, chapter 2 and 3 elaborate the
research question and the methodology, respectively.16 Chapters 4 17 and 5
describe the empirical studies and analyse the results, while chapter 6
discusses the findings and draws some conclusions about contemporary
housing design in the light of the findings of the empirical inquires. Based on
the findings achieved, the final chapter 7 reflects upon the context of
contemporary housing and on dwellings in times to come.

16
Some of the content of chapter 3 about methodology has been presented in a paper at the conference
“Methodologies in Housing Research” at KTH, Stockholm in 2003. (Manum, 2003)
17
The main results of the analysis of apartment layouts through time have been presented at the
“5th International Space Syntax Symposium” at TU Delft in June 2005. (Manum, 2005)

19
2 Focus and Limitations

The aim of this study has been to examine how the interior space of
Norwegian apartments has developed through time and, in the light of this, to
compare and discuss different floor plan layouts with respect to the daily
lives of the residents. An issue particularly focused on, is the dichotomy of
generality versus specificity as regards functions or use. The features of
interior space given particular attention, are the sizes of the rooms and the
spatial configurations of the apartments. These features are decisive for
domestic life, and drawbacks regarding them can hardly be compensated by
high quality in other features of the dwelling. Compared to such features as
colours, furniture and technical equipment, which the households might
easily change or upgrade according to their personal preferences at any time,
most spatial layouts are determined by the construction of the building or by
the technical infrastructure in ways that make them harder to alter. The
spatial layout is therefore important simply because of its permanence; the
floor plan layout has long-term consequences, positive or negative depending
on its quality.

There are two principal ways of describing a route. One is to describe where
precisely to go, such as which road to follow and exactly where to turn.
Another is to describe the route relative to its surroundings, such as to go
south of the mountain and not to cross the river. In the latter case, the
elements referred to are not the paths followed but more distant objects that
by being well known or characteristic are landmarks for recognising the
route. The following pages describe this study in both these way, first by a
review of how research-like studies of dwellings have developed over the
past 150 years and then by explicitly describing the subject focused upon in
this study.

20
2.1 A REVIEW OF HOUSING RESEARCH

Needs for dwellings and the importance of interior spaces appropriate for
everyday lives have been major concerns in much of the work referred to in
the following review, in the housing studies of the early modernist as well as
in the Scandinavian surveys of dwelling layouts and housing conditions in
the post-war period. Given that contemporary housing, even in wealthy
Norway, involves not only a question of identity and representation, but still
also of usability and needs, as is argued in the previous chapter, such
previous studies particularly of the layout of dwellings should represent a
basis of knowledge relevant for this study.

The interest in and the analytical approaches to housing and living

conditions, which developed during the first decades of the 20th century,
resulted from the extensive urban changes that had been going on in the most
industrialised countries for about hundred years. During the industrialisation
of the 19th century, the population in European cities grew rapidly.18 Due to
lack of dwellings and of technical amenities such as heating, clean water and
sanitary systems, the living conditions of the new working class were
appalling.19 For various reasons, ranging from fear of revolution to the
industries’ need for healthy workers, housing conditions became an issue of
wide political and social interest in the late 19th century England. This was
the background for the ideals of the “Garden Cities”. In order to handle the
problems of the over-crowded and densely built working-class housing areas,
Ebenezer Howard (1898) proposed decentralised and less densely built cities
inspired by the rural. The concept of “Garden Cities” reflected a belief in
architecture and urban planning as a means of improving health and social
conditions; a way of thinking that during the first decades of the 20th century,
were also brought to apartment buildings and to the individual dwellings and
their interiors.

18
The population of London more than doubled from 950,000 in 1800 to 2,300,000 in 1850 and doubled again
to 4,500,000 in the year 1900. In Norway the population of Oslo grow almost exponentially during the 19th
century, from 9,000 in 1800 to 112, 000 in 1880 and to 227,000 in 1900.
19
This was the subject of “Die Lage der arbeitenden Klasse in England” by Friedrich Engels (1845), English
title: “The Condition of the Working Class in England”.

21
 In Norway, the theologian Eilert Sundt and the medical doctor Axel Holst
were pioneers in studying living conditions. Sundt carried out a series of
studies and several of these addressed particularly poverty in Oslo. (Sundt,
1858 and 1870). Holst (1895) documented that the housing condition for the
workers were unacceptably unhealthy and cramped.20 The concern of Nanna
Broch, who worked for the local authorities in Oslo from 1919 to 1945, was
that low quality dwellings made women’s housework and children’s lives to
a daily struggle. Her series of exhibitions21, which documented the bad living
conditions and argued for technical improvements and better dwellings,
achieved a broad public attention and were important for the political interest
in housing conditions that emerged during the 1920- and -30s. (Johansen,
Berge and Andresen, 1961)

The architects’ concern about housing was not only based on such social
interests but also on more operational approaches. In mass production within
industry, studies of efficiency had been essential in order to design optimal
lines of production. Methods from such studies were transferred to studies of
homes and housing. An example is “Efficient Housekeeping or Household
Engineering, Scientific Management in the Home” by Christine Frederick
(1925), with the sub-title “A correspondence course on application of the
principles of efficiency engineering and scientific management to the every
day tasks of housekeeping.” Her writings were based on studies of the time
spent and the movements made during work in the domestic kitchen. Similar
analyses were carried out by Margarete Schütte-Lihotzky in her development
of “Die Frankfurter Küche”. Her concern was to develop kitchens that were
comfortable and practical places in which to work. According to Noever
(1992), Schütte-Lihotzky was inspired by dining in trains where food for 100
persons was efficiently prepared in kitchens of only 4 m2. A publication by
Schütte-Lihotzky from 1921 had the informative subtitle “Wie kann man
durch richtigen Wohnungsbau der Frau Arbeit ersparen”.22 “Die
Frankfurter Küche” was published as a full-scale kitchen and demonstrated
through movies showing the kitchens in use.23
20
Holst applied the official norm (according to Christiania Sundhetskomission 1893), which defined less than
15m3 space per adult as very cramped living, and found that the living in 17% of workers dwellings were very
cramped and that this percentage was more than 50% in some areas. (Holst, 1895, p. 16-21)
21
The exhibitions were named “Østlandsutstillingen” and went on from 1928 to 1956.
22
This would in English be “How to reduce the work of the housewife by building the right dwellings”.
Margarete Schütte-Lihotzky (1897-2000) was the first female Austrian architect, she won prices for her
designs even before graduation and collaborated with several of the most famous architects of the time (Adolf
Loos, Peter Behrens, Ernst May and Bruno Taut). She was imprisoned by the Nazi-regime and later prohibited
from architectural practice due to membership in the Communist Party. She received the Architectural Reward
of the City of Vienna in 1980 and awarded membership in “Gesellschaft der bildenden Künste Ôsterreichs” in
1997. For more about Schütte-Lihotzky and her work, see Noever (1992) or Hirdina (1984).
23
Schütte-Lihotzky’s kitchen was exhibited at ”Die Frankfurter Frühjahrsmesse” in 1927. After being
presented at the Stockholm exhibition in 1930, “Die Frankfurter Küche” of Schütte-Lihotzky became a basis
for the studies and the design of kitchens takings place in Sweden, studies that were brought back to Middle
Europe termed as “The Swedish Kitchen” after the Second World War. (Hirdina 1984)

22
Figure 2.1 “Reduzierung der Schritte”, a study by Schütte-Lihotzky.

Figure 2.2 Two of Schütte-Lihotzky’s kitchens.

Walter Gropius, Alexander Klein and Otto Haesler were also architects who
carried out analytical studies as a background for their housing designs.24
Klein and Haesler were particularly concerned with the layout of the
apartment, doing studies like comparing the depth of the building and the
floor plan layouts to rental costs. These costs were later compared with the
income of the households in need of dwellings. The work of Klein became a
basis for the “Bauentwurfslehre” by Ernst Neufert (1936), a work that has
been published in numerous editions and that has been the main normative
reference for a generation of post-war-architects. The discussions and focus
of Scandinavian architecture were closely connected to what happened in
Germany, and the close relations between politics, social interest and
architectural practice were manifest at the Stockholm exhibition in 1930.25

24
See: Adler (1927), Bredsdorff (1938), Isaacs (1983), Klein (1931, 1934), Warhaftig (2000) and Plan:3
:”Erfaringene fra Tyskland”.
25
For those reading Danish, a review of the Stockholm Exhibition can be found in Hansen (1930).

23
Figure 2.3
A study by Otto Heasler. Different versions of a floor plan (to be selected in accordance with the
economy and the size of the household).

In the beginning of the 20th century, Norwegian architects also became

concerned about bad living conditions and the need for housing, and towards
the 1930s they became deeply involved in the political and social debate on
housing as well as in the planning and design of the buildings.26 In Norway
the magazine “Plan” contributed to information and debate on housing and
apartment layout in a series of articles, some of them very similar to the
works of Klein and Haesler. (Plan, 1933: 1, 2 and 1935: 3.) The analytical
studies of housing in the early 20th century constituted a basis for modernism
within architecture, firstly by representing a critique of traditional housing
design and secondly by developing innovative designs. The architecture of
the 1920s and -30s, particularly the architecture developed in Germany27
during the Weimar republic, has inspired housing and apartment design ever
since.28

26
A subject for discussion was what kind of housing to build, tiny apartments (which people could afford) or
larger apartments (which by architects were considered to be better but rarely could be afforded by those
households most in need of improving their housing conditions), two points of view emphasised by the
architects Rivertz and Hals, respectively. See Rivertz (1935).
27
Among the well known projects from this period are Siemensstadt in Berlin and Weissenhofsiedlung in
Stuttgart, projects consisting of works by numerous of the famous architects of the time such as Peter Behrens,
Le Corbusier, Walter Gropius, Hugo Häring, Mies van der Rohe and Bruno Taut. See for instance Pommer
and Otto (1991) about Weissenhofsiedlung and Rave and Knöfel (1968) about dwellings in Berlin.
28
Concerning items of interior design such as lamps and kitchen equipment, the designs developed during the
1920s and -30s are not only still today important references today, they are even highly fashionable. This is
interesting as the analytical kinds of studies that were the basis for these designs rarely any longer are taken
seriously. It could be that contemporary dwellings are excellent concerning any aspects of design where such
analytical studies might be relevant. However, as indicated in the previous chapter, this does not seem to be the
case. The situation is more that contemporary interior design represents a kind of post-modernism where icons
of early modernism appear in eclectic manners without the original references to use or to ease of life, being a
kind of “non-functional functionalism”.

24
Figure 2.4
An apartment from the Stockholm exhibition, 1930. Architect: Kurt von Schmalensee.

25
In the history of Norwegian housing research, one of the most important
works is the survey by architect Carsten Boysen and his colleagues. As the
publications presenting the results were edited by Odd Brochmann, the
survey is usually (and therefore also in this thesis) in short named the housing
surveys of Brochmann. This extensive piece of work started during the
Second World War with the aim to evaluate small apartments built in the
years before the war. At this time, it was well known that the standard of
living in many old apartments was low due to overcrowding in houses with
low technical and sanitary standards. The subject of the survey was to what
extent the new housing projects, i.e. those built during the 1930s, did (or did
not) manage to improve this situation. The aim of the survey was to establish
factual knowledge relevant for planning and designing future dwellings. The
survey consisted of extensive registrations of housing conditions and of the
residents’ evaluations and preferences. The data were analysed by architects,
doctors and psychologists. The survey was presented in a series of
publications (Brochmann 1948, 1952 1958 and 1961) that contained
historical reviews of housing conditions, detailed analyses of floor plans and
rooms with respect to the space needed for various domestic activities, in-
depth “case studies” of particular households and their living conditions as
well as statistical analyses of households and dwellings. From these analyses,
many guidelines for design were identified, such as that apartments designed
for more than 4 persons should have separate WCs and that kitchens should
not be less than 8 square metres. (Brochmann, 1948, p. 138)

There has always been an aspiration to teach people “the right ways of
living” inherent to architects’ interest in housing and dwellings.29 The
publications of Brochmann are representative for their time in this respect.
Brochmann (1948, p. 102) argues that, in order to prevent “wrong use”, such
as using the kitchen as a kind of living room, kitchens should not be too
large. Other examples are his critiques of “representative” living rooms
preferred even by cramped living households,30 of “heavy furniture” and of
“bad-taste” paintings and decorations. However, regardless of this belief in
the relevance of architects’ aesthetical taste, the surveys of Brochmann and
his colleagues are still impressive by the amount of relevant data collected
and by the broad approach of the analyses. Figure 2.5 illustrates some of the
content of the survey.

29
A discrepancy between architects’ preferences and “common taste” is at least as old as the architects’
interest in dwellings. The architects’ concern with the totality of domestic life and their well-meant teaching of
people about how to live, is described in the essay “The Rich Poor Man” by Adolf Loos already in 1900.
(Munz and Künstler, 1966, p. 223) Rolness (1995) has elaborated the somewhat more modest consern of
Norwegian architects in the same respect.
30
This kind of “representative” living room is a parallel to the traditional use of the English “parlour”
described by Hanson (1998, p. 121).

26
Figure 2.5
Some figures from the publications of Brochmann.
2.5a A particular household and their apartment; a family of four, with one room for let.
2.5b Layouts of entrances, how to hang the coat, where Brochmann recommends the alternative
on the right.
2.5c Drawing documenting a living room. In text, Brochmann criticises the “heavy” furniture.
He was also annoyed by the diversity of quality and exemplifies this by “good paintings
and the best porcelain placed beside rubbish”.

27
In “Bostadsplanering”, a pamphlet packed with information, the Swedish
architect and scholar Björn Klarqvist (1969) has summarised the current state
of Swedish housing research and dwelling design in the late 1960s. Among
Klarqvist’s references is Lennart Holm’s doctoral thesis about dwellings and
living condition is Sweden.31 (Holm, 1955) Figure 2.6 and 2.7 show some
illustrations from this Swedish housing research. Figure 2.7 is one out of the
many detailed results of Holm about domestic lives, showing where men and
women spend their time, while figure 2.6 shows his result about the
residents’ evaluations of different bathroom-layouts. Figure 2.8 shows
Klarqvist’s illustrations of how a bedroom that is not too small can be used in
different ways. The figure at bottom right (in figure 2.8) shows proposed
section of a kitchen.

Figure 2.6
Satisfaction with different bathroom layouts. (Holm, 1955)

31
The above-mentioned Odd Brochmann was among those who guided the survey of Holm.

28
Figure 2.7
Women’s and men’s activities through the day. (Holm, 1955)

Figure 2.8
Three figures from Klarqvist (1969), showing
alternative use of a bedroom that is not too small,
recommended sizes of living room and kitchen
and recommended section of a kitchen.

29
As intended, the survey of Brochmann and colleagues became a basis for
planning of housing as well as for design of dwellings in more detail, issues
that have been major interests of Norwegian architects, planners and
politicians in the post-war period. During this period, until the late 1970s, the
governmental institutes of building research in Norway, as well as in
Denmark and Sweden, have focused on the layout of dwellings in terms of
alternative floor plans and the space needed for various domestic activities. In
all three countries the governmental institution for building research
developed normative guidelines for designing and evaluating apartments,
represented by Svennar (1975) in Norway, Bredberg (1978) in Sweden and
Ranten and Vedel-Pedersen (1982) in Denmark. Svennar’s guidelines
proposed appropriate sizes for different rooms and captured essential
information about the space needed for specific domestic activities, but they
are less useful when it comes to evaluation or design of entire dwellings. The
works of Bredberg as well as those of Ranten and Vedel-Pedersen describe
many and detailed parameters of the dwelling layout relevant to the quality of
living and represent thereby a large body of knowledge. However, since they
are extremely extensive, they are useful for evaluating apartments already
designed rather than applicable to the process of creating new layouts. The
normative guidelines developed by research institutions were implemented
into planning and design through Norwegian building legislation and as terms
to be followed in order to achieve favourable loans. The latter was managed
through “Husbanken”, the Norwegian Governmental Bank of Housing.32

Figure 2.9.a
Recommended place needed for living room furniture. (Ranten and Vedel-Petersen, 1982)

32
The guidelines of Husbanken, which are called “Husbankens minstestandard”, still exist. (Husbanken, 2000)
The aim of this minimum-standard, which concerns sizes of rooms and other features of the floor-plan, was
originally to prevent low quality dwellings. Today, it is usual that new apartments do not meet this “minimum
standard” where sizes of rooms and daylight conditions are concerned.

30
Figure 2.9.b
Recommended size a living room, depending on standard of living and number of persons.
(Ranten and Vedel-Petersen, 1982)

In 1967, NBI (the Norwegian institute of building research) carried out a

comprehensive survey about housing and living, a survey published in two
reports by Ås (1971). These reports contain extensive information about the
Norwegian stock of dwellings, about the households and about their living
conditions. The results of Ås about spaciousness of living referred to in
section 1.2.2 are among the results of this survey. Figure 2.10 is another
example of results from this survey about the households’ preferences,
showing percentages of respondents that considered their rooms too small.
(In order to please 80 % of the households, the living rooms, the kitchens and
the bedrooms should be 20, 8 and 12 m2, respectively.)

31
Figure 2.10
Room-sizes and residents’ satisfaction.
(Percentages of households considering their rooms too small. (Ås, 1971))

During the 1970s, extensive criticism of the large buildings, the monotonous
neighbourhoods and the segregation of functions that characterised housing
projects built in the 1960s and 1970s emerged.33 A consequence of this
criticism was that residents’ influence on the own living condition became a
subject of major interest.34 The focus of research changed from the buildings
and the individual dwellings to the broader aspects of living and domestic
life; to an interest in finding out how people best could participate in the
planning and the design of their dwellings.35 Another consequence of the
critique of large and repetitive housing projects was that alternative kinds of
buildings became a subject of interest. Particularly in Denmark this lead to a
renewed interested in “low and dense” housing. The research report “Tæt lav
– en boligform” by Statens Byggeforskningssinstitut (1971) 36 argues that this
principle, which is an old historical tradition in urban settlements, is efficient

33
An origin of this critique on an urban level was “The Death and Life of Great American Cities” by Jane
Jacobs (1961). She argued that urban planning, all since Ebenezer Howard, were trapped in the
misunderstanding that well functioning cities could be intentionally designed in detail.
In Norway, the housing projects constructed during the 1960s and 1970s were less criticised than in Britain
and in Sweden. To what extent this can be explained by the smaller size of Norwegian housing projects or by
the layouts of Norwegian dwellings, is an interesting issue not followed here.
34
The residents did not have more direct influence on the planning of their dwellings in the decades before, but
the need for “user’s participation” in the design process was then not on the agenda. The focus on individual
freedom and right of choice that was a part of the political changes of the 1960s and 1970s can explain some of
this emerging demand for user’s participation. However, it is hard to argue that the demand was not caused
also by the housing projects built during the 1960s and 1970s.
35
In “The Production of Houses” Christopher Alexander (1985) extended his earlier proposals for architectural
design (Alexander, 1977) into guidelines also for the design processes. “In the modern world, the idea that
houses can be loved and beautiful has been eliminated almost altogether.” Alexander states (p.14) and
describes his aim of constructing a housing process that results in houses where “.. people feel proud and
happy to be living in them and would not give them up for anything, because they are their houses, because
they are the product of their lives, because the house is everything to them, the concrete expression of their
place in the world, the concrete expression of themselves.“
36
Statens Byggeforskningsinstitut is the governmental institute of housing research in Denmark.

32
with respect to costs as well as site-areas while it avoids the problems of
large high-rise housing projects. Even though several “dense and low”-
projects were built,37 the concept became a critique of high-rise housing
rather than the new guideline for housing design.

As the 1970s passed by, there were still many pre-war dwellings of very low
technical and sanitary standard while it had become clear that the large high-
rise housing projects were the solution. Parallel to the interest in users’
participation and the search for alternative principles for new dwellings, a
new subject of interest was how to improve the dwelling standards by
renovation of the existing stock of dwellings rather than by demolition of
existing houses and construction of new dwellings from scratch.

Other aims of the research during the 1970s and –80s was to find out to what
extent the apartments built were in accordance with the normative guidelines
and to compare the actual living conditions to what was assumed by
researchers and architects. The research institutions therefore carried out
interview-based surveys in order to gather data about real households and
their preferences. One such study was the Swedish research from the late
1970s presented in a series of reports entitled “Bostadsutformning,
bostadsanvändning” (Statens Institut för Byggnadsforskning, 1979 and
1980). This survey examined 308 dwellings of altogether 33 kinds built
during the period 1942-1975. By collecting detailed information about the
households, their daily living and their dwellings, the survey established an
empirical basis applied in order to evaluate the normative guidelines. Among
the findings were that a wider range of activities than assumed went on in the
kitchen, that the kitchen should be larger for this reason and that the main
activity in the living room was watching TV. Concerning bedrooms, it was
discovered that many parents moved into one of the smaller bedrooms in
order to let teenagers move into the largest bedrooms. Figure 2.11 is an
illustration from this survey that documents the furniture of the dining rooms,
(showing the dominant positions of the TVs) while figure 2.12 shows the
sizes of circular kitchen tables. The latter is an example of one of the very
detailed results of this study.

37
The project “Skjettenbyen” showed in figure 2.22 is such a “low and dense” housing project.

33
Figure 2.11 Examples of the “TV-based” furniture in living rooms. (M80:3)

Figure 2.12
The sizes of round kitchen tables.
(M80:3)

In Norway, similar studies of households and their living conditions were

carried out by Guldbrandsen (1973) and by Guttu, Jørgensen and Nørve
(1985). The former examined the living in “one-rooms-apartments” in the
suburb Ammerud, concluding that such apartments were too small to be
acceptable for permanent living (see figure 2.13), while the latter was an in-
depth study of 3 kinds of apartments built during the 1970s, carried out by
visits and interviews in altogether 30 apartments. Similarly to the Swedish
research “Bostadsutformning, bostadsanvändning” mentioned above, they
found that households preferred larger kitchens and that many parents moved
into the second largest bedroom. They also found that people disliked
bedrooms accessed only through the living room; particularly when this was
the case for the largest bedroom. Some conclusions of Guttu, Jørgensen and
Nørve were that the size of the kitchen should be increased to about 15m2;
becoming a kind of second living room, and that the second largest bedroom
should be as big as the largest, which means about 12 m2.38 (See figure 2.14.)

38
Guttu, Jørgensen and Nørve (1985, p. 64).

34
Figure 2.13
The dwellings examined by
Gulbrandsen (1973), an apartment
that few residents, due to size and
to lack of a separate bedroom, found
acceptable for permanent living.

Figure 2.14
One example from the survey of Guttu, Jørgensen and Nørve (1985), an apartment where the
parents have moved into the second largest bedroom. The figure also illustrates the kind of floor
plan where the residents were dissatisfied with through-passage in the living room.

Since the 1990s, housing research has again been paying more attention to
the building, but this time focusing on sustainability and economy rather than
on floor plan layouts. The issues of interest have been the use of energy and
materials on the one hand and profit and efficiency of construction on the
other. The segment of research that is still about layout of dwellings is
“universal design”.39 Because normative guidelines for “universal design” are
terms for achieving favourable governmental loans, universal design is a
concern where research still influences the layout of new housing projects.

39
Universal design encompasses not only concern about those permanently disabled but also the understanding
that most people will have periods when their ability or ease of movement is restricted.

35
The tradition of Scandinavian housing research described above, has become
part of the architectural research that is now developing into an academic
field at the institutions teaching architecture. Some of the doctoral theses that
concern dwellings are Støa (1995) and Guttu (2003), respectively from
NTNU and from AHO in Norway, and Farah (2000), Johansson (1997) and
Nylander (1998) from Sweden; Farah and Nylander from Chalmers,
Gothenburg and Johansson from KTH, Stockholm.40 Støa has studied
Norwegian one-family houses from the 1980s. This kind of housing
represents the majority of Norwegian dwellings and there is a striking
discrepancy between the ideals of architects and those of the households
living in this, the most common, kind of Norwegian dwellings. Støa has
examined the households’ ideals about living as well as the layout of the
houses and the gardens. She describes how the households’ preferences are
due to the need for “personalisation” of the building and the building
process. Ideals about dwellings are the subject also of Guttu’s thesis, but
while Støa focused on the ideals of the households, Guttu has studied how the
professionals’ ideals about dwellings have developed in Norway since the
Second World War. In Sweden, Farah has studied Sudanese dwellings,
comparing the spatial layouts of “traditional” and “architect-designed,
modern” houses. By applying space syntax analysis,41 she found that the
dwellings, regardless of being “traditional” or “modern”, consist of a “male-
zone” and a “family-zone”. According to Farah, a courtyard is the most
“central” space in the traditional dwellings (or “most integrated” if applying
space syntax terminology), while the “family hall” is the “core” of the
modern dwellings. The theses of Nylander and Johansson are both about
Swedish dwellings. The aim of Nylander has been to conceptualise “non-
measurable” architectural attributes. Nylander notifies altogether seven
“fields of attributes”, these are “materials and detailing”, “axiality”,
“enclosure”, “movement”, “spatial figure”, “daylight” and “organisation of
spaces”, and argues that essential “non-measurable” features of architecture
can be captured by applying these. He has studied 4 housing projects built in
the period from 1987 until 1993, by visiting the apartments and interviewing
the residents as well as the architects. His thesis presents analyses of the
apartments with respects to these “fields of attributes” and exemplifies the
“fields of attributes” by the apartments. The study of Johansson is another
thesis that presents theoretical elaborations related to surveys of dwellings.
While Nylander intends to develop an original methodological framework for

40
A study that has some relevance to this study without focusing on common dwellings is the one of
Margrethe Dobloug (2006). In her dissertation-lecture about contemporary buildings for elderly she pointed
out the new typology of dwellings being developed in Norway, a kind of dwellings particularly designed for
elderly needing some care. These dwellings for elderly enforce the tendency found by this thesis, the tendency
that new dwellings are tiny and appropriate only for particular kinds of households.
41
Space syntax analysis is elaborated in section 2.3 and 3.4.

36
analysing architecture, Johansson uses his 6 cases of housing research as
background for discussing case-studies and qualitative evaluation of artefacts
in general. While Nylander’s thesis is of the kind where architects aim at
developing theoretical approaches and methodologies specific to
architecture,42 Johansson’s thesis represents a useful introduction to existing
theory and methodology that are relevant for the evaluation of architecture in
general and for survey-based architectural research in particular.

Figure 2.15.
Nylander analyses “axiality” and “movement”, two out of his altogether seven “non-measurable
fields of attributes”.

This review illustrates the development of research-orientated studies of

dwellings from early international modernism, where innovative architectural
design was closely related to analytical studies, up to the more scholarly
housing research of today.43 In spite of the research now going on in the
institutions teaching architectural design, there is, at least in Norway, a
remarkable lack of interaction between research and analytically based
knowledge on the one side and architectural practice and teaching on the
other.44 In a knowledge-based society and in a research-based educational
system, this represents a challenge for architectural research as well as for
architectural practice and education.

42
Nylander attempts to grasp the “non-measurable”. Coolen (2005) has discussed this subject more in detail.
An interesting question is whether such attempts manage to reveal essentials of “the non-measurable” or if it
fails due to the problem of “measuring the immeasurable”. The “non-measurable” attributes of Nylander
represent interesting fields of discussion, as several of them, particularly “axiality” and “movement” are
studied in explicit and “measurable” manners through the methods of space syntax, see sections 2.3 and 3.4.
43
For those who are interested in Scandinavian housing research and design, but do not read Scandinavian
languages, a review of the subject is available in English. “Housing research and design in Sweden” edited by
Sven Thiberg (1990) is an English version of “Bostadsboken” from 1985. Besides reviewing history of
housing research and housing design, it contains an extensive listing of relevant references.
44
Jansson (1998) describes how theoretical knowledge as such rarely constitutes basis for architectural
practice; the main references for architectural design are knowledge of specific buildings and not theory in
written form.

37
2.2 VOID AND SOLID

“Housing” as well as the more specific “dwelling” are terms that can be
applied to cover a wide range of meanings. Rapoport (1990) argues that a
dwelling is more than the structure built.

“-- a dwelling is not just a structure; it is an institution, a social and cultural

unit of space created to support the way of life of people.”

A focus on space rather than on the building is also part of the approach in
this study. The question about what space is and how space might be studied,
can be elaborated as an architectural issue or more in general. Concerning
architecture and space particularly, Scruton (1979, p.43) has made the
following statement:

”..the theory that the experience of architecture is an experience of space is

obviously indefensible. If space were all that interested us, then not only must
a large part of the architect’s activity seem like useless decoration, but it is
even difficult to see why he should bother to build at all.”

Scruton here argues that architecture cannot be about space since it is about
building. However, by this logic, where interest in one issue (space) must be
solely about this, excluding any interest in other issues (such as the building),
the conclusion might as well be the opposite. It is a fact that living takes
place in space and not in the physical mass of buildings. By the logic Scruton
applies above, he could therefore as well conclude that architecture cannot be
about building since we know it is about space. However, such focus on one
by dismissing the other is like attempting to remove the vase from the faces
in figure 2.16; space and building represent void and solid and are thereby
mutually dependent upon on each other. The relationship between space and
building can be pointed out by referring to lectures of the Swedish scholar
Björn Klarqvist. His standard introduction when teaching space syntax is to
draw a person standing in an empty flat landscape; a human in a space that is
an infinite hemisphere. By placing one wall (or one box) in the space and
then one more, he points out how the purpose of building is to create space
by limiting an initial space. In brief, architecture is about defining space, and
to construct buildings is a means for doing this. The physical building and the
space within and surround it define each other similarly to the vase versus the
faces; without the faces there is no vase as without the physical building there
would be no interior space. However, if we are particularly interested in the
vase, it might be relevant to study the vase without having the faces explicitly

38
in mind at the same time; some enquiries about the vase might even need
such a focus on the vase. Similarly, it might make sense to focus on the
internal space, having the building in mind only in terms of being the
physical surrounding that defines the space. Based on the assumption that the
spatial layout is of major importance for human behaviour, preferences and
well-being, this kind of focus on the space is the basis of this study. This does
not imply that features of the physical building (other than those physically
defining the space) do not influence the same; it means that it should be
relevant to study characteristics of architectural space in particular. The
subject of this study, or the “unit of analysis” if applying a term of Yin45, is
the interior space, this as opposed to the building. The objects and features
that define, serve and characterise the space, such as structures, technical
installations, materials, colours and textures, are relevant in terms of being
the context of the interior space but are not explicitly examined by this study.
The subject is “the physiognomy of the void” in the sense that the focus is on
the space rather than on the building, and on particular features of the space
rather than on intentions, perceptions or experiences concerning space.

Figure 2.16 Void and solid; the vase and the faces.

45
Yin uses the term “unit of analysis” at two levels, both in the meaning of a precisely described main subject
of inquiry, where there should be just one, (Yin, 1994, p. 23) and as the possible subunits to study within a
case. (Yin, 1994, p. 39) The term is here applied in the former meaning. About the term “unit of analysis”, see
also Patton (1987, p. 50).

39
2.3 SPATIAL CONFIGURATIONS

A feature of “the physiognomy” of space focused upon in this study is the

spatial configuration. Configurational aspects of space are an issue often
implicitly touched upon when discussing floor plans or typologies of
buildings. In “Bauen Der Neuen Wohnbau” Bruno Taut (1927, p. 68)
criticised a contemporary dwelling for having “aufgeschwemmter Grundriss”
(a “spread-out floor plan”) and for consisting of “totes Raum” (“dead
space”). Further, he stated that the search for the best floor plan is the mission
of the real architect:

“..Wir wissen deshalb, dass auf dem Gebiet des Wohnungsbaues hierin
allein, d.h. im Suchen nach dem besseren Wohnungsgrundriss die Aufgabe
des wirklichen Architekten liegt.” 46

The architects’ major tool for designing spatial layouts is drawing floor plans.
Within housing research, basic features of the dwelling have been analysed
by examining floor plans and comparing these to knowledge and data about
households and the social, economic and technical contexts. Among the early
works that explicitly concerned the interior space, are those of the architect
Alexander Klein, who in the 1930s carried out extensive studies of spatial
layouts of houses and apartments. He analysed contemporary dwellings and
proposed guidelines for new designs. He studied daylight conditions, the
proportions of rooms and the positioning of rooms relative to each other.
Figure 2.17 illustrates how Klein (1934, p. 115), by applying simple graphics
on the floor plan drawing, pointed out how the positions of the staircase and
the doors affect the space of a living room. Klein’s marking of floor-areas
needed for convenient movements is a simple technique that illustrates some
basic properties of the spatial layout. Figure 2.18 shows this kind of marking
at the floor plan drawings of a new apartment (from the same project as the
one showed in figure 1.1). Most of the living-and-kitchen-room is “occupied”
by necessary access to other rooms and to the kitchen, leaving little space left
for dining, relaxing, watching TV or other activities that we know people
prefer to do in their homes. Klarqvist (1969) and Svennar (1975) are among
those who have applied graphics similar to Klein’s, when analysing floor
plans and when describing guidelines for designing the rooms of dwellings.

46
This in English would be something like “…we know, when construction of housing is concerned, that
solely in this, in the search for the best apartment-floor-plan, lays the mission of the real architect”.

40
Figure 2.17
Analysis of different layouts of the ground floor in a small single-family house. (Klein, 1934)

Figure 2.18
An apartment from “Badebakken”,
Oslo, where floor-areas required
for movement/access are marked
similarly to Klein’s drawings in
figure 2.17.

A theoretical approach that particularly addresses configurational aspects of

space is the one of space syntax. This is a field of architectural research
where easily applicable methods for studying spatial configurations have
been developed from a consistent theoretical framework. What it focused by
space syntax, is how spatial units relate to each other, how such spatial
relations correlate to social life and how such relations between spatial units
can be identified and analysed. Marcus (2000, p.49-50) describes the
importance of spatial configurations in architectural design as follows:

“Of necessity, we construct certain configurative patterns when we build,

whether we are aware of it or not. The important thing is that we then also
construct potentials for certain functional performances of the building.”

41
Marcus then describes the nature of spatial configurations more explicitly:

“What configurative descriptions do, as opposed to traditional descriptions

of architecture or urban form, is that they focus on the relation between the
elements in an architectural system, rather than the elements themselves. …
Qualities in the built environment, whether we talk about an building or a
city, are primarily given by the internal relations to one another of the
elements of which it is composed, rather than by the qualities of the elements
as such. What gives a building its specific character is not so much, for
example, the rooms within looked at one at a time, but how they are
combined with one another. In other words, the configurative captures how
something is composed rather than what it is composed of.”

Figure 2.19-1.3.a shows a spatial configuration consisting of two identical

elements “a” and “b” that are directly connected, or two rooms connected by
a door, if the figure is the floor plan of a building. The situation is
symmetrical in the sense that “a” relates to “b” as “b” relates to “a”. By
introducing access between the room “a” and the outside world “c”, as shown
in 1.3.c, the situation becomes very different. Room “b” has now achieved
access to the outside without itself being changed, and room “b” and room
“a” are no longer equal in terms of accessibility. Figure 1.3.e shows a so-
called “connectivity graph” of the new configuration.

Figure 2.19
Spatial configurations and their connectivity graph representations. (Hillier, 1996)

42
When it comes to social activities, movement and occupation are two
principle kinds of behaviour that relate to basic configurational aspects of
space.47 If we return to the example in figure 2.19 (1.3.c), the spatial
configuration will imply movement through space “a” simply due to the
situation of “through passage”. Hillier (1996, p. 38) argues that the
configurational aspects of real buildings and cities, where the patterns are
much harder to see than in the figure above, are “non-discursive” in the sense
that they are hard to explain explicitly. In their studies of offices and
interaction between the employees, where they have applied space syntax
analyses, Blombergsson and Wiklander (2006) have some illustrative results
with respect to this “non-discursivity”. They found that the positions in
offices mostly preferred as workplaces and the areas where most interaction
between employees takes place, correlate strongly to basic configurational
characteristic concerning visibility and access. For those familiar with space
syntax studies, this alone is not surprising. What is interesting is their further
comparison of real behaviour (as registered on site) and the answers given at
interviews and questionnaires; the respondents rarely mentioned spatial
qualities as important for their choice of workplace or for their ways of
working, and their conscious perception of their actions did not correspond to
the their actions in reality (as registered on site). The study of Blombergsson
and Wiklander illustrates that there are significant correspondences between
social activities and spatial configuration and that these patterns are not
consciously known. This kind of “non-discursive” knowledge, a knowledge
that we apply intuitively, is what Hillier (1996, p. 40) describes as:

“social knowledge”,-- “which purpose is to create, order and make

intelligible the spatio-temporal events through which we recognise the
presence of culture in everyday life.”

The aim of the theory and methodology of space syntax is to grasp patterns in
spatial configurations, patterns that we, according to Hillier, intuitively
perceive and that we, more or less unconsciously, behave in accordance with.
The methodology that makes space syntax easily applicable in analyses of
architectural space is described in section 3.4. In the course of the recent
decade, space syntax has become a well-established field of international
architectural research. In the study described by this thesis, space syntax has
been essential both in terms of the theoretical approach to architectural space
and in terms of methods of analysis.

47
From this basic distinction between occupation and movement Hillier (1996, p. 317-318) has developed a
typology consisting of four kinds of topological positions in a spatial structure. This is explained in section
3.4.7.

43
Figure 2.20 shows the floor plan of an apartment designed by Scharoun and
the corresponding connectivity-graph, which is a notation that represents the
internal position of rooms with respect to accessibility. As will be elaborated
further in section 3.4, space syntax analysis consists of calculations and
comparisons of such graphs or other modelling of spatial configurations. A
substantial reference concerning space syntax analyses of dwellings is
“Decoding Homes and Houses” by Julienne Hanson (1998), where space
syntax analysis is applied in order to examine historic as well as
contemporary dwellings. In her analyses of traditional English farmhouses
(Hanson, 1998, p. 56-), she has revealed patterns not found by previous
studies that have analysed the floor plan drawings only. Studies of dwellings
have developed into a “sub-field” within space syntax research, a sub-field
where several contemporary studies have parallels to the one described by
this thesis.48

In this study, where an intention has been to look for patterns in the
development of apartments over a period of time, a useful application of
space syntax is to describe the spatial configuration of each apartment by
graph notations similar to those in figures 2.19 and 2.20, and then analyse
and compare these graphs. This is an application of space syntax within the
“core” of space syntax methodology. The exact modelling of the rooms and
the particular space syntax analysis carried out, are described in section 3.5.

Figure 2.20
An apartment by Hans Scharoun, presented by a floor-plan drawing and by connectivity graph as
seen from the staircase. (An apartment from Siemensstadt, Berlin, 1929 - 1931)

48
Parallel to the study described by this thesis, similar diachronic space syntax analyses of dwellings have
been carried out by Cunha and Magalhães (2005) and by Guney (2005). Cunha and Magalhães have examined
middle class apartments in Rio de Janeiro built between 1930 and 1970, and compared the historical
development of dwellings to economical, social and demographic development during the same period. Guney
has examined upper-middle-class apartments in Ankara designed by architects and built in the period 1920 to
1990. Even though these studies differ from mine in terms of sampling as well as of the main subjects of
interest, the three studies constitute an empirical basis from which it should be possible to develop some
relevant comparisons.

44
2.4 GENERALITY VERSUS SPECIFICITY

From an overall point of view, the aim of planning, designing and

constructing dwellings is to constitute a stock of dwellings appropriate for
contemporary as well as future population of households. As buildings last
much longer than specific needs and preferences, the challenge of housing is
not only to fit the requests or the needs of the moment, but also to be
appropriate in the future. Due to the temporality of contemporary needs and
the longevity of buildings, the latter is at least as important as the former.
Brand (1994, p. 181) argues that all buildings are predictions and that all
predictions are wrong, and points out a contradiction between contemporary
needs as captured by architectural “programming”, and future and unknown
requests:

“The great vice of programming is that it over-responds to the immediate

needs of the immediate users, leaving future users out of the picture, making
the building all too optimal to the present and maladaptive for the future.”

According to Brand, there is a conflict inherent in the modern tradition of

architects’ design-practice; while we attempt to design buildings of
permanent excellence by responding to contemporary “programs”, the
requests that constitute the “programs” are continuously changing.
Concerning dwellings, the changes in needs and preferences are due to the
individual persons in the household as well as to the society; the need for
dwellings corresponds to the phases of life of the individuals as well as to
social, cultural, economical and demographic development of the society.
Due to the less dominant position of households consisting of two parents
and the children they have in common, and to the greater variety of social,
cultural and national backgrounds within the population, there is now an
increasing diversity of preferences and needs where housing and dwellings
are concerned. In such a setting, how do we design our dwellings, dwellings
not made only for tomorrow but also for many decades to come? When
Klarqvist (1969, p. 81-86) pointed out the need for housing capable of
handling the plurality of contemporary households as well as future changes
in needs and preferences, he described generality as one out of three relevant
strategies for housing design. Generality here means the capacity of a fixed
situation to serve a variety of demands or wishes. The other strategies pointed
out and exemplified by Klarqvist are flexibility and elasticity.49 The notion of
flexibility describes a situation that is easily changeable while elasticity
49
Generality, flexibility and elasticity are the same categories as described by Thiberg (1967) and later pointed
out by Cold (1984).

45
means the ability to adapt in accordance with changing demands by
modifying size. As these terms are mixed in common language as well as in
architectural discussions, they will on the following pages be commented
upon in more detail and illustrated by examples.

Figure 2.21 shows a building from about 1890 with a general floor plan in the
sense that the main rooms are not specified for particular functions; neither
the sizes of the rooms nor the positioning of them relative to each other
predict the use of each room.50 Originally, the use of the rooms changed
during the day; from being used for sleeping at night, for living, play, and
work during daytime and back to sleeping in the evening. The directors of
these never-ending cycles of changes were the women, - the housewives in
the workers’ families and the maid among those better off; they rearranged
the rooms in accordance with the changing activities throughout the day. A
post-war example of a general apartment-floor-plan is the one developed by
Lennart Holm during the 1950s and presented as “Lägenhet 1960”
(“Apartment 1960”).51 This dwelling was designed in order to be appropriate
for as varied a daily life as possible and consisted of a kitchen and two rooms
that were all of 14 m2 and a living room that was of about 23 m2. Thiberg
developed these ideas further by making the rooms even more similar in size.
Figure 2.22 shows Thiberg’s proposal; an apartment consisting of 6 spatial
units of about 12m2, where the units for kitchen and for WC/bath/washing
were the only ones with pre-determined use. The other four units could easily
be adapted in accordance with the preferences of different households. Figure
2.23 shows a more recent housing project where many of the apartments have
general rooms as regards usability: The project is a courtyard house in
Amsterdam by the architects Diener and Diener (2002).

Concerning generality and the sizes of rooms, floor areas between 12 and 15
m2 have in different contexts been recommended in order to achieve “general
rooms”. As mentioned in section 2.1, Guttu, Jørgensen and Nørve (1985)
recommended increasing the kitchen area to 15 m2, in order to make it
suitable for general daily living.52 When Lynch (1972, p. 109) discussed
adaptability and elaborated upon a concept of modules that should be
“permanently useful”, he referred to a study of hospitals by Peter Cowan who
had found that rooms between 120 and 150 square feet (11-14 m2) were
those most easily converted to a wide range of new uses.

50
The floor areas of the main rooms are about 11 and 18 m2.
51
In Norway, this apartment was presented in an article by Liv Schjødt (1957).
52
The average size of Norwegian kitchens was then, in 1985, about 8 m2.

46
Figure 2.21
Generality.
Seilduksgata 11,
Oslo.
(about 1890)

Figure 2.22
Generality.
Sven Thiberg.
(1967)

Figure 2.23
Generality.
Apartments in Amsterdam.
Diener & Diener (2002)

47
Figure 2.24 and 2.25 show two examples of “elastic dwellings”. The first is
the Norwegian project “Skjettenbyen” from 1970,53 which consists of
module-based detached houses designed in order to be expanded in
correspondence with the needs and economy of the particular households. In
despite of the variety of possible layouts in this concept, the dwelling units
have ended up rather similar as most dwellings are extended to the maximum
version.54 This is due to the simple fact that parts of the house that are no
longer needed are not likely to be torn down. This means that elasticity in
such cases is a one-way process only. The other example, shown if figure
2.25, is more a kind of “real elasticity”. This is an apartment that consists of
two dwelling-units, where one can be let out as a tiny separate apartment in
accordance with the “owner-household’s” economy and need for space. Due
to Norwegian tax-legislation, such incomes are tax-free and this kind of
dwelling profitable. When the years pass and the children grow up, a family
living in such as apartment can occupy also the small unit instead of being
forced to move because of cramped living conditions. When more years pass
and the children move out, the small unit might be a “guest-dwelling”, a
place to stay for grownup children on visit, or it might again be let out. A
pamphlet elaborating elasticity with respect to apartment layouts is the
Danish research-report “Elastiske etagebolige” (Statens
Byggeforskningsinstitut, 1978). Figure 2.26 shows how this report
categorises kinds of elasticity.

Figure 2.24
Elastisity.
“Skjettenbyen”. Hultberg, Resen and Throne Holst in association with Nils Ole Lund. (1970)

53
About ”Skjettenbyen”, see Byggekunst 1970: 4. p. 146-149 and 1979: 6, p. 404-405.
54
The same results are pointed out by Olivegren and Schulz (1985, p. 250, English version: 1990) as they
comment “Kallebäckhuset”, a housing project built in 1960 and designed by the Swedish architect Erik
Friberger. This house has a structure of concrete slabs that allowed the residents to build their own dwellings
on the floor area between the neighbouring dwellings. Contrary to the variety intended by the architects, the
result very soon was “maximum-versions” of most dwelling units, similar to what happened in “Skjettenbyen”.

48
Figure 2.25
Elastisity.
“Lakkegården”, an apartment
consisting of two dwelling units.
Arkitektskap. (2003)

Figure 2.26
Kinds of elasticity, according to Statens Byggeforskningsinstitut, DK (1978). A: a section to let,
B: supplemental rooms, C: small units that might be merged, D: movable borders between
dwelling units, E: unfinished dwellings to be extended or completed by the residents.

49
When it comes to the principle of flexibility, an illustrative example is the
Schröder House, one of the most famous icons of architecture, designed by
Gerrit Rietveld and constructed in 1924. Figure 2.27 shows how the interior
space can be changed from an “open” plan to a dwelling consisting of
separate rooms that are conventional in sizes as well as in positions. In this
case, flexibility is achieved by permanent but easily moveable sliding walls
and door-like elements. An alternative strategy of flexibility is to apply
lightweight walls; either movable walls or walls made in order to be torn
down easily and substituted by new ones in other positions. A housing
project that is flexible due to such lightweight internal walls in combination
with thoroughly planned positions of fixed elements such as the pipes for
kitchens and bathrooms, is “Järnbrottshuset”, a housing project from 1952
designed by Tage and Anders William-Olsson, see figure 2.28.

Figure 2.27 Flexibility. First floor of Schröder House. Gerrit Rietveld. (1924)

50
Figure 2.28 Flexibility. Järnbrottshuset. Tage and Anders William-Olsson. (1952)

51
When comparing these three strategies for achieving the same purpose,
generality differs from the strategies of elasticity and flexibility in that the
capacity for handling various preferences is permanent; generality does not
(or does to a far lesser extent) require the action of making physical changes.
This alone is a reason to strive for generality. In a world spending substantial
parts of energy and materials on building and construction, buildings that last
should be preferred rather than buildings that soon require demolition and
rebuilding. Permanence is also relevant as regards architectural qualities.
Regardless of the fact that architectural practice consists in responding to
particular and time-specific requests, the nature of architectural practice is to
construct buildings that should last.55 Where Norwegian apartment buildings
are concerned, architects are involved in the planning and design of most
projects. Concerning the changes made after the original construction of the
buildings, the situation is different; most such changes are done without the
influence of architects. Given that there is some correspondence between
architects’ influence and the quality of the result, which is hard not to assume
as an architect, the strategy of generality is advantageous compared to the
alternative strategies that are more likely to imply rebuilding carried out
without consulting architects.56

Specificity is a strategy contrary to generality in that it aims to satisfy some

particular out of the many possible demands; the nature of specificity is to
respond precisely to needs or preferences that are explicit and time-specific.
As the explicit and time-specific requests captured by the programming in the
process of architectural design, are not likely to correspond to long time
requests57, specificity is a risky design strategy if the intention is architecture
of long-term usability.58 An issue particularly focused on in this study, is the
dichotomy of generality versus specificity of dwellings concerning functions
or use.

Concerning generality versus specificity of dwellings, it is possible to

distinguish between two main aspects of the interior space. One aspect is
features of the individual room, such as size and shape of the room, daylight
conditions and technical equipment. The size of a room is a feature that
strongly influences its potential for use; a room of 1.2 m2 metres without a
window and containing a WC is highly specific regarding function, whereas a

55
Unless buildings are intentionally temporary, demolition of a building is rarely a sign of excellent design.
56
However, from the residents’ point of view, a lack of architects’ influence might not be considered a
problem!
57
As argued by Brand (1994) and discussed in the beginning of this section.
58
In his thesis about “functional performance” of urban space, Marcus (2000, p. 57) points out a similar
conflict between actuality and long-term usability. In brief, a conclusion of Marcus is that cities planned in the
th
20 century have “a high degree of actuality” and “low functional performance” compared with the generality
and thereby the long term usability of urban spaces evolved through the previous centuries.

52
room of 12 m2 with good daylight has a wide range of potential uses. The
other aspect is features of the context of the individual room; the surrounding
spaces might considerably affect the potential use of a room. These two
aspects of space can be distinguished as features internal versus external to
the individual room, where the size of the room is of the first kind while the
spatial configuration is of the latter.59 These two, sizes of rooms and spatial
configurations60 are the particular features of space examined in this study,
and generality versus specificity are properties of space focused upon.

What characterises this study and distinguishes it from many similar survey-
based studies is the attempt at representative sampling and the way that the
spatial layout of different apartments is the explicit subject. When the spatial
configuration and sizes of rooms are chosen as the particular features to be
examined in this study, these parameters are not themselves main subjects of
interest but means of identifying and comparing different layouts of
dwellings. This differs from most surveys referred to in section 2.1 in the
sense that the subjects of the latter are the particular features being analysed.
This difference corresponds to the different aims of the studies. While the
Swedish survey Bostadutforming, bostadsanvändning examined sizes of
rooms in order to evaluate normative guidelines about sizes of rooms, the aim
of this study is to compare different spatial layouts of apartments; while
Bostadutforming, bostadsanvändning examines similar dwellings inhabited
by similar households (in order not to disturb the focus on the features of
particular interest),61 this study aims at sampling apartments that are
representative for larger populations of dwellings.

59
This is a parallel to the distinction of “global” from “local” properties of space done by Hillier (1996, p. 317)
when discussing movement versus occupation. Hillier here distinguishes occupation from movement by
describing occupation as a static and local use of space, while movement relates to “global” complexes of
spaces. “Occupation uses the local properties of specific spaces, movement the more global properties of the
pattern of spaces.”
60
The following quotations shows how Marcus (2000, p. 57) distinguishes speciality from generality:
“... within urban planning and design during the 20th century there is a dominating category, characterised by
a high degree of specificity, as opposed to generality, in that it has more categorical differences between
spaces, more well-defined differences in the relation of spaces, and in general more definition of what can
happen where.”
61
The sample of apartments examined in this Swedish survey was not representative; the apartments were
selected in order to compare similar dwellings (by size) inhabited by similar households (families with
children). More explicitly, concerning the dwellings, the sample consisted of “large” apartments and “small”
houses. Where the households are concerned, all the following categories were excluded: elderly (>70years),
singles, foreigners (!!), single parents as well as households where the only child was a baby or older than 17
years. (Dahlgren and Westerberg, 1979, p. 7)

53
2.5 DWELLINGS AND TIME PERIOD

While basic human needs are not that different throughout the world,
dwellings as well as domestic lives show an impressive diversity.62 Out of
this diversity of dwellings, this study examines Norwegian apartments.
Apartments represent only a minor segment of the Norwegian “stock” of
dwellings, as most of our dwellings are single-family houses.63 Since
apartments (as opposed to most single-family houses) are the kind of housing
where the design usually is created by architects,64 apartments are the kind of
Norwegian dwellings to analyse when the interest is dwellings that are the
results of architectural practice.

The role of the architect designing apartment buildings is different from the
traditional role of the architect designing single-family houses. In the latter
case, the architect works at the request of the future residents and is paid by
them; designing a single-family house is like tailoring a home for one
particular household, responding to preferences specific for that household.65
In contrast to this, future residents rarely have a direct influence on the design
process of apartment buildings. Due to the priority of profit, estate-
developers, constructors and landowners might very well determine design
criteria that differ from the personal preferences of future residents. In
apartment housing, the mission of the architect is thereby not only to please
the employer but also to advocate the needs of all the residents that are
unknown at the time of designing the dwellings. Apartment houses are
therefore a kind of dwelling where evaluation in the light of needs and
preferences of the population in general is highly relevant.

In order to concentrate the focus of this study, the subject is limited to “urban
apartments”. This does not imply that apartments, in order to be included in

62
According to Rapoport (1990) this shows how built environment in general and housing in particular are
results of culture specific preferences or “wants” rather than of more basic human “needs”.
63
Only 18 % of Norwegian dwellings are apartments in buildings of three storeys or more. (According to
Statistisk sentralbyrå / Statistics Norway, http://www.ssb.no/fobbolig/tab-2002-09-23-01.html)
64
While architects have central positions throughout the design process of most Norwegian apartment houses,
less than 10 % of the single-family houses are actually designed by architects. (According to NAL, The
Norwegian Architects Association).
65
Such “tailoring” might be of several kinds, varying from “making what the owner wants” to challenging or
even confronting the expectations of the commissioner. “House III” of Peter Eisenman is an example of the
latter. Here the convention of the master bedroom is confronted by a column placed at what would be the
expected position of the master bed. (Progressive Architecture 1974: 5) Similar, but less confronting, is the
house in Austliveien 30, Oslo, by Carl Viggo Hølmebakk (Byggekunst 2005: 07), a dwelling where tall
windows all down to the floor level allow for more “walk-around-movements” within the rooms by restricting
the conventional positioning of furniture along the walls.

54
the study, must be located within city centres. It means that they should be of
types found in a dense urban fabric; i.e. they should have layouts appropriate
for horizontal and vertical juxtapositions into buildings of a certain density
with respect to the ratio of floor area versus site area.66 The housing concept
of such apartment buildings does not have a long history in Norway; even in
the cities, such dwellings were not common until the very late 19th century.
The layout, the methods of construction as well as the craftsmen were then
imported from abroad, - from Germany in particular. As few dwellings were
built in Norway between 1900 and the 1920s, it was not until the “boom” of
housing construction in the 1930s, that dwellings became a major arena for
Norwegian architects. With the intention of studying Norwegian apartments,
any period from the late 19th century onwards could have been chosen.
However, due to interest in dwellings as a branch of architectural practice and
as a pragmatic result of searching for an appropriate sample of apartments,
the period examined by this study is limited to the one since 1930.67

2.6 THE STANDARD OF HOUSING

The basic aim of architectural practice is to create “good” architecture. Even

though there is no consensus on exactly what this “good” architecture should
be like, the nature of architectural practice is to strive to improve the man-
made physical environment. By this implicit logic of the profession, “bad
housing” is bad by nature and something that either should be demolished or
improved. However, in a wider perspective, low standard housing is not
necessarily bad, at least not in Norway where the dwellings of the lowest
standard are fairly good by international comparison.

Within a liberal economic market of dwellings, as is the situation in Norway,

the cost of living is closely related to the standard of the dwelling. In such a
situation, low quality dwellings represent a positive potential for sub-groups
of the population; the existence of dwellings that are considered as
unacceptably bad by the majority, is a condition for a preferred way of living
for others. An example of such a housing area in Norway is “Svartla´moen”
in the city of Trondheim. Due to politically determined demolitions that were
never carried out, the dwellings of this area turned out to be municipally
owned and of a low standard. The area was not a slum, but consisted of

66
In Norway, this density is usually described by the floor area of a building as percentage of the site area, as
defined in NBI byggdetaljblad 310.220, a standard formulated by the Norwegian Building Research Institute.
67
The sampling of apartments is described in section 4.1.

55
simple older houses, poorly insulated and of low technical standard. As many
of these houses were empty, awaiting demolition, students, artists and others
managed, with or without the will of the local authorities, to get permission
to live there at low cost. Thanks to this development, “Svartla`moen” has
been an oasis for a generation of artists and musicians in Trondheim.68

For the majority of the population as well as for politicians, planners and
architects, such low standard housing areas are likely to represent problems
that should be solved. In order to prevent real and permanent slums, striving
for better standards in the worst housing areas is a reasonable aim even in
Norway. However, in order to keep the positive potential of low cost
dwellings, we should be not to be too eager to make such improvements. A
“balanced” municipal strategy in these respects, might be to prevent parts of a
town from developing into permanently bad areas without worrying if not all
dwellings are of the standard considered decent by the majority; any town
should at any time have some areas where different kinds of alternative living
are given a chance. Even though this perception of “bad housing” should be
born in mind when planning of housing is concerned, the subject of this
thesis on architecture adheres to the conventional aim of architects, striving
for dwellings that are “good” rather than “bad”.

Figure 2.29
“Svartla’moen”,
Trondheim.

68
Christiania in Copenhagen is another example of a low standard housing area that for some groups of the
population represents positive possibilities while for others it represents the bad and the ugly. Apart from not
necessarily being spacious, these kinds of “bad dwellings” have many features in common with what Brand
(1994) categorises as “the low road”: ..What these buildings have in common is that they are shabby and
spacious. Any change is likely to be an improvement. They are discarded buildings, fairly free of concern from
landlords or authorities: “Do what you want. The place can’t get much worse anyway. It’s too much trouble to
tear down.” (Brand, 1994, p. 24) “Freedom is cheap. Low rent equals high control if you’re comfortable fixing
up crude Low Road Space.” (ibid., p. 32)

56
3 Methodology

This study is carried out as two empirical surveys, where one has been a
diachronic69 survey of the development of apartment layouts and where the
other has been a synchronic70 survey of contemporary living. The first has
been a quantitative inquiry of room-sizes and spatial configurations in a
historical sample of apartments, while the second, aiming at shedding light
on living in apartments that differ by floor plan layouts, has been carried out
as an interview-based and more qualitative survey.

This chapter explains this methodological structure more in detail, the

particular methods that have been applied and the choice of case. After
commenting upon research methods in general and within studies of
architecture more in particular, it will be described how some different
methodological approaches are combined. An important approach to the first
of the two empirical analyses has been the methods developed within the
architectural research field space syntax. This chapter, in section 3.4, reviews
basic space syntax methodology, before explaining the application of space
syntax in this particular study. Besides applying existing space syntax
methods, this study has itself contributed to the methodology of space syntax
by developing the software AGRAPH, which is presented in appendix
A.3.1.71 As the two empirical inquiries constitute major parts of this study,
this chapter about methodology also contains a section (section 3.5) that
sheds some light on the relationships between theory and empirical data by
commenting on assumptions, findings and generalisations. The empirical
surveys and their results are presented in the following chapters 4 and 5.

69
Diachronic: through time; study of a historical development, of development through time.
70
Synchronic: at a given stage; study of a subject at a given time.
71
This appendix is a paper presented at the software session at “The 5th Space Syntax Symposium” at TU-Delft,
2005. (Manum, 2005)

57
3.1 THE QUANTITATIVE AND THE QUALITATIVE;
COMMENTS ON DISCIPLINES AND THEORIES

A distinction frequently applied within research methodology is the one

between the quantitative and the qualitative.72 Where actual research practice
is concerned, these stringent categories constitute a theoretical framework
rather than being appropriate descriptions of the methods applied. As most
research practices contain elements of quantitative as well as qualitative
character it is hard to identify “pure” quantitative or “pure” qualitative
research; few research questions of quantitative character are without any
qualitative aspects and few qualitative research issues are without any
quantitative elements.73 When characterising different academic disciplines
or when discussing research methods, a more nuanced characterisation than
the dichotomy of the qualitative versus the quantitative is useful.

Based on an understanding of intentionality according to von Wright (1975)

the Norwegian civil engineer and scholar Bjørn Sandaker (2000), discusses
building structures as intentional objects in his thesis “Reflections on span
and space”. When he describes different aspects of functions within
architecture, Sandaker suggests a “scale” from mere explanation (causality)
on the one end, towards mere interpretation (intentionality) on the other. He
positions natural sciences at the explanatory end of the scale, studies of icons
and meaning at the interpretational end and studies of building structures
somewhere in the middle. This principle of ordering according to the
applicability of interpretation, or according to potential for interpretation, can
be applied in order to discuss methods and data on a detailed level and in
order to describe research and academic disciplines in general.74 As most
academic disciplines are constituted by their particular theoretical and
methodological framework, the character of these frameworks is a way of
describing the disciplines. Even though a fixed potential for interpretation

72
According to Creswell (1994, p. 1-5) and Patton (1987, p. 9 and 17) a generalised description of the
quantitative and the qualitative paradigm, where a paradigm encompasses theories as well as methods, could
be as follows: Within the quantitative paradigm the reality is seen as something that can be measured
objectively; the researcher is independent of that being researched, while a qualitative paradigm is based on an
interpretative approach where the researcher interacts with what is being researched. Quantitative studies are
characterised by handling a limited set of questions on large samples, often by numerical data appropriate for
statistical analyses, while qualitative studies are more in-depth and detailed, usually on smaller samples or a
few cases, emphasising more holistic perspectives.
73
Glaser and Strauss (1967, p. 17) emphasise that “there is no fundamental clash between the purposes and
capacities of qualitative and quantitative methods and data”. Yin (1994, p. 15) argues about the same as he
writes “there is a strong and essential common ground between the two”. Creswell (1994, p. 173) describes
the combining of qualitative and quantitative research more in detail.
74
Rolf Johansson (1997, p. 41) describes the correspondence between kind of data and kind of research
methods. Data used in statistical surveys are usually large samples where a few quantitative elements are
focused on (elements where the potential for interpretation is low), while anthropological case studies intend to
synthesize a wide range of input (allowing for or requiring a high degree of interpretation).

58
within a discipline is not a perfect conceptualisation of the research character
of a discipline, some differences between disciplines can be illustrated by
ordering them according to the character of their theoretical and
methodological framework, as illustrated in figure 3.1.

Figure 3.1
Academic disciplines characterised by ”potential for interpretation”

Contrary to what is the case for most well established scholarly practices, the
discipline of architecture can hardly be identified by any particular theoretical
or methodological framework. What architectural research have in common,
is that buildings and cities, or theories about buildings and cities, are studied;
architectural research is constituted by the subjects of interest rather than by
any consensus about relevant theories or methods for studying these.75 The
kind of theoretical frame that is relevant for a particular research within the
field of architecture, depends on the aspect of architecture that is focused on
and can therefore be of very different kinds, ranging all along the scale of
“potential for interpretation”, from the mere quantitative to the mere
qualitative. This diverse nature of possible architectural research explains the
fact that a theoretical framework relevant for major aspects of practice and
research within the field of architecture has still not been found, and probably
never will be.

75
This difference between architecture and the more established academic disciplines can be elaborated by
applying the distinction of Hillier (1996, p. 40-42) between ”ideas to think of” and ”ideas to think with”, where
Hillier points out that the latter is what constitutes theory. According to Hillier, most theorising about
architecture has the character of ”ideas to think of” rather than ”ideas to think with”.

59
From the situation described above, different conclusions have been drawn.
One is to reject the relevance of scientific approaches as they cannot grasp
the diverse nature of architecture,76 another is to dismiss the relevance of
theory in general.77 If the aim is to develop knowledge, a more fruitful
strategy is to accept that the field of architectural research consists of a
variety of possible research questions with a corresponding variety of
relevant theoretical approaches, and then to discuss and develop theories and
methods that are appropriate within the different aspects of architectural
research. Even in housing research, a minor branch of architectural studies,
the possible approaches are diverse; one approach for studying standards of
housing can be statistical surveys of floor area per person, while to study the
home as a representation of personal identity, would require a very different
approach. Even this study, focusing on a few aspects of apartment layouts,
shows a wide range wide concerning the quantitative versus the qualitative;
the sizes and configurations of interior space, which are the features
examined in the diachronic survey, are of the former kind, while the
residents’ judgements of their homes are of the latter. As the character of
relevant methods corresponds to the subject and the data, the methods applied
are diverse in accordance with the character of the data, ranging from simple
statistics to case studies and interviews. The approach of this study has been
that the applied methods might very well be of diverse kinds as long as they
are adequate for the subject examined at the specific stage of the study and
that this subject relates to the main research question. The following sections
describe how this study has combined different methodological approaches.

76
One example of this is Jan Christiansen who claims scientific criteria to be useless for evaluating quality of
housing. ”Der kan ikke opstilles objektive, videnskabelige kriterier for måling af boligkvalitet. Dertil er der for
mange irrationelle forhold, der spiller inn. Forhold som ikke altid er målbare og slet ikke med nogen fælles
målestok.” (Christensen, 1996, p. 5) Another example is Scruton (1979) who dismisses any theory that does
not cover all aspects of architecture, as commented in section 2.3.
77
It is not rare among architects to regard theorizing on architecture as a “parasite” on the art of architectural
design-practice, similarly to the attitude that can be found among artists toward art historians.

60
3.2 THE RESEARCH STRATEGY

The empirical sections of this study consist in the initial diachronic analysis
of the development of apartments through time, followed by the synchronic
survey of contemporary living in some particular apartments.

The aim of the initial and diachronic analysis has been to shed some new
light on the development of apartment layouts through the recent 75 years.
By analysing room sizes and spatial configurations in an appropriate sample
of apartments, it should be possible to identify apartments that are different
with respect to spatial layout. This has two purposes. If significant changes
through time are found, this might itself contribute to the knowledge about
Norwegian housing in a historical perceptive. The other purpose is to
establish a basis for the synchronic study; the apartment layouts identified
through the diachronic analysis is what will be examined and compared in the
synchronic survey of different apartments as places for contemporary living.

The diachronic inquiry examines two features of “the physiognomy of the

void”; these are the sizes of the rooms and the spatial configuration of the
apartments. The assumption is that analysis of these parameters will bring
about findings relevant for discussing generality versus specificity. The sizes
of rooms are examined by measuring the floor areas of the rooms of the
selected apartments, while the spatial configurations are analysed by space
syntax methodology. The application of space syntax is particularly
explained in section 3.5. The fact that the floor areas as well as the space
syntax parameters can be measured or registered without much interpretation,
implies that the diachronic analyses of the rooms and the spatial
configurations (as these analyses are figured out here), can be carried out
without entering the field of qualitative discussions.78 This is advantageous in
the way that it makes it possible to analyse a larger number of apartments, as
intended in this study.

78
In this way of placing the evaluation of quality at particular stages of the research, I am inspired by Cuff
(1995), even though this study otherwise is not very similar to that of hers. Cuff has done a study of excellent
architecture without herself making any evaluation of architectural quality. In her study on the premises for
creating and producing excellent architecture, she has defined an architectural object to be excellent when it is
considered so by three defined “referees”: those involved in the production (the owner included), the public
and the architectural profession. By this strategy she manages to place her own research question outside (or
more correct: chronologically after) the actual evaluation of quality. By the research design of my study I have
intended to do the opposite by consciously carrying out a major section of work prior to discussions or
evaluation of quality.

61
What has been studied by the initial diachronic analysis is not people’s living
but some specific features of the interior space. The synchronic and more
qualitative study, attempting to examine different kinds of apartments as
contemporary dwellings, is carried out by comparing the living in apartments
that by the diachronic analysis are found typical or otherwise interesting.79 To
elaborate, the strategy has been to select a few kinds of apartments (which are
identified by the initial study), and then to carry out an interview-based
survey in some apartments of each kind. By registering who lives in these
apartments, how these residents have chosen to use their apartment and ask
for their opinions about living there, it should be possible to bring about data
relevant for discussing apartment layouts in relation to kinds of households80
and their preferences of living. Given that sizes, shapes and configurations of
interior spaces influence domestic life, which is a basic assumption of this
study, there should be some correspondence between spatially significantly
different apartments and the contemporary living in them. An assumption is
that apartments with layouts characterised as general will house a greater
range of households than apartments where the spatial configuration or the
individual rooms are more specific as far as functions or use are concerned.
The data for examining this will be questionnaires and interviews as well as
statistical data on population and demography.81

The aim of this methodological strategy, which consists in applying different

limited foci at different stages of the study, has been to bring about some
knowledge about specific aspects of floor plan layouts and living that would
probably not have been found by using approaches of wider scope throughout
the work.

79
According to Glaser and Strauss (1967, p. 62), the sampling of apartments for the synchronic survey will
thereby be “theoretical” rather than statistical. Glaser and Strauss (1967, p. 45) describe “theoretical” sampling
as a process where the selection of data is “controlled” by the emerging theory and not by preconceived
systematic procedures (as in statistics). They doubt the importance of ensuring statistical sampling unless
really intending to apply statistics; if intending to generate theory they claim “theoretical” sampling to be more
appropriate.
80
Different kind of households is here to be understood as differences concerning parameters like age, sex and
relationships. Due to Norwegian legislation concerning anonymity of ”personal information”, parameters like
cultural and national background or income are not included in this study. See section 5.2.3 for more detailed
information about kinds of households.
81
The questionnaires and the interviews are explained in the first sections of chapter 5.

62
3.3 THE CHOICE OF CASE

With the intention of examining the development of the interior space of

Norwegian apartments, an initial question was what apartments might be
appropriate as the empirical basis. What was looked for, was a sample that
was not only a particular case but also a case representing Norwegian
housing some more in general. The practical work of getting access to the
documents needed in order to select the apartments, had also to be born in
mind. After various models of cases and sampling had been evaluated, the
decision made was to select OBOS. OBOS, “Oslo Bolig Og Sparelag”, is the
main co-operative housing company in Oslo, it was founded in 1929 and is
the largest co-operative housing company in Norway. They have been in
charge of more than the half of the housing production in Oslo for long
periods. They are still a major developer of housing in the Oslo-region. Even
though a large number of architects have been engaged in the OBOS projects,
co-operation with leading and prize-winning architects has by OBOS always
been seen as an important strategy for developing quality housing within
reasonable cost. When intending to shed light on the development of
Norwegian urban housing in general by doing a case study of one single
actor, it is hard to find a better candidate than OBOS.82 Figure 3.2 is a brief
review showing the development of their housing projects since the 1930s.

1930s

Figure 3.2 a, b
A brief pictorial review of
housing projects by OBOS. 1950s
(Photos : OBOS)

82
A closer description of OBOS is published in the Norwegian architectural magazine Byggekunst (Sjølie,
2004).

63
 1960s

1970s

1980s

Figure 3.2 c- f
A brief pictorial review of
housing projects by OBOS. 2004
(Photos: OBOS)

64
When Yin (1994, p. 39) discusses kinds of case studies, he identifies four
basic kinds by applying two distinctions, one between holistic / embedded
and one between single-case / multiple-case. (See figure 3.3) According to
Yin, an embedded case study is advantageous when logical subunits within
the case can be defined and examined and when the phenomenon of interest
is really about these subunits. The problem might occur, he argues, when the
original phenomenon of interest is not about these subunits but about a larger
unit. If such a study fails to return from the subunits being examined back to
this larger unit, the original phenomenon of interest has become the context
instead of being studied.

single-case multiple-case
holostic
(single unit of analysis)
embedded
(multiple units of analysis)

Figure 3.3
This study of OBOS positioned according to Yin’s basic types of case study designs.

In principle, the methodology of case studies differs fundamentally from

statistics. However, as Rolf Johansson (1997, p. 67) points out, this does not
mean that they cannot be combined; statistical methods might very well be
useful in analysing the subunits within a case.

If this case study of the OBOS-apartments is to be characterised according to

Yin and Johansson, it would be as an embedded single-case study where
statistics are among the methods applied in order to analyse some of the
selected subunits within the case. The study is embedded in the way that the
research question concerns only some out of the many possible aspects of
OBOS’ housing projects. As the interior space of apartments, which is the
unit of analysis, really is a subunit of urban housing and of the OBOS
projects, this embedded kind of case study should, if using Yin’s criteria, be
appropriate for the research question. If the research subject had been a more
general approach to urban housing, then the case study should have been of
the holistic kind. When the single-case/multiple-case distinction is concerned,
the particular study of OBOS is of the single-case kind. However, as we will
come back to in chapter 6, additional information in terms of a comparative
case is useful in order to shed light on the general relevance of the result.

65
3.4 BASIC SPACE SYNTAX METHODOLOGY

As mentioned in section 2.3, space syntax is theory and methods for studying
space and spatial configurations. A basis of space syntax is the works of Bill
Hillier and Julienne Hanson at the Bartlett, University College London.83
Space syntax analyses internal positioning of spatial elements within a spatial
configuration. Extensive empirical studies have shown that analyses of
spatial configurations carried out by the methods of space syntax correlate
convincingly to many aspects of traffic, movements and use of in-door rooms
as well as outdoor urban spaces.84 This section summarises basic space
syntax methodology, first by showing how space syntax approaches
configurational features of space and how the basic space syntax parameters
are defined, then by describing the different kinds of space syntax modelling
and how these are applied.

Figure 3.4 Basic configurational properties. (Hillier, 1996)

83
The most basic references of space syntax are Hillier and Hanson (1984), Hanson (1998) and Hillier (1996).
84
Such empirical studies are counting of traffic and movements of people, in cities as well as buildings.
Examples of such studies can be found at the web-sites of UCL (http://www.space.bartlett.ucl.ac.uk/) and
Space Syntax Limited (http://www.spacesyntax.com/). See also the example in figure 3.23.

66
Some basic properties of spatial configurations can be represented by graphs
consisting of nodes and connections85 as shown in figure 2.19 and 2.20. Such
images of nodes and lines that represent spatial elements and the connections
between them, are called “connectivity graphs”. Figure 3.4 shows three
different configurations of the same elements a, b and c presented by images
of physical objects and by connectivity graphs, similarly to the “floor-plans”
of two spatial configurations and the corresponding connectivity graphs in
figure 2.19. As pointed out by the connectivity graphs in figure 3.5, the
spatial configurations might differ significantly even within a sample of very
simple dwellings.

Figure 3.5
Some simple dwellings represented by floor plans and connectivity graphs. (Hansson, 1998)

Numerical parameters describing configurational features of the single

elements as well as of the system of connected elements can be determined
by “counting” the internal distances between the elements. In space syntax
terminology, distance is to be understood as the number of “space-steps”
from one element to another, which rarely corresponds to metric distance
between the same elements. In basic space syntax, the distance between two
elements has the value of one when the elements are directly connected.86
The very basic space syntax calculation is to determine the shortest such
internal distances between all the elements of a system. The further space
syntax analyses consist in determining more elaborated space syntax
parameters by mathematical and statistical operations on these “internal
spatial distances”. For small systems, the space syntax parameters can easily
be determined by manual calculations. For systems of more than 5-10
elements the calculations become too time-consuming to be done without

85
Or ”vertexes” and “edges” according to the terminology of graph theory and combinatorics, which is the
mathematical basis of space syntax.
86
About more advanced modelling: see Figueiredo and Amorim (2005) about axial-line modelling with curved
lines or Turner (2005) about taking the angle between intersecting lines into consideration (by introducing
other “connection-values” than only 0 and 1).

67
computer-programs. In “axial-line analyses” of cities and in “visual-field
analyses” the numbers of elements easily become thousands. In such analyses
efficient computer algorithms are required in order to manage the
calculations within a reasonable time. Through elaborating the space syntax
parameters and comparing them to empirical studies, several kinds of
modelling and analyses have been developed. The major kinds of modelling,
which are “nodes and connections”, “axial lines” and “visual-fields”, will be
described after explaining the most basic space syntax parameters.

3.4.1 Justified Graphs

“Justified graphs” is a way of picturing a configuration of elements “as seen

from one particular element”. The element, from which the configuration is
“seen”, is termed the “root” of the justified graph. Figure 3.6 shows four
justified graphs for one configuration and points out how one configuration
appears significantly different depending on which node that is selected as
“root”. Figure 3.7 shows two justified graphs of the same system, one with
element A and the other with element B as “root”. Each vertical level of
nodes in the justified graphs represents the distance of one “space-step”
further away from the root node, i.e. the distance from the root increases by
1.0 for each level, as showed in figure 3.7.

Figure 3.6 Justified graphs as seen from different “root-nodes”. (Hillier,1996)

Figure 3.7
”Space-steps” or “depths”
from the root-note.
(Klarqvist, 1991)

68
3.4.2 Total Depth and Mean Depth;
Measurements of Space Syntax Integration

Figure 3.8 shows a linear configuration of seven elements where an 8th

element is added in different positions. The number within each element
shows the “total depth” of each element, TD(n), which summarises the
shortest distance from an element to all the other elements. The number listed
below each figure is the total depth of the system. This is found by simply
summarising the TD(n) of all elements in the system. The figures illustrate
how the position of an additional element influences not only the depth of
this added element but also the depths of all other elements and the total
depth of the system; when adding an element to an existing system, the total
depth of the system increases less when the new element is placed in a central
position than is the case when it is connected to the far ends of a system. The
“mean depth” from an element “n”, MD(n), is the mean of the shortest
distance from that element to all the others. If “k” is the number of elements
in the configuration, k-1 is the number of internal distances. The mean depth
of an element is therefore found by the formula MD(n)=TD(n)/(k-1). In space
syntax terminology the positioning according to internal distances is termed
“integration”; the element with the lowest depth (with the shortest total
distance to all the other elements in the system) is the one most integrated
while the element with highest depth is the one least integrated (or the one
most segregated).

Figure 3.8.
A line of 7 elements where an 8th element is added in different positions. (Hillier, 1996)

69
3.4.3 “Relativised” Measurement of Space Syntax Integration

As the size of a system in terms of number of element grows, the “depths “

between the elements increase. The value of the integration parameter “mean
depth” therefore depends on the size of the system. In order to achieve a
parameter that describes “integration” without this direct correlation to the
number of elements in the system, the parameter “relative asymmetry” (RA)
has been developed. RA describes the “mean depth” of an element by a value
between (or equal to) 0 and 1, where 0 represents the most integrated position
(“the shortest depth”) possible and 1.0 represents the most segregated
position possible.87 If “k” still is the number of elements in a system, then
such a parameter RA for a node “n” can be described by the formula
RA(n)=2*(MD(n)-1)/(k-2).88 Figure 3.19 shows the TD, the MD and the RA
for all elements of the three basic configurations a “hub”, a “line” and a
“ring”. The nodes are coloured in accordance with their integration in the
way that red represents the highest integration possible (RA=0.0) and dark
blue the lowest integration possible (RA=1.0).

Figure 3.9
Three fundamental 5-node-graphs (coloured by integration) and some of their basic space syntax
parameters. (Graph-images and tables are output from AGRAPH)

87
RA=1.0 represents the value of the end node of a linear sequence of elements (which is the most
“segregated” position possible), while RA=0.0 represents the value of a node that is directly connected to all
other nodes a system (which is the most “integrated” position possible), se figure 3.9.
88
This formula is explained in appendix A.3.1, note 4.

70
3.4.4 “Normalised” Measurements of Space Syntax Integration

In theory, the RA-values characterise the integration of an element by a “size-

independent scale” ranging from 0.0 to 1.0, which are the maximum and
minimum integration possible. However, in despite of the “relativisation”
with respect to the size of the system that is captured by the RA, RA-values
of real configurations are not directly comparable across systems of different
sizes.89 The reason is that the level and the distribution of RA-values in
configurations correlate to the size of the configuration; configurations tend
to be relatively less deep as they grow.90 In order to handle this, the
parameter Real Relative Asymmetry, RRA, has been introduced. RRA
represents a linear scaling of the RA values, where the scaling factor depends
on the number of elements in the system. The scaling factor is defined as the
RA-value of the root element in a “reference configuration” of the same size
(size in terms of number of elements) as the configuration being analysed.
(See figure 3.10.) The RA-values of all elements in a system are scaled into
RRA-values (RRA=RA/D), where the value of the scaling factor D is found
by a table91 or by calculation92. As the RRA-values represent a linear scaling
of the RA-values, the RRA-values capture no information not already present
in the RA-values when analysing one particular configuration. In any space
syntax study that either concerns on one particular case (which is the case in
most urban analyses) or in studies that examine order of integration-value of
the elements (which is often the case when studying many dwellings), to
apply the RA or the RRA makes no difference. In such studies, nothing is
achieved by transforming the easily explainable RA into the less obvious
RRA. However, if comparing exact values of space syntax parameters of
configurations of different sizes, normalisation must be taken into account. 93

89
This is argued by Hillier and Hanson (1984, p. 111) and Hillier (1996, p. 52).
90
As the size of configurations grows, it is more likely that an addition element connects to highly integrated
elements (elements in the “centre” of the configuration) than to segregated ones; in small systems, the
probability for a linear-like configuration (and thereby high RA-values, i.e. close to 1.0) is higher than what is
the case for large systems.
91
A table of D-Values for systems up to 250 elements is shown in Hillier and Hanson, 1984, p. 112.
92
The reference-configuration for determining the D-value is a ”diamond-shaped graph” as shown in figure
3.10. For more about this graph or for the formula of the D-values, see Krüger (1989), Asami, Kubat,
Kitagawa and Iida (2003) or (for the formula) see the users guide of the software “Mindstep”, by Figureiredo
and Amorim, http://www.mindwalk.com.br/papers/Figueiredo_2005_Space_Syntax_Software_en.pdf.
93
How the levels of RA-values vary according to size of the system, which is the problem intended to be
solved, and how this is solved by the D-values, is not easy to find out though the main syntax literature. Space
syntax publications usually refer to Krüger (1989), who describes two kinds of “standardisations”, one where
the scaling factor is the root of a “diamond shaped pattern” and one where it is the “corner of a grid”, see
figure 3.10. Krüger proposes to apply the latter on RA-values of “node-grid-graphs”, while the one to apply on
“axial-line-graphs” depends on the “significant level of the difference of their mean RA”. (Ibid, p. 33) Among
those who are interested in the theoretical details of space syntax, several have questioned the validity of the
RRA and have therefore applied or proposed other kinds of normalisations. Thaler (2005) applies
normalisation according to Teklenburg (instead of the RRA according to Hillier/Hanson and Krüger), while
Park (2005) more in detail discusses measurements of space syntax integration.
The “normalisation” described here, is sometimes termed “standardisation” for instance by Krüger (1989).

71
Figure 3.10 The two kinds of “reference configurations” according to Krüger (1989), where
the D-value is derived from the “diamond-shaped-pattern” on the left.

3.4.5 Integration Value

The fact that RA (and thereby the RRA) is of low value when an element is
highly integrated and vice versa might cause misunderstanding. By
introducing the parameter integration-value, “i”, which is defined as the RA
(or the RRA) inverted (i=1/RA or i=1/RRA), the logic is more obvious in the
sense that a high value means high integration.94 As all these parameters (RA,
RRA, 1/RA and 1/RRA as well as other normalisations of RA) describe the
phenomenon space syntax integration, which one that is applied in a
particular space syntax analysis should be stated explicitly.95 The figure 3.11
– 3.13 illustrates the integration value of the elements in a grid; in figure 3.11
all elements are connected to their “grid-neighbours”, in figure 3.12 some
connections are cut, while figure 3.13 shows the integration when some new
connections are added. This is done by applying the software AGRAPH that
automatically generated the coloured images. The nodes are coloured by the
option “integration, relative”, which means that the most integrated element
is read and the most segregated is dark blue.96

94
Due to this definition, the integration value (”i”-value) has a disadvantage compared to the RA; in “hub-
like” configurations, which is the configuration of many small dwellings, “i-values” might be infinite (as
RA=0.0 for a node that is directly connected to all other nodes in a system). This is not elegant and could be
avoided by instead defining integration as i(n)=1-RA(n), which means i(n)=(K-2*MD(n))/(k-2); achieving a
parameter of integration that (similarly to the RA) ranges between (or equals to) 0 and 1. However, as
i(n)=1/RA(n) is the established definition within the space syntax methodology, the i-value proposed above
will not be applied here.
95
Integration (by RA, RRA or “i”) can be calculated within different “radii”. What is described here so far, is
“global integration” were the distances “counted” are the shortest ones from one element to all other elements
in the system. Other “radii” of integration means to count the distances only to the element within other
distances (within other “radii” of space syntax steps), such as radius-3 or radius-10. Integration of various radii
is often applied in urban space syntax analysis since the results of calculating within different radii have shown
to correlate to different kinds of activities in the city; to put it simple, radius-3 or radius-5 correlates well to
pedestrian movements while larger radii correspond better to car-traffic. As this is not the part of space syntax
that I know best and is not at all applied in this thesis, details of urban analyses will not be described any
further here.
96
The alternative “integration-colouring” is “integration-absolute”, which implies that red is the highest
integrated value possible (RA=0,0) and dark blue is the lowest (RA=1.0).

72
Figure 3.11
A regular 7x7 grid,
all nodes connected
to their “neighbours”.
(“coloured by integration”)

Figure 3.12
Some connections cut.
(“coloured by integration”)

Figure 3.13
Some connections added.
(“coloured by integration”)

73
3.4.6 Control Value

Control Value (CV) is a parameter describing to what extent one element is

“controlling” the access to the elements to which it is connected. The value is
found by letting each element “give away” the total value of 1.0 equally
distributed to its “neighbouring” elements. The CV-value of a node is the
value received by the node through this operation, as showed in figure 3.14.
As shown in figure 3.15 control values do not have the “centre-effect” as is
the case for integration. (Compare figure 3.11.)

Figure 3.14
Calculating “control values”. (Klarqvist, 1991)

Figure 3.15
7x7 grid “coloured by
control-values”.
(AGRAPH)

74
3.4.7 Spatial Topology; Basic Kinds of Positions in a Configuration

When describing positions within a spatial configuration, Hillier (1996,

p.318-327) has distinguished between four kinds of positions and termed
these A-, B-, C- and D-types of spaces. In short, the A-type is a space
connected to one other space only. B is a space that is the connection to a
space or a sub-complex of spaces without rings and without itself being on a
ring, while C and D categorise different positions on rings. C is positioned on
one ring, while D is positioned on two or more rings. (See figure 3.16) This
typology of positions captures features of space that are essential concerning
movement versus occupation. Generally, when a space B must be passed
through in order to access a “dead-end” space A, the use of either of these
spaces strongly influences the potential use of the other. The potential use of
space A depends on the space B not restricting the passage to A. Similarly,
the potential use of space B depends on the kind of through-passage
generated by space A. Spaces of C- or D-types differ from B-spaces with
respect to through-passage; through passage is not necessary in any particular
space on a ring, as there are alternative routes.

Figure 3.16
Basic topological properties described by basicl positions in a configuration.
(According to Hillier, 1996)

75
3.4.8 Kinds of Space Syntax Analyses

The modelling of rooms by nodes and connections (or vertices and edges), is
one of several ways of modelling spatial configurations when applying space
syntax methodology. If we distinguish space syntax analyses according to the
spatial units applied in the modelling, there are three kinds or three “modes”
of analyses; in addition to the “node-mode”, there are an “axial-line mode”
and a “visual-field mode”. Figure 3.17 shows the visual images that are
typical for analyses carried out by each of these three kinds of space syntax
modelling. Most spaces can be analysed by applying any of these kinds of
modelling. However, as the different modelling-modes capture different
aspects of space, some modelling is likely to be more relevant than others.
Which kind of modelling that might be best in a particular case, depends on
the kind of space that is studied as well as on the subject of interest.
Regardless of the kind of modelling, a kind of visualisation often applied in
order to present the results of space syntax calculations is colouring of the
elements according to the value of selected space syntax parameters. The
illustrations in figure 3.17 are a node-graph, an axial-line map and a visual-
field image. These kinds of coloured images clearly illustrate basic features
of spatial configurations and have become the graphic icons of space syntax.
(See also the figures 3.9, 3.11, 3.12, 3.13 and 3.15.) In order to communicate
results of space syntax analyses to the public or to other professionals these
kinds of illustrations are highly useful. When it comes to more detailed space
syntax analyses, the data are examined more closely by comparing the exact
values, by looking at the rank of integration across the spatial elements or by
doing statistical calculations on the space syntax parameters.

Figure 3.17 (a, b, c)

The three modes of space syntax modelling: nodes, axial-lines and visual fields / isovists.

76
In “node-mode” the modelling is done by identifying the connectivity graphs
directly in terms of drawing nodes and connections. The nodes (vertexes)
represent spaces while lines (edges) represent connections between the
spaces. The figures 3.9, 3.11 and 3.13 are such “node-and-connection”
representations. A node might represent a real room in a building (an
enclosed space connected to other rooms by doors or door-like openings) or a
“subspace” within a room, such as a “convex space”. A “convex space” is a
space where no straight line between any two points in the space crosses the
surface of the space, or more literally: a space where the entire space is
visible from any point in the space. (Hillier and Hanson, 1984, p. 98) Figure
3.18 shows a “convex” and a “concave” space and how a concave space can
be subdivided into several convex ones. Due to the shape of the rooms and to
the perception of the spaces, node modelling is often applied in analyses of
dwellings; as most dwellings consist of enclosed spaces connected by doors
or door-like openings, a modelling by nodes (representing enclosed rooms)
and connections (representing the access between rooms) is likely to capture
relevant features of the spatial configuration. Concerning the conception of
space, the residents of a dwelling know the interior space well. This makes
node-modelling favourable in most analyses of dwellings, as the alternatives
axial-line or visual-field modelling both capture perceptions of space that
depend on views rather than on a priori knowledge. Figure 3.19 shows a floor
plan drawing, a subdivision of the rooms into “convex-spaces” and the
connectivity “node-graph” of the convex spaces as “seen from” the entrance
of the dwelling.

Figure 3.18
Convex spaces. (Klarqvist, 1991)
Left: “convex” and “concave” space.
Right: examples of how to divide a concave space into several convex ones.

77
Figure 3.19
Floor plan, convex-space
and connectivity graph.
(Klarqvist, 1991)

Figure 3.20
Axial lines represented
by a “graph”.
(Klarqvist, 1991)

In the second mode, - the axial-line mode, the space is represented by straight
lines. By basic axial-line modelling the space of interest is modelled by
“fewest and longest straight lines covering all convex spaces”. In the
mathematical graph-representation each line is represented by one node. For
lines crossing each other, the nodes representing the respective lines are
defined as connected. This is shown in figure 3.20. The axial-line modelling
captures essential features of the continuous outdoor space between buildings
in a city, a spatial configuration where long spaces of streets intersect each
other. This is a kind of space very different from enclosed interior rooms
connected by doors. The activities of peoples in streets are also very different
from the activities taking place in the home; the pattern of human movements
in streets has more the character of free-floating movement than is the case
inside private dwellings. As such free-floating-movements corresponds very
well to space syntax integration of axial line models, this modelling is the one
usually applied for analysing continuous spaces such as outdoor urban space.
Even though more advanced linear modelling97 correlates better to real life
situations in some particular cases, the basic, very simple axial-line
modelling is still the line-modelling most frequently applied. Figure 3.21
shows a map of a town and the corresponding axial-line-map, while figure
3.22 shows a part of an “axial-map” of London.

97
Such as applying curved line-elements or taking the crossing angle and the length of lines into consideration,
see footnote 86.

78
Figure 3.21 Axial-line modelling of streets and public spaces by drawing ”fewest and
longest lines covering all convex spaces”. (Hanson and Hillier, 1984)

Figure 3.22 “Axial-map”, part of London, coloured by integration. (Hillier, 1996)

79
In the third mode, the spatial element on which the calculation is based, is the
“visual-field” or the “isovist”. The “isovist” of a particular point in a space is
the part of the space that is visible from that point. The space that can be seen
from all points in this “first isovist” is at “distance one” from the first isovist
(being “connected” to the first isovist, if using the terminology earlier
applied) and so forth. For spaces not known in advance by the persons being
present, “visual fields” are likely to be a relevant modelling; in such spaces,
“what you know is what you see”. Visual fields analyses are therefore often
applied in studies of larger buildings where movements have the character of
“free float” of persons, such as museums or shopping malls, where many of
the people present are likely to not know the space well in advance. In studies
of domestic spaces, visual-fields are less relevant as the consciousness of the
space in a dwelling to a far lesser extent depends on what you actually can
see from each position in the rooms. Visual-field-analyses are now usually
done by the software DEPTHMAP.98 Figure 3.23 shows how the result of a
space syntax visual field analysis of the Tate Gallery corresponds to the
movement of visitors as registered on site.

Figure 3.23
Tate Gallary, space syntax visual field analysis (left) and movement of visitors in reality (right).
(Space Syntax Ltd.)

98
DEPTHMAP is available from Space Syntax Limited, http://www.vr.ucl.ac.uk/research/vga/.
A more advanced visual-field-modelling is applied in the software SPATIALIST developed by Peponis and
Wineman (See Peponis et al. (1998), http://undertow.arch.gatech.edu/homepages/jpeponis/FormalModels.htm
or http://www-personal.umich.edu/~jwineman/formalModels.html), where the visual-fields are subdivided in
accordance with their boundary-walls. This kind of modelling captures some essentials of human perception;
the modelling captures the fact that a new wall occurring in the isovist of a real person represents an additional
step of cognitive information. For more about “visual-field graph analyses” see Do and Gross (1997) or
http://depts.washington.edu/dmachine/PAPER/CF97-ISO/spatial.html.

80
The fact that space syntax parameters correlate well to many kinds of social
activities in buildings and cities, does not imply that space syntax is a way of
achieving successful studies of architecture without knowledge of
architecture. In most cases, it is not obvious what would be the most relevant
space syntax modelling. Even what might be the most relevant subdivision of
real space into space-elements, for instance how to draw the “axial-lines”, is
not obvious. Different modelling captures different aspects of space and
correlate to different aspects of human lives and activities.99 Knowledge
about both the physical and the social context is important at many stages of
a study also when applying methods of space syntax. Firstly, knowledge and
creativity are the basis for figuring out the subject to study. Given that space
syntax is a relevant approach to the subject, knowledge about the physical
situation (the site, the building or the interior) as well as the cultural and
social setting is needed in order to do a relevant space syntax modelling.
Whether the kind of people’s movements most relevant for the subject is on
foot, by bicycle, by private car or by public transport and whether the site is a
suburb of the 1970s or a dense historical city-centre, influences what kind of
space syntax modelling is most likely to be appropriate. Finally, other
knowledge than space syntax is necessary in order to interpret the results;
whether a new layout that makes a “segregated” street into a highly
“integrated” one, is positive or negative, depends on the situation. If the aim
is to increase the activity of shopping, then the more integrated street is likely
to improve the situation. If the street should rather be kept as a quiet,
residential neighbourhood, the change to becoming highly integrated is not
likely to be an improvement. Relevant knowledge about space syntax as well
as about the cultural and social context is needed in order to achieve space
syntax analysis that is relevant for a particular subject in a particular context.
The following section explains how space syntax analysis has been applied in
this particular study.

99
When Klarqvist (1998) argue for studying dwellings by space syntax analyses, he shows all the three kinds
of space syntax modelling applied on dwellings.

81
3.5 THE SPACE SYNTAX ANALYSIS OF THIS
STUDY

This section describes the kind of space syntax modelling applied in this
study and explains how the analyses have been carried out.

3.5.1 The Kind of Space Syntax Modelling

The kind of space syntax analysis applied in this study is a consequence of

the kind of space being studied, which is the interior space of apartments, the
research question, which is about similarities and differences in a historical
sample of floor-plan layouts, and the size of the sample, which is somewhat
more than 150 apartments. As elaborated in the previous section, three major
kinds of space syntax modelling can be distinguished by the elements applied
for representing the space. These three elements are “nodes and connections”,
“axial-lines” and “visual fields”. In this study, where one aim has been to
look for patterns in the historical development of apartments, a useful space
syntax approach is to apply “nodes and connections” in terms of connectivity
graphs. As shown by Hanson in “Decoding Homes and Houses” (Hanson,
1998), these simple figures that represent a higher level of abstraction from
the architectural object than the floor plan drawing, makes it possible to
discover similarities and differences between spatial layouts that are difficult
to identify by comparing floor plan drawings.100

When looking at small systems such as the dwellings examined in this study,
a node-graph-representation does not capture other information than what can
be found by studying a floor-plan-drawing. A node-graph adds no
information as to what is present in a floor plan drawing. What it does is to
point out the spatial configuration by leaving out many features of space that
are captured by the floor plan drawing, features such as metric distances (and
thereby size and proportions of rooms), views and daylight conditions.
Concerning small apartments, apartments where the spatial configuration can
easily be recognised without the closer focus of a connectivity-graph-image,
such a graph-image might be considered as a loss of relevant information.
The same can be said about doing the further space syntax calculation that
consists in determining the values of the space syntax parameters of the
rooms. In the case of such small systems, the most interesting feature of
space syntax is not to examine one single dwelling, but to compare larger
100
If we take a floor plan drawing, change the proportions of the individual rooms, and mirror the floor plan, it
can be hard to recognise that the spatial configuration is unchanged.

82
numbers, looking for similarities and differences. Due to the level of
abstraction achieved by connectivity graphs and space syntax parameters,
features not concerning the spatial configuration are left out with the
consequence that patterns of the spatial configuration appear more clearly.
The graph-images as well as the space syntax parameters are therefore useful
in order to identify fundamental kinds of layouts and in order to reveal
similarities and differences among larger samples of dwellings. These are the
reasons for applying connectivity graph analysis in this study.

3.5.2 Identifying the Spatial Units

When modelling spatial syntaxes by “nodes and connections”, it must be

determined what units of the real space are represented by the nodes. As most
of the apartments analysed in this study consist of separate convex rooms
connected by doors, each room can usually be modelled by one node, without
having to discuss whether to subdivide a “concave room” into several convex
spaces. By the concept of “Raumplan versus Plan Libre”, according to
Risselada (1988), distinguishing between “Loos-space” and “Corbusier-
space”, the great majority of the rooms are of the “Loos-kind”. Circulation
spaces like entrances and corridors, which because of various kinds of “L-
shapes” are often “concave rooms”, are not subdivided into several spatial
units unless they really are separated by doors. The rooms, of which the
modelling has been least obvious, are living rooms that are more openly
connected to the kitchen or dining room than by a door. In such cases, the
part of the room containing a sink and a cooker is usually termed “the
kitchen” by the residents, even in cases where it is hardly separated from the
living room. It is difficult to know whether such naming represents the
kitchen as furniture or the kitchen as a spatial unit. In the cases, where the
kitchen or the dining area is more or less “openly” connected to the living
room, such as the apartments 70-2, 76-1 or 04-3 (see figure 3.24) we decided
to apply two criteria in order to model the kitchen (or the dining) as a
separate node. Firstly, the kitchen (or the dining) area must have the character
of being a separate space, i.e. being a space partly enclosed from the rest of
the living room. Secondly, it must have daylight otherwise than through the
living room, i.e. it must have its own window. By this definition the kitchens
of 76-1 and 70-2 are modelled as separate spaces, while the kitchen of 04-3, a
kitchen that is neither separated from the living room nor has its own
window, is modelled as included in the living room.

83
Figure 3.24
Examples of node-modelling,
floor plans and space syntax connectivity graphs of 70-2, 76-1 and 04-3.

84
3.5.3 The Analysis, Applying the Software AGRAPH

Even though the space syntax analysis carried out in this study is of a
conventional kind and in the core of space syntax methodology, software
appropriate for the task was hard to find. What was particularly needed when
intending to analyse more than 150 apartments, was software that made
coloured graph-images with the explanative texts required, graph-images that
should be suited for visual comparison and for presentation without any
further editing. The software should make different calculations of the same
configuration, such as calculations with and without exterior, without
requiring redrawing the graph. The available software was either only for
Mac (not for PC), the calculated values were laborious to handle or the
graphs-images were not of the graphical quality needed. The decision made,
was therefore to develop the software required, as a part of this PhD-study.
This software, AGRAPH, is presented in the paper appended as appendix
A.3.1.

The first step of the space syntax analysis was to model the spatial
configuration of each apartment by connectivity graphs consisting of nodes
and connections. This modelling, which was done by applying the AGRAPH-
software, consisted in drawing nodes and connections representing the rooms
and their internal connections (by importing the floor plan as “background
image” when drawing the graphs of the largest apartments) and then to
“code” the nodes by colour as well as by abbreviation.101 The room, from
which the apartment is accessed, is the “root” of the justified graph; i.e. the
graph pictures the spatial configuration, as “seen” from the entry into the
apartment.102 (For more about “justified graphs”, see section 3.4.1.) For each
apartment, the justified and coloured graph was then “exported” from
AGRAPH as a picture-image (see figure 3.25 and Appendix A.3.1).
Appendix A.4.2 lists these justified graphs of all the apartments in
chronologic order, as illustrated in figure 3.26. In order to examine some
connections between rooms in detail, two sequences have been extracted
from the connectivity graphs and listed individually. The room-sequences
examined in this way are the accesses from entrance to kitchen and from
bedrooms to bathroom. Figure 3.27 and 3.28 shows examples of some room
sequences and some results of the space syntax calculation, respectively.

101
The colours and the abbreviations are explained in section 4.2.
102
The justified graphs are topological rather than mathematical in the sense that the positioning of nodes
corresponds topologically to the apartment that they represent; i.e. the order of internal positioning of rooms
(in terms of left and right) has been kept. This captures some information about the positioning that would not
necessarily be kept by a “mathematical” graph. In a “mathematical terms”, the graphs of 32-1 and 33-1 would
be identical and would not explain internal positioning apart from which rooms are connected.

85
Figure 3.25
Drawing a connectivity graph and doing a space syntax calculation by applying the software
AGRAPH. (See Appendix A.3.1 for more about this software.)

Figure 3.26
A selection of the justified graphs ordered chronologically. (See Appendix A.4.2 for the graphs
of all apartments.)

86
A.4.3 Some room sequences, all apartments Page 1/3

E-K Bd1 - Bt/Wc Bd2 - Bt/Wc Bd3 - Bt/WC

31-1
32-1
33-1
33-2
36-1
36-2
36-3-1
36-3-2
36-4
38-1
39-1
39-2
40-1
40-2-1
40-2-2
40-3
41-1
41-2
41-3
41-4-1
41-4-2
47-1
47-2-1
47-2-2

Figure 3.27
Some rooms sequences of some apartments, example.
(See Appendix A.4.3 for all apartments.)

Space Syntax Calculation, Typical Apartments, Every Third Case

Case R L MD(E) CV CV Integration (by Relative Asymmetry, RA) Order (E-L-K-Bd1)
E L/D

33-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
36-2 A 1,16 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
38-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
40-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-4-1 1 C 1,20 2,83 1,75 0,10 0,20 0,30 0,60 0,60 0,80 0,10 0,20 0,30 0,60
48-1-2 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,28 0,38
48-5-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
49-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
50-2 1 c 1,42 2,83 1,75 0,14 0,19 0,28 0,33 0,38 0,42 0,42 0,47 0,14 0,19 0,28 0,42
51-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
52-2-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
53-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
54-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
55-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
56-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53

Figure 3.28
The space syntax parameters of some apartments, example.
(See Appendix A.4.4 for all apartments.)

87
When it came to the space syntax calculations, these were done “without the
exterior” (which here means without the staircase or the corridor outside the
apartment), without storage-rooms (unless the storage room was a through-
passage-room) but with the balcony. The reason for including the balconies is
that they, at least in the summer, often function as a kind of second living
room. Based on the model of nodes and connections AGRAPH generates a
table listing the values of the space syntax parameters of all rooms (except
those explicitly excluded, as described above). The further analysis consisted
in importing these tables in EXCEL and ordering the data by apartment
numbers, by kind of rooms and by value of the space syntax parameters. The
space syntax parameters examined are the integration-values (by relative
asymmetry, RA), the control values (CV), the spatial depths from entrance,
the existence of “rings” and the position of the living room according to the
room-categories described in section 3.4.7. Figure 3.25 shows the output-files
listing the results from the space syntax calculation, while figure 3.28 shows
how these results have been order for comparing the sample of apartments.

Despite the application of newly developed software, the space syntax

analysis of this study is conventional, using well-known methodology in
order to reveal patterns among a sample of spatial layouts. What is less
conventional in this study is the extended use of colouring, in space-syntax-
graphs (which are made in colours by applying AGRAPH) as well as in
tables. The colouring of graphs in accordance with the functions of the
rooms, as shown in figure 3.26, made it easy to identify patterns compared to
having to read the room-functions as text. Similarly, the colouring by
function in the tables of calculated results, as shown in figure 3.25, simplifies
the identification of changes through time. The results from analysing these
results are described in chapter 4.

88
3.6 COMMENTS ABOUT ASSUMPTIONS, FINDINGS
AND GENERALISATIONS

As a substantial part of this study is the two empirical inquires, this section
comments upon the relationship between empirical data, assumptions,
findings and generalisations.

Even when intending to analyse empirical data with an open mind, pre-
conceptions in terms of assumptions or hypotheses about findings establish a
basis for doing the analyses. Findings might be of various characters; they
might correspond to expectations, they might lead elsewhere than expected
and they might even be contrary to what was expected. The conclusions to be
drawn are not obvious in any of these cases.

Findings that are contrary to expectations do not necessarily falsify a

hypothesis. There might be relevant explanations other than those captured
by the hypothesis, explanations that “overrule” what is assumed by the
hypothesis. However, if the findings are contrary to the hypothesis and
alternative explanations cannot be found even when they are explicitly
looked for, the hypotheses should be reconsidered critically. When various
layouts of apartments are compared in the synchronic survey, the location of
the apartments represents a bias that might make the effect of the interior
space negligible. This has been borne in mind when selecting the apartments
to study; in order to make the results comparable, it has been attempted to
select apartments located similarly or close to each other.

Even when findings are as expected, the hypothesis might be wrong as the
explanations might still be different from those assumed by the hypothesis.
This might be the case in several questions studied by the synchronic survey
as there are many parameters potentially influencing the results, parameters
which due to complex situations are hard to keep separate. The interior space,
which is the subject of this study, is only one out of many conditions that
influence who lives in the apartment, how these households live, and how
they evaluate their homes. Due to a variety of locations in the city, various
floors, various views and all other parameters not being equal for any of the
apartments examined, the frame of “other conditions being equal” hardly
exists among any of the apartments in the selected sample, even when
looking only at the apartments as “objects”. When further comparing these
non-identical apartments inhabited by households of all kinds, such an ideal
methodological framework is even harder to construct. Due to this lack of
89
“other conditions being equal” and to all the relevant parameters that have
not been studied, results of the synchronic survey corresponding to the
assumptions about general versus specific layouts coming out of the
diachronic analysis, indicate that these assumptions and tentative
explanations might be right but do not prove them to be true. However, this is
not a crucial methodological problem; according to Popper (1959, p. 40) the
“criterion of demarcation” is not verification but falsification; a hypothesis
cannot be proven to be true, it can only (by falsification) be indicated that a
hypothesis is wrong.

In reality, findings are likely to be of various kinds when it comes to their

correspondence with a priori assumptions or hypotheses. Whatever the
character of the findings might be, analysis consists in comparing the results,
the context and explanative hypotheses, looking for patterns of similarities
and differences among the results that correspond to the hypothesis, as well
as among those that do not. If results of the latter kind are numerous or
otherwise convincing, the hypothesis is not likely to be a good one. It might
then be falsified, which, even if uncomfortable, is a way of increasing
knowledge, or it might be revised. In the latter case, when a hypothesis is
replaced by a revised one that also captures findings not corresponding to the
earlier hypothesis, the hypothesis has simply become a better one. In this
study, the typology of apartments identified in sections 4.3.3 is the result of
such a process. The conclusion from the diachronic survey was the
identification of an apartment typology consisting of three kinds of floor-plan
named the A-, B- and C-type. When analysing the results of the synchronic
survey later on, some results of the A- the C-type seemed to be without any
patterns.103 The typology was therefore examined critically. By a closer
comparison of the result from the synchronic survey and the typology of
apartments, it became clear that some of the apartments had a significant
spatial characteristic not captured by the typology developed so far. By
“refining” the definition of the A-type of apartments and thereby identifying
some apartments as hybrids between this revised A-type and the C-type,
being of an “AC-kind”, a correspondence between the typology of apartments
and distinct categories of results was achieved.104 This way of developing
categories according to empirical data rather than sticking to predefined
categories is a major concern of “grounded theory” as described by Glaser
and Strauss (1967).

103
For these results in detail, see table 5.5.c, listing the kinds of households living in apartments of three
rooms.
104
The “revised A-type” of apartments differs from the former one in that all (and not only most) bedrooms
have direct access from a “neutral” entrance or corridor, see footnote 127 and section 5.1.1.

90
When Yin (1994, p. 36) discusses how to generalise from case studies, he
makes a distinction between statistical and analytical generalisations. In
statistical analyses, the results are considered valid for the total population (or
“universe”) out of which only a selected sample (a sample that should be
statistically representative) is analysed. Yin claims that analogy to statistical
samples and “universe” is irrelevant when case studies are concerned; there
exists no representative case or set of cases and to add further cases to a case
study does not turn a case study into statistics. The generalisation from case
studies, named analytical generalisation by Yin, is different from statistical
generalisations and works by generating theory from the findings of one or a
few particular cases.105 The analyses in this thesis are of both these kinds; for
some subjects statistical generalisations should be valid while for others, due
both to the questions asked and to the data that are analysed, generalisations
are of the “analytical” kind according to Yin. An example where the subject
as well as the available data is appropriate for simple statistics is the floor
area of rooms.106 However, in most questions of this study the size of the
sample (by numbers of apartments) is not large enough for hypotheses to be
made probable through statistics; as there are many dependent variables the
sample is rarely of the size needed to constitute statistical representations of
the total population. This does not imply that the explanations and hypotheses
are of little interest nor that they are wrong; in such cases, the purpose of a
hypothesis is not to claim truths by the means of probability but to develop
tentative explanations. Such generalisations are of the analytical kind
according to Yin.

In this study, kinds of generalisations vary in accordance with the questions

focused upon and the data being analysed. Some conclusions are based on
simple statistics while others come from analyses of a few particular cases.
The kind of generalisation will not be commented upon explicitly for every
question discussed through the thesis; having the comments above in mind, it
should be possible to have an understanding of the kinds of generalisation
behind the hypotheses and explanations that are offered through the
following chapters.

105
This is a parallel to the distinction made by Glaser and Strauss between statistical and theoretical sampling.
(Glaser and Strauss, 1967, p. 45-.)
106
Simple statistics is here to be understood as very basic statistics, such as average values and percentages of
a limited sample to be representative for the population out of which the sample is selected.

91
3.7 SUMMARY

The methodological strategy of this study has been to do a “quantitative”

inquiry of the development of apartment layouts through time prior to a more
qualitative survey on contemporary living in selected apartments of different
layouts. Due to the different time-perspectives, these two empirical studies
can be distinguished as diachronic and synchronic, respectively. The
diachronic inquiry, where the co-operative housing company OBOS is the
case examined, consists of analysing floor areas and spatial layouts, the latter
by applying space syntax methods. This diachronic inquiry has two aims. The
first is to identify significant changes in Norwegian apartments built during
the period 1930-2005, findings that themselves might represent a contribution
to the knowledge about Norwegian dwellings in a historic perspective. The
second aim is to establish an empirical basis for selecting the apartments to
examine in the following, more qualitative synchronic survey. The latter is an
in-depth study of a few typical or otherwise interesting apartments, a study
carried out by visits and interviews, asking about kinds of households, the use
of room and the residents’ preferences and evaluations about their dwelling.
The main aim of this synchronic survey is to find out about the potential
usability, or generality versus specificity, of different floor plan layouts. By
doing so, this PhD-project might provide some informed premises for further
discussions about dwelling layouts with respect to the population’s
preferences and needs.

The research strategy described through this chapter has been clarifying for
planning as well as for carrying out the work. However, when looking at the
work in retrospect, the diachronic and the synchronic surveys have been
carried out less separately than intended. As the diachronic study was only
half-complete at the time when the synchronic survey had to start due to its
own tight schedule, the selection of apartments for the synchronic survey had
to be based on a brief look at results from the diachronic study rather than on
sure findings. However, this lack of clear division between the initial
quantitative diachronic study and the following more qualitative synchronic
survey has not been disadvantageous; being forced (by lack of time) to start
the latter before the first was finished had some positive effects. One example
is the typology of apartments where findings from the interviews made it
possible to distinguish between different apartment layouts more clearly. 107

107
This is explained in section 5.1.1.

92
The conclusion to draw from this is that a consistent “a priori” strategy is a
useful guideline. However, when carrying out a job in reality, the pragmatic
adaptations needed in order to solve problems that turn up, do not necessarily
weaken the work. Unpredicted adjustments of direction might bring the work
into important fields that would not have been discovered if sticking to the
planned course. 108

As regards the space syntax analysis, which is the major section of the
diachronic inquiry, this differs from most otherwise similar space syntax
studies by the sample of apartments and by the rather extensive interview-
based survey that has been carried out on the basis of the space syntax
analysis. Due to the large sample of apartments, the results of the space
syntax analysis are likely to have a greater extent of general validity than is
the case in studies of smaller or less representative samples. Where the
synchronic and interview-based survey is concerned, this might not only shed
light on the apartment layouts identified by the space syntax analysis, it
might also provide general knowledge about correspondences between
particular space syntax features and real domestic lives. Concerning the
diachronic analysis as well as the synchronic survey, further information
about samples, data and results is given in the respective chapters: in chapter
4 about the diachronic study of apartment layouts and in chapter 5 about the
synchronic survey of contemporary living.

108
According to Glaser and Strauss (1967, p. 61) analyses of data should not be rigorously predetermined;
when saturation is reached (i.e. the moment has come when further analyses of the data only confirm what is
already found), the best use of time is not to go on analysing more of the same data, but to look for other kinds
of data or to examine data that do not correspond to the patterns found or the hypotheses developed so far.

93
4 Apartment Layouts;
a Diachronic Inquiry

The basis of this thesis is the two empirical inquiries, where the first, which is
the subject of this chapter, is an analysis of the interior spaces of a
chronological sample of apartments. The aim of this analysis is to identify
changes of the apartment layouts through time; to find out what the changes
have been and when they have happened. By studying a presumably
representative sample of apartments, the findings might also bring about
knowledge about the total stock of dwellings.

4.1 THE CASE AND THE SAMPLE

As explained in section 3.3, the co-operative housing company OBOS is the

particular case examined by this study. With the intention of studying a kind
of representative sample of apartments, the next question was how many
apartments to examine and how to select them. On the one hand, bearing in
mind the intention of capturing some general tendencies and not only
information about a few particular apartments, the number of apartments
could not be too restricted. On the other hand, in order to carry out the
analysis within a reasonable time, there could not be too many apartments.
The decision made was to select three OBOS-projects from every year of the
period 1930-2005.109 The selection was done according to two parameters;
size, looking for projects containing the largest number of apartments, and
location, trying to achieve a sample representing most parts of town.

The information needed in order to do this selection was found in the

archives of OBOS. Figure 4.1 shows all the selected projects marked on a
map of Oslo. When deciding which apartment to select from each project,

109
As there are years, such as during the Second World War, where very few dwellings were built, the number
of apartment selected by this strategy of sampling, are less than number of years multiplied by three.

94
this was done by looking for the typical110 apartment within each project. In
most projects, this was an apartment of three rooms, while in some cases it
was of two or four rooms.111 In the projects where the typical apartment is not
of three-rooms but where there are apartments of three-rooms, the most
typical three-room apartment was selected as a second apartment from the
project. This was done in order to have a possibility to compare equally sized
apartments from as many projects as possible. 112 Each apartment is identified
by a number according to the following system:

33-1 - the year of finishing construction is 1933

33-1 - number distinguishing projects built in the same year

In the projects where two apartments are selected, a third number is added in
order to distinguish apartments from the same project:

31-1-2 -number distinguishing apartments from the same project

Figure 4.1 Map of Oslo showing the OBOS-projects examined.

110
Most typical apartment here means the most numerous apartment-floor-plan in the project, considering
mirrored floor plans being equal.
111
When describing apartment sizes by the number of rooms, this is done according to Norwegian terminology
where the rooms being counted are the “habitable rooms”, i.e. rooms for living and sleeping except the
kitchen.
112
Such a study of three-room-apartments has not yet been carried out, but is among the more detailed
analyses that can be done with the data collected.

95
As already pointed out, one aim of this study is to examine the development
of Norwegian dwellings. This is different from studying “interesting”, “nice”,
“excellent” or other categories of dwellings where the samplings are based on
qualitative evaluations. What characterises the sample of this study is that it
has been selected in order to represent a “population of dwellings” rather than
qualitative categories. Due to this representativeness, the sample of
apartments should be relevant as data also for studies other than the one
carried out here.113 This distinguishes the sample of this study from most
samples of apartments presented in existing publications of Norwegian
dwellings and is the reason for presenting the apartments in a separate
catalogue-like volume.

113
A Norwegian publication that appears similar to the appendix of this thesis but is very different in terms of
representativeness is “Århundrets norske boligprosjekter 1900-2000” (Ditlef-Martens, 2000). The book
presents 156 housing projects from all over Norway by photos, drawings and some further information and
represents thereby an interesting documentation of Norwegian dwellings. However, due to the open criteria of
sampling, the publication is not very useful if intending to study Norwegian dwellings in general. (The criteria
for projects to be selected are diverse; the dwelling should be particularly excellent with respect to functional
or architectural qualities or they should be typical for particular time-periods or they should otherwise be
important with respect to Norwegian cultural history.)

96
4.2 THE DATA; COLLECTING, EDITING AND
ANALYSING FLOOR PLANS

Floor plan drawings have been the main source of information for the space
syntax analysis as well as for the analysis of the sizes of rooms. Concerning
the potential of the floor plan drawing in representing architecture, Bruno
Taut has made the following statement:

”Wer Sinn dafür hat, kann die Architektur am Grundriss ablesen, ohne den
Bau zu sehen, so wie ein Musiker die Noten liest.” (Taut, 1927, p. 68) 114

Even if one might doubt such a belief in the potential of the floor plan
drawing, it is hard to dismiss it from being an abstraction of the building that
captures essential features of the interior space. Even though drawings are
secondary sources, as the apartments built would be the primary ones, they
are a reliable source. As long as there have not been any illegal building
constructions, which are rare where Norwegian apartment housing is
concerned, the drawings in the municipal archives illustrate the apartments as
they were when built. With few exceptions, the municipal archive contains
floor plan drawings of all dwellings in Oslo. Floor plan drawings are
therefore a relevant, reliable and easily accessible source for this study.

After selecting the projects by using the archives of OBOS, information and
drawings for each project were gathered at the Oslo municipal archives and
in some cases from the architects of the projects. Based on this information,
projects that did not fit the criteria for being examined, such as those
consisting of detached houses, were excluded.115 Appendix 4.1 lists the
projects finally selected. The projects are presented more extensively in a
volume separate from this thesis, a volume that presents every apartment with
copies of the original drawings and with key data such as the year of
construction, address, architect and number of apartments. In order to make
the apartment floor plans easily comparable, they have been drawn by CAD
and printed in identical typography.116 From these floor plan drawings, the
sizes of the rooms and the spatial configuration of the apartments have been
analysed.

114
This in English would be something like “Those who have the skill, are able to read the architecture from
the floor-plan drawing, without watching the building, similarly to how a musician read notes.”
115
See section 2.5 about the kind of dwellings examined.
116
The original drawings are not only of different kinds, some of them are also discoloured or for other reasons
hard to read.

97
The analysis of the room sizes has been done by looking at the development
through time of the floor areas of the individual rooms, while the spatial
configuration has been examined by space syntax analysis consisting in
drawing the connectivity graphs and doing some basic space syntax
calculations on these by applying the software AGRAPH, as explained in
section 3.5. The appendix A.4.2 presents the space syntax representations of
all the apartments. The nodes, which represent the rooms of the apartment,
are “coded” by colour as well as by abbreviation in accordance with the
function (or presumed use) as listed below. The same colours are applied in
the table of the room-sequences in appendix A-4-3 and in the tables that list
the space syntax parameters (the tables 4.2 and 4.3 and appendix A.4.4).

Room In short Colour

Entrance E
Corridor C
Hall H
Staircase (internal) Sc
Living room L
Dining room D
"Room" (un-specified habitable room) R
Kitchen K
Lagest bedroom Bd1
Second largest bedroom Bd2
Third largest bedroom Bd3
Bathroom Bt
Wc Wc
Washing room Wr
Balcony Bc

Table 4.1
Legend (Rooms, abbreviation and colours)

98
4.3 RESULTS

This section summarises the results of the diachronic survey, first the results
from the analysis of room sizes across the sample of apartments and then the
results from the space syntax analysis of the same apartments. As a
conclusion, section 4.3.3 identifies three “generations” of apartments that
differ significantly with respect to the principle layout of the floor plans.

4.3.1 Sizes of Rooms and Apartments

Figure 4.2 illustrates the floor areas of some main rooms in the sample of
apartments examined. Concerning the development of floor areas over time,
it is possible to distinguish between two categories of rooms. The first
category is living rooms and bathrooms117. These rooms follow the
development of the apartments’ total floor areas with increasing areas until
about 1980 followed by a decrease. Concerning the size of the living rooms,
there is a recent tendency that deserves some elaboration. Subsequent to the
declining tendency since 1980, the size of the living room has been
increasing since about the year 2000. However, contrary to this recent
increase in the size of the living room, the total floor area of
living+dining+kitchen has decreased continuously since the early 1980s. As
the increase in floor-area of the living rooms is far less than the floor-area
lost by no longer having separate kitchens and dining rooms, the recent larger
living rooms do not imply increased space for daytime living.118 As we will
come back to in section 6.3, this reduction in the total floor areas of the
individual apartments is likely to be a main reason for merging previously
separate rooms for living, cooking and dining into one room.

The total area of WC, washing and bathroom shows a development similar to
the area of living+dining+kitchen just described. Since the period around
1970-1980, when these functions were usually separated into three separate
rooms and their floor areas thereby were the largest, the situation is now
similar to what was the case until the mid-1950s. Within the sample being
analysed here, only 3 out of the 27 apartments built during the last 15 years
have a WC separate from the bathroom and there is no apartment with a
separate washing room. The floor area of bathroom/WC/washing-room was
117
When WCs or washing-rooms are separate rooms, the floor areas of these rooms are added to the floor area
of the bathroom.
118
However, as the numbers of bedrooms as well as the number of persons in the household are lower in these
new apartments than it was in the large apartments of the 1960s and -70s when these were new, the decreasing
floor area of “living+dining+kitchen-room” does not necessarily imply that living in new apartments has
developed towards becoming more cramped.

99
3.5 m2 on average in the period 1940-1949. From the peak of 8.1 m2 in the
1980s, the size has decreased rapidly towards 6.3 m2 in the 1990s, down to
4.7 m2 on average since the year 2000. When comparing these new
apartments with those built as much as 50 years ago, the new ones have only
slightly more space for WC and bathroom. But there is one major difference
between the new and the 50-year-old apartments where floor areas and
usability are concerned: in recently built apartments, the tiny bathrooms are
also the place for washing and drying clothes, while the pre-war apartments
had (and still have) separate rooms for laundry. These laundries, which were
for common use for several apartments, were located in the basements and
were well equipped for washing as well as for drying clothes. The recent
decrease in standards in these respects is remarkable considering how
advertisements emphasise the new housing projects as modern, functional
and of high technical standard.

The second category of rooms concerning the development of size through

the period of time, is the bedrooms and differs significantly from the first as
the average size of the bedrooms has declined more or less continuously
throughout the whole period.119 The change towards more differentiated and
less “general” rooms is most significant when comparing the size of the
living room with the size of the “main bedroom” (or the second largest room
of the apartment, not necessarily being used as bedroom). Between 1930 and
1940 the average floor area of the “main bedroom” was 15 m2 while the
largest room, the living room, was usually between 18 and 20 m2; with
respect to size, the living room and the main bedroom were then of the same
category. In recently built apartments, this is very different; the average floor
area of the main bedrooms since 1990 is 11.3 m2. This is a kind of space
fundamentally different from the recently built living rooms, which are on
average 31 m2 and include the kitchen; the bedroom is a room for sleeping
while the living room is now the place for all daytime activities.

119
The sizes of the second and the third largest bedrooms show similar but less significant pattern than the size
of the main bedrooms. Because far from all apartments selected for this study have a third or even a second
bedroom, a more continuous chronological sample of large apartments is needed in order to generalise about
the development of the second and the third bedrooms.

100
F l o o r ar ea
( m2 )
40
Li vi ng
Bedr oom1
35
Bath/ Wc/ Wash

30

25

20

15

10

0
C hr o no g r ap hy ( 19 3 0 - 2 0 0 5)

Figure 4.2
Floor areas of living room, largest bedroom and bathroom+WC+washing.

101
F lo o r a re a
(m2)
140

Total

120 Living+dining+kit chen

100

80

60

40

20

0
C hro no gra phy ( 19 3 0 - 2 0 0 5 )

Figure 4.3
Floor areas of the apartment and of living+dining+kitchen.

102
4.3.2 Spatial Configurations

The spatial configurations are examined by comparing the justified

connectivity graphs and the space syntax parameters that are determined by
calculation. The following sections describe the results one by one, first by
explaining the patterns found by comparing the graph-images and some
particular sequences of rooms, then by going in detail into some of the
calculated parameters. As presentation of the more than 150 graph-images
requires numerous pages, these as well as the tables of the space syntax
parameters of all apartments are edited as appendices at the end of this thesis,
as appendices A.4.2 and A.4.4, respectively. In order to point out the main
patterns, a one-page short version of the space syntax parameters is presented
in table 4.2. This table lists the data for every third apartment of the total
sample. The results in terms of numbers and percentages (in this text as well
as table 4.3) refer to the analyses of the total sample of apartments and might
therefore differ slightly from what would be found if analysing the 50 cases
in the short-version list only.

Figure 4.4
A selection of floor plans and their corresponding justified connectivity graphs.

103
4.3.2.1 The Connectivity Graphs

Figure 4.4 shows a selection of floor plans and the corresponding justified
connectivity graphs, while appendix A.4.2 shows the graphs of all
apartments. The space from which you enter the apartment is usually a
staircase, while in a few cases it is a corridor or an outdoor gallery. Until
1955, the typical graph is a “bush” where all rooms except the balcony (if
there is one) have access directly from the entrance. (See 33-1 and 52-2 in
figure 4.4.) The main differences among the floor-plan-layouts from this
period are whether there is access between the living room and a bedroom as
well. From the late 1950s, until the 1970s and -80s, there was a development
towards an increasing number of separate rooms organized in a “deeper”
spatial layout. A remarkable pattern is how the apartments constructed by the
company “Selvaag” differ from others by their very large “rings”, rings that
might contain as many as 7 rooms and any kind of rooms except the WC.
(See 84-3 in figure 4.4.) 120 These rings are a means of achieving access to
many rooms with a minimum of space solely for transition, a layout that most
architects have probably considered unacceptable due to all through-passing
generated. When it comes to more recently built apartments, the number of
rooms has decreased and the spatial structure has returned to a simpler type.
A pattern that clearly appears from the connectivity graphs is that the space
connecting the other rooms no longer has to be the entrance but very often is
the combined living-and-kitchen-room. (See 04-3 in figure 4.4.)

4.3.2.2 Some Room Sequences

Some changes in the spatial configuration can be revealed by studying spatial

sequences such as the one from the entrance to the kitchen and the one from
the main bedroom to the bathroom. Appendix A.4.3 list these room
sequences. Until 1965, most apartments (95%) had a kitchen with direct
access from the entrance. This changed around 1965. Since then as many as
50% of the apartments have kitchens that can only be accessed through the
living room. The spatial sequence from bedroom to bathroom shows a similar
development towards less direct access and more through passing. Until the
late 1950s, when the connectivity graph was usually a “simple bush”, the
entrance was the only room linking bathroom and bedroom in 90% of the
apartments. In the period 1958-1995, this percentage was reduced to an
average of 65. Since 1995, only 55% of the apartments have a main bedroom
that can be accessed without passing through the living room.

120
See also 77-1, 82-2 or 83-2 in appendix A.4.2.

104
Table 4.2 Space Syntax Calculation, Typical Apartments, Every Third Case
Case R L MD(E) CV CV Integration (by Relative Asymmetry, RA) Order (E-L-K-Bd1)
E L/D

33-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
36-2 A 1,16 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
38-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
40-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-4-1 1 C 1,20 2,83 1,75 0,10 0,20 0,30 0,60 0,60 0,80 0,10 0,20 0,30 0,60
48-1-2 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,28 0,38
48-5-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
49-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
50-2 1 c 1,42 2,83 1,75 0,14 0,19 0,28 0,33 0,38 0,42 0,42 0,47 0,14 0,19 0,28 0,42
51-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
52-2-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
53-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
54-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
55-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
56-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
57-2 1 C 1,20 2,83 1,75 0,10 0,20 0,30 0,50 0,50 0,60 0,10 0,20 0,30 0,50
58-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
59-3 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
60-3 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
61-3 1 C 1,14 4,83 1,66 0,04 0,19 0,23 0,33 0,33 0,33 0,33 0,47 0,04 0,19 0,23 0,33
62-3 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
64-1 1 C 2,25 0,20 2,70 0,10 0,14 0,28 0,32 0,35 0,35 0,35 0,39 0,39 0,14 0,32 0,35 0,39
65-1 1 C 1,16 3,83 1,70 0,06 0,20 0,26 0,40 0,40 0,40 0,53 0,06 0,20 0,26 0,40
66-1 1 C 1,28 3,66 1,53 0,09 0,19 0,19 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,19 0,38
66-2-1 1 C 1,42 2,75 2,75 0,14 0,14 0,33 0,33 0,42 0,42 0,42 0,42 0,14 0,14 0,33 0,42
67-2 B 1,42 3,25 3,35 0,14 0,14 0,42 0,42 0,42 0,42 0,42 0,42 0,14 0,14 0,42 0,42
68-2 1 C 1,14 4,83 1,66 0,04 0,19 0,23 0,33 0,33 0,33 0,33 0,47 0,04 0,19 0,23 0,33
69-2 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
70-2 B 1,40 2,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,20 0,60 0,60
71-3 1 C 1,75 1,75 2,66 0,14 0,21 0,21 0,32 0,32 0,39 0,39 0,46 0,46 0,14 0,21 0,21 0,32
72-3 1 C 1,37 3,66 1,70 0,10 0,21 0,21 0,32 0,35 0,35 0,35 0,46 0,46 0,10 0,21 0,32 0,35
74-1 1 C 2,00 2,16 1,58 0,18 0,21 0,22 0,24 0,24 0,28 0,34 0,37 0,37 0,40 0,45 0,45 0,54 0,18 0,21 0,24 0,34
75-3-1 B 1,62 3,25 2,75 0,14 0,17 0,32 0,39 0,39 0,42 0,42 0,42 0,57 0,14 0,17 0,32 0,39
77-1 1 C 1,80 2,66 1,75 0,17 0,22 0,22 0,31 0,31 0,37 0,37 0,37 0,37 0,42 0,42 0,17 0,22 0,22 0,31
78-1 B 2,11 1,20 2,20 0,11 0,22 0,27 0,27 0,33 0,33 0,44 0,44 0,50 0,50 0,22 0,27 0,33 0,44
81-1 2 C 1,63 1,95 2,25 0,12 0,14 0,18 0,23 0,25 0,27 0,30 0,32 0,32 0,32 0,32 0,36 0,12 0,18 0,32 0,32
82-2 1 C 1,90 2,66 1,75 0,18 0,20 0,23 0,25 0,30 0,32 0,34 0,36 0,36 0,38 0,41 0,43 0,18 0,20 0,23 0,30
83-2 1 C 1,90 2,66 1,75 0,18 0,20 0,23 0,25 0,30 0,32 0,34 0,36 0,36 0,38 0,41 0,43 0,18 0,20 0,23 0,30
84-2 B 2,33 0,20 2,20 0,11 0,22 0,22 0,33 0,33 0,33 0,44 0,44 0,44 0,44 0,22 0,33 0,33 0,44
87-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
88-3 B 1,28 4,33 2,20 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
90-1 1 C 1,66 0,70 4,00 0,06 0,26 0,26 0,40 0,40 0,40 0,46 0,06 0,26 0,26 0,40
91-4 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
98-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
00-2-2 A 1,90 2,16 1,16 0,08 0,20 0,24 0,24 0,28 0,28 0,28 0,40 0,40 0,44 0,44 0,20 0,24 0,24 0,28
02-2 B 1,40 2,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,20 0,60
03-3 1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50
04-3 B 2,00 0,20 4,50 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,40 0,40
05-3-1 B 2,20 0,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,60 0,60

Legend
bedroom 1 / main bedroom
R Rings (see section 3.4.7) other bedrooms
L Living room's kind of space kitchen
MD(E) Mean depth from entrance living room, dining room (or non-spesified room for daytime living)
CV Control Value bathroom
WC, washing room
E Entrance entance/corridor/hall
L Living room internal staircase
D Table
Dining4.2 Space
room syntax calculation, in brief (every 3. case)
balcony
(white) stortage room (with through-passage)

105
Table 4.3 Space
p Syntax
y Calculation,, Some results poited
p out
Case R L CV CV Order (E-L-K-Bd1)
E L/D

33-1 1 C 3,00 0,75 0,00 0,33 0,33 0,50

36-2 A 4,00 0,25 0,00 0,50 0,50 0,50
38-1 A 4,00 0,25 0,00 0,50 0,50 0,50
45 % with rings
40-1 A 4,00 0,25 0,00 0,50 0,50 0,50
41-1 A 4,00 0,25 0,00 0,50 0,50 0,50
41-4-1 1 C 2,83 1,75 0,10 0,20 0,30 0,60

55 % A-type
48-1-2 1 C 3,83 1,70 0,09 0,19 0,28 0,38
48-5-1 A 3,50 1,25 0,10 0,30 0,50 0,50

CV(E) >> CV(L)

49-1 1 C 3,83 1,70 0,09 0,19 0,38 0,38
50-2 1 c 2,83 1,75 0,14 0,19 0,28 0,42
51-2 B 3,33 2,25 0,13 0,20 0,46 0,53

The entrance is most integrated.

i
52-2-1 A 3,50 1,25 0,10 0,30 0,50 0,50
53-2 A 3,50 1,25 0,10 0,30 0,50 0,50
54-1 B 3,33 2,25 0,13 0,20 0,46 0,53
55-1 1 C 3,83 1,70 0,09 0,19 0,38 0,38
56-1 B 3,33 2,25 0,13 0,20 0,46 0,53
57-2 1 C 2,83 1,75 0,10 0,20 0,30 0,50
58-2 B 3,33 2,25 0,13 0,20 0,46 0,53
59-3 B 3,33 2,25 0,13 0,20 0,46 0,53
60-3 A 4,50 1,20 0,06 0,26 0,40 0,40
61-3 1 C 4,83 1,66 0,04 0,19 0,23 0,33
62-3 A 3,50 1,25 0,10 0,30 0,50 0,50
64-1 1 C 0,20 2,70 0,14 0,32 0,35 0,39
65-1 1 C 3,83 1,70 0,06 0,20 0,26 0,40
66-1 1 C 3,66 1,53 0,09 0,19 0,19 0,38
66-2-1 1 C 2,75 2,75 0,14 0,14 0,33 0,42
67-2 B 3,25 3,35 0,14 0,14 0,42 0,42
70 % with rings

70 % C-type

68-2 1 C 4,83 1,66 0,04 0,19 0,23 0,33

69-2 1 C 3,83 1,70 0,09 0,19 0,38 0,38
70-2 B 2,33 2,33 0,20 0,20 0,60 0,60
71-3 1 C 1,75 2,66 0,14 0,21 0,21 0,32
72-3 1 C 3,66 1,70
The living room is often most integrated. 0,10 0,21 0,32 0,35
CV(L)>CV(L) in 50% of the apartments

74-1 1 C 2,16 1,58 0,18 0,21 0,24 0,34

i

75-3-1 B 3,25 2,75 0,14 0,17 0,32 0,39

77-1 1 C 2,66 1,75 0,17 0,22 0,22 0,31
78-1 B 1,20 2,20 0,22 0,27 0,33 0,44
81-1 2 C 1,95 2,25 0,12 0,18 0,32 0,32
82-2 1 C 2,66 1,75 0,18 0,20 0,23 0,30
83-2 1 C 2,66 1,75 0,18 0,20 0,23 0,30
84-2 B 0,20 2,20 0,22 0,33 0,33 0,44
87-1 B 3,33 2,25 0,13 0,20 0,46 0,53
88-3 B 4,33 2,20 0,09 0,19 0,38 0,47
5 % with rings

90-1 1 C 0,70 4,00 0,06 0,26 0,26 0,40

65 % B-type

91-4 B 3,33 2,25 0,13 0,20 0,46 0,53

98-1 B 3,33 2,25 0,13 0,20 0,46 0,53
00-2-2 A 2,16 1,16 0,20 0,24 0,24 0,28
02-2 B 2,33 2,33 0,20 0,20 0,60
03-3 1 A 3,50 1,25 0,10 0,30 0,50
04-3 B 0,20 4,50 0,06 0,40 0,40
i

05-3-1 B 0,33 2,33 0,20 0,60 0,60

Table 4.3
Space syntax calculation, in brief (every 3. case). Some patterns pointed out.

106
4.3.2.3 The Rooms Characterised According to Spatial Topology

The third column of table 4.2 (and of appendix A.4.4) lists positioning of the
living rooms according to Hillier’s typology explained in section 3.4.7. Table
4.3 points out the most significant changes through time. Until 1955, about
55% of the living rooms were of the A-type while 35 % were of the C-
type.121 In the early 1960s, the living room changed to being mainly of the C-
type; in the period 1965-1985, 70% of the living rooms were C-type space.
During the 1980s, there was again a change, and since 1995, the living room
has been a B-type space in 65% of the apartments. When the living room is
an A-type space entered from a neutral entrance, it is a space without social
limitations on use caused by the spatial layout. The C-type of space, which is
the common type of the living rooms from the period 1965-1980, is a
through-passage-space, but being on a ring, not all movements have to be
through the space as there are alternative routes. Concerning the most recent
period, where most living rooms are B-type of spaces, the living rooms and
the rooms accessible only from the living room are closely dependent on one
another where potential use is concerned, similarly to the figure 2.19-1.3.c.
This indicates a limited potential use for both the living room and the
connected rooms; contemporary apartments can therefore be characterised as
less general than the earlier ones.

4.3.2.4 Rings

Not surprisingly, the existence of rings corresponds to the type of rooms

described in previous section. The occurrence of rings is listed in the second
column of table 4.2 and pointed out in table 4.3. Until 1950, half the
apartments in the sample (46%) had an internal ring of rooms. Most of these
rings were “short” ones consisting of only three rooms. During the period
1950-1965, the frequency of apartments with internal rings was much lower,
only 30% of the apartments from this period had such a ring. About 1965,
there was a change again, and as many as 70% of the apartments built
between 1965 and 1980 had an internal ring. As commented in section
4.3.2.1, many of these were large rings consisting of 4-7 rooms. The next
significant change happened in the early 1980s; the frequency of rings
decreased rapidly during few years and among the apartments built since
1990 there is only one that has an internal ring.

121
The balcony is here defined as an extension of the living room, not as a separate connected space.

107
4.3.2.5 Rank of Integration

Integration is here described by Relative Asymmetry (RA), which is a value

between 0 and 1, where 0 represents the highest possible integration. Table
4.3 lists the rank of integration for all rooms in every third of the apartments.
In order to point out the range of integration among the main rooms, the table
also lists explicitly the integration-rank of the entrance (E), the kitchen (K),
the living room (L) and the main bedroom (B).122 As expected from the
connectivity graphs, the entrance is usually the most integrated room. Until
1955, there was no exception to this. During the entire period until the late
1970s, the dominant order of the main rooms regarding RA was
E<L<=K<=Bd1 ; this was the case in 80% of the apartments. Since then, the
rank of integration has been more varied. In the period 1980-2000, the main
bedroom was more integrated than the kitchen in 50% of the apartments.
Concerning the most integrated space, a change seems to be going on just
now; in 4 out of the 7 most recently built apartments, the living room is the
most integrated space. This rank of integration does not appear in any of the
apartments built before 1957 and in only 10% of the apartments built
between 1957 and 2000.123

4.3.2.6 Depth from Entrance

The fourth column of table 4.2 lists the mean depths from the entrance,
MD(E), while figure 4.5 points out the changes through a period of time. In
apartments built up to 1950, all rooms except the balcony and the bathroom
have direct access from the entrance (or from a hall) in 75% of the
apartments. The mean depths were thereby 1.0124 in many of the apartments
from this period. As the number of rooms in the apartments increased
towards the 1970s, the spatial configuration became deeper. Whereas the
average Mean Depth of the entrance was 1,27 during the period 1930-1965, it
was 1,66 in the period 1970-1985. Since then, the number of rooms has
decreased and the average MD(E) since 1990 is 1,43. This is significantly
higher that the MD(E) in the first period due to the bush-like spatial layout of
the first period, where most rooms are directly accessible from the entrance.

122
This way to compare ranks of integration of selected rooms is a common way of comparing spatial layouts
across dwellings of different sizes (in terms of the number of rooms). As shown in tables 4.2 and 4.3, such
comparison of selected room makes patterns appear more clearly then the table listing the integration values of
all rooms.
123
This change towards a highly integrated living room, is parallel to the findings of Hanson (1998, p.129) that
the previously segregated and representative “parlours” of small traditional London houses have changed into
highly integrated living rooms as traditional working-class residents are replaced by the new middle-class.
124
1.0 is the lowest mean depth possible, see the “hub” in figure 3.9, section 3.4.3.

108
 M e a n D e p t h f r o m En t r a n c e , M D ( e )

3,0

2,8

2,6

2,4

2,2

2,0

1,8

1,6

1,4

1,2
Figure 4.5
Mean depth from 1,0
entrance, MD(E). Chronology (1930-2005)

4.3.2.7 Control Value

Where the Control Value (CV) is concerned, the most significant changes are
the ones of the living room compared to the entrance. Until 1955, the CV of
the entrance was far higher than the CV of any other room. In the whole
period of 1930-1960, the entrance was the room with the highest CV in 95 %
of the apartments. Since 1960, due to the more central position of the living
room, the number of apartments where the entrance has the highest CV is
lower (63% of the apartments). The Control Value indicates to what extent a
space is “controlling” the access to its neighbouring spaces. However, what
“amount of control” this represents in real life depends on the use of the
“controlling room”; a corridor having a high CV does not mean that one or
some of the residents have much control over what is going on in the
apartment. If a living room or a kitchen has the same high CV, the situation is
different. As the residents occupy kitchen and living rooms more frequently,
one can expect a high CV for these rooms to indicate “high real-life control”
of what is happening in the apartment. When the typical CV has changed
from 3.0 for the entrance and 0.75 for the living room to being on average
more than 2.0 for the living room, the social control of life in the apartment
can be assumed to have increased significantly; a change that represents a
development from generality to specificity.

109
4.3.3 Conclusion; Three Generations of Apartments

From the analysis described above, it is possible to distinguish between three

kinds of floor plan layouts with respect to the potential usability of the
dwellings. Although there are exceptions, these kinds of layouts have a
chronological order.

Until about 1955, most apartments had a character of generality with respect
to the size of the rooms as well as to the spatial configuration. Compared to
newer apartments, the bedrooms were large while the living rooms were
small; they both had the potential for various uses. This generality was
enforced by the spatial configuration where the kitchen, the living room and
the “bedrooms” were similarly situated within a simple spatial configuration,
a configuration where all the rooms were accessible from the entrance. This
kind of spatial configuration, which has the justified graph figure of a
shallow bush (when seen from the entry) is a layout that in principle is
general.125 The fact that the most integrated room is a neutral entrance makes
these apartments even more general. The other rooms are usually equally
integrated. Where living rooms from this period are more integrated than the
kitchens, this is due to the balcony (which is accessed from the living room)
or to a “ring” made by a door between the living room and a “bedroom”. The
naming of the rooms confirms that generality was intended or at least
expected. In many of these apartments, the habitable rooms are not termed by
function-specific terms such as living room and bedroom but by the function-
independent (and thereby general) Norwegian terms “værelse” or “rom”,
which simply mean “room”. (See figure 4.6.)

Figure 4.6
The original floor plan drawing
of apartment 32-1, where the
habitable rooms are termed
“værelse” (English “room”)
instead of being determined
for sleeping or living.

125
Hillier (1996, p. 314) describes how “depth minimalising forms” (such as a “bush”, which is as “depth-
minimalised” as a form can possibly be) seem to be suited to a large number of possible functions.

110
From the 1950s, there was a change towards thoroughly planned functional
specificity; the interior spaces became highly specialised both in terms of the
design of the individual rooms and in terms of the spatial configuration. The
different rooms were tailored according to specific functions; the number of
rooms increased, the size of the rooms became more varied and the
apartments became larger. In the period 1970-1980, more than half the
apartments had WC, bathroom and washing room as three separate rooms. In
most of the large apartments of this period, the kitchen was a space only
partly separated from the living room. (See apartment 69-2 in figure 4.4.)
Compared to the previous period, the apartments had a “deeper” spatial
configuration where each room was situated according to its very specific
function. With few exceptions, the entrance was still the most integrated
space, but on average the living rooms had nearly the same RA-value as the
entrances.

In the 1990s, the typical floor plan changed towards something in-between
the two earlier main types. The number of rooms has decreased and the
spatial layout is neither strictly general nor strictly specific. Where the living
room is concerned, this room is general in the way that it has the size, shape
and daylight conditions that make it appropriate for a wide range of activities.
In spite of this, the living room is not very general; it is a room specialised
for “living”. Most living rooms are now through-passage-rooms and do
therefore not have the potential for rest and quiet activities as is the case for
the living rooms of the typical pre-1950 apartment. Compared to the complex
spatial layout of the 1970-1980 apartment type, the spatial configuration has
recently become more similar to the early general ones, but due to the much
more central positioning of the living room, the apartments are in reality
highly specific concerning the use of space. This can be illustrated by the
control values listed in table 4.2; in the typical pre-1950 apartment, the
neutral entrance had the highest control value, much higher than any other
room, while in many of the new apartments the living room has the highest
control value. The latter is a feature decisive for usability in the sense that
access to (and activities in) other rooms are “controlled” from the living
room.126 Where the kitchen is concerned, the typical recently built apartment
does not have a separate kitchen; it is included in the living room.

126
Hillier (1996, p. 323-324) explicitly points out that the B-type of space, which is the kind of space of the
living rooms in these new apartments, is the most constraining, as movements are required to pass the space,
which has a powerful effect on the usability. (See section 3.4.7 for more about Hillier’s typology of spaces.)

111
These three generations of apartments are the conclusion of the diachronic
analysis.127 The three generations, which are significantly different in terms
of room sizes and spatial layouts, are hereafter labelled the A-, the B- and the
C-type of apartments, in chronological order. In brief, the type-A is an
apartment consisting of kitchen and 2 or 3 rooms,128 where all the rooms are
accessible from the entrance. This was the usual layout until the 1950s.
Compared to the apartments of this A-type, the apartments of the B-type are
larger in floor area as well as in number of rooms; most of them have 3 or 4
rooms. Concerning functions or use, the rooms of B-type apartments are
more specific than the rooms of A-type apartments; the living rooms are
larger, the bedrooms are smaller and the functions of WC, bathing and
washing are often separated into three rooms. The B-type is the common
layout of suburban housing projects built during the period from the late
1960s until about 1985. The third category, the C-type, is smaller than those
of the B-type. They are usually of 2 or 3 rooms, the kitchen is included in the
living room and the WC, the bathroom and the washing room are no longer
separate rooms. Compared to the A-type, the “living and kitchen room” of
the C-type has a more central129 position in the spatial layout. The “type of
space” of the living rooms, according to Hillier’s typology of spaces
described in section 3.4.7, is significantly different among the three
generations or types of apartments. Unfortunately, this correspondence does
not follow the labelling; in the A-type of apartment the living room is an a-
type of space, in the B-type of apartment it is a c-type of space, while it in
the C-type of apartment the living room is a b-type of space. Table 4.4 at
next page summarise the characteristics of the three generations of
apartments and figure 4.7 shows floor plans and the connectivity graphs for
three apartments representing the three generations.

In the diachronic analysis described in this chapter, basic space syntax

methods have been applied in order to identify changes in the development of
Norwegian apartments built during the last 75 years. Awareness of this
historical background as well as of the spatial features that characterise
contemporary apartments should constitute a basis of knowledge relevant for
further discussions on contemporary and future housing design.

127
Through the synchronic interview-based survey following this diachronic inquiry, it was found that this
typology of apartments could be improved by applying a more precise definition of the A-type, in the sense
that all rooms of the apartment, not only most of them, should be directly accessible from the entrance. This
improvement of the typology is commented upon in sections 3.6 and 5.1.1.
128
As described in footnote 111, describing apartment size by the number of rooms is done according to
Norwegian terminology where the rooms being counted are the “habitable rooms”, i.e. rooms for living and
sleeping except the kitchen.
129
Which means that they are more integrated, if using space syntax terminology.

112
Table 4.4 The three types (or three generation) of apartments

Type (or generation ) A B C

Time period 1930 - 1955 1965 - 1980 2000 - ?
Average floor area (m2) 62 95 59
Number of "bedrooms" 1-2 2-3 1-2
Living room, kind of space A C B
Internal rings some often few
-E-L-Bd- (large rings usually
including washing room)
"potential usability" general spesific (but some flexible due specific
to lightweight interior walls)
Brief descriptions of the floor plans
A The bedroom/second largest room is not much smaller than the living room.
All rooms have access from a “neutral” entrance.
B The apartments are planned for families with children
The rooms are individually tailored to particular functions.
The bathroom, washing room and WC are often three separate rooms.
The living room is by far the largest room of the apartment.
The kitchens are often openly connected to the living room
Not one particular layout (due to the many rooms and thereby many possibilities)
C There is no separate kitchen; the kitchen is included in the living room.
Table 4.6
The living room is by far the largest room of the apartment.
The three
Onetypes, or three
bedroom generations,
is usually of flooronly
accessible planthrough
layoutsthe living room.

A B C

33-1 69-2 04-3

Figure 4.7 Apartments representing the three types of layouts,

floor plans and justified connectivity graphs.

113
5 Spatial Layout and Contemporary
Living; a Synchronic Survey

The conclusion of the diachronic inquiry was to describe a historical

development of apartments by a typology consisting of three kinds of floor
plans. This typology was found by examining the “physiognomy” of interior
space and is not based on any empirical data about domestic lives. The
synchronic survey, which is the content of this chapter, examines whether
this typology of apartments is relevant to the residents’ experiences and
actions, and not only represents some features of the floor plans. In order to
make it possible to grasp the main results of this survey without examining
all the details presented through this chapter, the final section 5.2.9
summarises the results.

5.1 THE SURVEY

The typology identified in the diachronic survey consists of the floor plan
layouts called the A-, B- and C-type of apartments. The intention of this
synchronic survey has been to study and to compare these three layouts with
respect to the contemporary domestic lives taking place in them and to the
respective households` evaluations of their dwellings. The survey, which has
been carried out by ringing on doorbells and visiting people at home, focuses
on two hypotheses that emerged from the diachronic study. Firstly, if the A-,
B- and C-types of apartments are significantly different concerning the
degree of generality, as concluded in the previous chapter, then there should
be some correspondence between apartment type and households.
Apartments of the general A-type should house a larger variety of households
than apartments of types characterised as more specific with respect to
functions or use. Secondly, looking closer at the separate rooms, rooms of the
kinds characterised as general should house a larger variety of domestic
activities than rooms characterised as more specific regarding function or use.
This interview-based survey attempts to shed light on these hypotheses.

114
5.1.1 The Sample

The strategy has been to do a pilot survey of three housing projects first and
then, as the main survey, to study nine housing projects, three of each of the
apartment types A, B and C. The pilot-survey was done in the spring 2004
and the main survey was carried out during the autumn of the same year.
Altogether twelve projects were selected from among the 150 projects that
were studied in the diachronic analysis. From each of these twelve projects, a
number of identical apartments were selected for the interviews.130 Based on
the experience from the pilot-survey and with the capacity for doing the
planning, the fieldwork and the analysis in mind, the intention was to
examine about 20 identical apartments from each project.

The subject of interest, or the unit of analysis, is the interior space of

apartments. However, when it comes to people’s choices of dwellings, the
geographic location is a decisive feature. Due to this importance of location,
it is hard to study the effect of various floor plans by comparing apartments
with very different locations; in such cases the strong influence of the
location is likely to make the effect of the spatial layouts negligible.
Similarly, if comparing apartments that differ in size as well as in spatial
layout, it is hard to point out the influence of the latter. This has some
consequences for the survey, both as concerns which projects to compare and
which aspects of the apartments to examine. Generally, apartments of the B-
type differ from those of the A- and the C-type in size and location, while
apartments of the A-type and of the C-type are more often similar in size and
location. The latter made it possible to select projects of the A-type and C-
type that are comparable. Compared to this, the projects of the B-type have
been selected in order to be different and presumably interesting or
representative examples of the B-type, not in order to be directly comparable
to apartments of the A- or of the C-type; a frame of “other circumstances
being equal” between the B-type of apartments and those of the A- and the C-
type has not been attempted.

Due to a tight schedule, the projects were selected in advance of finishing the
diachronic analysis, at a time when the apartment types A, B and C were not
actually distinguished. All the selected apartments are therefore not as
typologically clear as intended. The projects 51-3 and 55-1 were selected in
order to represent the A-type of apartment, while 00-1 was selected in order

130
Identical apartment is to be understood as apartments with identical or mirrored floor plans.

115
to represent the C-type. These three apartments are characterised by direct
access from the entrance to all rooms except the largest bedroom where the
access is from the living room, see figure 5.1. According to the typology of
A, B and C described in section 4.3.3, the apartments 51-3, 55-1 and 00-1 are
all hybrids between the A- and the C-type. While the intention was to select
one project of each type (A, B and C) for the pilot-survey and three of each
type for the main survey (altogether 4 projects of each apartment-type), the
types of apartments actually selected came out differently due to these three
apartments being hybrids of an “AC-kind”. Table 5.1 shows the types and
some basis information of the selected apartments, while appendix A.5.1
contains some further information. Figure 5.1 present the floor plan drawings
of the apartment, while figure 5.2 lists the space syntax graphs. 32-1, 55-1
and 74-1 are the projects of the pilot-survey while the other 9 have been
examined in the main survey. Each apartment is identified similarly to the
principle described in section 4.1:

33-1 A4 -the year of completion of construction is 1933

33-1 A4 -number distinguishing projects built the same year
33-1 A4 -number distinguishing apartments from the same project,
the last number (here: 4) refers to the floor number131

Table 5.1 Apartments selected for the survey

Size (m2)
Project

Rooms

Type *

Name and address Location (by part of Oslo)

32-1 2 57 A Maridalsveien 64, Maridalsveien 64 Ila Centre-N/E
33-1 2 50 A Fagerheimen, Fagerheimgata 2-12 Dælenenga Centre-E
41-2 3 75 A Table 5.1
Bøkkerløkka, here
Brockmannsgate 8-10 Bjølsen Outer-centre-N
51-3 3 74 AC Akersbakken, Frederikke Qvamsgate 13-21 Ila / Gamle Aker Centre-N
55-1 3 67 AC Hovin, St.Jørgens vei 41-47 Hovin East
70-2 2 56 C Vosseløkka, Vossegate 18, 20, 20B Torshov / Lilleborg Outer-centre-N/E
74-1 4 83 B Orremyr, Odvar Solbergs vei 28-30 Romsås North/East
76-1 4 95 B Svarttjern, Odvar Solbergsvei 126-128 Romsås Nort/East
77-2 4 104 B Sandaker, Åsengata 2-4-6-8 Sandaker Outer-centre-N
83-1 4 98 B Casinetto, Gustav Vigelandsvei 42-44-46 Skøyen Outer-centre-W
98-2 3 62 C Frydenlundsgate, Frydenlundsgate 5-7 Bislett Centre
00-1 3 70 AC Byhagen, Nordbygata 3-11 Grønland Centre-E
* : Apartment layout according to the typology described in section 4.3.3.
Projects in italics are studied by the pilot survey

131
Floor number counted as in Norwegian, i.e. “ground floor” is floor number one.

116
Figure 5.1.a Floor plans of the selected apartments, 1 : 300.

117
Figure 5.1.b Floor plans of the selected apartments, 1 : 300.

118
Figure 5.2.a
Connectivity graphs of the selected apartments.

119
Figure 5.2.b
Connectivity graphs of the selected apartments.

120
5.1.2 The Interviews and the Questionnaires

When ringing doorbells for interviews or when distributing questionnaires to

be returned by post, a feedback percentage higher that 50 is hard to achieve.
Since those who give feedback are rarely a representative sample of those
receiving the requests, such answers are an unreliable basis for generalisation
unless further information is known. Visiting people at home and doing
interviews that include looking into all the rooms of their dwelling, is
intruding on people’s privacy and cannot be expected to give a higher
response than usually achieved by simple questionnaires. In order to increase
the rate of response on some basic questions, this survey has been carried out
as interviews of “two steps”.

A few questions were selected for “Step1”-interviews to be done immediately

after ringing the door bell; it should be easier to answer the questions than to
explain that one has not time to participate. In these Step1-interviews we
asked about the kind of household (in terms of number of persons, gender,
age and relationship), about having the space needed or not (did the
households consider their living cramped, appropriate or spacious) and about
well-being in general. Finally, the residents were asked about their inclination
to participate in a more in-depth “Step2”-interview. By this strategy, it should
be possible to achieve a response rate close to 100% for the Step2-interview,
as all households asked for Step2-interviews had declared their willingness to
participate in advance. Appendix A.5.2 shows the forms applied for the
interviews; page 1 shows a form for the Step1-interview while pages 2-7
show a form for the Step2-interview.

The issues asked about in the Step2-interviews, were the residents’ reasons
for choosing their apartment, their evaluation of their apartment, their use and
disposition of rooms and their preferences about selected issues concerning
their dwelling. The interviews were done by visiting people in their homes
and consisted of both open questions and questions with predefined
categories of answers. Information about the disposition or use of rooms was
procured in two ways; firstly, (while being guided through the apartment by
one of the residents) by sketching the furniture in on a floor-plan-drawing,
and secondly, by asking explicitly where various activities took place. The
latter was done by giving each activity a sum of 12, and asking the residents
to distribute this sum of 12 among the rooms of the apartment in accordance
with the extent to which this particular activity usually took place in each
room. In a case where TV was watched only in the living room, the activity
of watching TV is described by the value of 12 for the living room. In a case
where breakfast was eaten in the kitchen as well as in the living room, and as

121
often in the kitchen as in the living room, the activity of eating breakfast is
described by the value of 6 for the kitchen as well as for the living room.132
Page 4 in the “Step2-form” (see A.5.2) shows this “activity-versus-room-
form” of the interview in apartment 83-1-9.

The experience from the pilot-interviews became a basis for planning the
number of interviews to be carried out in Step1 and in Step2 of the main
interviews. Among the roughly 15-20 identical apartments (from each of the
nine projects) selected for the Step1-interview, about six apartments were
intended to be studied by the Step2-interviews.133 This would give a total
number of about 200 apartments to visit by ringing doorbells doing the
Step1-interviews and about 70 visits into apartments doing the Step2-
interviews.

The interviews were done by myself and 4 students of architecture, with two
of us working in partnership during each interview. Being two persons
working together was important both in order to manage an interview within
a reasonable time and in order to reduce the misinterpretations that can easily
occur during an interview. Contrary to our expectations; being two did not
seem to be an invasion in the apartments or into the privacy of the residents;
in fact it was easier to achieve a relaxed atmosphere of conversation being
two visitors than being only one. The schedule was to do the interviews for
one housing project a week, which means to do 15-20 Step1 interviews and
about 6 Step2 interviews a week. The fieldwork for each project was carried
out on two evenings between 5 and 9 p.m. Parallel to this fieldwork, we had
to summarise and edit the answers of the interviews done the previous week
and to prepare for the interviews to be done the following weeks. These
preparations consisted in selecting the apartments, getting permission to do
the interviews, editing the forms and informing all households in advance by
mail. We carried out the fieldwork by doing the Step1-interview in all the
selected apartments plus one or two Step2-interviews on the first evening. On
the second evening, we did the rest of the Step2-interviews and attempted
once again Step1-interviews in the apartments where nobody was at home on
the first evening.

132
In the pilot-interviews, we only asked whether an activity took place in a room. To achieve more nuanced
information, the form was developed into having the possibility of “weighting”. A similar way of analysing
domestic activities was applied in the Swedish housing research “Bostadsutformning, bostadsanvändning”.
133
In the projects where the number of interviews is lower, such as in the project 55-1 (see table 5.2), this is
rarely due to people changing their mind about being willing to participate but to our lack of time or to people
not being at home.

122
5.1.3 Response

Table 5.2 lists the number of apartments requested and the response rates in
each of the 12 projects. Due to the number of identical apartments really
existing in each housing project, the numbers of selected apartments came
out somewhat different from the intended 20, being 19 on average. The
response rates on the Step1-questions varied from 29 to 81 % and were 57 %
on average. As expected, only a few people who were at home did not answer
the Step1-questions. Concerning the question about kinds of households, we
managed to get the relevant information about more apartments than those
with people at home. Because a name-plate picturing and naming mother,
father and two children or a house porter describing the households on the
third floor as an old widow on the right and a young couple on the left, were
sufficient information for determining the categories of the households, we
did not depend on people being at home in all apartments. The “response
rates” about kinds of household were therefore as high as 91 % on average.

As mentioned in the previous section, the plan was to carry out Step2-
interviews in about six apartments of each project and to select these
apartments among those where the residents had confirmed their interest in
participating when asked about this in the Step1-interview. However, this
selection was rarely required as the number of people at home and interested
in participating in the Step2-interview exceeded the intended six Step2-
interviews only in a few cases.

Table 5.2 Interviews , Step1 and Step2, samples and responses

Project Type * Number of Step 1 Step 2
number apartments Answers Kind of households known Willing to Interviews
selected number % number % participate done
32-1 A 12 8 67 12 100 6 5
33-1 A 20 10 50 19 95 6 5
41-2 A 20 9 45 20 100 7 6
51-3 AC 28 8 29 28 100 6 4
55-1 AC 14 9 64 10 71 8 3
70-2 C 24 12 50 20 83 7 4
74-1 B 18 8 44 18 100 4 4
76-1 B 20 14 70 20 100 7 5
Table 5.2
77-2 21
B interviews;
The 15 and
samples 71responses.
21 100 5 4
83-1 B 21 17 81 19 90 7 6
98-2 C 18 6 33 11 61 4 3
00-1 AC 15 12 80 14 93 6 6
Total 231 128 212 55
Average 19 11 57 91 5
* : Apartment layout according to the typology described in section 4.3.3.

123
5.1.4 Questions and Answers

From the time a question is figured out until an answer is received, there are
many possibilities for misunderstandings and misinterpretations. Thanks to
the pilot-survey, questions that did not work, could be improved or left out in
advance of the main survey.

Initially we intended to find out both what had been important for the
households when they chose their apartment and what the households`
qualitative evaluation of their apartment was.134 First, we asked about the
importance of the apartment, the building and the nearby surroundings, by
applying a five-step scale from “no importance” to “very important”. Then
we asked about qualitative evaluations of the same (the apartment, the
building and the nearby surroundings). This did simply not work; just a few
persons managed to distinguish between describing the importance and doing
a qualitative evaluation. Even in the cases where people understood the
difference between the questions, it was hard to find interesting information
by analysing the answers. This is due to a very close correlation between the
evaluation of quality and the evaluation of importance; features evaluated as
of high quality are likely to be considered important.135 The conclusion to be
drawn from this is that there is no reason to separate a question into sub-
questions that are neither understood by the person being asked, nor possible
to keep separate when analysing the answers. We therefore changed the
questions towards more open ones. Such open questions might very well be
combined with more guided questions with pre-defined categories of
answers. In order to keep an open question open, this open question should
be asked in advance of more guided ones on the same subject. By such
structured questions, it is possible to capture detailed information about the
subjects that are the main focus, without loosing information about the
importance of the subject (as perceived by the person being asked). In these
interviews the open question “Why did you choose this dwelling?” (question
1.4, see A.5.2) was followed first by some more focused questions such as

134
Even in cases where the subject is of no importance for the person being interviewed, it is possible to get
seemingly informative answers describing the quality by a scale from good to bad when asking explicitly for
such an evaluation. In principle, what is interesting is therefore not such an evaluation of quality but the
evaluation of quality “weighted” in accordance with the importance of the subject (as considered by the person
being asked). This is what we attempted to grasp by distinguishing between the evaluation of importance and
the evaluation of quality.
135
When asking for evaluation of features that are chosen by the person being interviewed, this
correspondence is not hard to understand. As such choices often is based on personal judgements, a negative
evaluation would imply to disqualify this personal judgement. A reasonable way to make life comfortable is
therefore to reduce the importance of the aspects of life where you have made failures and to focus on your
more successful choices.

124
“Was there anything about the apartment that was important for your
choice?- If so, please specify” (question 1.4.d) and then, finally, by questions
with predefined categories of answers.

The forms applied in order to capture how various activities take place in the
different rooms of an apartment were not straightforward to answer.136 Quite
surprisingly, these questions were a section of the interviews that worked
very well; almost without exceptions the residents were deeply involved and
corrected us immediately if they considered our notes differed from their
answers. Our conclusion from this is that the challenge of designing
interviews is not the complexity of questions; problems arise when even
seemingly simple questions are slightly unclear. Pilot surveys are extremely
important in order to identify questions that are easily misunderstood as well
as questions on which the answers do not provide information relevant to the
subject of interest; a bad working pilot survey might be the best basis for a
successful subsequent main survey.

136
This form is page 4 (side 4) in the Step2-interview-form, see A.5.2. For further explanations see section
5.1.2 and 5.2.5.

125
5.2 THE RESULTS

5.2.1 General

This section sheds light on the typology of apartments identified in chapter 4

by describing the results from the synchronic survey. The analyses have been
carried out by looking for general patterns in the answers as well as by
examining some out of the many possible combinations of information
concerning kind of household, layout of apartments and the residents’
preferences in detail. The main information analysed is as listed below.

Information about apartments and households:

x floor plan of the apartment (by projects, see the floor plans in A.5.1)
x principle layout of the apartment (by the types A, B and C, see 4.3.3)
x apartments size (by number of rooms and by floor area)
x sizes of rooms (by floor area)
x household by kind of households (by category 1-5, see section 5.2.3)
x household by number of persons

Information about use and about the residents’ preferences:

x evaluation of well-being
x evaluation of spaciousness (in terms of cramped versus spacious living)
x use of rooms (according to furniture and naming of rooms)
x daytime spent in the different rooms
x kind of activities taking place in the different rooms
x changes of the interior space (made after the building was constructed)

The following sections present a selection of the analyses that have been
carried out and are ordered in accordance with the subjects listed below. The
results come from the pilot-survey as well as from the main survey.137

5.2.2 Spaciousness of living

5.2.3 Kinds of households
5.2.4 Time spent in different rooms
5.2.5 Rooms and activities
5.2.6 Changes made by the residents
5.2.7 Preferences about alternative layouts

137
As some questions were improved after the pilot-survey, some data from the pilot-survey are less complete
than similar data from the main survey. However, in most questions there have been minor or no changes
between the pilot-survey and the main survey, and the data from all twelve projects are comparable. ”The
distribution of daytime living” is an example of information that is collected only in the main survey. (See 32-
1, 55-1 and 74-1 in table 5.3.)

126
5.2.2 Spaciousness of Living

The size is an important feature of the dwelling as regards floor area as well
as number of rooms. The size of dwelling preferred by a particular
households depends on the number of persons as well as on the composition
of the household. Where Norwegian housing conditions are concerned,
Dagfinn Ås (1971) has categorised to what extent living conditions were
cramped or spacious. He compared the number of persons to the number of
“habitable rooms”138, and defined the size of a dwelling to be appropriate
when the number of persons in the household was equal to the number of
habitable rooms.139 By this definition, a couple lives in an apartment of
appropriate size when this has “two rooms”, this usually means that the
dwelling, in addition to kitchen and “service spaces”, has a living room and
one bedroom. By the same definition, a couple with one child have an
apartment of appropriate size when they have one more room, a room usually
used as a bedroom for the child. The data collected in this survey is an
opportunity to examine if such an evaluation of spaciousness still makes
sense today.

Based on the definitions of Ås, a five level “index of spaciousness”, ranging

from “very cramped” to “very spacious”, would be as follows.

“index” description number of persons versus rooms

1 Very cramped140 P>R+1

2 Cramped P=R+1
3 Appropriate size P=R
4 Spacious P=R-1
5 Very spacious P<R-1

Where P is the number of persons in the household, and R is the number of

“habitable rooms” in the apartment.

138
Number of habitable rooms does not count “service spaces” such as kitchen, bathrooms, washing, storage or
circulation.
139
Ås (1971, p. 25). The same definition is used by Guldbrandsen (1973, p. 8). For another evaluation of
cramped living conditions, see footnote 20, which describes the “official definition” in Oslo in 1893.
140
By this definition single person households live “very cramped” only when the number of rooms is “0”. In
common terminology there is no such thing as a dwelling of no rooms. However, if the traditional Norwegian
terminology is applied (which means that the kitchen is not counted), an apartment where the kitchen is in the
living room and where there is no separate bedroom would be an apartment of “0” rooms. Unless such few
rooms are compensated by a very large living room, “very cramped” is not an inappropriate description of
sleeping in your only room that includes the kitchen, a kind of layout that is common for new apartments now.
(As will be commented in chapter 6 and 7.)

127
Due to the increasing standard of living141 and the increasing number of
single-family households since 1970,142 one can expect the preferred size of
apartments according to Ås to be out of date. Given that the Ås’ results were
representative in 1973, households of to-day would be expected to describe
their apartments as more cramped than indicated by the index above.
Particularly where single person households are concerned, evaluations of
spaciousness as defined by Ås is likely to differ from evaluations done by
residents to-day. According to Guldbrandsen (1973) single-person-
households accept “one-room-apartments” (a dwelling with living room and
kitchen but without a separate bedroom) only for limited periods; as a
student’s home or in other temporary phases it is acceptable, while at more
permanent stages of life adults were not satisfied with homes without a
separate bedroom. According to Guldbrandsen single-person-households can
be expected to prefer more space than described by the categories of Ås.

Table 5.3 lists this “index of spaciousness” versus spaciousness as evaluated

by the residents of the apartments examined by the Step2 interviews. In the
interviews, the residents were asked to characterise the spaciousness of their
apartment by the 5-level scale listed on previous page; ranging from very
cramped to very spacious. (Question 5 in the Step1-interview, see A.5.2.)
When this “real-life-evaluation” by the residents is compared with the “index
of spaciousness” according to Ås, the average deviation is “minus 0.4”,
which in literal terms means that the households evaluate their living as
somewhat more cramped than indicated by the index. Such relatively good
correspondence between this index and the evaluations means that the simple
and 30 years’ old definition of what is an appropriate number of rooms is less
outdated than expected above.143 However, the deviations between real life
experiences and the index depend on the kind of household. For singles the
deviation is -0.7 on average, which means that singles prefer almost one more
room than what according to the index should be an apartment of appropriate
size. This corresponds well to the above-mentioned conclusions of
Guldbrandsen (1973). For couples with children, the average deviation is
+0.2, which means that family-households feel comfortable with slightly less
place than what would be an appropriate number of rooms according to the
index.

141
Concerning spaciousness of living, the average number of person per room in Norwegian apartments was
0.6 in 2001 compared to 0.8 in 1980. (According to Statistisk Sentralbyrå / Statistics Norway,
http://www.ssb.no/fobbolig/tab-2002-09-23-01.html)
142
In 2001 the number of single-households in Norway was 740 000 (38%) compared to 426 000 (28%) back
in 1980. (According to Statistisk Sentralbyrå / Statistics Norway, see table 1.3.
143
If keeping a scale of integers, the evaluation of spaciousness should differ from the index by more than 0.5
before an adjusted scale (i.e. a scale where P=R+1 was the appropriate size) would be more descriptive.

128
Table 5.3 Summary of survey
Spaciousness Daytime living Spaciousness

Area/person
Apartment
(by evaluation

Bedroom 2
Bedroom 3
Well being

Evaluation

Differance

Bedroom1
Floor area

By Index
and by floor area

Kitchen
Persons
Project
Rooms

Living
per person)

Type

Age
1 2 3 4 5
32-1 2 A 57 A3 30 29 4 3 1 29
A4 30 29 4 3 1 29
B1 60 57 2 4 -2 57
C1 60 57 5 4 1 57
C4 40 29 2 3 -1 29
Average 1,6 44 40 3,4 3,4 0,0
33-1 2 A 50 A4 30 50 5 5 4 1 7 3 2 - - 50
B1 30 25 3 2 3 -1 8 3 1 - - 25
C5 30 50 5 4 4 0 6 3 2 - - 50
D2 30 25 5 3 3 0 8 2 1 - - 25
D5 20 25 5 3 3 0 5 3 4 - - 25
Average 1,6 28 35 4,6 3,4 3,4 0,0 6,8 2,8 2,0
41-2 3 A 75 A4 30 25 5 4 3 1 6 2 1 3 - 25
C1 30 37 5 4 4 0 11 0 1 0 - 37
C3 40 19 5 2 2 0 5 3 1 2 - 19
D1 30 25 4 3 3 0 4 4 2 1 - 25
D2 30 37 5 3 4 -1 4 3 2 1 - 37
D4 60 25 5 4 3 1 6 1 1 4 - 25
Average 2,8 37 28 4,8 3,3 3,2 0,2 6,0 2,2 1,3 1,8
51-3 3 AC 74 C2 30 25 4 4 3 1 5 5 0 0 - 25
C5 60 74 5 4 5 -1 2 5 0 5 - 74
D2 30 74 5 4 5 -1 10 2 0 0 - 74
F1 30 37 5 3 4 -1 6 2 3 1 - 37
F4 30 37 5 4 4 0 8 2 1 0 - 37
Average 1,8 36 49 4,8 3,8 4,2 -0,4 6,2 3,2 0,8 1,2
55-1 3 AC 67 C1 30 22 4 3 3 0 22
C2 70 67 5 4 5 -1 67
D2 60 33 5 3 4 -1 33
Average 2,0 53 41 4,7 3,3 4,0 -0,7
70-2 2 C 56 A4 60 56 5 4 4 0 8 2 0 - - 56
B1 80 56 5 3 4 -1 12 0 0 - - 56
C3 40 56 5 3 4 -1 10 1 1 - - 56
G2 30 28 5 3 3 0 6 0 4 - - 28
Average 1,3 53 49 5,0 3,3 3,8 -0,5 9,0 0,8 1,3
74-1 3(4) B 83 A2 20 41 5 5 4 1 41
B4 60 41 5 3 4 -1 41
D3 20 21 4 3 2 1 21
Average 2,7 33 34 4,7 3,7 3,3 0,3
76-1 4 B 95 A1 40 19 5 2 2 0 5 3 0 1 2 19
A4 80 95 5 4 5 -1 7 2 1 1 0 95
B1 30 47 5 4 5 -1 6 3 0 0 2 47
B4 30 16 3 3 2 1 4 4 0 0 0 16
D4 80 95 5 5 5 0 6 2 0 3 0 95
Average 3,0 52 54 4,6 3,6 3,8 -0,2 5,6 2,8 0,2 1,0 0,8
77-2 4 B 104 A4 70 52 4 3 5 -2 6 2 0 2 0 52
A5 80 104 5 5 5 0 6 2 1 2 0 104
A7 70 52 5 3 5 -2 4 2 - 2 2 52
C6 70 52 4 4 5 -1 4 4 1 1 2 52
Average 1,8 73 65 4,5 3,8 5,0 -1,3 5,0 2,5 0,7 1,8 1,0
83-1 4 B 98 3 40 33 5 3 4 -1 7 3 0 1 0 33
4 50 49 5 5 5 0 4 3 4 0 1 49
7 50 49 3 2 5 -3 4 4 0 2 2 49
8 40 33 5 3 4 -1 4 2 2 1 2 33
9 30 33 4 4 4 0 4 5 0 0 2 33
Average 2,6 42 39 4,4 3,4 4,4 -1,0 4,6 3,4 1,2 0,8 1,4
98-2 3 C 60 A1 40 30 5 5 4 1 8 - 4 0 - 30
A4 50 60 4 4 5 -1 9 - 3 0 - 60
B3 30 60 4 3 5 -2 10 - 1 0 - 60
Average 1,3 40 50 4,3 4,0 4,7 -0,7 9,0 2,7 0,0
00-1 3 AC 70 A2 30 35 4 4 4 0 5 3 0 3 - 35
A4 50 70 4 4 5 -1 2 8 2 0 - 70
B3 60 70 5 3 5 -2 8 2 0 0 - 70
C2 30 35 5 4 4 0 7 2 3 0 - 35
C4 30 35 5 4 4 0 4 2 3 0 - 35
E3 50 35 5 4 4 0 4 4 2 1 - 35
Average 1,7 42 47 4,7 3,8 4,3 -0,5 5,0 3,5 1,7 0,7
Average deviation evaluation - index -0,4 Average of all 33 38 46 61
Table 5.3
Average of A type 2,1 36 33,9 4,7 3,4 3,3 0,1 32 28 32 53
Average of B type 2,5 51 48,9 4,5 3,6 4,2 -0,6 34 35 57 72
Average of C type 1,3 47 49,2 4,7 3,6 4,1 -0,6 50 58 30
Average of AC-type 1,8 42 46,3 4,7 3,7 1294,2 -0,5 41 49
If we examine the results listed in table 5.3, we find some correspondence
between the principle apartment-layout and the evaluation of spaciousness.
As pointed out by table 5.4, households living in the A-type apartments
evaluate their living as about as spacious as described by the “index”, while
those living in the other types of apartments consider their living significantly
more cramped than indicated by the index. We find a similar pattern when
comparing the floor areas. Households who consider their living conditions
appropriate where spaciousness is concerned, have on average 28 m2 per
person in the apartments of the A-type, while those living in the B-type and
the C-type apartments need on average 35 and 50 m2, respectively, in order
to consider their apartment appropriate. However, as the apartments to
compare becomes few when going into such details, it is hard to generalise
about the influence of the spatial layout on the evaluation of spaciousness
without examining further samples.144

In conclusion, returning to the evaluation of spaciousness in general, the very

simple index based on Ås (1971) is still relevant. Even though such an index
of spaciousness is far from being a scientific measurement, the index still
makes sense today as an indicator of real-life experience where the sizes of
Norwegian dwellings are concerned. However, the fact that the preferred
number of rooms depends on the size of the rooms must be kept in mind. Due
to the very different sizes of rooms and the move towards small apartments
with tiny rooms apparent now (see section 6.3), evaluations of spaciousness
of living by simply comparing the number of persons to the number of
rooms, should be done with more care for newer apartments than for older
ones.

Table 5.4 Apartment type and spaciousness of living.

Apartment type Area/ Spaciousness Area/person in apartments

person by evaluation by index * deviation evaluated by the residents
to be of "appropriate size"
A 34 3,4 3,3 0,1 28
Table 5.4 B 48 3,6 4,2 -0,6 35
C 49 3,6 4,1
Kind of floor-plan layout and evaluations of spaciousness. -0,5 50
AC 46 3,7 4,2 -0,5 41

* : about "index of spaciousness" , see section 5.2.2

144
There is for instance a correspondence between “floor area needed” and the age of the households, which
need larger samples in order to be examined. (See footnote 175.)

130
5.2.3 Kinds of Households

In order to examine the distribution of households in different apartments,

some categories of households must be determined. The categories should
cover the entire “population” of households, they should distinguish between
households presumed to be different concerning the subject of interest, they
should not be too numerous145 and they should have some correspondence to
the categories applied in official Norwegian statistics of housing and
population. The latter makes it possible to compare the results of this survey
to existing relevant statistics. The five categories listed below were defined
thereafter.146 Most people pass through many of these categories of
households in their lifetime.

1. Single person
2. Couple
3. Single person with children 0-18 years
4. Couple with children 0-18 years
5. Several adults, with or without children
(more than one adult, where at least one is not part of a couple) 147

Kinds of households distinguished by five categories.

5.2.3.1 Comparing all apartments

Table 5.5 lists the distribution of households according to the categories just
listed. Some patterns appear clearly when comparing the distribution of kind
of households in all projects (see table 5.5.a). Most significant and worth
further comment is the many “several-adults-households” in projects 41-2,
74-1 and 00-1 and the few couples and many households with children in
project 83-1.

145
When samples are small, the categories should be few. If there are many, even simple statistics such as
applying percentages and comparing this to larger populations do not make sense.
146
This is two categories fewer than those applied in the official Norwegian statistics by the SSB (Statistisk
Sentralbyrå / Statistics Norway). However, the categories of SSB can easily be merged into these five.
147
Category 5 contains households where there are other adults than one single or one couple. Some examples
of such households are:
x two (or more) adults (for instance students) sharing an apartment (without being a couple)
x households with adult children living at home
x households with a room for rental
x households with adult relatives other than the mother and father

131
Table 5.5 Distribution of households, by kind of household

Table 5.5.a All projects, ranked by chronology

Kind of households Kind of households

Households known (numbers)

(numbers) (% of known)
Case, by project number

Average time of living (**)

1 2 3 4 5 1 2 3 4 5

Average "well-being" (*)

Unknown (of sample)

Couple + child(ren)

Couple + child(ren)
Type of apartment

Single + child(ren)

Single + child(ren)
Number of rooms

Floor area (m2)

Several adults

Several adults
Couple

Couple
Single

Single
32-1 A 2 57 12 6 4 1 1 0 50 33 8 8
33-1 A 2 50 19 3,2 4,6 13 5 1 1 68 26 5
41-2 A 3 75 20 11,4 4,9 6 4 2 5 3 0 30 20 10 25 15
51-3 AC 3 74 27 9,4 4,9 13 8 1 3 2 1 48 30 4 11 7
55-1 AC 3 67 12 4,5 6 4 2 4 50 33 17
70-2 C 2 56 19 7,0 4,7 15 4 4 79 21
74-1 B 4 83 18 4,3 3 7 3 2 3 0 17 39 17 11 17
76-1 B 4 95 19 14,1 4,6 6 5 2 5 1 1 32 26 11 26 5
77-2 B 4 104 21 16,9 4,4 12 6 3 0 57 29 14
83-1 B 4 98 19 6,2 4,7 4 4 10 1 2 21 21 53 5
98-2 C 3 62 11 3,0 4,3 5 4 1 1 7 45 36 9 9
00-1 CA 3 70 14 2,9 4,8 3 6 1 1 3 1 21 43 7 7 21
Average 43 28 8 14 7
Total of Oslo (percentage of total population of households) 52 18 5 16 9
Table 5.5.b Apartments of two rooms
32-1 A 2 57 12 6 4 1 1 0 50 33 8 8
33-1 A 2 50 19 3,2 4,6 13 5 1 1 68 26 5
70-2 C 2 56 19 7,0 4,8 15 4 4 79 21
Average 66 27 5 0 3
Table 5.5.c Apartments of three rooms
41-2 A 3 75 20 11,4 4,9 6 4 2 5 3 0 30 20 10 25 15
51-3 AC 3 74 27 9,4 4,9 13 8 1 3 2 1 48 30 4 11 7
55-1 AC 3 67 12 4,5 6 4 2 4 50 33 17
98-2 C 3 62 11 3,0 4,3 5 4 1 1 7 45 36 9 9
00-1 CA 3 70 14 2,9 4,8 3 6 1 1 3 1 21 43 7 7 21
Average 39 32 6 12 11
Table 5.5.d Apartments of four rooms
74-1 B 4 83 18 4,3 3 7 3 2 3 0 17 39 17 11 17
76-1 B 4 95 19 14,1 4,6 6 5 2 5 1 1 32 26 11 26 5
77-2 B 4 104 21 16,9 4,4 12 6 3 0 57 29 14
83-1 B 4 98 19 6,2 4,7 4 4 10 1 2 21 21 53 5
Average 32 23 12 26 7
(*) : Average well being for all households participating in Step1-interview, the scores
are therefore not exactly the same as "well-being" in table 5.3 that summarises the
fewer apartments visited by the "Step2-interviews".
(**) : Table 5.5 time
Average Kinds
ofof households
living in the apartment, among those living there.

132
The “Several-Adults-Households”

The projects 41-2, 74-1 and 00-1 seem to be particularly preferred by

“several-adults-households” as the percentages of such households are 15, 17
and 21 in these projects, while it is less than 10 in all the other projects.
Given that the spatial layout strongly influences the kinds of households
preferring an apartment, which is a basic assumption of this study, these three
apartments should have some features in common that distinguish them from
the others. When looking at the layout in terms of the typology distinguished
in section 4.3.3, no such pattern is found as the apartment 41-2 is of the A-
type, 74-1 is of the B-type while 00-1 is a hybrid of the AC-kind. However,
when looking more closely at the rooms and their configurations, some
patterns that might be relevant occur. Firstly, all these three apartments are
among the five (out of the total twelve) where there are two bedrooms and
where there is a kitchen separate from the living room and with direct access
from the entrance. Not surprisingly, households consisting of adults that are
not couples, prefer apartments with more than one bedroom. Concerning the
kitchen separate from the living room, this is an advantage for similar
reasons. As such a kitchen can be used without disturbing what goes on in the
living room and vice versa, this is a layout likely to be appreciated by
households consisting of two (or more) adults not closely related. A look at
the positions of the bedrooms reveals another feature being characteristic for
these three apartments; the main bedroom (or a large second-largest bedroom,
which is the case in 00-1 where a storage room works as an extension of the
second largest bedroom) is accessible directly from the entrance and located
as far from the other bedroom as possible. (See floor plans in figure 5.1 and
connectivity graphs in figure 5.2.) In conclusion, two features of the floor-
plan-layouts seem to be important if intending to attract households
consisting of adults which are not couples: one is that there should be a large
bedroom positioned by the entrance and not close to the other bedroom, the
other is that the kitchen and the living room should be separate from each
other. By such a floor plan layout, adults not intimately related might have
some privacy within the one apartment.

133
Children or Not; Contradicting Preferences

Where households with children are concerned, the number of such

households in the apartments 83-1 is as high as 74 % (21+53). In comparison,
the percentage of households with children in the whole of Oslo is 21
(5+16)148 while the percentages among the other large apartments of this
study (74-1, 76-1 and 77-2) are 28 (17+11), 37 (11+26) and 14, respectively.
Apartment 83-1 differs from the others in this study by having a spatial
layout very similar to common single-family houses, in the way that three
bedrooms, the bathroom and the WC constitute a part of the dwelling located
farthest from the entrance and separated from the rest of the apartment by a
corridor. This spatial layout might explain a higher that average rate of
households with children, but hardly an increase to 250 %, which is the rate
of households with children in 83-1 compared to 74-1 and 76-1. By looking
at the situation outside the apartments, a more plausible explanation of the
very high rate of households with children in project 83-1 can be found. This
project consists of nearly 300 apartments and is located in the west of Oslo
among mainly single family houses and attached houses. It is close to
Frognerparken, which is the largest park in Oslo. The areas between the
houses have the character of playgrounds, the use of cars is restricted within
the area and school as well as kinder garden are located nearby. As long as
the apartments were not particularly unsuitable for households with children,
anything but a high rate of such households in this project would be
surprising. The rate of couples without children is converse to the rate of
families with children. Among the 21 households in this project known
through this survey, there are no couples without children, while the
percentages of such households in the other 11 projects vary between 20 and
43. A likely explanation is that couples without children do not appreciate the
crowding inevitable with the numerous children and their equipment; this is
simply not a pleasant place to live for people who prefer to sit in peace and
quiet on their balconies. The fact that the apartments as well as the out-door
areas are preferred by households with children, makes the place unattractive
for couples without children. This project shows how the feature of being
highly appreciated by one particular kind of household may imply the
exclusion of others.

148
According to Statistisk Sentralbyrå / Statistics Norway, see table 1.3.

134
5.2.3.2 Comparing Apartments of Similar Sizes

The presumed effect of apartment layout on the kind of households choosing

an apartment can be examined more closely when comparing apartments of
the same number of rooms. In this study, there are three projects where the
typical apartments have “two rooms”. These are 32-1 and 33-1, which are of
the A-type, and 70-2, which are of the C-type. (See table 5.5.b.) 70-2 is
inhabited by singles and couples only149, while 32-1 and 33-1 have other
categories of households as well. Even though the A-type of apartments
house a wider range of households than the apartments of the C-type, the A-
type apartments do not have a high frequency of households with children.
As mentioned earlier, the number of rooms must be borne in mind when
studying the layout of apartments compared with the kinds of households
living in the apartments; whatever the spatial layout might be, an apartment
with only one bedroom cannot be expected to house many large families. As
these apartments have “two rooms”, which usually means only one bedroom,
the small number of households with children is not surprising.

Where apartments of three rooms are concerned, the sample contains

altogether five cases. 41-2, which is of the A-type, 98-2, which is of the C-
type and the cases 51-3, 55-1 and 00-1, which are all hybrids described as
being of an AC-kind150. (See table 5.5.c.) The A-type apartment 41-2 has the
most even distribution of kind of households; all kinds being represented,
varying from 10% of the apartments housing singles with children to 30% of
the apartments housing single-person households. Looking more closely at
the rate of households with children, the percentages among the A-type and
the C-type apartment are 35 (10+25) and 9, respectively.151 The AC-hybrid-
type of apartments, which are 51-3, 55-1 and 00-1, have a rate of children
families in between the two others; the percentages being respectively 15, 17
and 14.

149
The C-type apartment 70-2 is the apartment with the highest rate of single person households (79%).
150
About this AC-kind, see section 4.3.3 and 5.1.1.
151
This is the same pattern as among the apartments of two rooms. Among the apartments of two rooms, the
C-type apartments 70-2 are the only ones where there live no children. (See table 5.5.b.)

135
When it comes to the apartments of four rooms, all are of the B-type. (See
table 5.5.d.) Due to the sizes of the apartments as well as the locations within
the city, these apartments differs from the A- and the C-type of apartments;
the B-type apartments are larger than the others and are usually located in the
suburbs built in the period around the 1970s. 74-1 is the B-type apartment
examined by the pilot-survey while 76-1, 77-2 and 83-1 are the B-type
apartments selected for the main survey. 74-1 and 76-1 are both located at
Romsås, a suburb in the north of Oslo, while 77-2 and 83-1 are located in the
north and west of central Oslo, respectively. The sizes of the four B-type
apartments vary from 83 to 104 m2. When built, the 74-1 had only two
bedrooms, the fourth room of this apartment was in original a dining room.
As expected, due to the floor plan of the apartment as well as to the suburban
locations, both 74-1 and 76-1 house many families with children; the
percentages being respectively 28 (17+11) and 37 (11+26). The distribution
of households in the two other projects, 77-2 and 83-1, differs significantly
from those in 74-1 and 76-1 in the sense that the rate of households with
children is remarkably high in 83-1 and very low in 77-2. The high rate of
households with children in project 83-1 is commented upon in the previous
section. What therefore should be examined more closely here, is the project
77-2 where the percentage of households with children is as low as 14. 77-2
is the largest apartment in the survey, with a lift, underground car parking
and a floor area of 104 m2. Given the layout of the apartments, the case 77-2
has a low rate of households with children and a high rate of singles.152
However, the explanation is not that the apartment or the building is
particularly appropriate for singles or particularly inappropriate for
households with children. 77-2 is the case of this survey where the residents
have been living in their apartments for the longest time, on average they
have been living there for as long as 17 years. Among the 15 households
which answered this question, 7 have been living there all the 28 years since
the building was constructed. To put it simple; those who bought these
apartments as dwellings for family-living back in the 1970s still live here;
having no garden to worry about, a spacious apartment, a nice view, access
by lift and no more loan to pay, there is no reason to move when getting
older. Despite what by first thought might appear to be an apartment
preferred by limited categories of households only, 77-2 is appropriate for a
large variety of households, ranging from young families with children to
elderly.

152
Only the cases 33-1 and 70-2, which are the smallest apartments in the survey (being only half the size of
77-1), have a higher percentage of single person households.

136
The results of this section particularly worth further comment, are those
found by the comparison of the three-rooms apartments. First, the kinds of
households living in the different apartments correspond almost too well with
the hypotheses about generality described in section 4.3.3; the A-type of
apartment, which is the one characterised as most general, houses the greatest
diversity of households, the C-type house the least, while the hybrid-type AC
has a diversity of households in between the two. Second, the deviation
between the results from the three projects of the latter kind is remarkably
low. Larger samples should be examined in order to find out if there really in
general is such a significant correspondence between distribution of
households (by kinds of households) and the spatial layout of apartments.153
If this should be the case, the spatial layout is of even higher importance to
dwelling preferences than initially assumed in this thesis.

As just described, the survey seems to confirm the assumption that the
general A-type of apartments should house a diversity of households.
However, this does not imply that such apartments are necessarily better
dwellings than those that are more specific with respect to functions or use;
apartments chosen by limited kinds of households might be highly
appreciated by those particular households preferring them. The project 70-2
is such an example, housing only singles and couples and achieving a high
score on well-being, see table 5.6. The distribution of households living in
the project 41-2 is very different, representing a wide range with respect to
kinds of households. This diversity of households does not imply that the
apartment necessarily is highly appreciated by everybody living there. The
higher than average rate of households with children might for instance be
explained by the fact that many other dwellings and housing areas are not
suitable for households with children; giving a higher rate of such households
in the kind of dwellings that are fairly suitable. Where singles and couples are
concerned, the percentage living in 41-2 is far below the total in Oslo (being
50 % (30+20) compared to 70 % (52+18) in the whole of Oslo, see table 5.5).
It could therefore be that this apartment is not suitable for singles and
couples, even though some such households live there. A way to examine this
more closely is to compare the well-being in 41-2 to the well-being in 70-2.
All the ten households in 70-2 participating in the Step1-interview were
singles or couples and they evaluated their well-being as 4.7 on average,
where score 5.0 is the maximum possible. In the apartments 41-2 two
households of singles or couples were interviewed; both evaluated their well-

153
Among the three-rooms apartment examined here, there is for instance some differences in apartments’
floor areas, ranging from 62 to 75 m2, differences that partly corresponds to the apartments types A, C and
AC. In order to analyse the spatial configurations more closely, the sample examined should consist of
apartments even more equal by size or of apartments were the floor areas (if possible?) do not corresponds
with the spatial configuration of the apartment.

137
being to the score of 5.0. (See table 5.6.) If we compare the well-being in all
apartments of all the 12 projects, 41-2 is the one with the highest score, being
4.9 on average. (See table 5.5) The high rate of households with children is
therefore unlikely to be explained by the apartment being inappropriate for
others; 41-2 is highly appreciated by all the kinds of household living there.
Another result indicating that the residents in 41-2 are very well being is how
long they have been living there. As shown in table 5.5, 41-2 is the project
with the second highest score in this respect, only beaten by the project 77-2
described earlier in this section. This closer look at the apartment 41-2
thereby indicates that generality represents a potential not only for housing a
wide range of the existing variety of households, but also for being highly
appreciated by such a variety of households.

Table 5.6 Spaciousness of living and well-being in 41-2 and 70-2

Floor area (BRA)

Spaciousness Spaciousness
Apartment type

Spatial layout

(by evaluation
Area/person

Well being

Evaluation

Differance

and by floor area

By Index
Rooms

per person)
Age

Nr. 1 2 3 4 5
41-2 3 A 75 A1 30 75 5 4 5 -1 75
A4 30 25 5 4 3 1 25
B3 40 15 5 3 1 2 15
C1 30 37 5 4 4 0 37
C3 40 19 5 2 2 0 19
D1 30 25 4 3 3 0 25
D2 30 37 5 3 4 -1 37
D3 40 19 5 3 2 1 19
D4 60 25 5 4 3 1 25
Average 37 31 4,9 3,3 3,0 0,3 19 24 41
70-2 2 C 56 A1 30 56 5 2 4 -2 56
B1 80 56 5 3 4 -1 56
Table 5.6 D1 90 56 5 3 4 -1 56
and well-being in projects 41-2, 70-2 and 77-2.
A2 20 28 5 4 3 1 28
C2 30 28 4 2 3 -1 28
D2 40 56 5 3 4 -1 56
G2 30 28 5 3 3 0 28
H2 40 56 3 3 4 -1 56
C3 40 56 5 3 4 -1 56
A4 60 56 5 4 4 0 56
E4 40 56 5 3 4 -1 56
Average 50 48 4,7 3,0 3,7 -0,7 42 52 42

138
5.2.4 Time Spent in Different Rooms

Information about the amount of daytime spent in the different rooms was
found similarly to the method applied when asking about the different
domestic activities.154 We asked the households to give their total daytime
spent at home the sum of 12 and to distribute this sum among the rooms in
accordance with the daytime spent in each room. The results in terms of the
“relative amount of daytime spent in different rooms” are listed in table 5.3
and summarised in table 5.7 below. As expected, the living room is the room
of all apartments where most daytime is spent. Some more interesting
patterns can also be found. If the results are ordered according to the time
spent in living rooms, as shown in the lower table in table 5.7, three
categories can be distinguished. The first category is the apartments with the
lowest scores (varying from 4.7 to 5.6), the second is those with medium
score (6.0 and 6.8) while the third is those with by far the highest score (both
being 9.0).155 These categories corresponds with the typology of apartments
in the way that the B-type apartments have least daytime spent in the living
room, those of the A type have the medium while the C-type apartments have
the most. These results are in the following examined more in detail.
Table 5.7 Relative amount of daytime spent in different rooms.

Ranked by project number ( = chronography )

Project Apartment type Living room Kitchen Bedroom 1 Bedroom 2 Bedroom 3
33-1 A 6,8 2,8 2,0
41-2 A 6,0 2,2 1,3 1,8
51-3 AC 6,2 3,2 0,8 1,2
70-2 C 9,0 0,8 1,3
76-1 B 5,6 2,8 0,2 1,0 0,8
77-2 B 5,0 2,5 0,7 1,8 1,0
83-1 B 4,7 3,4 1,2 0,8 1,4
98-2 C 9,0 2,7 0,0
00-1 AC 5,0 3,5 1,7 0,7
RankedTable 5.7 spent in living room
by time
Time spent in the living room
83-1 B 4,7
77-2 B 5,0
76-1 B 5,6
41-2 A 6,0
33-1 A 6,8
70-2 C 9,0
98-2 C 9,0

154
See section 5.1.2.
155
When in the interviews asking about daytime living, the kitchens of 70-2 and 98-2 are “modelled” as a
separate room and as included in the living room, respectively. (See the floor plans in A.5.1 and “daytime
living” in table 5.3.) This different modelling can theoretically be questioned (as the kitchen in 70-2 is just
slightly more separated from the living room than is the case for the kitchen in 98-2) and it is not consistent
with the space syntax analysis where both these kitchens are modelled as separate rooms (due to both being
partly enclosed from the living and both having windows). However, as there is hardly any daytime living in
the kitchen in 70-2, this different modelling has no influence on the results discussed here.

139
The first category consists of the large apartments 76-1, 77-2 and 83-1, which
are the apartments of the B-type. Despite the fact that living room in the B-
type of apartments is a room particularly designed for daytime living, these
apartments are those with the lowest relative amount of time spent in their
living rooms. This can be explained by “demand” as well as by “supply”. In
77-2, households consisting of few persons live in large apartments. In such
cases, all bedrooms are not needed for sleeping and are thereby available for
other purposes. On the other hand, households consisting of many persons
(households of which there are several in 76-1 and 83-1) can be expected to
have a “demand” for places appropriate for simultaneous and “not easily co-
existing activities”. As these large apartments have more rooms (compared to
the smaller apartments) where such “parallel activities” might possibly take
place, daytime living becomes less concentrated in the living room.

The second category, - those where the living rooms have the medium score
concerning daytime spent, are 33-1 and 41-2, which are both apartments of
the A-type. The relative amount of time spent in the living rooms is greater
here than is the case for any of the B-type apartments described above, even
though the A-type of apartments is described as most general due to several
rooms that should be appropriate for daytime living. Explanations of this are
not obvious, but if we take a closer look at the spaciousness of living, some
patterns occur. Figure 5.3 shows how the spaciousness of living differs
between apartments of different layouts, and particularly between those of the
A-type and those of the B-type. The average “spaciousness of living” is
“appropriate” for the A-type while it is about “spacious” for the B-type of
apartment156 and the deviation of spaciousness is far smaller among the A-
type apartments than among the B-type of apartments.157 As will be
elaborated in section 5.2.5, when examining where different domestic
activities take place, spacious living seem to increase the use of the rooms
other than the living room. When the apartments of the A-type are compared
with those of the B-type, their more “appropriate” size on average, as well as
the fact that very few A-type apartments are spacious or cramped, might
explain that more time is spent in the living rooms in the apartments of the A-
type than in the living rooms of the apartments of the B-type.158

156
The average “spaciousness by index” for the A- and the B- type of apartments are 3.4 and 3.7, respectively.
The “spaciousness by evaluation” shows the same pattern, being 3.3 and 4.2, respectively. (Found from table
5.3, see also figure 5.3 on next page.)
157
Most apartments of the A-type (9 out of 16) are “appropriate” (which means that they achieved the score of
3 on “index of spaciousness”) while only one is “cramped” and not a single apartment is “very cramped” or
“very spacious”. Out of the 17 B-type apartments there is not one “appropriate” while 2 are “cramped” and as
many as 9 are “very spacious”, see figure 5.3.
158
As explained in section 5.1.1, the apartments of the B- type differ from the others by size and by location.
Too strong conclusions should therefore not be drawn from direct comparisons of apartments of the B-type and
the others, such as the comparison attempted here.

140
 A-type apartments
75

Percentage
50

25

0
1 2 3 4 5

B-type apartments
75
Percentage

50

25

0
1 2 3 4 5

C-type apartments
75
Percentage

50

25

0
1 2 3 4 5

"Spaciousness of living" A-type apartments B-type apartments C-type apartments

(by index, see section 5.2.2) number % number % number %
1 very cramped
2 cramped 1 9 3 18
3 appropriate 6 55 1 14
4 spacious 4 36 5 29 4 57
5 very spacious 9 53 2 29
total 11 100 17 100 7 100
Figure
average5.3"spaciousness" 3,3 4,2 4,1
Spaciousness of living in A-type and B-type of apartments.
Figure 5.3 "Spaciousness of living" in the differnt types of apartments

The third category, - the apartments where the amount of daytime spent in the
living rooms is highest, are the projects 70-2 and 98-2. These are the C-type
of apartments, where the kitchens are not separate rooms but tiny appendices
to the living rooms. In conclusion, these results about time spent in the living
rooms correspond to the typology of apartment layouts, but the C-type is the
only one where this correspondence can be explained by generality or
specificity of the floor plans. The result that the living rooms in the C-type
apartments are those with most day-time living is due to the fact that these
apartments have few or no other places where daytime living can possibly
take place. The pattern that more time is spent in the living room of the
“general” A-type apartments than in the living room of the “specific” B-type
apartment can be explained by taking into account the effects of spaciousness
of living. As will be commented further in the next section, spacious living
seems to reduce the time spent in the living room.

141
5.2.5 Rooms and Activities

When it comes to the use of individual rooms in the apartments, information

was found by the “activity-versus-room”-form described in section 5.1.2. The
legend of numbers of activities is listed below, where the numbers 1-13
represent daytime activities while numbers from 14 to 17 represent sleeping.

Activietes, legend

1 “Everyday” breakfast
2 “Everyday” dinner
3 Dinner with guests
4 Other kinds of visits (simple or no serving of food/drink)
5 Children’s play

6 Handwork, small repairs

7 Reading for entertainment (newspapers, magazines, books)
8 Reading / studies / work
9 Watching TV
10 Radio / stereo listening
11 PC
12 Children's homework
13 Hobbies
14 Night-time sleeping (except guests)
15 Relaxing / sleeping at daytime, adults
16 Relaxing / sleeping at daytime, children
17 Night-time sleeping, guests

Figure 5.4 Rooms and activities. Legend and example of diagram.

Activities in “bedroom” in apartment 33-1-D5.

142
While table 5.3 illustrates daytime spent in different rooms, figures 5.5 – 5.7
summarises some answers about where different activities take place. Two
kinds of “table-pages” are applied. The first is one that lists the results of one
kind of room across all apartments in one project (such as figure 5.5.a on next
page, which shows the use of bedrooms in project 32-1), while the other
summarises the results of one particular room across all projects (such as
figure 5.6 and 5.7, summarising living rooms and largest bedrooms,
respectively). If we look at the first kind (the figures 5.5), the table on the
upper left summarises the data from the interview159, while the diagrams on
the right illustrate the results for each apartment visited. Each column in these
diagrams represents one activity. As described in section 5.1.2, the value of
12 (which is the maximum value on the Y-axis) represents the total of an
activity; a score of 12 means that the activity takes place in one room only.
When the columns are few, few activities take place in the room, while many
columns means that a wide range of activities takes place there. In order to
compare results from several projects, there are two kinds of “summarised
diagrams; one showing the “average” and another showing the “maximum”
(or “all”) scores of each activity. These summarised diagrams, showing the
results of particular rooms of all apartments of one project, are shown at the
bottom left of the “project-by-project-pages” in figure 5.5. These two show
similar patterns but capture somewhat different information; while
“average/gj.snitt” illustrates the mean, “all/alle” illustrates the variety of use
going on among all apartments in one project better. These summarised
diagrams of all projects are what is presented in the second kind of “table-
pages”, such as the figures 5.6 and 5.7, summarising the results of living
rooms and largest bedrooms, respectively.160

These tables with the results for particular rooms in all projects (figures 5.6
and 5.7) show some patterns. As expected, the range of activities in a room
corresponds to the amount of time spent; living rooms where much time is
spent are the living rooms where many activities take place.161 Where the
bedrooms are concerned, the apartments of the A-type (32-1, 33-1 and 41-2)
have the bedrooms (or the second largest rooms) with the widest range of use
(see figure 5.7). This is also as expected since general bedrooms (i.e.
bedrooms that may not only be used for sleeping, but are also appropriate for
various daytime activities) is a basic characteristic of the A-type of
apartments. The two following sections analyse the results more in detail.

159
A.5.2 (“side 4” in “skjema 2”) shows how the results were captured in the interviews.
160
Where rooms with much activity are concerned, patterns are most easily found by the “average diagrams”
while the “maximum / all” diagram (which lists the maximum values of the activities instead of calculating the
mean values) is most useful for rooms with fewer activities. “Average” is therefore applied when examining
the living rooms (see figure 5.6) while “all” is applied in order to examine the bedrooms (see figure 5.7).
161
Compare 33-1, 70-2 and 98-2 in figure 5.6 with table 5.7.

143
Figure 5.5.a
Bd1-32-1

144
Figure 5.5.b
Bd1-33-1

145
Figure 5.5.c
Bd1-70-2

146
Figure 5.5.d
Bd2-74-1

147
Figure 5.5.e
Bd1-76-1

148
Figure 5.5.f
Bd2-76-1

149
Figure 5.6 Activities in living rooms - “average”

150
Figure 5.7 Activities in largest “bedroom”, “all”
151
5.2.5.1 Activities and Spaciousness of Living

The spaciousness of living affects the use of the rooms and makes it difficult
to compare use of rooms where the spaciousness of living differs across the
sample of dwellings examined.162 When looking more closely at the result
from the individual projects (the figures 5.5), the widest ranges of activities
in the bedrooms seem to occur among the apartments with spacious living.163
According to the categories of spaciousness described in section 5.2.2,
“spacious living” implies that there are more bedrooms than needed as rooms
for sleeping; all bedrooms are not occupied as private rooms for the
individuals of the household. In dwellings, which in this way house few
persons compared to the number of rooms, it is therefore “bedrooms” that are
“freed”; they are available for other purposes and a variety of daytime
activities can take place in them. Also when the living is cramped, there is
examples of many activities in the “bedrooms”.164 However, as the sample of
apartments examined here contains few cramped living households, it cannot
be generalised about the latter.

5.2.5.2 Activities and Sizes of Rooms

A feature that obviously influences the activities possible in a room is the

size of the room. In order to examine particular parameters, it is useful to
look for frames of “other conditions being equal”. As the number of rooms in
a dwelling is another parameter that strongly influences the use of the rooms,
a way to focus on room-sizes in particular, is to compare apartments with the
same number of rooms.165 There are three projects where the apartments have
the size of “two rooms”; these are 32-1, 33-1 and 70-2, where the two former
are apartments of the A-type while the latter is of the C-type. Where the
activities in the individual rooms of these apartments are concerned, the
bedrooms of 70-2 are rarely in use during daytime, while the bedrooms of 32-
1 and 33-1 are rooms where many daytime activities take place.166 The
bedroom-size of 70-2 differs significantly from the size of the bedroom (or

162
This problem is reduced by the approach of studies like Holm (1955) and “Bostadsutforming
bostadsanvändning”, both these surveys focus on apartments that house families.
163
Such as in “bedroom1/Sov1” in 32-1-B1, “bedroom1/Sov1” in 33-1-C5 and “bedroom2/Sov2” in 76-1-D4,
see figures.5.5.a, b and e.
164
See “bedroom2/sov2” in 74-1-B4.
165
Even more “equal conditions” (giving even closer focus on the spatial configuration and the sizes of rooms)
would be achieved by comparing similar households living in apartments of similar size (and of course still in
similar locations). Although interesting, such studies on the sample of this survey do not make sense as the
number of apartments left to examine becomes few (or none!) as the number of fixed variables increases. For
such more detailed studies larger samples have to be selected for the particular purpose.
166
See 32-1, 33-1, 70-2 in figure 5.7. For data of each project, see A.5.5.a, b and c.

152
the second largest room) of 32-1 and 33-1, being 9.3 m2 compared with 18.2
and 15.2 m2, respectively. The size of the living room is also likely to be of
importance. The living rooms of 32-1 and 33-1 are relatively small, being
19.5 and 17.6 m2, respectively, while the living room of 70-2 has a floor area
of 29.4 m2. On the one hand, the wider range of activities in the “bedrooms”
of the A-type apartments is a necessity in the sense that the relatively small
living rooms result in a demand for other spaces for daytime living. On the
other hand, it shows how a larger bedroom has the capacity to become a
second living room that increases the potential use of the apartment.

Another example concerning the correspondences between room-sizes and

activities is project 98-2 where the second bedroom is as small as 6.8 m2.
This room has no “score” at all as concerns daytime activities or daytime
living (see table 5.3); a room this small is inappropriate for any kind of day-
time activity, very different from the large bedrooms of the A-type
apartments (32-1, 33-1 and 42-1) where a variety of daytime activities take
place. These results support the ideas of Holm and Thiberg almost 50 years
ago, where, in order to achieve dwellings appropriate for a diversity of
households and preferences, they argued for increasing the size of the
“bedrooms”.167

167
About the floor plans of Holm and Thiberg, see the section about generality in 2.4 and figure 2.22.

153
5.2.6 Changes Made by the Residents

During the lifetime of most apartments, the residents have modified the
interior spaces in one way or another.168 Such changes made after the end of
construction are of two main categories. One is changes concerning internal
connections between rooms; conditions of access are altered by making new
doors or by blocking existing ones. The other consists of more structural
changes where sizes and shapes of rooms are changed by tearing down
existing walls or by constructing new ones. There are at least two parameters
that strongly influence the number of changes made after finishing the
construction of the building; one is the needs or wishes to make changes (as
considered by the residents or by the owner); another is the possibility of
carrying out such changes. Regardless of what a household might consider
the most preferable spatial layout of their apartments, the existing building
determines the potential for making changes. In buildings where walls within
the apartments are load-bearing, which is the case in most Norwegian
apartments built before the 1970s, extensive changes of interior space are
much harder to do than in buildings where such walls are light-weight and
not load-bearing.

Whether changes made by the residents indicate positive features of the

original floor-plan layout is not obvious. A high rate of change can be
considered positive in the way that the apartment is flexible; it has the
potential to be easily altered according to various needs and preferences.
Alternatively, a high rate of change can be understood as negative because it
indicates that the original layout does not function well; the apartment has to
be changed in order to satisfy the needs of the residents. More extensive
surveys, - in terms of a larger number of apartments as well as more detailed
information about the history of each apartment, are needed in order to
analyse this in detail. However, even the limited sample of apartments
examined in this survey, gives some information about changes made after
the apartments were originally built, information about what the changes
have been and why they have been done. This section describes and
comments upon the most significant changes found by this survey. Table 5.8
summarises the results. The floor plan drawings of the particular apartments
referred to in this section are attached in appendix A.5.3.

168
Whether a change is done by the contemporary or by a previous residents is not known in this survey, as
this was not explicitly asked about in the interviews.

154
Where WCs and bathrooms are concerned, there are two significant changes.
One is that separate WC and bathroom are merged into one in the cases
where there is “through passage” in bathroom. There are 4 projects where the
selected apartments in original had separate WCs and washing-rooms/
laundries, these are 41-2, 74-1, 76-1 and 77-2. In project 41-2 as well as in
project 77-2, both with “through passage” in bathroom originally, WC and
bathroom are merged into a larger bathroom in all the apartments examined.
Where there is no through passage, the number of such merging seems to
depend on the sizes of the rooms. In 76-1, where the WC is only 1.5 m2, 2
out of 5 WCs are merged with the bathroom, while in 77-2, where the WC is
2.6 m2, none is merged. The other common change is that bathtubs have been
removed and replaced by showers. This has happened in 26 out of the 43
apartments that originally had a bathtub. This is often done in order to
achieve space for washing machine in the bathroom.169 (See table 5.8.)

Figure 5.9 (left)

Apartment 41-2-C1, an example where the access between living room and bedroom is no longer
in use.

Figure 5.10 (right)

Apartment 41-2-D2 is the exception where “breaking of rings” is concerned. Here, the closed
access to “bedroom” is not the one from the living room but the one from the entrance. This
particular “three-room-apartment” is exceptional in having as many as five rooms used for
daytime living: the living room, the kitchen, the child’s room, the mother’s room and the
“library-like” entrance that contains a 500-litre aquarium and a sofa.

169
Washing machine is installed in 8 out of the 18 apartments where bathtubs are replaced by showers and
where there was no place for washing machine in the apartment in advance.

155
Another significant change made by the residents is the “breaking of
rings”.170 Among the 32 apartments that originally had an internal ring, only
14 rings remain. (See table 5.8, showing that 18 rings are broken.) In the
projects 41-2 and 55-1, none of the apartments we visited has the internal
ring of spaces any longer. Among projects 32-1 and 33-1, two out of five
internal rings were “broken”.171 In all cases but one, the “breaking of rings” is
done by obstructing the access between “bedroom” and living room. This
common preference for breaking the ring in these A-type of apartments, and
thereby changing the spatial configuration into a “pure bush”172 means that
the “bush-like” configuration is “general” in terms of being preferred by a
wide range of households. By obstructing “unnecessary doors”, the residents
achieve more walls without openings and (at least where rings consisting of
few rooms are concerned) more floor area without through-passage. This
gives other possibilities for use and furniture, such as the dining table in the
living room and the wardrobe in the bedroom in 33-1-B1. (See A.5.3.) In
apartment 41-2-D2, which is the exception from the pattern as the closed
access is the one between “bedroom” and entrance, the change makes it
possible to use the entrance as a “library-like” hall. As explained in figure
5.10, this particular “three-room-apartment” has as many as 5 rooms used for
daily living. Where the apartments of B-type are concerned, the rings are kept
more often; they still exist in 8 out of 13 apartments. There are probably two
explanations to this difference between the A- and the B-type of apartments
concerning the breaking of rings. One is that the access between rooms in the
larger apartments becomes long and with much through passing if rings are
broken (as opposed to the A-type apartment where there is access from the
entrance to all rooms anyway). The other explanation is that the doors in the
rings of the B-type apartments are rarely into habitable rooms, which means
that blocking them does not improve the usability of habitable rooms (in
terms of more walls without doors or more floor area without through
passage).

In dwellings where there are internal rings, children without knowledge of

the spatial layout are likely to find such rings immediately; for them, rings
represent highly appreciated possibilities for endless running and “hide-and-
seek” that are impossible in “non-ring-floor-plans”. Given that children’s
play is a parameter relevant to evaluating interior space, the existence of
internal rings is an essential feature of a floor plan layout. From an architect’s
point of view, internal rings can represent positive spatial qualities as they

170
For more on “rings”, see section 4.3.
171
The size and number of rooms and the effect on the daylight conditions might explain this difference
between the cases 32-1/33-1 and the case 41-2. See 5.2.7.1 for more on this.
172
“A bush” is a spatial configuration where there is one room from which all the other rooms can be accessed.

156
make the internal movements less restricted.173 However, despite the
popularity among children and at least some architects, this survey indicates
that preferences for internal rings are more nuanced and depend on the floor
plan layout. In the A-type of apartments, where rings represent a possible
rather than a necessary access, the rings are not considered worth keeping,
while the rings in the B-type of apartments represent necessary access or a
potential for movement that the households seem to appreciate.174

Figure 5.11
Two examples of the changes made in the apartments 74-1 (left: 74-1-A2, right: 74-1-D3).

Due to the lightweight internal walls, it is not surprising that the greater
changes are most often found among the apartments of the B-type. These
changes are various, with few general tendencies. Depending on the
preferences of the households, a variety of changes have been made. A
family of four persons, such as in 74-1-D3 (see figure 5.11), has created a
third bedroom by enclosing part of the dining room. In 74-1-A2, a couple
living in the same kind of apartment, does not need as many separate rooms
and has therefore changed the floor plan into a larger master bedroom, a
studio/guest-room and a kitchen that includes a dining area. (See figure 5.11.)
The apartments in projects 76-1 show a similar variety of layouts

173
Particularly in dwellings where the living is cramped, internal rings reduce obstructions of movements.
However, if rings are large and the only possible access, such “through-passage” might be a problem.
174
This preference of “breaking” rings that include habitable rooms was also found by Holm (1955, p. 133-
134) and by Lindquist, Orrbeck and Westerberg (1980, p. 130). Holm found that 13 out of 17 households in
the two-room apartments of his sample preferred to remove the not-needed-door between living room and
bedroom. Lindquist. Orrbeck and Westerberg found that 2/3 of the households in their three-room apartments
preferred to remove such not-needed doors to bedroom.

157
corresponding to the different households. (See A.5.3.) The B-type
apartments are originally specific with respect to usability but have some
flexibility due to the large floor areas and the lightweight interior walls that
can easily be changed.

The frequencies of changes in the projects 77-2 and 83-1 are lower than in
74-1 and 76-1. This might be explained by the kind of households. The
households in 74-1 and 76-1 are of various kinds while those living in 77-2
and 83-1 are elderly and families with children, respectively. The elderly in
77-2 have spacious apartments and no immediate need to make
improvements175 while the apartments in 83-1 seem to be appropriate for
households with children without making any changes. As changes of the B-
type apartments are usually easily made, the fact that such changes are
frequent does not prove that the original layout of these apartments is worse
than the layout of apartments where changes are more rare but also more
difficult to make. However, amongst the apartments examined here, there are
some changes indicating that the original layout has not been a success; the
kitchens in 74-1 have been altered in one way or another in all the apartments
visited.176 It is not obvious which of the alternative layouts is the best; what
seems clear is that the original separation of storage room from kitchen and
kitchen from dining areas have not been considered worth keeping.177

Among the apartments in 70-2 and in the more recently built projects 98-2
and 00-1, no changes at all have been made. This could have several
explanations. Where 98-2 and 00-1 are concerned, one plausible explanation
is that these apartments have been built very recently. When you have chosen
to buy a brand-new apartment and you have paid the additional price for this,
you are less likely to tear it down and pay for alterations than when you buy
an older, well-used apartment. However, concerning the interior spaces, it is
not easy to imagine that frequent changes of these new apartments will occur
even in the future. As all existing connections are needed and possible new
ones are few, the potential for the easy kind of changes are limited. Due to
the sizes and positions of the rooms, it is also hard to imagine that many
changes by merging or dividing existing rooms will be carried out in these
newer apartments.

175
When becoming elderly, people rarely make changes of their apartment due to their own changes in
household category; the children’s rooms or their husband’s office/studio are often kept unchanged even after
the children have moved out and after the husband is dead. This corresponds to the evaluation of spaciousness
as listed in table 5.3; elderly singles or couples in large apartments rarely consider their dwellings spacious; the
rooms for the dead husband or for the children no longer living at home are not considered rooms free for other
purposes.
176
See A.5.1 for the original floor plan and A.5.3 for the altered ones.
177
However, the original layout might have worked well in the 1970s, when the kitchen was more of a
working place for the mother and where there was a mother working full time at home in most households.

158
Table 5.8 Changes made after construction
Baths Dividing Separating rooms New Merging two rooms into one

Wash.mach. in original

Comments/other changes
Replaced with shower
Apartment number
one room without by doors

- and wash.machine
Bathtub in original
Apartment type
Project number
into two closing obstruct- made

"ring" broken
connec- ing WC+Bt Others
tion doors

> Bt St Wr Bd
32-1 A A3 x x x Bd>Bd+O
A4 x x x Bd - L x
B1 x Bd - L x
C1 x
C4 x x x
33-1 A A4 x x
B1 x Bd - L x
C5 x x
D2 x x Bd - L x
D5 x x
41-2 A A4 x x x Bd - L x x
C1 x x Bd - L x x
C3 x x Bd - L x x
D1 x x Bd - L x x
D2 x x Bd - E x x
D4 x x Bd - L x x
51-3 AC C2 x x -
C5 x Bd2 - E - Bd2-L
D2 x x x -
F1 x -
F4 x x x -
55-1 AC C1 x St - E x
C2 x x x St - Bd1 x Bd1-L
D2 x x x St - Bd1 x
70-2 C A4 -
B1 -
C3 -
G2 -
74-1 B A2 x x x K-D x x St+K Bd2+D > Bd1
B4 x x x K-Wr x St+E Wr-Bt
D3 x x x D>C+Bd3 K-Wr x St+K Wr-Bt
B1x x x K-Wr x
76-1 B A1 x x
A4 x x x
B1 x x x K-Wr x x St+Bd1 Bd2+Bd3 > O
B4 x x K-L x x
D4 x x x K-L 2
77-2 B A4 x x Bd1-L x
A5 x x
A7 x x x Wr-Bd1 x Bd1+L > L
C6 x x x
83-1 B 3 x x -
4 x x -
7 x x -
8 Table 5.8x Changes
x made by the residents -
9 x x -
12 x x - 1
98-2 C A1 x -
A4 x -
B3 x -
00-1 AC A2 x -
A4 x - 1
B3 x -
C2 x -
C4 x -
E3 x -
1 : previous "cold loft" made into a second floor of the apartment
2 : still opening (and access) between kitchen and living room
159
5.2.7 Preferences about Alternative Layouts

5.2.7.1 “Through-Light” or not

The possibility for light to pass through an apartment, from one facade to the
other, improves the daylight conditions remarkably. This is a rare layout of
recently built apartments for at least two reasons. Firstly, in order to have this
possibility the apartment must stretch from one facade to the other. Until
around 1980, most Norwegian apartments had such a layout as only two or
three apartments on each floor were accessed from the same staircase. As
recently built apartments are more often placed towards one facade only,
even the theoretical possibility of through-light is lost for many apartments.
Secondly, due to the greater depth of new buildings, through-light situations
are hard to achieve even for apartments reaching from facade to façade, as
storage rooms and other enclosed rooms not necessarily requiring daylight
are placed in the middle of the apartment, obstructing the possibility of
“through-light”.

Figure 5.12
Two apartments from project 33-1, with and without “through-light”.

160
The project 33-1 is interesting where “through-light” or not is concerned.
Originally, these apartments consisted of two rooms for daytime living, - two
rooms of almost the same size, both accessible directly from the entrance. In
addition, there was direct access between the two rooms. Five apartments of
project 33-1 have been studied through interviews. In two out of these five,
the direct connection between the two rooms is now blocked. Figure 5.12
shows two apartments from 33-1, one with and one without “through-light”.
Having this door or not, influences the usability of the rooms as well as the
daylight conditions; this door makes it easy to connect the activities of the
two rooms while not having the door gives more space for furniture and
makes the rooms more independent of each other.

Table 5.9 Through-light or not, well-being in apartments 33-1

Households Floorplan Satisfaction (residents' evaluations) Daytime

Case Door L-Bd "Through- Well- Size of Floor- View Day- living
33-1 Age permanently lit" being in apartment plan in light L Bd
closed general general
A4 30 x 5 5 3 4 4 7 2
B1 30 x 3 4 4 2 3 8 1
C5 30 5.9
Table x 5 5 5 4 5 6 2
D2 30 x or not, the residents 5well-being3in project233-1. 3
Through-light 4 8 1
D5 20 x 5 3 5 4 5 5 4

Scored that are "filled" : the apartments with the highest scores

The results about well-being in project 33-1 are summarised in table 5.9.
Some information showed in this table, for instance that households of
singles are more satisfied with the size of the apartment than couples and that
the view and daylight conditions achieve higher scores in the upper floors,
could be predicted without carrying out interviews.178 However, there are
some interesting and less obvious correspondences between the floor plan
(with respect to having or not having direct connections between the “living
room” and the “bedroom”) and the residents’ well-being. Among all answers
about well being and satisfaction (see table 5.9) the highest scores are found
in the “open” apartments (A4, C5, and D5)179 while the lowest scores are
found among those that are more closed (B1 and D2)180. A similar patter is
found when looking at the amount of daytime living (table 5.3); the
bedroom181 of B1 and D2 “scores” only 1, while it scores nearly 3 on average

178
As described in section 5.1.1 the last number in the “apartment-name” refers to the floor number as counted
in Norwegian, which means that apartment 33-1-A4 is located at floor number four, where ground floor is
counted as floor number one.
179
The only exception from this is “size of apartment” in apartment D5.
180
The only exception from this is “well being in general” in apartment D2.
181
Or the second largest room, not necessarily being used as bedroom.

161
in the “open” apartments A4, C5 and D5.182 When answering the more open
questions183 the residents of C5 and D5 commented upon the spatial layout in
positive terms. The information given by the residents in addition to
answering explicit questions enforces the impression that the households in
the “open plan” feel more comfortable about their apartments than those who
have similar flats without any direct connection between the two largest
rooms. The residents of B1 and D2 (which are the apartments without the
direct connection between living room and bedroom) both complained about
floor area “spent” in the bedroom. Even though the residents of the “open”
apartments also preferred to reduce the area by taking it from the bedroom,
(when they in the interviews were “forced to” decide from which room a
hypothetic reduction of the floor area should be taken), they expressed fewer
complaints about the existing situation.

Among these apartments in project 33-1, there is a correspondence between

apartment layout and floor level in that apartments with the open layout (A4,
C4 and D5) are on the upper floors, while the others (B1 and D2) are on
lower floors. Concerning issues where floor plan as well as floor level matter,
it is therefore difficult to distinguish the importance of the two. However, by
looking at some of the other interviews, we can find more information about
this. While there is a strong correspondence between floor level and
satisfaction concerning view and daylight, there is no such correspondence
between well-being in general and floor level.184 It therefore seems that the
high scores with respect to well-being in A4, C4 and D5 are affected by the
open floor plans allowing the through-light situation rather than by the floor
level.

As pointed out when describing spatial configurations in section 2.3, a door

between two rooms is a simple physical difference that affects the spatial
configuration significantly. Whether or not two adjacent rooms are directly
accessible to each other, influences real life situations in terms of the use of
the two rooms.185 From the interviews in project 33-1, access and light across
the apartment also seem to influence the residents’ well-being.186

182
This pattern can be seen more in detail when looking at the various activities, see the bedrooms of 33-1 in
figure 5.5.b.
183
For those reading Norwegian, see A.5.2, page 2, questions 1.4 and 1.4.d.
184
See tables 5.3 and 5.9.
185
As elaborated in section 3.4.7.
186
A causality between apartment layout and evaluation of own apartment is not straight forward. It might be
that people who are more interested in design have an affinity towards open spatial layouts and that they (as a
matter of fashion, rather than of features of the apartment) are more conscious about interior space and
therefore tend to value their interior more positive than people less interested in fashion and design. Following
this logic, the well-being of the households in A4, C4 and D5 concerning spatial layout might not depend on
the features of the spatial layout as such, but might instead be a secondary effect caused by satisfaction of
feeling fashionable and updated when interior design is concerned.

162
5.2.7.2 The Storage Room

Due to building legislation, new Norwegian dwellings must have a minimum

area of storage rooms. A certain amount of this area must be in-doors but not
necessarily within the apartment.187 Even though a storage room within the
apartment is convenient, there are arguments for placing some of the storage
area elsewhere.188 One argument concerns some basic architectural features
of the interior space. The storage room is a closed room. When this “box” is
placed within the apartment, it is hard to achieve an open floor plan. Another
argument concerns economy. The building cost per floor area apartment is
higher than for underground areas of simpler standard. By placing storage
rooms underground, some of the more valuable area within the apartment is
made available for other purposes than storage. In this survey, the households
were asked explicitly about their opinion on storage rooms. There are two
categories of apartments in this respect; those with storage rooms inside the
apartment and those without. In those with such storage room inside the
apartment, we asked whether they were willing to “give away” this storage
room, if the area was made available for alternative purposes189. Those
without such a storage room were asked whether they would have liked to
have such a room. Those of the latter answering “yes” were asked if they
were willing to give the required floor area away (by taking this area from
somewhere in the rest of the apartment) or if they were willing to pay the cost
of adding the area of the storage room to the apartment. Table 5.10 lists the
answers. Among those with storage inside the apartment, 6 out of 15 190
would like to convert the storage into other kinds of room (which means that
60% prefer to keep the situation with a separate storage room inside the
apartment). Among those without a separate storage room inside the
apartment, only 5 out of 33 would like to change the situation (which means
that 85% prefer to keep the situation of not having a separate storage
room).191 Having a separate storage room or not, most people wanted the
existing situation to continue.192 However, when we compare having a
storage room inside the apartment or not, the interest in keeping the existing

187
This size and location of storage-rooms are stated in “Ren veiledning til teknisk forskrift til plan- og
bygningsloven 1997”, §10-34. Kommunal- og Regionaldepartementet, Oslo, 2003.
188
My experience from working as architect designing apartments is that housing developers and estate agents
advocate storage rooms within the apartment, arguing that this is a preference of the potential buyers.
189
In the way that the floor area of the storage room hypothetically could be added to any other room.
190
Out of the 17 apartments with such a storage room, 15 households were asked this question. (Due to
changes of questions made during the pilot-survey all 17 were not asked about this.)
191
This given that the storage room had to either be paid for as an additional area or achieved by taking floor
area from existing rooms. Not surprising most households (21 out of 33, or 66 %) would prefer such a storage
room if they were given it free, without any costs or disadvantages!
192
Whether this is because both categories have chosen their dwellings, and they both therefore tend to be
satisfied with it, or if it is due to a tendency of adapting to given situation is not known or examined here.

163
situation is significantly higher among those without such a storage room
(being 85 % (36+49) compared to 60 %); among the sample examined by this
survey, the layout without a storage room inside the apartment is most
preferred. In conclusion, this survey does therefore not support the apparently
well established belief that a storage room within the apartment is the most
preferable layout.

Table 5.10 Preferences about storage rooms

Apartments with a separate Apartments without a separate storage room

storage room inside the apartment inside the apartment
Total Being willing to remove Total Missing such If missing this, willing to
number the storage room? number storage room? change other area into
Number of
interviews

(and threreby increase storage (or to pay for

other floor areas) additional floor area)?
Case No Yes No Yes No Yes
32-1 5 5 1 4 2 2
33-1 5 5 1 4 4
41-2 0
51-3 5 5 1 4 4
55-1 3 3 1 2
70-2 4 4 2 2 2
74-1 4 2 (*) (*) 2 1 1 1
76-1 5 5 3 2
77-2 4 4 3 1 1
83-1 6 1 1 5 2 3 1 2
98-2 3 3 1 2 1 1
00-1 6 6 4 2
Total 0 17 9 6 33 12 21 16 5
Table 5.10
Preferring Preferences
change about storage rooms.
40,0 15,2
Preferring existing 60,0 36,4 48,5
(*) : Answers missing as this was not asked about in all cases of the pilot-survey.

5.2.7.3 Access by Gallery

The principle of access to the apartments defines the premises for the layout
of the building as well as for the floor plans of each apartment. Due to the
fact that indoor rooms are more expensive to build than outdoor galleries, the
principle of access by galleries is efficient in terms of building costs and is
therefore a layout often argued for by constructors and estate developers.
From the point of view of architects, the evaluation of access by galleries is
dual; on the one hand, galleries have been argued for by architects and have
been applied in highly esteemed projects,193 on the other hand, external
galleries is the principle often applied when building as cheaply as possible.

193
Gifu Kitagata Apartments by Sejima & Associates (GA Architect: 18, 2006) and Schwitter-building by
Herzog & de Meuron (GG (1989) Herzog & de Meuron. Barcelona) are examples of such projects.

164
Two projects of this survey, 83-1 and 00-1, have access by galleries. In
project 83-1 the households were explicitly asked for their opinion about the
galleries. The opinions were very positive; there were no explicit negative
remarks. Several considered the gallery a pleasant second balcony and a
place for meeting neighbours. However, it cannot be denied that galleries
have a disadvantage due to neighbours passing by the windows; two out of
six households answered that the gallery was fine as they were living at the
end (at the end of the gallery, - where neighbours do not pass). In the project
00-1, we did not explicitly ask about the galleries in the interviews. However,
in two out of the six apartments where we did interviews, the residents
commented on the galleries when they answered open questions, one was
very positive while the other was negative. The fact that the galleries in these
two projects look out over park-like backyards and not towards dirty, noisy
roads is likely to influence the evaluation of them. In conclusion, the answers
of the interviews do not prove that galleries are the best principle of access,
but indicate that access by galleries works well and may be highly
appreciated if the galleries are well designed and look towards pleasant
views.194

Figure 5.13.
Casinetto (project 83-1 in this survey), a housing project were the residents appreciate the access
by galleries.

194
Due to the disadvantage of passing by windows, too many dwellings should not be accessed from each
staircase. In the project 83-1, where the residents appreciate the access by galleries, there are rarely more than
two apartments between stairs and any apartment accessed from these stairs.

165
5.2.8 Some Additional Findings

This survey has captured information about many issues beside what is
described in previous pages. This section summarises some of these.

5.2.8.1 Time of Construction

Several results of this survey show how the time of construction might
represent a bias with respect to the kind of household living in a particular
housing project. The case 77-2, described in section 5.2.3.2, illustrates how
the time of construction might affect the contribution of households living in
a particular building for a long time. Now, 30 years after construction, many
of the households living in this project are elderly people that have lived
there since the building was new nearly 30 years ago. If the first residents that
move into a new housing project are of similar category and age, which they
might well be, the times of moving out are not randomly distributed through
time. For housing projects where people are well satisfied, the time of
construction might determine whom that are living there for many decades.

By looking more closely at the project 00-1 we can probably see another
effect from time of construction upon kinds of households. 00-1 is the project
with the highest rate of “several-adults-households”. (See table 5.5.a.) The
answers from the interviews indicate that this might be a consequence of the
time of construction. Among the six 00-1 apartments where we interviewed
the households, four were the first apartment owned by the respective
households. Two out of these four households were singles renting out one
bedroom for an extra income; both described the rental situation as
temporary. By Norwegian tax-legislation, such rental is a favourable way of
financing the purchase of a “self-owned” apartment. As times goes by and
some loan is paid off, this rental will no longer be necessary for managing the
costs. As project 00-1 was built in 2001, all households have bought their
apartments recently. As time goes by, the rate of households that have
recently moved in will decrease and the number of apartments with a room to
let, and thereby the number of “several-adults-households”, can be expected
to decrease. However, due to the floor plan that allows for a room to let, as
described in section 5.2.3, 00-1 will probably still be among the housing
projects with more that average number of “several-adult-households” .

166
5.2.8.2 Disadvantage of Popularity

The reasons for people’s moving are diverse, as for example growing up and
moving from the childhood home, having children and needing more rooms
or being forced to move because of a new job. Only 2 out of the total 55
households that have been interviewed, explicitly described dissatisfaction
with their previous homes as the reason for moving; both lived previously in
the part of Oslo named Grünerløkka and have moved because of noise from
streets and restaurants at night time. Grünerløkka is a former East-side-
working-class-area that has become popular, fashionable and relatively
expensive during the last 10 years. The increasing prices of apartments and
the fact that many new bars and cafés are doing well, do not imply that the
area must be a pleasant place to live. This effect is worth bearing in mind
before classifying “popular” parts of town as entirely successful.

5.2.8.3 Annoying New Technology

Technical standard has not been a focus in this survey, but it is a subject
complained about by several households when answering the open questions
of the interviews. What is remarkable is that the strongest complaints are not
from people living in older and well-used apartments. The issue, on which
residents most often complain, is the ventilation-system in apartments built
during the last 20 years. The problem is the more or less advanced ventilation
systems; some cannot be regulated in each apartment, some make a
remarkably loud noise and some simply do not work. The most annoying
ones have all these features; they do not work as expected, they are noisy and
they cannot be turned off. This is interesting as advanced technical systems
are an important field of both engineering and business and a feature that has
been emphasised in the marketing of new dwellings during the recent couple
of decades.

5.2.8.4 The Sizes of New Apartments

The fact that most questions in the interviews concerned the particular
apartment where we did the interview did not prevent people from
commenting upon housing projects in their neighbourhood or in general.
Some such comments shed light on some consequences of the fact that many
small apartments are being built now. As small apartments now are the most
profitable ones to build and the regulations by government and local

167
authorities are weak, new apartments are small despite intentions of planners
and politicians to build apartments for families as well. In areas of Oslo
where most existing apartments are small, the small size of new apartments
causes some problems worth commenting upon. In project 41-2, the
households are very satisfied with their part of town, with their close
neighbourhood, with their own building and with their apartments. What
several households with children here worry about, is the size of the
apartments recently built and of those planned in the neighbourhood. The fact
that these new housing projects contain mostly small apartments and rarely
any with more than two bedrooms, leads to a high rate of moving; in order to
find apartments of appropriate size, households with children must move to
other parts of town. For those who prefer to stay, this destabilizes the social
lives of children as well as adults. Parents were seriously worried about this
on behalf of their children. When children experience their best friends
moving to other parts of the town several times (due to lack of larger
apartments in the area), the size of new apartments has a negative influence
on their socialisation. This illustrates how the size of new apartments is of
importance for others than those moving into them. Previously, until the
1980s, the local authorities were in charge of and responsible for the planning
of housing. Today, the construction of housing is more usually organized in
individual housing projects managed by private companies that have little
responsibility for the influence their projects might have upon the existing
city. A challenge within the current political and economic framework for
planning and building of housing, is how to advocate the interests of all those
who are not involved in these privately driven building projects but whose
lives are strongly affected by them.

168
5.2.9 Summary of the Synchronic Survey

The typology of apartments distinguished in chapter 4, was the result of a

diachronic study of floor plan drawings, not a result of visiting the sites or of
any other specific knowledge about living in the dwellings. The survey of
contemporary living described in this chapter, constitutes an empirical basis
shedding light on the typology of apartments, a typology consisting of A-, B-
and C-types of floor-plans.195 As the spatial layout is the main subject of this
study, this typology of apartments is a frame applied for discussing the
results. The survey has focused on three issues; one is the kinds of household
preferring the apartment, the second is the activities or the use of the different
rooms of the apartments, and the third is changes made by the residents.
What characterises this survey and distinguishes it from many similar studies
of households and their living conditions,196 is that the particular parameters
examined (such as time spent, activities in rooms and households
evaluations) are not the main subjects of interest but means for comparing
apartments of different spatial layouts.

Concerning the projects of the A- and of the C-type, these were selected in
order to be comparable, trying to achieve some kind of “other conditions
being equal”. Among these projects, the variety of households (variety of
households living in apartments of identical layouts) corresponds to the
generality of the apartment as described by the floor-plan-typology.
Apartments of the A-type (which should be the most general type), house the
widest range of households, while few others than singles and couples choose
the less general apartments of the C-type. The distributions of households in
the apartments of the AC-type are in between what is the case for the A- and
the C-type apartments.197 Concerning the apartments of the B-type, which are
the four-room apartments from around the 1970s, the projects were selected
in order to represent similar spatial layouts in various settings with respect to
outdoor-areas, location in the city and time of construction. These apartments
function as intended in the sense that they house a large number of families
with children. The apparent exception is project 77-2, where a majority of the
households are now elderly singles or couples. Contrary to what we might
suppose, this shows how apartment 77-2 is appropriate for many kinds of
households; those who moved into these apartments as young adults in 1977

195
Where the A-type is the general apartments common until the 1950s, the B-type is the larger apartments
built in the 1970s and -80s, while the C-type is the smaller apartments without separate kitchens that are
common now. See section 4.3.3 about this typology in detail.
196
Such as Holm (1955), Dahlgren and Westerberg (1979) and Guttu, Iversen and Nørve (1985), which have
focused on particular kinds or households or particular kinds of apartments, see footnote 61.
197
See figure 5.3.c.

169
have been living there throughout their lives. Without moving to other
dwellings, they have passed most categories of households, such as being a
couple, having children, bringing up teenagers, having grown-up children
living at home and now living there as elderly singles or couples.

When it comes to floor plan layouts versus dwellings in use, many other
parameters intertwine with these in ways that make it hard to draw
conclusions in terms of simplified statements. A parameter that particularly
influences the way households use their dwelling, is the spaciousness of
living. It is therefore hard to examine effects of the floor-plan-layouts by
comparing apartments that differ with respect to spaciousness of living.
However, by studying the results in detail, it is possible to reveal some
patterns in the effects of “spaciousness of living” and thereby manage to
point out some general correspondence between principle floor-plan-layouts
and well-being, preferences and domestic activities. If we summarise the
results from comparing apartment-sizes (in terms of floor areas as well as
numbers of rooms) to spaciousness of living, the households consider
apartments of the general A-type more spacious than similarly sized
apartments of the C-type. When it comes to the activities taking place in the
individual rooms of the apartments, there are correspondences with the
typology of apartments in the way that apartments of the A-type have a larger
range of activities in “bedrooms”/”second-largest-rooms” than apartments of
the C-type. In the latter, the rooms are more specific with respect to functions
or use and there are few places for daytime activities other than the living
rooms. Concerning changes of the apartment-layouts carried out by residents,
the changes in A-type apartments are of a simple kind, usually made by
obstructing a door, changes in B-type apartments are more structural ones
made by removing walls or constructing new ones, while changes in C-type
apartments are rare so far.

In conclusion, the types of apartments defined by the diachronic analysis are

not just a theoretical classification of different floor plan layouts but also a
typology that captures features of spatial layout that correspond to real
domestic lives. The result that kinds of households and use of rooms
correspond to basic features of the spatial layout is interesting in light of the
fact that very few households explicitly mention the spatial layout as being
important when they were asked in interviews about their reasons for
choosing their dwelling. This corresponds to the results of Blombergsson and
Wiklander (2006), showing that people behave in accordance with
configurational features of space without necessarily being aware of them.
(See section 2.3.) If applying Hillier’s terms this is “the non-discursive social
logic of space”. (Hillier, 1984 and 1996.)

170
6 The Contemporary Dwellings

The typology identified by the study of OBOS and described in section 4.3.3
represents a historical development of dwellings and captures aspects of
domestic space relevant for peoples’ choices of apartments and for their daily
living. This chapter 6 focuses on the recent tendencies of apartment layouts
found by the study of OBOS and discusses whether these are representative
for contemporary dwellings.

6.1 BACKGROUND; OBOS THROUGH 75 YEARS,

THE THREE GENERATIONS OF APARTMENTS

By analysing OBOS’ apartments built since 1930, this study has identified
patterns in the development of apartment layouts through time, patterns
described by three generations or three types of apartments according to the
apartments’ degree of generality concerning functions or use. The first, the
A-type, can be described as general due to sizes of the individual rooms as
well as the spatial layout. The second, the B-type, is highly specific both in
the individual rooms and in the spatial configuration, while the third, the C-
type, is by spatial configuration less specific than the second but due to very
specific individual rooms; it is still highly specific with respect to functions
or use. This history of apartments not only describes a development of
layouts, it also gives insight into the contemporary situation as concerns
Norwegian housing in that it points out typical features of a large segment of
the existing stock of dwellings. The historical development described by the
three generations of apartments thereby represents a basis from which
contemporary and future apartment design can be discussed. The following
sections will more closely examine what is going on at present by comparing
the recent changes found in the diachronic study of OBOS to a brief analysis
of other apartments recently built.

171
6.2 A COMPARATIVE CASE

This thesis has so far described a single-case-study examining dwellings of

OBOS. Even though OBOS is a case that is likely to shed some light on the
development of Norwegian urban housing more in general,198 it is not
obvious to what extent the patterns found can be generalised.
Methodologically this can be handled in several ways where the simplest is
not to suggest any generalisations outside the particular case of OBOS.
Alternatively, the general validity of the results has to by examined more
closely. Concerning the validity and relevance of a theory, Glaser and Strauss
have distinguished between what they have called “substantial” and “formal”
theory. A substantial theory is applicable in limited empirical issues,
sometimes only in the case from which it is developed, while a formal theory
is conceptual and has a general relevance. They have emphasised the study of
comparative cases as the strategy for developing a substantial theory into a
formal one. (Glaser and Strauss, 1967, p. 31-35.) For this particular study,
such comparative cases might be apartments of co-operatives or companies
other than OBOS, apartments located in other places or dwellings of other
kinds.199 As the decision to examine a comparative case was taken at a late
stage of this study, there was no time to do comparative studies similar to the
diachronic study of the OBOS-apartments. Instead, a minor comparative
study has been carried out, a study not at all comparable to the OBOS-survey
in terms of strategy of sampling or in terms of number of apartments. Even a
small comparative survey, particularly if independent from the main survey
in all respects, might give information relevant for discussing the validity of
making generalisations from the main survey.

Therefore, before assuming that the patterns captured by the apartment

typology described in section 4.3.3 are of general validity, we shall take a
look at a minor comparative case. A Saturday in October 2005, the Architects
Association of Oslo (OAF) arranged an excursion to some recently built
housing projects. The apartments visited on this excursion were chosen as a
case comparative to the main diachronic study of the OBOS-apartments. In
order to be distinguished from the OBOS-apartments, the apartments visited
on the OAF-excursion are hereafter termed the “OAF-apartments”. Due to
the great difference in the sample (being few; consisting of only 7 housing

198
As argued in section 3.3.
199
This way of collecting data is by Yin (who further refers to Patton (1987, p. 60)) termed data triangulation
(one out of four kinds of triangulation where the other are investigator-, theory- and methodological
triangulations). Data triangulation “— encourage you to collect information from multiple sources but aimed
at corroborating the same fact of phenomenon.” (Yin, 1994, p. 92)

172
projects) as well as the sampling (it is not known to what extent the selected
OAF-apartments are representative for contemporary apartments), the OAF-
excursion does not represent a case directly comparable to the historical and
much wider study of OBOS. However, a brief analysis of the OAF-
apartments sheds some light on the layout of contemporary apartments other
than those built by OBOS. To what extent the results from examining these
OAF-apartments correspond to the recent tendencies found by examining the
OBOS-apartments, indicates the validity of drawing general conclusions
from the OBOS study.

Figure 6.1 shows the floor plan drawings of 6 apartments visited on the OAF-
excursion.200 For projects where the excursion visited several apartments, the
one most typical for the respective project has been selected. The apartments
are labelled from OAF-1 to OAF-6. Figure 6.2 shows the connectivity graphs
of the apartments.

The OAF-apartments have some features in common. Compared to the

OBOS-apartments, the OAF-apartments contain fewer enclosed rooms
connected by doors. This implies that space syntax modelling by nodes,
which is the kind of modelling applied when drawing the connectivity
graphs,201 is less descriptive for the OAF-apartments than for the OBOS-
apartments in the main study. OAF-3 and OAF-5 are apartments where it can
be discussed whether the narrow space in the middle of the floor plan (by the
bathroom) should be modelled as a separate space (as they are now, in the
way that they are represented by a node named C for corridor), see the graphs
in figure 6.2.202 The alternative would be to consider these parts of the
apartments as extensions of other spatial units of the apartments. If following
the latter logic strictly, these two apartments could be modelled as one room
and a bathroom. Even though it can be discussed which alternative would be
the best modelling of these two apartments, it should be possible to point out
some relevant information from the graphs in figure 6.2.203

200
One out of the 7 projects visited by the OAF excursion is not analysed as the floor plan was neither sent
from the architect nor available on the web. The total number of projects is therefore 6.
201
About kinds of space syntax analysis, see section 3.4.8.
202
This modelling is partly a “convex space modelling”, see section 3.4.8, figure 3.18.
203
How the total space of an apartment is divided into spatial sub-units that are considered separate rooms,
does of course also influence the comparison of sizes of rooms.

173
 OAF- 2

OAF- 1 (a two-floor apartment)

OAF- 3 OAF- 4

Figure 6.1
The “OAF-apartments”
1 : 200 OAF- 5 OAF- 6

174
Figure 6.2
The “OAF-apartments”, space syntax connectivity graphs.
175
A feature that the OAF-apartments have in common is that there are no
separate kitchens; all kitchens are included in the living rooms. As regards
the spatial configurations, these are presented in terms of connectivity graphs
in figure 6.2. In four out of the six OAF-apartments, in all except OAF-1 and
OAF-4, the entrance and the habitable rooms are positioned in a linear rather
than in a bush-like configuration.204 In two of the apartments (OAF-3 and
OAF-5) a “spatial extension” by the entrance is most likely to be the part of
the dwelling used as “bedroom”; if these apartments shall have their beds
elsewhere than in the living room, the “bedroom” must be passed through
before accessing the “living-and-kitchen area”.205 The apartments OAF-1 and
OAF-4 differ from the other four in that their main bedrooms have direct
access from the entrance.

Table 6.1 lists the floor areas of the OAF-apartments while table 6.2 lists the
average of the OAF-apartments and the averages of the OBOS-apartments for
comparison. By a look at the floor areas of each apartment and the floor plan
drawings, OAF-1 differ from the others in terms of being somewhat more
generous where amount of apace is concerned. This apartment has a large
bedroom and a “living-and-kitchen-room” that has the proportions needed in
order to be appropriate for two separate groups of furniture. The apartment
with the largest “living-and-kitchen-room”, which is apartment OAF-4, is
different in this respect; due to the layout of the kitchen, it is difficult if not
impossible to place conventional furniture.206

Table 6.1 OAF-apartments, floor areas

Floor Areas (m2)

Project Total L+K Bt Bd1 Bd2
OAF-1 H16 54,4 27,4 4,7 12,5
OAF-2 Grüner Garden 30,0 16,9 3,2 5,9
OAF-3 CK30 42,6 23,7 3,6 7,0
OAF-4 Table 6.1 Hage
Waldemars 64,0 31,2 5,4 10,0 6,8
Floor areas of the OAF-apartments and their rooms.
OAF-5 Waldemar Thranes gate 3 32,5 16,3 2,4 6,7
OAF-6 Sporveisgata 35 26,1 13,0 2,2 5,2
Average 39,0 20,2 3,4 7,0 6,8

204
The exceptions are OAF-1 and OAF-4.
205
Such rooms where through-passage is required are B-type of space according to Hillier’s room types.
Among the OBOS-apartments examined in this study, all bedrooms are separate rooms of the A- or the C- type
according to Hillier’s room types. This means that the bedroom is a “dead-end-space” or on a ring, implying
that there is no through-passage (the A-type of space) or that through-passage can be avoided (the C-type of
space). (See 3.4.7 for more about Hillier’s typology of spaces. This typology of spaces should not be mixed
with the apartment typology identified and applied in this thesis.)
206
The floor plan drawing is misleading as the sofas are 1.4 meter long and placed without considering
common furniture such as television and audio-equipment.

176
6.3 THE NEW LAYOUT

By comparing the results of the diachronic study of OBOS to the results of

the minor contemporary “OAF-case” just described, this section points out
what seem to be recent changes and the common layout just now.

Figure 6.3 illustrates the development of room sizes of the OBOS apartments
during the period 1930-2005 and the average room sizes of the recently built
apartment visited by the OAF-excursion, while table 6.2 lists the average
floor areas by numbers. According to figure 6.3, the room sizes of the OAF-
apartments seem to “accelerate” the recent tendencies found by the OBOS-
study. While the total floor area of the OBOS apartments has decreased
towards becoming 70.1 m2 on average for the last 5 years (see table 6.2), the
average of the 6 brand new apartments of the OAF-excursion is as small as
41.6 m2. Where the sizes of the individual rooms are concerned, the areas of
the living-kitchen-room, the main bedroom and the bathroom are respectively
33.1, 11.4 and 4.7 m2 for the OBOS-apartments built since 2000 while they
are 21.4, 7.9 and 3.6 m2 for the OAF-apartments. A significant difference
between the OAF- and the OBOS-apartments relates to balconies; among all
the OBOS-apartments examined, there are only 2 built after the Second
World War that do not have a balcony207 while 3 out of the 6 OAF-
apartments are without balconies. The apartment-type with only one room
and a bathroom, such as OAF-3 and OAF-5, is not found among the OBOS-
apartments.
Table 6.2 Floor areas, OBOS- and OAF-apartments

Total L K L+K+D Bd1 Bd2 Bd3 WC+ St Bc

Wash.+
Bath
OBOS 1930-1940 60,8 21,8 8,5 30,3 14,7 2,8 1,9 3,6
1941-1950 66,1 20,4 8,7 28,5 13,0 9,8 3,5 1,8 3,5
1951-1960 63,7 20,9 9,2 30,6 12,7 9,4 3,3 1,7 2,7
1961-1970 70,3 22,7 10,4 33,7 11,8 9,3 7,3 4,4 3,0 4,9
Table 6.2
1971-1980 91,1 25,2 12,4 39,8 12,1 9,7 7,9 7,7 3,9 12,5
Floor areas of the OBOS- and the OAF-apartments.
1981-1990 86,3 26,6 12,2 38,8 12,6 10,0 8,1 3,6 10,8
1991-2000 74,9 24,3 11,3 35,6 11,2 9,0 6,3 4,7 5,8
2001-2001 70,1 28,5 11,8 33,1 11,4 8,9 4,7 2,7 9,0
OAF 2005 41,6 21,4 - 21,4 7,9 6,8 7,0 3,6 - -

207
These two OBOS-apartments are 00-2-1, which has a “French-balcony”, and 67-3-2, which has no balcony.
The project 67-3 consists of several kinds of apartments, where all the large ones, such as 67-3-1, have
balconies while the 1- and 2-rooms (such as 67-3-2) have not.

177
 Figure 6.3 Floor are as
O BO S(1930-2005) and "O AF-apartme nts"(2005)
(OAF: large marks)
110
Living+kitchen+dining
105
Largest B edro o m
100
B ath+Wc+Wash

95 To tal

90

85

80

75

70

65

60

55

50

45

40

35

30

25

20

15

10

0
Chronography (1930 - 2005)

Figure 6.3 Floor areas of the OBOS- and of the OAF-apartments.

178
If the apartments of the OAF excursion are representative for the urban
dwellings now being built, then the features of only one room for general
living and tiny bathrooms and bedrooms are much more significant than
found by the study of OBOS. The apartments in the OAF-sample, might
represent a new generation of Norwegian dwellings, a new generation in
terms of the apartments’ layouts (such as consisting of only one room in
addition to the bathroom) and in terms of layout of the entire buildings (such
as apartments facing north or east only and apartments without balconies).
These kinds of layouts have previously been rare.208

Figure 6.4 shows a sample of new apartments presented by Vegard Ramstad

and Kristian Ribe. (Ramstad and Ribe, 2006) According to Ramstad and
Ribe, these apartments are typical for new urban dwellings in Oslo. Where
their average floor areas are concerned, the total is 40,0 m2, the
living/kitchen is 21,9, the bedroom 9,3 and the bathroom 3,9m2. They have
almost identical spatial layout and are of the C-type if characterised by the
typology of section 4.3.3, with the same spatial configuration as apartment
04-2-1 in this study. (See A.4.2.)

When the OAF-apartments as well as the sample of Ramstad and Riebe are
taken into consideration, it can be imagined that the C-type of apartment
described in section 4.3.3 is only a predecessor of the “real-C-type”, which
might be characterised as below.

x The apartments as well as the individual rooms is tiny.209

x There is one or no separate bedroom.
x There is no separate kitchen; the kitchen is in the living room.
x There is not necessarily a balcony.
x The apartment has only one façade, which might face any direction.

208
Small apartments are not new inventions; they are a kind of apartment that have been built previously and
that (after evaluation) have been considered unacceptable for permanent dwelling. In Norway the critical
report by Guldbrandsen (1973) about small apartments at Ammerud in Oslo, a report that pointed out problems
of living within such limited space, has been a foundation for this opinion. A similar survey of living in the
small apartments now built should be carried out. These apartments will be likely to achieve a low score
concerning the apartments as such, but might in conclusion be evaluated as better than the Ammerud-
apartments due to their location closer to the city centre and urban attractions.
209
The floor areas of the “OAF-apartments” are 42 m2 on average. There are new apartments that are much
smaller. In “Living in a box”, a housing project that is among the “branded” ones commented upon in section
7.3 ,the floor area of the dwelling unit is 15 m2.

179
Figure 6.4
Examples of the apartment layout that now is typical. (Ramstad and Ribe, 2006).

180
The small apartments that seem to be common now, is a subject that can be
discussed in many contexts. What will be commented briefly here, is the
relation between sizes of the rooms and the principle floor plans.

In theory as well as in practice there are some limits to the minimum sizes of
the individual rooms of a dwelling. According to Norwegian legislation, a
separate room for sleeping or living must have a volume of no less than 15
m3 and a window allowing a certain amount of daylight into the room.
Regardless of such legislation, a room must have a certain amount of space
simply due to the intended use of a room; a bedroom should at least have the
space needed for a bed and a combined bathroom/WC should be large
enough to contain at least a shower, a basin and a WC. The conclusion from
this study is that the apartments as well as the individual rooms have
continuously become smaller since the 1980s. (See figure 6.3.)210 When the
individual rooms have become as small as possible with respect to legislation
as well as to usability, it is no longer possible to reduce the apartment-sizes
by means of the size of any individual room. A continuing of the
development toward smaller apartments is therefore achieved by another
strategy. This strategy is the “open-plan” layout. Instead of dividing the space
of the dwelling into rooms individually designed for specific activities such
as cooking, dining, relaxing and sleeping, the entire space of the dwelling
unit is kept undivided as far as possible. The only functions, for which a
separate room is made, are those of the toilet and bath. While the last
generation of apartments identified by the OBOS-study consisted of several
rooms but only one room appropriate for various domestic activities, the new
type of apartment occurring in the OAF-sample has only one room in
addition to the bathroom.211 The future will tell whether the two layouts
common among new dwellings, the one in figure 6.4 (similar to OAF-2 and -
6) and the one consisting of only one room and a bathroom (OAF-3 and -5),
are peculiarities of the years just after 2000 or if they represent a new
generation of flats in a historic perspective. If the latter is the case, then their
typical time-period has just started.

210
About the development of apartment sizes and room-sizes more in detail, see section 4.3.1.
211
Two out of six apartments (OAF-2 and OAF-5) consist of only one room in addition to the bathroom.

181
7 Final Reflections

On the background of the empirical surveys described in the previous

chapters, this chapter reflects upon contemporary housing and its contexts.

7.1 THE HOUSEHOLDS IN THE NEW DWELLINGS

The recent tendency in apartment design is remarkable in that the criteria for
design seem to be other than concern for daily domestic life. The floor plans
with bedrooms accessed from the living room and no more than one place for
daily living differ from the earlier modernist tradition of housing design. The
sizes of the apartments as well as of the individual rooms have decreased
rapidly since the late 1990s and are now far below the recommendations from
the Scandinavian housing research institutions. This discrepancy between the
traditional architectural guidelines for housing design and the layouts of new
dwellings is not a theoretical mismatch between theory and practice, but a
simple fact that has some consequences for who that might live in these new
dwellings, - or at least for who that might live well in them. To-days’ typical
apartments with only one room for general daytime living are well suited for
some categories of households. These households are young couples or
singles (a category that has been increasing due to a later start of family life)
and older couples (whose children no longer live at home) who prefer to sell
their house and garden. These households can be termed “uniform” in the
meaning that their daily living can go on in one room without necessarily
causing conflicts. For other households, households where daily lives consist
in a wide range of “not easily co-existing activities”, the new apartments are
inappropriate. Such “less uniform” households are of many kinds. One is all
those who by cultural background prefer separate men and women’s spaces,
another are the very common but not traditional families (such as couples
with small children where one of the parents also has teenagers living at
home) and a third is the different kinds of “several-adults-households” (such

182
as students or others sharing one apartment). In the current state of
Norwegian economic and demographic conditions, there are enough wealthy
people among the households above termed “uniform” as to constitute the
entire population of customers purchasing new apartments. As the profit from
building and selling four or five apartments of 20 m2 is now far above what
can be achieved by selling one apartment of 100 m2, most new apartments
are tiny. These “uniform” households do not only influence the size of new
apartments becoming small, they also influence the layout of the larger
apartments. The pattern, as apartments become larger, seems to be that the
“living-and-kitchen-room” becomes larger and that tiny bedrooms are added.
Large bedrooms or other rooms for living than the “living-kitchen-dining-
room” seem to be unusual even in large apartments; regardless of size, new
apartments tend to have only one room for daytime living. 212 This implies
that the new larger apartments are not for other kinds of households than the
new small apartments, they are for the more wealthy ones among the same
kind of households.

7.2 THE EXTENT OF THE RECENT CHANGES

New dwelling layouts are not a change forced upon society by “bad guys”
among constructors and estate agents; it is as much a rational response to
preferences or requests of the population. What here is pointed out so far is
that these requests or preferences are those of some kinds of households and
that other households (without being noticed in the liberal market) might
have other and even contradictory preferences.213 The tiny apartments now
common, are appropriate for limited kinds of households and differ from
what architects until very recently considered good housing design. It might
be argued that this new layout, regardless of its quality, does not represent
any housing problem, as the dwellings of this new kind are still few
compared to the total stock of Norwegian dwellings. Unfortunately, the
situation is not this simple, as the change of layouts going on in new built
apartments goes on in the existing stock of dwellings as well. Such changes
of existing dwellings are carried out by numerous individual residents as well
as by professional estate-developers. By buying second-hand apartments,
modifying them according to what considered most profitable and then
selling them, professionals contribute to the same kind of changes in existing

212
This description of new larger apartments is based on looking at advertisements and other presentations, not
on any case studies or statistical inquiries.
213
To what extent requests or preferences are guided by basic “needs” or by more luxury “wants” (as
distinguished by Rapoport (1990)) is not a subject followed here.

183
dwelling as in new apartments. Regardless of the exact rate of such changes
made to existing dwellings, the changes of apartment layouts pointed out in
this study is a phenomenon influencing our total stock of dwellings much
more than by the new built dwellings only.

7.3 MARKETING; THE CONCEPT OF BRANDING

A new tendency as concerns Norwegian dwellings is the intensive marketing

of specific housing projects as brands.214 Figure 7.1 shows an advertisement
of this kind. Different from the more traditional estate-advertisements, which
emphasise inherent properties of the dwelling such as floor areas, number of
rooms, views, materials and technical standard, these new and “branded”
projects are marketed by fashion-like images that characterise and identify
the brand and a lifestyle rather than the dwelling.

Figure 7.1 “Branding” of a housing project, an example.

214
Examples of such projects from Oslo are “Grüner Garden” and “Living in Box”.

184
The concept of branding is well known in the fields of fashion and design.
Branding is a way of distinguishing between otherwise similar products and a
way of increasing profit without adding quality to the product itself.215
Within clothing and accessories there is often little correspondence between
technical quality or cost of production and the retail price of a product; the
price is determined by the fashion of the brand rather than by inherent
properties of the item. In such cases, the price paid for an exclusive product is
not primarily for the item as such but for the image of the brand. In other
kinds of products such as cars, sporting goods and audio/video-equipment,
the identities of the brands relate more to features inherent to the product,
features such as performance and reliability. Whether a brand represents “real
qualities” of the product or not, the brand is a feature connected to the
product. What is worth some attention when it comes to “branding” of
specific housing projects is the permanence of the brand. Regardless of
whether the brand Mercedes-Benz represents explicit qualities in the
Mercedes cars, Mercedes-Benz is likely to be an expensive brand of car also
in the future. The added cost and value represented by the brand Mercedes is
therefore a permanent feature of a Mercedes; the additional cost of buying a
Mercedes represents an additional value of the car, a value that more or less
can be realised in the future when selling the car.

The new and branded individual housing projects are very different from the
Mercedes in that the value added by the brand cannot be expected a similar
permanence. Contrary to what is the case for famous and presumably long-
lasting multinational brands, the marketing of a particular housing project
lasts only for a limited period of time; the marketing and thereby the
maintenance of the brand is completed as soon as the apartments are sold for
the first time. As the most speculative “housing-brands” scarcely represent
any permanently attractive features inherent to the dwellings, there are few
reasons to expect that the additional price achieved by the branding will be a
value that endures with the apartment. The value of the brand thereby
represents an income for the seller of the apartment when it is new, but not
any permanent value that can be realised by the residents when they sell their
apartment sometime in the future. Such lack of permanence of a brand is not
a problem when clothes and accessories are concerned as these objects have
limited costs and are not expected to last long. A dwelling is very different,
both because of the amount of money involved and because of the longevity
of the item.

215
As described by Naomi Klein (2000), “branding” is a substantial part of contemporary marketing.

185
Brands that represent permanent and attractive features inherent to a product
are less dependent on marketing. In such cases, the product might keep the
value represented by the brand even after the marketing has ended. However,
this is not likely to be the case of the “hyped” new housing projects in Oslo.
The situation is almost the opposite in that fashion-like branding is a
conscious strategy applied in order to take attention away from the low
quality of permanent features inherent to the apartments such as daylight
conditions, room sizes and different aspects of usability, - features that in the
modern tradition of housing design have been considered highly important.
The project “Grüner Garden” has been marketed explicitly as “not for your
mother”; the feature of not being appreciated by adult households (due to
being extremely narrow) is here turned into a positive rebel-like attraction.
This kind of apartments has the character of being temporary dwellings
except for the eccentric (in terms of not bothering about material goods);
living there requires access to the homes and houses of parents or others in
order to store common basic belongings. These apartments are not for young
people in terms of being tailored particularly for them; they are for young
people in the sense that they are unacceptable for most other households.

7.4 SOME SCENARIOS

From the contemporary situation where new apartments are appropriate for
limited kinds of households only and where the number of such households is
so great as to constitute almost the entire market for new urban dwellings,
several scenarios can be imagined. One question is whether to-days’ common
apartment layouts are likely to be continued. At first glance the answer is no.
Unless the number of households for which the contemporary layout is
appropriate increases rapidly, a time will soon come when there will be more
of these apartments than there are households preferring them. Through basic
effects of supply and demand, the layout of apartments would then
presumably change towards some that correspond to the needs of other
households as well. However, this logic can be questioned. Given that the
Norwegian economic wealth continues, the future will continuously
“produce” new wealthy households of the same kind as those who now buy
most of the new apartments. Given that apartments also in the future will be
designed in accordance with continuously updated fashion-preferences of this
kind of households, it is hard to see why wealthy singles in the year 2015
should buy a worn out, outdated apartment from 2006 if they are offered a
new one corresponding to the trends in 2015. Within a free housing market, it
is therefore possible to imagine that most new apartments in the future might
be designed for the same kinds of households as they are today.

186
One argument against the mechanisms of the liberal market continuing as
described above concerns the second-hand value of the apartments. Among
the households constituting the market for second-hand apartments, many are
of the kinds preferring more than one room for daytime living. This
represents a limit to the number of potential buyers when the small new
apartments are offered for sale as second-hand apartments sometime in the
future. Within a free market, it is hard to imagine that this will not influence
the price. If this assumption about second hand prices becomes real, then the
reputation about low second-hand value will have negative influence on the
demand for new ones. When this is added to the effect of branding suggested
in the previous section, the kind of apartments now being built are unlikely to
be the most profitable ones to buy when they are new.

Regardless of changes in demand and prices as suggested here, estate

developers will probably soon offer apartments alternative to the
contemporary ones designed for singles and couples in particular. This is due
to the simple fact that there are also many wealthy people among other kinds
of households; in the long term, it will not be profitable to exclude these from
the opportunity of buying new urban dwellings.

As argued in section 2.6, low standard dwellings are not necessarily bad since
a consequence of not being appreciated by the majority is that the dwelling
might represent positive potentials for subgroups of the population. Most of
the small apartments now being built will achieve a low score if traditionally
appreciated features such as spaciousness, usability and daylight conditions
are evaluated. It is therefore not hard to imagine that these dwellings, within
20 years, when they have lost their attractions of being new and fashionable,
will be considered low quality dwellings by most of the population and
thereby become less costly compared with other dwellings in the same
locations. According to the possibly positive potential of having some low
standard dwellings, these apartments might represent a positive contribution
to the stock of dwellings in the future; the new and “unacceptable” tiny
dwellings might in the future make urban living possible also for people who
would never afford ordinary dwellings in central parts of town. To what
extent such a scenario would be a problem, - for those who now buy them
expensive, or for the neighbourhoods with many such “bad dwellings” are
worth further discussions but will not be elaborated upon here. What seems
clear, is that the future of these new dwellings are uncertain, and that there
are some correspondences between dwelling quality, market and prices that is
not mentioned in the marketing of these new dwellings and that have not
been discussed much so far.
187
7.5 DWELLINGS AND GENERALITY

A brief conclusion that can be drawn from this study is that generality works;
the apartment characterised as general due to sizes of rooms and spatial
configuration are dwellings that house a large variety of households who are
well being. When it comes to what to learn from this, or when discussing
general versus specific floor plan layouts in general, it should be
distinguished between small and larger dwellings. Larger dwellings do not
depend on generality in order to house different kinds of households. In the
1970s, a large number of new apartments were function-specific in terms of
being designed for a limited range of households; they aimed particularly at
housing families with children. However, this specificity with respect to
“range of households”, is to lesser extent a problem than it is for the small
apartments now being built. While the large apartments of the 1970s have the
potential for becoming spacious dwellings for other kinds of households than
traditional families,216 it is hard to imagine that the recently built apartments
will become attractive for other households than singles and couples. As
commented in section 7.4, it might be that these small apartments in the
future will not even be attractive for the kinds of household now buying
them.

Where smaller apartments that these family apartments of the 1970s are
concerned, these are not likely to be appropriate for a diversity of households
unless this has been an explicit concern in the design process. The result from
this study is that general apartments of “two rooms” house all categories of
households except couple with children while those of “three rooms” house
many families as well. What characterises the general apartments found and
examined in this study is that they have separate kitchen and at least two
other rooms appropriate for different kinds of daytime living.217 The latter
means that the “bedrooms” are not too small and that they have access
otherwise than through the living room. This layout corresponds to a proposal
that has been made in different versions throughout the history of post-war
Scandinavian housing research: a dwelling aiming at housing different kinds
of households should have at least two habitable rooms where the smallest
should be no less than 10 -12 m2.218 The strategy of designing general

216
See for instance table 5.3 (in section 5.2.2) about single’s and couples’ well-being in the B-type apartments
or see figure 5.11 (in section 5.2.6), showing how apartment 74-1-A2 has become a spacious dwelling for a
couple.
217
Separate kitchen is an issue that should be examined more closely; is separate kitchen just a feature in
common for these general apartments or is a prerequisite for being general (in terms of being preferred by a
wide range of households)? This can be examined by comparing kinds of households living in similarly sized
apartments with and without separate kitchen.
218
See Brochmann (1952, p. 165), Holm (1955, p. 222), Thiberg (floor-plan proposal, 1967, see figure 2.2.2),
and Guttu, Jørgensen and Nørve (1985, p. 64).

188
dwellings do not imply to aim at one layout fitting all; it is hard to imagine
any particular dwelling that would be the best dwelling through all stages of a
person’s life. Similar to change of car in accordance with changing needs
(sport-cars are rarely customised into station wagons), some moving is
necessary in order to achieve a good match between dwelling and the living
preferences that alter due to age, health, economy and family circumstances.
However, in a stock of housing consisting of apartments designed with
generality in mind, people are less forced to move; moving can be a
possibility rather than a necessity. A master bedroom with room for a child’s
bed, allows for years of flexibility as concerns when to move to a larger
apartment with a separate room for the child.

The kind of dwellings built in Oslo in the late 19th century (see figure 2.21)
represent a kind of spatial layout that deserves to be commented upon where
generality is concerned. The generality of these dwellings is remarkable.
Regardless of being located in the traditional wealthy centre-west parts of
Oslo or in the previous working-class and now popular centre-east, and
despite of a low technical standard, these 120 years old buildings are today
expensive and attractive for a diversity of purposes, dwelling included. With
spacious rooms that are not determined exclusively for particular activities,
these old apartments are still highly appreciated dwellings but they might as
well be shops, offices or restaurants. Such attractiveness and variety of
activities rarely exist in housing projects built during the last 70 years and is
even less likely to be found in the ones now being built. In conclusion, there
are at least three lessons to learn from the buildings constructed in the 19th
century where generality is concerned. The first is that generality not only in
theory but also in reality is a permanent feature. The second is about
attractiveness. Similarly to the results concerning the general apartments
examined in this study, the feature of being appropriate for a diversity of
preferences do not prevent a building from being highly appreciated by the
different, specific users. The third is about generality on a more overall level.
Instead of discussing how to achieve general dwellings, the subject could as
well be how to design general buildings, buildings that have the potential to
accommodate diverse activities, dwellings included; as shown by these more
than 100 years old dwellings, such generality does not prevent the buildings
from being highly attractive places to live.

189
8 Further Research

When it comes to the influence of housing research on architectural practice,

the expectations should not be too high. As mentioned in section 2.1, the
main references for architectural design are buildings rather than theory in
written form. However, assuming that housing research still might affect
what dwellings to built, this final section mentions some subjects related to
this study that are worth to be examined. These are of two kinds, the first
concerns the historical development of dwellings, while the other concerns
alternative strategies for dwelling design. As mentioned in chapter 5, many of
the detailed subjects discussed there should also be worth further studies.
These subjects will not be repeated here.

Where the historical development of dwellings is concerned, several issues

are worth examinations similar to what has been carried out in this study.
One is other kinds of dwellings, the single-family house in particular. This is
the most common kind of dwelling in Norwegian. Regardless of being mass-
produced or being individually designed by architects, the layouts of single-
family houses seem to have little variations in spatial configuration; very few
single-family houses have a spatial layout different from the early modernist
template, no matter how fashion or avant-garde the house might appear as a
visual image. The first question to examine is whether this is the case. If it is,
there are at least two possible explanations. One is that the modernist scheme
is still a good one; another is that interior space of the single-family house
has not been focused by substantial criticism in the same way as other
disciplines of architecture.219 By both explanations, the development of the
interior space of single-family houses should be worth a survey similar to the
one of this study. Another issue worth further inquiry is the general relevance

219
Some of the criticism of urban-scaled functionalism, for instance of the subdivision of human behaviour
into functions that in the design process are treated as separate issues, is easily transferable onto dwellings in
general and onto the single family house in particular.

190
of the recent changes identified in this thesis. The new kind of small
apartments found in this study of apartments in Oslo, seems to be common
elsewhere in Norway as well. If, or to what extent, the same development
happens elsewhere has not been looked upon in this study, but should be
examined. The small apartments that now are common here, cannot be
identified as a general pattern among housing projects published in the
international architectural magazines. However, as these magazines are not
representative for the dwellings built, this does not imply that the pattern
found in this study cannot be found internationally as well.

Where strategies for housing design are concerned, this thesis has focused on
the dichotomy of specificity and generality. As elaborated in section 2.4,
generality can be considered one out of three alternative strategies for
achieving dwellings capable of serving a diversity of preferences, where the
other strategies are elasticity and flexibility. A plausible question from the
results of this study is how generally works compared to the alternative
strategies aiming at the same purpose. Flexibility is briefly touched in this
study by the analyses of the B-type of apartments, which are the large
apartments built around the 1970s. As described in section 5.2.6, these
apartments of the B-type are somewhat flexible due to large floor areas and
lightweight internal walls. The kind of dwellings not examined in this thesis
but worth analyses is the apartments designed in order to be “elastic”. The
first question is what elastic dwellings are like after some time have passed;
are they elastic in terms of really altering between different layouts, or do
they tend to “converge” towards particular layouts?220 The next question
concerns their “functional performance”221. By surveys similar to the one
carried out here (which would imply to examine different floor-plan
alternatives in the “stages of elasticity” and to compare these with the
residents and their daily living) it should be possible to compare the
alternative strategies for achieving dwelling appropriate for a diversity of
preferences. The fact that generality works, which is a conclusion of this
study, does not imply that other strategies might not be useful as well. This
issue about how to design in order to handle a diversity of requests concerns
not only dwellings but also buildings and architecture in general.

The kinds of research really carried out are rarely a result of proposals as
those mentioned above; the issues examined by Norwegian housing research
are determined by governmental dispositions in terms of founding of research
projects. In accordance with the governmental dispositions, a major concern

220
Similarly to what has happened in the project Skjettenbyen (see figure 2.24) where most dwellings have
been extended to the maximum size and thereby become equal.
221
”Functional performance” is a term applied by Marcus (2000).

191
for Norwegian housing research is now “universal design”. A conclusion of
this thesis is that new apartments differ significantly from what architects,
planners and researchers until very recently considered a minimum standard
where usability is concerned. As described in section 7.1, a consequence of
this is that new apartments are appropriate for a limited range of households
only. It is thereby a discrepancy between the ideals about housing (as
represented by the issues focused upon in academic research and by
governmental institutions) and the reality of dwelling construction. While a
large number of new apartments are inappropriate even for people who are
functioning fully when health and ability to move are concerned, the
governmental focus is set on how to design dwellings appropriate for all.
While planners and researchers discuss how to increase the accessibility for
persons in wheelchairs, a large number of the apartments actually built do not
even have room for physically functional people to sit and eat. The aim in
pointing out this is not to disavow “universal design”, but to address that the
problem of usability in new dwellings is now of another scale than what is
handled by traditional “universal design”. Concerning the quality of
Norwegian dwellings in terms of usability, the greatest potential for
improvement is not likely to be achieved by studying how dwellings should
be designed in order to fit all, but in finding out how to influence a housing
production where new apartments are well suited only for a few.

192
References:

Adler, Leo (1927) “Fragen und Vorbedingungen der Typisierung.

Bemerkungen zu Studien und Entwürfen von Alexander Klein,
Berlin”, Wasmuths Monatshefte für Baukunst, 1927, p. 252-257.
Alexander, Christopher (1977) A Pattern Language. New York
Alexander, Christopher (1985) The Production of Houses. New York
Asami, Yasushi, Kubat, Ayse Sema, Kitagawa, Kensuke and Iida, Shin-ichi
(2003) “Introducing the third dimension on Space Syntax:
Application on the historical Istanbul”, Proceedings 4th International
Space Syntax Symposium London 2003 (volume II, paper 48).
London
Blombergsson, Magnus and Wiklander, Johanna (2006) “Spatial support for
key usability factors: spatial influence on interaction patterns for
800 office worker”, Proceedings, Trondheim international
symposium, Changing user demands on buildings. Trondheim
Bostadsutformning bostadsanvändning, see:
Dahlgren and Westerberg (1979)
Dahlgren (1980)
Eriksson, Lindquist and Westerberg (1980)
Gaunt (1979)
Lindquist, Orrbeck, and Westerberg (1980)
Brand, Steward (1994) How Buildings Learn, What Happens After They’re
Built. New York
Bredberg, Ulf (1978) Metod för analys av planegenskaper vid prosjektering
av bostäder. Stockholm
Buhl, Ole (1965) Le Corbusier, Menneskenes bolig. København (p. 15)
(Danish version of Le Corbusier (1942) La Maison des
Hommes. Paris)
Bredsdorff, Peter (1938) “Boligtyper for børnerige Familier” Arkitekten
1938, p. 5-8.
Brochmann, Odd (1948) Mennesker og boliger. Oslo
Brochmann, Odd (1952) “Hus Grete”, Bonytt, 1952, p.161-165.
Brochmann, Odd (1958) By og bolig. Oslo
Brochmann, Odd (1952) Livsform og Boligform. Oslo
Brochmann, Odd (1961) Bedret boligbruk. Oslo
Christiansen, Jan (1996) 1975-1995 Arkitektonisk kvalitet, om dansk
boligbyggeri. København
Cold, Gunnarshaug, Hiortøy og Raaen (1984) Nye boligformer, en
eksempelsamling. Trondheim

193
Coolen, Henny (2005) “Measurement and analysis of less structured data in
housing research”, Methodologies in housing research, Vestbro,
Dick Urban, Hürol, Yonca and Wilkinson, Nicholas. Gateshead
Cunha, Viviane and Magalães, Fernanda (2005) “Domesticy and middle class
in Rio de Janeiro”, 5th International Space Syntax Symposium, Delft
June 2005, Proceedings Volume I (p. 635). Delft
Cuff, Dana (1991) The Story of Practice. Massachusetts Institute of
Technology.
Dahlgren, Stefan and Westerberg, Ulla (1979)
Bostadsutformning bostadsanvändning 1, Metodredovisning
(Statens institut för Byggnadsforskning, Meddelande M79:6). Gävle
Dahlgren, Stefan (1980)
Bostadsutformning bostadsanvändning 4, Sammansällning av
intervjusvar
(Statens institut för Byggnadsforskning, Meddelande M80:3). Gävle
Do, Ellen Yi-Luen and Gross, Mark D. (1997) “Tools for visual and spatial
analysis of CAD models, implementing computer tools as a means
to thinking about architecture”, CAAD Futures '97. Munich
(Proceedings, p. 189-202) Paper available at:
http://depts.washington.edu/dmachine/PAPER/CF97-
ISO/spatial.html)
Dobloug, Margethe (2006) Bak verket. Kunnskapsfelt og formgenererende
faktorer i nyttearkitekturen 1935-1985. Oslo
Engels, Friedrich (1845) Die Lage der arbeitenden Klasse in England.
Leipzig (in English versions: The Conditions of the Working Class
in England)
Eriksson, Birgitta, Lindquist Margareta and Westerberg, Ulla (1980)
Bostadsutformning bostadsanvändning 6, Möblerade
lägenhetsplaner från 308 hushåll
(Statens institut för Byggnadsforskning, Meddelande M80:5. Gävle
Farah, Eman (2000) House Form and Social Norms. Spatial Analysis of
Domestic Architecture in Wad-Nubbawi, Sudan. Gøteborg
Figueiredo, Lucas and Amorim, Luiz (2005) “Continuity lines in the axial
system”, 5th International Space Syntax Symposium, Delft June
2005, Proceedings Volume I (p. 161). Delft
Frederick, Christine (1925) Efficient Housekeeping or Household
Engineering, Scientific Management in the Home. London (Home
Economics Association (1915 – 1925) and Routledge and Sons)
Gaunt, Louise (1979) Bostadsutformning bostadsanvändning 3,
205 skolbarns bostadsanvändning
(Statens institut för Byggnadsforskning, Meddelande M79:7). Gävle

194
Glaser, B. G. and Strauss, A.L. (1967) The Discovery of Grounded Theory:
Strategies for Qualitative Research. New York
Guldbrandsen, Ole (1973) Å bo på ett rom i blokk. Oslo
Guney, Yasemin Ince (2005) “Spatial types in Ankara Apartments”, 5th
International Space Syntax Symposium, Delft June 2005,
Proceedings Volume I (p. 623). Delft
Guttu, Jon, Jørgensen, Ivar and Nørve, Siri (1985) Bovaner. Oslo
Guttu, Jon and Ditlef-Martens, Johan (1998) Sentrumsnære byboliger,
NIBR Prosjektrapport 1998:10. Oslo
Guttu, Jon (2003) “Den gode boligen, Fagfolks oppfatning av boligkvalitet
gjennom 50 år. Oslo
Grønvold, Ulf (2000) Priset arkitektur 1904-2000. Oslo
Hansen, Willy (1930) “Stockholmsudstillingens boligafdeling”, Architekten
(DK) 1930, p. 147.
Hanson, Julienne (1998) Decoding Homes and Houses. Cambridge
Hillier, Bill and Hanson, Julienne (1984) The Social Logic of Space.
Cambridge
Hillier, Bill (1996) Space is the Machine. Cambridge
Hirdina, Heinz (editor) (1984) Neues Bauen Neues Gestalten, Das Neue
Frankfurt / Die Neue Stadt, Eine Zeitschrift Zwishen 1926 und 1933.
Berlin (West)
Holm, Lennart (1955) Familj och bostad. Stockholm
Holst, Axel (1895) Undersøgelser og Forslag angaaende Arbeiderstandens
Boliger i Christiania. Christiania (Oslo)
Howard, Ebenezer (1898) To-morrow: A Peaceful Path to Real Reform (in
later print (1902) Garden Cities of To-morrow. London)
Husbanken (1985) God bolig. Oslo
Husbanken (2000) Husbankens minstestandard, HB 7.B.1.3. Oslo
Isaacs, R.R (1983) Walter Gropius, Der Mensch und sein Werk. Berlin
Jacobs, Jane (1991) The Death and Life of Great American Cities. New York
Janson, Ulf (1998) Vägen till verket. Göteborg
Johansen, Thora, Berge, Marit and Andresen, Lucie (1961) Nanna Broch.
For Boligsak og hjemmets trivesel. Oslo
Johansson, Rolf (1997) Utvärdering av bostadshus i bruk. Stockholm
Lindquist, Margareta, Orrbeck, Kenneth and Westerberg, Ulla (1980)
Bostadsutformning bostadsanvändning 5, Bostaden i norm och
verklighet
(Statens institut för Byggnadsforskning, Meddleande M80:4). Gävle
Lynch, Kevin (1972) What time is this place?. Cambridge, Massachusetts
Klarqvist, Björn (1969) Bostadsplanering. Gøteborg
Klarqvist, Björn (1991) Manual för rumslig analys av städer och byggnader.
Gøteborg
195
Klarqvist, Björn and Thiberg, Sven (2003) Acceptera inte. Göteborg
Klein, Alexander (1931) “Wirtschaftliche Grundrissbildung und
Raumgestaltung”, Wasmuths Monatshefte für Baukunst, 1931, p.
538.
Klein, Alexander (1934) Das Einfamilienhaus, Südtyp. Wohnbau und
Städtebau, Beiträge zur Entwurfslehre, In Verbindung mit dem
Internationalen Verband für Wohnungswesen. Berlin
Klein, Naomi (2000) No Logo: Taking Aim at the Brand Bullies. Toronto
Krüger, M. (1989) On Node and Axial Grid Maps: Distance measures and
related topics, University College, London. London (Paper
available at
http://eprints.ucl.ac.uk/archive/00001011/01/Kruger_1989_OnNode
AndAxialGrid.pdf )
M79:6 : see Dahlgren and Westerberg (1979)
M79:7 : see Gaunt (1979)
M80:3 : see Dahlgren (1980)
M80:4 : see Lindquist, Orrbeck and Westerberg (1980)
M80:5 : see Eriksson, Lindquist and Westerberg (1980)
Manum, Bendik (2003) “Combining Quantitative and Qualitative Methods:
A PhD Study on the Interior Space of Apartments”,
(Paper presented at the conference “Methodologies in Housing
Research”, KTH, Stockholm)
Manum, Bendik (2005) “Generality versus Specificity; a Study on the
Interior Space of Apartments”, 5th International Space Syntax
Symposium, Proceedings, Volume II, (p. 3). Delft
Manum, Bendik, Rusten, Espen and Benze, Paul (2005) “AGRAPH,
Software for Drawing and Calculating Space Syntax Graphs”, 5th
International Space Syntax Symposium, Proceedings, Volume I,
(p. 97). Delft
Marcus, Lars (2000) Architectural Knowledge and Urban Form, The
Functional Performance of Architectural Urbanity. Stockholm
Markelius, Sven (1935) “Olika husbredder och deras byggnadskostnader",
Plan: 3, 1935
Martens, Johan-Ditlef (2000) Århundrets norske boligprosjekter 1900-2000.
Oslo
Münz, Ludwig and Künstler, Gustav (1995) Adolf Loos Pioneer of modern
architecture. London
Neufert, Ernst (1936) Bauentwurfslehre. Berlin (later published in many
languages and many editions)
Noever, Peter (1992) Die Frankfurter Küche von Margarete Schütte-
Lihotzky. Berlin

196
Nylander, Ola (1998) Bostaden som arkitektur. Gøteborg (English version:
(2002) Architecture of the home. Chichester)
Olivegren, Björkman and Schultz, Solveig (1990) “Increase users’
influence”, Housing research and design in Sweden (p. 237-),
Thiberg, Sven (editor). Stockholm
Park, Hoon Tae (2005) “Before integration: a critical review of integration
measure in space syntax”, 5th International Space Syntax
Symposium, Delft June 2005, Proceedings Volume II (p. 555). Delft
Peponis, J., Wineman, J., Rashid, M., Bafna, S and Kiw, S., H. (1998)
”Describing plan configuration according to the covisibility of
surfaces.”, Environment and Planning (B): Planning and Design,
25, p. 693-708.
Pommer, Richard and Otto, Christian F. (1991) Weissenhof 1927 and the
Modern Movement in Architecture. Chicago and London
Plan (articles where name of authors is not published):
“Hybeleiligheter”, Plan: 1, 1933
“Erfaringene fra Tyskland”, Plan: 2, 1933
“Forfallsperioden (1929-?)”, Plan: 2, 1933
“Familien og boligen”, Plan: 3, 1935
“Ned med boligstandarden?”, Plan: 3, 1935
“Veien til lavere husleier”, Plan: 3, 1935
“Hvor meget koster lys og luft?”, Plan: 3, 1935
Popper, Karl R. (1959) The Logic of Scientific Discovery. New York
(original: Popper (1934) Logik der Forschung. Wien)
Ramstad, Vegard and Ribe, Kristian (2006) Kontemporær boligproduksjon i
indre Oslo, (Diploma-project / final project, Master degree at The
Oslo School of Architecture and Design). Oslo
Ranten, Karen and Vedel-Petersen, Finn (1982) Bedre boligplaner,
Beskrivelse af en metode til prosjektering og vurdering. Hørsholm
Rapoport, Amos (1990) “Housing and culture”, Housing: symbol, structure,
site, Taylor, Lisa (editor). New York
Rave, Rolf and Knöfel, Hans-Joachim (1968) Bauen seit 1900 in Berlin.
Berlin
Rievertz, Kr. (1935) “Bolignød, boligforhold, boligkultur”, Byggekunst,
1935, tillegg, p. 3-4.
Risselada, Max (1989) Raumplan versus Plan Libre. New York
Rolness, Kjetil (1995) Med smak skal hjemmet bygges: innredning av det
moderne Norge. Oslo
Sandaker. Bjørn Normann (2000) Reflections on span and space. Oslo
Schjødt, Liv (1957) “Svensk utnyttelse av kvadratmetre 1960”, Bonytt 1957,
p. 254-257.
Scruton, Roger (1979) The Aesthetics of Architecture. Princeton
197
Sjølie, Mette (2004) “Hus i bruk – OBOS 75 år”, Byggekunst, 2000: 7, p. 24-
35.
Statens Byggeforskningsinstitut (1971) Tæt lav - en boligform, SBI-rapport
71. København
Statens Byggeforskningsinstitut (1978) Elastiske etageboliger, Vilkår,
konsekvenser, eksempler, SBI-rapport 103. København
Statens Institut för Byggnadsforskning (1979 and 1980), see:
Dahlgren and Westerberg (1979)
Dahlgren (1980)
Eriksson, Lindquist and Westerberg (1980)
Gaunt (1979)
Lindquist, Orrbeck, and Westerberg (1980)
Støa, Eli (1995) Boliger og kultur, Norske boligfelt på åtti-tallet sett
i lys av beboernes boligidealer. Trondheim
Sundt, Eilert (1858) Arbeiderklassens kår i Christiania 1858. Oslo
Sundt, Eilert (1870) Om fattigforholdende i Christiania. Oslo
Svennar, Elsa (1975) Boligens planløsning, Håndbok 25, Norges
Byggforskningsinstitutt. Oslo
Taut, Bruno (1927) Bauen der neue Wohnbau. Berlin
Thaler, Ulrich (2005) “Narrative and Syntax, new perspectives on the Late
Bronze Age palace of Pylos, Greece”, 5th International Space Syntax
Symposium, Delft June 2005, Proceedings Volume II (p. 327). Delft
Thiberg, Sven (1967) Rommål, plantyper, hustyper for industriell bygging –
orientering om svenskeforskningsprosjekter,
Norges Byggforskningsinstitutt. Oslo
Thiberg, Sven (1990) Housing Research and Design in Sweden. Stockholm
(Swedish original: Thiberg (1985) Bostadsboken)
Tucker, Alan (2002) Applied Combinatorics. New York
Turner, Alasdair (2005) “Could A Road-centre Line Be An Axial Line In
Disguise?”, 5th International Space Syntax Symposium, Delft June
2005, Proceedings Volume I (p. 145). Delft
Warhaftig, Myra (2000) “Wohnungen der kurzen Wege: Alexander Klein”
Der Architect 2000: 6, p. 20- 22.
Yin, Robert K. (1994) Case Study Research. Thousand Oaks, California
Øvergaard, Jørgen Gunnar (1938) “Bedre boligtyper, billigere boligbygging”
Byggekunst 1938:7, p. 129.
Ås, Dagfinn (1971) Norske boliger 1, Resultater fra boligundersøkelsen
1967, Om undersøkelsen, Boligmassen 1967, boligens rom, NBI
Rapport 69-1. Oslo
Ås, Dagfinn (1971) Norske boliger 1, Resultater fra boligundersøkelsen
1967, Boligstandard: Utstyr og plass, NBI Rapport 69-2. Oslo

198
Appendices

Contents

A.3.1 “AGRAPH; Software for Drawing and Calculating Space Syntax

Connectivity Graphs” (paper presented at “The 5th International
Space Syntax Symposium”, TU Delft, 2005)

This software is free download from http://www.aho.no/ahograph/.

A.4.1 List of the projects examined in the diachronic survey.

A.4.2 Space syntax connectivity graphs, all apartments.
A.4.3 Some room sequences, all apartments.
A.4.4 Results from space syntax calculations, all apartments.

A.5.1 Floor plans and key information about the projects examined in the
synchronic survey.

A.5.2 The interview-forms, an example.

A.5.3 Floor plan drawings with furniture, selected apartments.

NOTE:
In originals, these appendices were in A4-format. The floor plans in the
printed appendices are therefore not in original sizes and scales. For floor
plans in original scales, see the separate appended volume.

199
 1

AGRAPH,
Software for Drawing and Calculating Space Syntax Graphs

Bendik Manum, PhD student, The Oslo School of Architecture and Design
Software programming: Espen Rusten and Paul Benze

Keywords: Space Syntax graphs, Space Syntax calculations

Abstract
AGRAPH is a PC-application for drawing Space Syntax graphs and for doing Space Syntax
calculations. AGRAPH is made for drawing graphs on the basis of imported background
images such as floor plan drawings. By a “snap to grid”-option and “click-and-drag”-
commands justified graphs are easily made. AGRAPH calculates the basic Space Syntax
parameters of the graph-nodes but has also the option of excluding selected nodes from the
calculation. AGRAPH has several colouring modes. One is the “custom mode” that is useful
for coding nodes by colour. Another is the mode for colouring nodes according to the Space
Syntax parameters Relative Asymmetry or Control Value. By the export-command the graph-
image can be stored in formats appropriate for further editing and publishing. The tables
containing the Space Syntax parameters of the nodes in the graph can be opened in common
software such as EXCEL for further analysis. The program structure of the AGRAPH is
designed for further development. At the time of writing this paper, several additions are being
developed.

1. Introduction

AGRAPH is developed for being applied to a PhD-study at the Oslo School of Architecture and
Design. In this study a PC-application was needed for doing Space Syntax analysis of a large
number of apartments. There exist a number of software applications for Space Syntax
analysis, ranging from tools for simple analysis of connectivity to advanced visual field
analysis. Nevertheless, when in this PhD-study looking for software capable of handling both
the Space Syntax calculations (giving easily accessible output data) and the intended drawing
and editing of graphs, appropriate software for the PC was not found. It was then decided to
develop software for the actual task. The outcome of this is AGRAPH. This paper explains the
basic features of the software and describes briefly how it is used.

In the fields of graph-theory and combinatorics, which are the mathematical basis for Space
Syntax calculations, the elements of the graphs are usually described as vertexes and edges.1 In
this paper, which is on Space Syntax analysis and not on mathematics, the more literal terms
nodes and connections are preferred.

2. The AGRAPH Application in General

AGRAPH is developed in C# using VisualStudio.NET. The compiled .EXE file runs on PCs
operated by WINDOWS. AGRAPH works on files of a format named .AGX. This file contains
all information of the graph. From the graph-image drawn on the PC-screen, AGRAPH
generates a “connectivity matrix” (listing whether nodes are connected or not) and an “internal
distance matrix” (listing the shortest distance between the nodes in the graph)2. By simple
calculations on these matrixes, the Space Syntax parameters of the nodes are determined.
 2

3. The Space Syntax Parameters

AGRAPH calculates the parameters Control Value (CV), Total Depth (TD), Mean Depth
(MD), Relative Asymmetry (RA) and the integration value (i).3

Control Value
The Control Values (CV) are found by letting each node give the total value of 1 equally
distributed to its connected nodes. The Control Value of node n, CV(n), is the total value
received by node n during this operation.

Total Depth
Total Depth of a node n, TD(n), is the total of the shortest distances from node n to the other
nodes in the systems, i.e. TD(n) is the total of line n (or column n) in the distance matrix.

Mean Depth
Mean Depth for a node n is the average depth (or average shortest distance) from node n to all
the other nodes. If k is the total number of nodes in the system, then MD(n)=TD(n)/(k-1).

Relative Asymmetry
The Relative Asymmetry (RA) describes the integration of a node by a value between (or equal
to) 0 and 1, where a low value describes high integration. RA is calculated by the formula
RA=2*(MD-1)/(k-2).4

Integration Value
A parameter that (contrary to RA) describes integration by a high number when a node is
highly integrated is the “integration value” (i). The integration value is found by inverting the
RA, i=1/RA.5

4. Drawing and Justifying the Graph

Unless the spatial structure to be analysed is very simple, a convenient way of drawing the
graph is to use an appropriate background image. Such a background-image is imported into
AGRAPH as a file of .TIF or .JPG format. In the case of studying apartments this background
image would be the floor plan drawing. By setting the “snap to grid” to “off”, nodes and
connections are easily drawn on the background image. Especially when drawing larger
systems it is useful to name the nodes before removing the background image and before
justifying the graph. (Naming of nodes is done by “right click” on node and then using the “set
node name”-option.) By turning on the “snap to grid”, justified graphs are made by simple
“click and drag”. The nodes are automatically numbered from 0 and upwards in the order they
are drawn.6 The results of the calculation are listed according to node number. As it might be
preferred to have the nodes listed in a given order, the numbering of the nodes can be edited
after drawing. Having the node numbers visible in the centre of each node is an option that can
be turned on/off. The name of the node is placed beside the node on the graph-image and can
be edited as text at any stage. The file-name automatically appears at the down left of the
graph. (See Figure 2) The graphs-image can be exported in formats such as .JPG or .BITMAP
for printing or for further analysis.
 3

5. The Colouring Modes

There are several modes for colouring the nodes. The simplest is to keep the default colour
(light grey) and to identify the nodes by name only. One alternative is to use the “custom
colour”-mode for coding the nodes by colour. When intending to compare many graphs or to
study graphs consisting of a large number of nodes, such coding by colour simplifies the
further analysis.

Similar to the well known coloured graphics used for presenting the integration values of axial
line analysis, AGRAPH has the option of colouring the nodes according to the integration or to
the control value of the individual nodes. When using “Colour by Control Value” or “Colour
by RA, relative”, the total scale of colour is set to range between the minimum and the
maximum value occurring in the graph. Using “Colour by RA, absolute” the range of colour
spans from the value RA=0 (red) to RA=1 (dark blue/purple), and the nodes are coloured
according to their RA-value within this scale. When applying “Colour by RA” for comparing
different graphs, the “Colouring by RA, absolute” is useful. If analysing only one graph the
“RA-relative” is better as this gives a larger range of colour and thereby a more differentiated
information. In addition to being useful for analysing given architectural spaces, these
“colouring by calculation”-modes are helpful for visualising the effects of making changes to a
system by deleting or adding nodes and connections.

When nodes are coloured by “custom colour”, this information is stored in the .AGX file. This
means that different “colour by calculation” can by applied without loosing a previous custom
colouring such as nodes coloured by function.

6. The Calculation

For each calculation AGRAPH generates three output files. These are tables in the .HTML-
format that are stored in the same file-catalogue as the .AGX file. The file “--summary.HTML”
lists the CV, MD, RA and the 1/RA for all nodes of the graph. The files “--connections.HTML”
and “--distances.HTML” contain respectively the connectivity matrix and the matrix of the
internal distances between nodes. In most cases the summary file is the only information
needed for further Space Syntax analysis. All these output files can be opened in common
software such as EXCEL for further editing or analysis.

In Space Syntax analysis there are often cases where all of the existing spaces are not included
in the calculation. The “calculation without exterior” is a well known such case. Depending on
the actual study, it might be relevant to exclude other spaces. AGRAPH has the option of
“deactivate node” which excludes the selected nodes from the calculation.

7. Further Development of AGRAPH

AGRAPH now works well as the tool searched for in the on-going PD-study. However, for a
wider range of use, some additions to the software would be useful. Some improvements are
being developed at the time of writing this paper and several more are planned.
 4

Figure 1. Opening AGRAPH. Choosing Background Image.

Drawing nodes and connections. Naming the nodes.

Figure 2. Clearing the background. Justifying the graph. Colouring the nodes.
 5

Figure 3. Doing the calculation. Saving the graph in an image-format.

Figure 4. Some AGRAPH-images edited for presentation, an example.

 6

Notes:
1
For more on the mathematics of graphs, see for instance Tucker (2002).
2
Distance not as “metric distance”, but as “Space Syntax distance” (i.e. number of “space-
steps”).
3
For more on the basics of Space Syntax calculations, see:
Hillier, 1996, p.33, p.88-.
Hiller and Hanson, 1984, p. 92-123,p. 147-155.
Hanson, 1998, p. 22-31.
Those who are reading a Scandinavian language might also see Klarqvist (1991).
4
The RA-value is defined to be 0 when a node is as integrated as possible. The most integrated
position possible is the root of a pure “bush”-graph. As all distances from such a root are 1, the
MD is 1. By this RA=0 when MD=1. The RA might therefore be of the form RA= a*(MD-1).
Contrary, RA is defined to be 1 when a node is as segregated as possible. The most segregated
positions possible are the end nodes in a pure linear sequence. For a linear sequence of k nodes
the MD for the end nodes is half the number of nodes in the line, MD=k/2. By this: RA=1
when MD=k/2. This implies 1=a*(k/2-1) which gives a=2/(k-2). By this RA=2*(MD-1)/(k-2).
5
This is the integration value of RA. Integration might be defined as the inverse of other
asymmetry parameters than the RA. The most usual is to invert the RRA, the Real Relative
Asymmetry, as described by Hillier and Hanson, 1984, p.111-113.
6
By this the highest node number is one less than the total number of nodes in the graph.

References:

Hanson, J., 1998, Decoding Homes and Houses, Cambridge, Cambridge University Press.

Hillier, B., Hanson, J., 1984, The Social Logic of Space, Cambridge, Cambridge University
Press.

Hillier, B., 1996, Space is the Machine, Cambridge, Cambridge University Press.

Klarqvist, B., 1991, Manual för rumslig analys av städer och byggnader, Göteborg, Chalmers
Tekniska Högskola

Tucker, A., 2002, Applied Combinatorics, New York, John Wiley & Sons.
Appenix A.4.1 Total sample of apartments Page 1/4
Number Name Address Apartment Architects
m2 rooms

31-1-1 Etterstad I BRL Etterstadsletta 4 55 2 Jacob Christie Kielland

31-1-2 Etterstad I BRL Etterstadsletta 4 76 3 Jacob Christie Kielland
32-1 Maridalsveien 64 BRL Maridalsveien 64 57 2 Einar Engelstad
33-1 Fagerheimen BRL Christies gt. 7-13/ Fagerheimgt. 2-14 50 2 Henrik Nissen og Gunnar Brynning
33-2 Bentsebakken BRL Bentsegata 15, 17, 19 62 2 H. E.Brevik

36-1 Galgeberget BRL Galgeberget 3 54 2 Fritjof Rojahn

36-2 Trondheimsveien 170 BRL Trondheimsveien 170 52 2 Andreas Nygaard
36-3-1 Frydenberg BRL Grenseveien 9 61 2 Victor Schaulund
36-3-2 Frydenberg BRL Grenseveien 9 81 3 Victor Schaulund
36-4 Munkegården BRL Munkedamsveien 86B 74 3 Ernst Motzfeldt og Leif Wiersholm
38-1 Myrahagen BRL Bentsebrugata 16C 49 2 Jens Dunker og Georg Rohde
39-1 Øvre Tøyen BRL Frydensgate 5 59 2 Sverre Wiik
39-1 Presidentgaten BRL Anna Sethnesgate 2 55 2 Klingenberg & Klingenberg arkitekter
40-1 Tøyenhus BRL Sarsgate 44-62 55 2 Th. Chr. Hauff
40-2-1 Trasopløkken BRL Treschows gate 17 51 2 Ragnvald Tønsager
40-2-2 Trasopløkken BRL Treschows gate 15 79 3 Ragnvald Tønsager
40-3 Maridalsveien BRL Maridalsveien 225-237 59 2 Einar Engelstad
41-1 Engelsborg BRL Sarsgate 64-80 56 2 Thr. Chr. Hauff
41-2 Bøkkerløkka BRL Brockmannsgate 8-10 75 3 Gunnar Gregaard Jørgen
41-3 Sandefjordgaten BRL Sandefjordgata 5 55 2 Victor Schaulund

41-4-1 Tøyen II BRL Helgesensgate 76 77 2 Victor Schaulund

41-4-2 Tøyen II BRL Helgesensgate 78 70 3 Jens Selmer
47-1 Brockmannsgata BRL Brockmannsgate 14 69 4 Arne Pedersen & Reidar Lund
47-2-1 Flaen BRL Nordtvetveien 44-46 57 2 S.Narve Ludvigsen
47-2-2 Flaen BRL Flaengrenda 1, 3, - - -, 15 67 3 S. Narve Ludvigsen
48-1-1 Nordre Tøyen BRL Frydensgate 1, 3 70 3 Hougen, Pedersen & Viksjø
48-1-2 Nordre Tøyen BRL Frydensgate 1, 3 71 3 Hougen, Pedersen & Viksjø
48-2 Gløtt BRL Søren Jaabæks gate 10 61 3 Ragnar Nilsen
48-4 Utsyn BRL Valhallveien 63-65 64 3 Ulf Nyquist
48-5-1 Storo BRL Birch Reichenwaldsgate 3 53 2 Hoel, Mugaas og Aastorp

48-5-2 Storo BRL Birch Reichenwaldsgate 35-45 62 3 Hoel, Mugaas og Aastorp

48-7 Høybråten BRL Slåttenveien 2 70 3 Ola M. Sandvik & co
48-8 Gaustadfunksjonærenes BRL Gaustadveien 15 70 3 H. D. Elgethun
49-1 Etterstad Nord BRL Etterstadsletta 7, 9, 11 60 3 Morseth & Wiel Gedde
49-2 Lille Ekeberg BRL Enoksvei 2-8 66 3 S. Narve Ludvigsen

50-1 Keyserløkka Nord BRL Bergljotsvei 9 A, B, C 68 3 Rinnan,Tveten & Colbjørnsen

50-2 Etterstad Øst BRL Etterstadsletta 47-49 76 3 Morseth & Wiel Gedde
50-3 Nygård BRL Nygård Terrasse 2-4 70 3 Henrik Nissen & Gunnar Brynning
51-1 Nordre Åsen BRL Kjelsåsveien 10 A, B, C 40 1 Rinnan,Tveten & Colbjørnsen
51-2 Etterstad Sør BRL Etterstadsletta 55 67 3 Morseth & Wiel Gedde

51-3 Akersbakken BRL Fredrikke Qvamsgate 13-21 74 3 Edgar Smith Berentsen

52-1 Søndre Åsen BRL Edvard Griegs allé 71 3 Oslo byarkitektkontor
52-2-1 Keyserløkka Øst BRL Einars vei 2-4 55 2 Rinnan,Tveten & Colbjørnsen
52-2-2 Keyserløkka Øst BRL Einars vei 2-4 70 3 Rinnan,Tveten & Colbjørnsen
52-3 Ensjøsvingen BRL Ensjøsvingen 4 57 2 Victor Schaulund
Appenix A.4.1 Total sample of apartments Page 2/4
Number Name Address Apartment Architects
m2 rooms
53-1 Teisen BRL Klosterheimveien 16, 18, 20 68 3 Rinnan,Tveten & Colbjørnsen
53-2 Myrer BRL Kurvveien 50, 52, - - , 60 57 2 T. Narve Ludvigsen
53-3 Stjerneblokkveien BRL Stjerneblokkveien 16-18 69 3 Bernt Heiberg & Ola Mørk Sandvik
53-4-1 Mogata 11 BRL Mogata 11 59 2 H. D. Elgethun
53-4-2 Mogata 11 BRL Mogata 11 73 3 H. D. Elgethun
54-1 Marmorberget BRL Antenneveien 34, 36, 38 64 3 Erling Viksjø
54-2 Valle BRL Vallefaret 1 - 11 66 3 Thams & Gaare arkitekter
54-3-1 Vetlandsveien BRL Vetlandsveien 81 81 4 Torgeir Alvsaker & E. Vaardal Lunde
54-3-2 Vetlandsveien BRL Vetlandsveien 81 74 3 Torgeir Alvsaker & E. Vaardal Lunde
55-1 Hovin BRL St.Jørgensvei 41-47 67 3 Rinnan,Tveten & Colbjørnsen

55-2 Oppsal BRL Håkon Tvetersvei 38, 40, 42 52 2 Helge Thams & Arne Bjønness
55-3 Engsletta BRL Kalbakkstubben 1-9 63 3 S.Narve Ludvigsen
56-1 Blåfjellet BRL Gråsteinsveien 8, 10, 12 66 3 Knut Knutsen
56-2 Åsenbygg BRL Borger Withsgate 19-25 51 2 Th. Chr. Hauff
57-1 Frederik Gladsgate BRL Frederik Gladsgate 23 B, C 60 2 Hoel, Mugaas og Aastorp

57-2 Etterstad Vest BRL Biskop Jens Nielsensgate 14 51 2 Morseth & Wiel Gedde
57-3-1 Husebybakken BRL Ullernschausseen 46A 32 1 Morseth & Wiel Gedde
57-3-2 Husebybakken BRL Ullernschausseen 46A 66 3 Morseth & Wiel Gedde
58-1 Manglerudjordet BRL Plogveien 33, 35, 37 77 4 Engh & Qvam
58-2 Mogaten BRL Mogata 12 A, B, C 68 3 Chr. Fredrik Størmer
59-1 Nils Huusgate Birch Reichenwaldsgate 29 69 3 S. Narve Ludvigsen
59-2 Nylænde BRL Nylænde 6 68 3 Byarkitekten i Oslo
59-3 Skøyenåsen BRL Solbergliveien 104 63 3 Helge B. Thams
60-1 Storgården BRL Byggveien 1, 3, 5, 7 73 3 Rinnan,Tveten & Colbjørnsen
60-2 Lillo Terrasse BRL Betzy Kjeldsbergsvei 15 66 2 Christiansen & Rosland
60-3 Vossegata BRL Vossegata 46 68 3 Ragnar Nilsen
61-1 Fuglemyra BRL Kampheimveien 28 69 3 T. Alvsaker & E. Vaardal Lunde
61-2 Hauktjern BRL Nøkleveien 2-8 69 3 Helge B. Thams
61-3 Rognerud BRL Otto Sognsvei 11-17 73 3 S. Narve Ludvigsen
62-1 Manglerudvangen BRL Rugveien 38-44, 53-87 69 3 Engh & Qvam

62-2 Bjerke BRL Refstadsvingen 7 73 4 Heiberg & Sandvik

62-3 Laura Gundersensgate 3 BRL Laura Gundersensgate 3 55 2 Kjell Colbjørnsen & Ulf Colbjørnsen
63-1 Pynten BRL Mellombølgen 50-52-54 68 3 Frode Rinnan & Olav Tveten
63-2 Solrabben BRL Kranveien 20-24 71 3 Engh & Qvam
64-1 Enerhaugen BRL Enerhauggaten 7 78 3 Sofus Hougen

64-2 Stallerudåsen BRL Stallerudveien 91 54 2 Preben Krag & Jens Selmer

64-3 Stormyra BRL Solbergliveien 101-103 74 4 Helge Thams
65-1 Rustadåsen BRL Rustadsaga 4-6 71 3 E. S. Berentsen & E. A. Berentsen
66-1 Bogerud BRL Martin Lingesvei 1, 3, 5 72 4 USBL
66-2-1 Rustad BRL Welding Olsensvei 5-7- -19 69 3 E. S. Berentsen & E. A. Berentsen

66-2-2 Rustad BRL Rustadgrenda 27-29 74 3 Preben Krag & Jens Selmer
66-3-1 Lønnealleen BRL Sigrid Undsetsvei 3 79 3 John Engh
66-3-2 Lønnealleen BRL Sigrid Undsetsvei 3 61 2 John Engh
67-1 Munkelia BRL Langbølgen 56 73 3 Helge Thams
67-2 Smalvollskogen BRL Lavransvei 10, 12, 14 67 3 Preben Krag & Jens Selmer
Appenix A.4.1 Total sample of apartments Page 3/4
Number Name Address Apartment Architects
m2 rooms

67-3-1 Ammerudlia BRL Ammerudgrenda 166-176 92 4 USBL v/ Th. Tostrup

67-3-2 Ammerudlia BRL Ammerudgrenda 166-176 49 2 USBL v/ Th. Tostrup
68-1 Lohøgda BRL Kristinsvei 36-42 73 3 Odd Brochmann
68-2 Ammerudfaret BRL Ammerudhellinga 26-50 78 3 Håkon Mjelva
68-3 Haugerud BRL Haugerudveien 38, 40, 42 93 4 Frode Rinnan & Olav Tveten
69-1 Tveita BRL Nåkkvesvei 1 66 2 Hans Backer Fürst
69-2 Haugenstua BRL Ole Brumms vei 18-24 78 3 Ungdommens selvbyggerlag
69-3 Smedstua BRL Kristoffer Robins vei 38-40 77 3 Preben Krag & Jens Selmer
70-1-1 Solfjellet BRL Haugerudveien 82 47 2 Frode Rinnan & Olav Tveten
70,1,2 81 3
70-2 Vosseløkka BRL Vossegata 18, 20, 20B 56 2 S. Narve Ludvigsen
71-1 Ravnkollen BRL Ravnkollbakken 72, 74 95 4 Alex Christiansen
71-2 Sandbakken BRL Selvbyggerveien 124-128 76 4 Alex Christiansen
71-3 Karihaugen BRL Harald Sohlbergsvei 14 93 3 Helge Thams & Per Gaare
72-1 Etterstadsletta 46 BRL Etterstadsletta 46 87 4 Kjell Colbjørnsen & Ulf Colbjørnsen
72-2 Fagerholt BRL Dr. Dedichensvei 38-46 82 4 Frode Rinnan & Olav Tveten
72-3 Meklenborg BRL Hovseterveien 48 A-B 89 3 A/S Ungdombygg
73-1 Karlstua BRL Hagapynten 9-11 99 4 A/S Ungdombygg
73-2 Ellingsrudåsen BRL Henrik Sørensensvei 32 74 3 Helge Thams & Per Gaare
74-1 Orremyr BRL Odvar Solbergsvei 28-30 83 3(4) Alex Christiansen
74-2 Stubberudlia BRL Tvetenveien 231 59 2 Frode Rinnan & Olav Tveten
75-2-1 Setra BRL Hovseterveien 66 80 3 Frode Rinnan & Olav Tveten
75-2-2 Setra BRL Hovseterveien 66 95 4 Frode Rinnan & Olav Tveten
75-3-1 Orebakken BRL Landingsveien 74-78 87 3 Preben Krag & Jens Selmer
75-3-2 Orebakken BRL Landingsveien 74-78 96 4 Preben Krag & Jens Selmer
76-1 Svarttjern BRL Odvar Solbergsvei 126-128 95 4 Alex Christiansen
76-3 Bjørnheim BRL Høybråtenveien 21 A, B, C 112 4(5) Klippgen, Holm & Halvorsen
77-1 Lindebergskogen BRL Lindebergveien 45 A, B 88 4 Selvaag-Bygg
77-2 Sandaker BRL Åsengata 2, 4, 6, 8 104 4 John Enghs arkitektkontor
77-3 Østre Lindeberg BRL Jerikoveien 57 125 4 Selvaag-Bygg

78-1 Lutvannkollen BRL Ole Reistadsvei 5 A, B, C 99 4 Frode Rinnan & Olav Tveten
80-1-1 Frydenlund BRL Pionerstien 8 82 3 Helge Thams & Per Gaare
80-1-2 Frydenlund BRL Pionerstien 8 99 4 Helge Thams & Per Gaare
80-2 Sloreåsen BRL Sloreåsen 16 92 4 Selvaag-Bygg
81-1 Valdresgata BRL Valdresgata 13 A-D 115 5 F. S. Platou A/S

81-3 Hoffsgrenda BRL Skøyen Terrasse 25, 26, 27 95 4 Selvaag-Bygg

82-1 Skansen BRL Dragonstien 5 A, B 84 3 Helge Thams & Per Gaare
82-2 Nordskrenten BRL Lusetjernveien 42, 44, 46 102 4 Selvaag-Bygg
82-3-1 Nedre Silkestrå BRL Nedre Silkestrå 24 98 4 Anker & Hølaas
82-3-2 Nedre Silkestrå BRL Nedre Silkestrå 24 89 3 Anker & Hølaas

83-1 Casinetto BRL Gustav Vigelandsvei 42, 44, 46 98 4 Telje- Torp- Aasen
83-2 Vestskrenten BRL Nordåssløyfa 14-16 102 4 Selvaag-Bygg
83-3 Verksgata BRL Fjellgata 2 A-E 97 4 Borgen & Bing Lorentzen
84-1 Øvre Ravnåsen BRL Ravnåsveien 71-79 66 2 P. J. Eriksen & B. E. Knutsen
84-2 Stolmakergata BRL Stolmakergata 9 78 3 Truls Thorenfeldt
Appenix A.4.1 Total sample of apartments Page 4/4
Number Name Address Apartment Architects
m2 rooms
84-3 Måltrostskogen BRL Arnulf Øverlandsvei 1-5 97 4 Selvaag-Bygg
85-2 Korsgata BRL Korsgata 12 B 77 3 Truls Thorenfeldt
87-1 Beverlia BRL Østmarkveien 62-102 76 3 Asplan prosjekt
87-2 Jordal BRL Totengata 12, 14, 16 76 3 Alf Bastiansen
88-1 Nordbygata 3-9-13 BoligsameieNordbygata 3, 9, 13 92 3 Anker & Hølaas
88-3 Tirillsletta BRL Høgåsveien 37-41 84 3 Klippgen, holm & Halvorsen
89-1 Smestaddammen BRL Hoffsveien 60 B 103 4 Helge Thams & Per Gaare
89-2 Lilleparken BRL Pilestredet 53 65 2(3) 4B Arkitekter v/ Henrik Poppe
89-3 Leirfallsgata BRL Leirfallsgata 9-11 81 3(4) Truls Thorenfeldt
90-1 Sportveien BRL Lysehagen 50 59 3 Arkitektskap

90-3 Nittedalsgata 1-7 BRL Nittedalsgata 1-7 64 2 Terjesen, Kjellstad, Horn

91-2 Sætrehjørnet BRL Østerdalsgata 2 A-C 79 3 Gasa
91-4 Liakollen BRL Høgåsveien 50-148 88 4 HRTB
92-1 Welhavensgate 10 BRL Welhavensgate 10-14 81 3 Fosse & Aasen
96-1 Wesselsgate 16 BRL Wesselsgate 16 77 3 Dahl & Myrhol

98-1 Pontoppidan BRL Pontoppidansgate 9B 70 3 Arcasa

98-2 Frydenlundsgata 5/7 BRL Frydenlundsgate 5-7 62 3 Thorenfeldt
00-1 Byhagen BRL Tøyengata 3-11 70 3 Dahl & Myrhol ANS
00-2-1 Gøteborggata BRL Gøtebotggata 31 B-C 59 2 Selvaagbygg AS
00-2-2 Gøteborggata BRL Gøtebotggata 31 B-C 109 4 Selvaagbygg AS
01-1-1 Jacob Meyer BRL Wesselsgate 15 57 2 Dahl & Myrhol ANS
01-1-2 Jacob Meyer BRL Wesselsgate 15 90 4 Dahl & Myrhol ANS
02-1 Nedre Steensgaarden BRL Stensgata 30 65 3 Thorenfeldt AS
02-2 Øvre Steensgaarden BRL Stensgate 34, 36 72 3 Thorenfeldt AS
03-1 Høgåsen BRL Høgåsveien 74-94 80 3 Spor Arkitekter AS
03-2 Hospitalhaven Boligsameie Pilestredet Park 25 62 3 Lund og Slaato / GASA
03-3 Lakkegården BRL Heimdalsgata 19-23 65 3 Arkitektskap AS
04-1-1 Vålen BRL Biskop Jens Nilsensgate 13-19 49 2 Arkitektskap AS
04-1-2 Vålen BRL Biskop Jens Nilsensgate 13-19 78 3 Arkitektskap AS
04-2-1 Hausmann BRL Hausmannsgate 10-14 42 2 Dyrvik Arkitekter AS

04-2-2 Hausmann BRL Hausmannsgate 10-14 64 3 Dyrvik Arkitekter AS

04-3 Marienlyst Park Boligsameie Gydas vei 16-28 80 3 Lund/Hagem Arkitekter AS
05-1 Freidigkollen BRL Gamle Bygdevei 168-204 80 3 Alex Christiansen AS
05-2 Sogn Terrasse BRL Klaus Torgårdsvei 10 A-B 82 3 Selvaagbygg AS
05-3-1 Karenlyst Plass Boligsameie Drammensveien 154 59 2 LPO og 4B Arkitekter
05-3-2 Karenlyst Plass Boligsameie Drammensveien 154 69 3 LPO og 4B Arkitekter
Appendix A.4.2 page 1/29
Appendix A.4.2 page 2/29
Appendix A.4.2 page 3/29
Appendix A.4.2 page 4/29
Appendix A.4.2 page 5/29
Appendix A.4.2 page 6/29
Appendix A.4.2 page 7/29
Appendix A.4.2 page 8/29
Appendix A.4.2 page 9/29
Appendix A.4.2 page 10/29
Appendix A.4.2 page 11/29
Appendix A.4.2 page 12/29
Appendix A.4.2 page 13/29
Appendix A.4.2 page 14/29
Appendix A.4.2 page 15/29
Appendix A.4.2 page 16/29
Appendix A.4.2 page 17/29
Appendix A.4.2 page 18/29
Appendix A.4.2 page 19/29
Appendix A.4.2 page 20/29
Appendix A.4.2 page 21/29
Appendix A.4.2 page 22/29
Appendix A.4.2 page 23/29
Appendix A.4.2 page 24/29
Appendix A.4.2 page 25/29
Appendix A.4.2 page 26/29
Appendix A.4.2 page 27/29
Appendix A.4.2 page28/29
Appendix A.4.2 page 29/29
A.4.3 Some room sequences, all apartments Page 1/3

E-K Bd1 - Bt/Wc Bd2 - Bt/Wc Bd3 - Bt/WC

31-1
32-1
33-1
33-2
36-1
36-2
36-3-1
36-3-2
36-4
38-1
39-1
39-2
40-1
40-2-1
40-2-2
40-3
41-1
41-2
41-3
41-4-1
41-4-2
47-1
47-2-1
47-2-2
48-1-1
48-1-2
48-2
48-4
48-5-1
48-5-2
48-7
48-8
49-1
49-2
50-1
50-2
50-3
51-1
51-2
51-3
52-1
52-2-1
52-2-2
52-3
53-1
53-2
53-3
53-4-1
53-4-2
54-1
54-2
54-3-1
54-3-2
55-1
55-2
55-3
56-1
A.4.3 Some room sequences, all apartments Page 2/3

E-K Bd1 - Bt/Wc Bd2 - Bt/Wc Bd3 - Bt/WC

56-2
57-1
57-2
57-3-1
57-3-2
58-1
58-2
59-1
59-2
59-3
60-1
60-2
60-3
61-1
61-2
61-3
62-1
62-2
62-3
63-1
63-2
64-1
64-2
64-3
65-1
66-1
66-2-1
66-2-2
66-3-1
66-3-2
67-1
67-2
67-3-1
67-3-2
68-1
68-2
68-3
69-1
69-2
69-3
70-1-1
70-1-2
70-2
71-1
71-2
71-3
72-1
72-2
72-3
73-1-1
73-1-2
73-2
74-1
74-2
75-2-1
75-2-2
75-3-1
A.4.3 Some room sequences, all apartments Page 3/3

E-K Bd1 - Bt/Wc Bd2 - Bt/Wc Bd3 - Bt/WC

75-3-2
76-1
76-3
77-1
77-2
77-3
78-1
80-1-1
80-1-2
80-2
81-1
81-3
82-1
82-2
82-3-1
82-3-2
83-1
83-2
83-3
84-1
84-2
84-3
85-2
87-1
87-2
88-1
88-3
89-1
89-2
89-3
90-1
90-3
91-2
91-4
92-1
96-1
98-1
98-2
00-1
00-2-1
00-2-2
01-1-1
01-1-2
02-1
02-2
03-1
03-2
03-3
04-1-1
04-1-2
04-2-1
04-2-2
04-3
05-1
05-2
05-3-1
05-3-2
Appendix A.4.4 Space Syntax Calculation, Typical Apartments, All Page 1/3

Case R L MD(E) CV CV Integration (by Relative Asymmetry, RA) Order (E-L-K-Bd1)

E L/D

31-1-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
32-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
33-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
33-2 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
36-1 1 C 1,16 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
36-2 A 1,16 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
36-3-1 1 C 1,33 3,00 0,75 0,00 0,33 0,33 0,50 0,50 0,00 0,33 0,33 0,50
36-4 A 1,66 1,20 0,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,40 0,40 0,40
38-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
39-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
39-2 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
40-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
40-2-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
40-3 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-1 A 1,00 4,00 0,25 0,00 0,50 0,50 0,50 0,50 0,00 0,50 0,50 0,50
41-2 1 C 1,16 2,83 1,70 0,10 0,25 0,28 0,28 0,28 0,35 0,50 0,53 0,53 0,10 0,25 0,28 0,35
41-3 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
41-4-1 1 C 1,20 2,83 1,75 0,10 0,20 0,30 0,60 0,60 0,80 0,10 0,20 0,30 0,60
47-1 1 C 1,28 3,75 2,70 0,09 0,14 0,23 0,38 0,38 0,38 0,42 0,42 0,09 0,14 0,23 0,42
47-2-2 1 C 1,20 2,75 2,75 0,13 0,13 0,26 0,46 0,46 0,46 0,46 0,13 0,13 0,26 0,46
48-1-2 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,28 0,38
48-2 A 1,28 4,00 1,20 0,09 0,28 0,28 0,38 0,38 0,38 0,57 0,57 0,09 0,28 0,38 0,38
48-4 1 C 1,20 3,00 0,75 0,10 0,30 0,30 0,50 0,50 0,50 0,10 0,30 0,30 0,50
48-5-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
48-7 1 C 1,28 3,75 2,70 0,09 0,14 0,23 0,38 0,38 0,38 0,42 0,42 0,09 0,14 0,23 0,38
48-8 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
49-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
49-2 1 C 1,28 3,75 2,70 0,09 0,14 0,23 0,38 0,38 0,38 0,42 0,42 0,09 0,14 0,23 0,38
50-1 A 1,14 5,50 1,16 0,04 0,23 0,33 0,33 0,33 0,33 0,33 0,52 0,04 0,23 0,33 0,33
50-2 1 c 1,42 2,83 1,75 0,14 0,19 0,28 0,33 0,38 0,42 0,42 0,47 0,14 0,19 0,28 0,42
50-3 1x A 1,28 4,00 1,20 0,09 0,28 0,28 0,38 0,38 0,38 0,57 0,57 0,09 0,28 0,38 0,38
51-1 A 1,25 2,50 1,33 0,16 0,33 0,66 0,66 0,83 0,16 0,33 0,66
51-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
51-3 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
52-1 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
52-2-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
52-3 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
53-1 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
53-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
53-3 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
53-4-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
54-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
54-2 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
54-3-1 1 A 2,83 0,25 1,25 0,12 0,16 0,19 0,25 0,28 0,28 0,30 0,33 0,33 0,33 0,36 0,36 0,46 0,16 0,19 0,25 0,33
55-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
55-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
55-3 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
56-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
56-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
57-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
57-2 1 C 1,20 2,83 1,75 0,10 0,20 0,30 0,50 0,50 0,60 0,10 0,20 0,30 0,50
57-3-1 B 1,50 1,33 2,50 0,16 0,33 0,66 0,66 0,83 0,16 0,33 0,66
58-1 A 1,55 4,25 2,70 0,13 0,13 0,30 0,36 0,36 0,36 0,36 0,36 0,36 0,52 0,13 0,30 0,36 0,36
58-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
59-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
59-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
59-3 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
Appendix A.4.2 Space Syntax Calculation, Typical Apartments, All Page 2/3

Case R L MD(E) CV CV Integration (by Relative Asymmetry, RA) Order (E-L-K-Bd1)

E L/D

60-1 1 C 1,33 2,75 2,75 0,13 0,13 0,26 0,46 0,46 0,46 0,46 0,13 0,13 0,26 0,46
60-2 1 C 1,33 2,66 1,58 0,13 0,20 0,22 0,46 0,46 0,53 0,53 0,13 0,20 0,22 0,53
60-3 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
61-1 A 1,16 4,50 1,20 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,26 0,40 0,40
61-2 A 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
61-3 1 C 1,14 4,83 1,66 0,04 0,19 0,23 0,33 0,33 0,33 0,33 0,47 0,04 0,19 0,23 0,33
62-1 1 C 1,33 2,75 2,75 0,13 0,13 0,26 0,46 0,46 0,46 0,46 0,13 0,13 0,26 0,46
62-2 A 1,25 5,33 2,16 0,07 0,17 0,32 0,32 0,32 0,32 0,32 0,42 0,42 0,07 0,17 0,32 0,42
62-3 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
63-1 A 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
63-2 1 C 1,33 2,75 2,75 0,13 0,13 0,26 0,46 0,46 0,46 0,46 0,14 0,32 0,35 0,39
64-1 1 C 2,25 0,20 2,70 0,10 0,14 0,28 0,32 0,35 0,35 0,35 0,39 0,39 0,14 0,32 0,35 0,39
64-2 B 1,60 1,25 3,50 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
64-3 1 C 1,33 4,75 2,66 0,08 0,13 0,25 0,30 0,30 0,30 0,30 0,30 0,36 0,36 0,08 0,13 0,25 0,36
65-1 1 C 1,16 3,83 1,70 0,06 0,20 0,26 0,40 0,40 0,40 0,53 0,06 0,20 0,26 0,40
66-1 1 C 1,28 3,66 1,53 0,09 0,19 0,19 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,19 0,38
66-2-1 1 C 1,42 2,75 2,75 0,14 0,14 0,33 0,33 0,42 0,42 0,42 0,42 0,14 0,14 0,33 0,42
66-2-2 B 1,50 2,25 3,33 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
66-3-1 1 C 1,25 4,83 1,66 0,07 0,17 0,25 0,32 0,32 0,32 0,32 0,35 0,42 0,07 0,17 0,32 0,35
67-1 B 2,00 1,25 2,83 0,14 0,23 0,33 0,42 0,42 0,52 0,52 0,61 0,23 0,33 0,42 0,52
67-2 B 1,42 3,25 3,35 0,14 0,14 0,42 0,42 0,42 0,42 0,42 0,42 0,14 0,14 0,42 0,42
67-3-1 1 C 1,25 4,83 1,66 0,07 0,17 0,25 0,32 0,32 0,32 0,32 0,35 0,42 0,07 0,17 0,32 0,35
68-1 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
68-2 1 C 1,14 4,83 1,66 0,04 0,19 0,23 0,33 0,33 0,33 0,33 0,47 0,04 0,19 0,23 0,33
68-3 1 C 1,33 4,75 2,66 0.08 0,13 0,25 0,30 0,30 0,30 0,30 0,30 0,36 0,36 0.08 0,13 0,30 0,36
69-1 1 C 1,50 1,83 1,83 0,20 0,20 0,33 0,33 0,40 0,53 0,53 0,20 0,20 0,33 0,53
69-2 1 C 1,28 3,83 1,70 0,09 0,19 0,28 0,38 0,38 0,38 0,38 0,47 0,09 0,19 0,38 0,38
69-3 B 1,42 3,25 3,25 0,14 0,14 0,42 0,42 0,42 0,42 0,14 0,14 0,42 0,42
70-1-1 B 1,60 1,25 3,50 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,50
70-2 B 1,40 2,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,20 0,60 0,60
71-1 1 C 1,81 3,53 3,03 0,12 0,16 0,20 0,23 0,30 0,30 0,34 0,34 0,34 0,38 0,38 0,56 0,16 0,20 0,23 0,34
71-2 1 C 1,25 4,83 1,66 0,07 0,17 0,25 0,32 0,32 0,32 0,32 0,35 0,42 0,07 0,17 0,32 0,35
71-3 1 C 1,75 1,75 2,66 0,14 0,21 0,21 0,32 0,32 0,39 0,39 0,46 0,46 0,14 0,21 0,21 0,32
72-1 1 C 1,33 4,75 2,66 0,08 0,13 0,25 0,30 0,30 0,30 0,30 0,30 0,36 0,36 0,08 0,13 0,25 0,36
72-2 2 C 1,37 2,75 2,70 0,10 0,14 0,28 0,32 0,32 0,32 0,35 0,39 0,39 0,10 0,14 0,32 0,39
72-3 1 C 1,37 3,66 1,70 0,10 0,21 0,21 0,32 0,35 0,35 0,35 0,46 0,46 0,10 0,21 0,32 0,35
73-1-2 1 C 1,37 3,66 1,70 0,10 0,21 0,21 0,32 0,35 0,35 0,35 0,46 0,46 0,10 0,21 0,32 0,35
73-2 1 C 1,75 1,75 2,66 0,14 0,21 0,21 0,32 0,32 0,39 0,39 0,46 0,46 0,21 0,14 0,32 0,39
74-1 1 C 2,00 2,16 1,58 0,18 0,21 0,22 0,24 0,24 0,28 0,34 0,37 0,37 0,40 0,45 0,45 0,54 0,18 0,21 0,24 0,34
74-2 B 1,40 2,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,20 0,60 0,60
75-2-1 B 1,57 1,33 2,33 0,19 0,28 0,38 0,38 0,57 0,57 0,66 0,66 0,19 0,28 0,38 0,57
75-3-1 B 1,62 3,25 2,75 0,14 0,17 0,32 0,39 0,39 0,42 0,42 0,42 0,57 0,14 0,17 0,32 0,39
76-1 1 C 1,60 3,58 1,53 0,13 0,15 0,20 0,22 0,33 0,33 0,33 0,35 0,35 0,40 0,42 0,13 0,20 0,22 0,33
76-3 1 C 1,90 0,41 2,33 0,18 0,18 0,21 0,34 0,36 0,36 0,36 0,36 0,36 0,36 0,40 0,52 0,18 0,21 0,34 0,36
77-1 1 C 1,80 2,66 1,75 0,17 0,22 0,22 0,31 0,31 0,37 0,37 0,37 0,37 0,42 0,42 0,17 0,22 0,22 0,31
77-2 1 C 1,55 2,50 2,75 0,13 0,19 0,19 0,25 0,36 0,36 0,41 0,41 0,41 0,41 0,13 0,19 0,25 0,41
77-3 A 2,40 1,20 1,20 0,15 0,22 0,31 0,31 0,33 0,35 0,35 0,51 0,51 0,53 0,53 0,22 0,31 0,31 0,35
78-1 B 2,11 1,20 2,20 0,11 0,22 0,27 0,27 0,33 0,33 0,44 0,44 0,50 0,50 0,22 0,27 0,33 0,44
80-1-1 B 1,28 4,33 2,20 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
80-2 1 C 1,60 3,66 1,70 0,13 0,20 0,20 0,31 0,31 0,33 0,33 0,33 0,37 0,40 0,40 0,13 0,20 0,31 0,33
81-1 2 C 1,63 1,95 2,25 0,12 0,14 0,18 0,23 0,25 0,27 0,30 0,32 0,32 0,32 0,32 0,36 0,12 0,18 0,32 0,32
81-3 1 C 1,60 3,66 1,70 0,13 0,20 0,20 0,31 0,31 0,33 0,33 0,33 0,37 0,40 0,40 0,13 0,20 0,20 0,31
82-1 B 1,28 4,33 2,20 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
82-2 1 C 1,90 2,66 1,75 0,18 0,20 0,23 0,25 0,30 0,32 0,34 0,36 0,36 0,38 0,41 0,43 0,18 0,20 0,23 0,30
82-3-2 1 C 1,28 3,75 2,70 0,09 0,14 0,23 0,38 0,38 0,38 0,42 0,42 0,09 0,14 0,23 0,42
83-1 B 2,44 0,25 3,16 0,08 0,13 0,30 0,30 0,30 0,30 0,30 0,36 0,36 0,36 0,13 0,30 0,36 0,36
83-2 1 C 1,90 2,66 1,75 0,18 0,20 0,23 0,25 0,30 0,32 0,34 0,36 0,36 0,38 0,41 0,43 0,18 0,20 0,23 0,30
83-3 C 1,77 2,75 2,08 0,13 0,19 0,25 0,30 0,36 0,36 0,41 0,41 0,47 0,47 0,13 0,19 0,25 0,30
Appendix A.4.2 Space Syntax Calculation, Typical Apartments, All Page 3/3

Case R L MD(E) CV CV Integration (by Relative Asymmetry, RA) Order (E-L-K-Bd1)

E L/D

84-1 1 C 1,60 0,75 3,00 0,10 0,30 0,30 0,50 0,50 0,50 0,10 0,30 0,30 0,50
84-2 B 2,33 0,20 2,20 0,11 0,22 0,22 0,33 0,33 0,33 0,44 0,44 0,44 0,44 0,22 0,33 0,33 0,44
84-3 1 C 1,60 3,66 1,70 0,13 0,20 0,20 0,31 0,31 0,33 0,33 0,33 0,37 0,40 0,40 0,13 0,20 0,20 0,31
85-2 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
87-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
87-2 B 1,28 4,33 2,20 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
88-1 A 1,85 1,20 1,20 0,09 0,28 0,28 0,38 0,38 0,38 0,57 0,57 0,28 0,28 0,38 0,57
88-3 B 1,28 4,33 2,20 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
89-1 1 C 1,44 3,25 2,70 0,11 0,16 0,27 0,27 0,33 0,33 0,33 0,38 0,38 0,50 0,11 0,16 0,27 0,33
89-3 B 1,28 4,33 2,22 0,09 0,19 0,38 0,38 0,38 0,38 0,47 0,47 0,09 0,19 0,38 0,47
90-1 1 C 1,66 0,70 4,00 0,06 0,26 0,26 0,40 0,40 0,40 0,46 0,06 0,26 0,26 0,40
90-3 B 1,60 2,40 0,83 0,30 0,30 0,50 0,70 0,70 0,90 0,30 0,30 0,50 0,70
91-2 B 1,50 2,25 3,33 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
91-4 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
92-1 A 1,28 4,00 1,20 0,09 0,28 0,28 0,38 0,38 0,38 0,57 0,57 0,09 0,28 0,28 0,38
96-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
98-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,53
98-2 B 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50 0,70
00-1 B 1,50 3,50 0,75 0,20 0,26 0,46 0,53 0,53 0,53 0,80 0,20 0,26 0,46 0,53
00-2-2 A 1,90 2,16 1,16 0,08 0,20 0,24 0,24 0,28 0,28 0,28 0,40 0,40 0,44 0,44 0,20 0,24 0,24 0,28
01-1-1 B 1,50 1,33 2,50 0,16 0,33 0,66 0,66 0,83 0,16 0,33 0,66
02-1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50
02-2 B 1,40 2,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,20 0,60
03-1 B 1,42 2,83 2,25 0,14 0,23 0,33 0,42 0,42 0,52 0,52 0,61 0,14 0,23 0,33
03-2 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50
03-3 1 A 1,20 3,50 1,25 0,10 0,30 0,50 0,50 0,50 0,70 0,10 0,30 0,50
04-1-1 A 1,25 2,50 1,33 0,16 0,33 0,66 0,66 0,83 0,16 0,33 0,66
04-2-1 B 1,50 1,33 2,50 0,16 0,33 0,66 0,66 0,83 0,16 0,33 0,66
04-3 B 2,00 0,20 4,50 0,06 0,26 0,40 0,40 0,40 0,40 0,60 0,06 0,40 0,40
05-1 B 1,33 3,33 2,25 0,13 0,20 0,46 0,46 0,46 0,53 0,53 0,13 0,20 0,46 0,46
05-2 A 1,33 3,00 1,25 0,13 0,33 0,33 0,46 0,46 0,66 0,66 0,13 0,33 0,33 0,46
05-3-1 B 2,20 0,33 2,33 0,20 0,20 0,60 0,60 0,60 0,60 0,20 0,60 0,60

Legend

R Rings (see section 3.4.7)

L Living room's kind of space
MD(E) Mean depth from entrance
CV Control Value

E Entrance
L Living room
D Dining room

bedroom 1 / main bedroom

other bedrooms
kitchen
living room, dining room (or non-spesified room for daytime living)
bathroom
WC, washing room
entance/corridor/hall
internal staircase
balcony
(white) stortage room (with through-passage)
0DULGDOVYHLHQ%5/ 
$GUHVVHJQUEQU  0DULGDOVYHLHQ,OD
$UNLWHNWW  (LQDU(QJHOVWDG
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
$6)DJHUKHLPHQ 
$GUHVVHJQUEQU  &KULVWLHVJDWH)DJHUKHLPJDWD' OHQHQJD
$UNLWHNWW  +HQULN1LVVHQ *XQQDU%U\QQLQJ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
%¡NNHUO¡NND%5/ 
$GUHVVHJQUEQU  %URFNPDQQVJDWH%M¡OVHQ
$UNLWHNWW  *XQQDU*UHJDDUG-¡UJHQ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
$NHUVEDNNHQ%5/ 
$GUHVVHJQUEQU  )UHGHULNNH4YDPVJDWH,OD*DPOH$NHU
$UNLWHNWW  (GJDU6PLWK%HUHQWVHQ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
+RYLQ%5/ 
$GUHVVHJQUEQU  6W-¡UJHQVYHL+RYLQ
$UNLWHNWW  5LQQDQ7YHWHQ&ROEM¡UQVHQ
/HLOLJKHWVIRUGHOLQJ  K\EHOOHLOLJKHWHU
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
9RVVHO¡NND%5/ 
$GUHVVHJQUEQU  9RVVHJDWD %7RUVKRY/LOOHERUJ
$UNLWHNWW  61DUYH/XGYLJVHQ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
2UUHP\U%5/ 
$GUHVVHJQUEQU  2GYDU6ROEHUJVYHL5RPVnV
$UNLWHNWW  $OH[&KULVWLDQVHQ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP


6YDUWWMHUQ%5/ 
$GUHVVHJQUEQU  2GYDU6ROEHUJVYHLEORNN 5RPVnV
$UNLWHNWW  $OH[&KULVWLDQVHQ5RPVnVWHDPHW
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
6DQGDNHU%5/ 
$GUHVVHJQUEQU  cVHQJDWD$% &6DQGDNHU
$UNLWHNWW  -RKQ(QJKVDUNLWHNWNRQWRU
/HLOLJKHWVIRUGHOLQJ  HQNHOWURPSUEORNN
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
&DVLQHWWR%5/ 
$GUHVVHJQUEQU  *XVWDY9LJHODQGVYHL 6N¡\HQ
$UNLWHNWW  7HOMH7RUS$DVHQ
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
)U\GHQOXQGVJDWH%5/ 
$GUHVVHJQUEQU  )U\GHQOXQGVJDWH%LVOHWW6W+DQVKDXJHQ
$UNLWHNWW  7KRUHQIHOGW
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP
%\KDJHQ%5/ 
$GUHVVHJQUEQU  7¡\HQJDWD1RUGE\JDWD*U¡QODQG
$UNLWHNWW  'DKO 0\UKRO$16
/HLOLJKHWVIRUGHOLQJ  
7HJQHWE\JJHPHOGW  
7\SLVNOHLOLJKHW%5$  URPP