Imperatives Key Topics in Semantics and

Pragmatics Jary M. pdf download

https://ebookgate.com/product/imperatives-key-topics-in-semantics-and-pragmatics-jary-m/

★★★★★ 4.9/5.0 (45 reviews) ✓ 116 downloads ■ TOP RATED

"Amazing book, clear text and perfect formatting!" - John R.

DOWNLOAD EBOOK
 Imperatives Key Topics in Semantics and Pragmatics Jary M.
pdf download

TEXTBOOK EBOOK EBOOK GATE

Available Formats

■ PDF eBook Study Guide TextBook

EXCLUSIVE 2025 EDUCATIONAL COLLECTION - LIMITED TIME

INSTANT DOWNLOAD VIEW LIBRARY

Instant digital products (PDF, ePub, MOBI) available
Download now and explore formats that suit you...

Key Topics in Urology Key Topics Series BIOS 1st Edition

M. Underwood

https://ebookgate.com/product/key-topics-in-urology-key-topics-series-
bios-1st-edition-m-underwood/

ebookgate.com

Current Trends in Diachronic Semantics and Pragmatics Maj

https://ebookgate.com/product/current-trends-in-diachronic-semantics-
and-pragmatics-maj/

ebookgate.com

Key Terms in Semantics 1st Edition M. Lynne Murphy

https://ebookgate.com/product/key-terms-in-semantics-1st-edition-m-
lynne-murphy/

ebookgate.com

Where Semantics meets Pragmatics Volume 16 Current

Research in the Semantics Pragmatics Interface 1st Edition
Klaus Von Heusinger
https://ebookgate.com/product/where-semantics-meets-pragmatics-
volume-16-current-research-in-the-semantics-pragmatics-interface-1st-
edition-klaus-von-heusinger/
ebookgate.com
Key Terms in Pragmatics 1st Edition Nicholas Allott

https://ebookgate.com/product/key-terms-in-pragmatics-1st-edition-
nicholas-allott/

ebookgate.com

Little Words Their History Phonology Syntax Semantics

Pragmatics and Acquisition Ronald P. Leow

https://ebookgate.com/product/little-words-their-history-phonology-
syntax-semantics-pragmatics-and-acquisition-ronald-p-leow/

ebookgate.com

Key Topics in Otolaryngology 2nd Edition N. J. Roland

https://ebookgate.com/product/key-topics-in-otolaryngology-2nd-
edition-n-j-roland/

ebookgate.com

Key Topics in Sports Medicine Kindle Edition Lynn B.

https://ebookgate.com/product/key-topics-in-sports-medicine-kindle-
edition-lynn-b/

ebookgate.com

Key Topics in Landscape Ecology 1st Edition Jianguo Wu

https://ebookgate.com/product/key-topics-in-landscape-ecology-1st-
edition-jianguo-wu/

ebookgate.com
Imperatives

Imperative sentences usually occur in speech acts such as orders,

requests and pleas. However, they are also used to give advice, and to
grant permission, and are sometimes found in advertisements, good
wishes and conditional constructions. Yet, the relationship between
the form of imperatives, and the wide range of speech acts in which
they occur, remains unclear, as do the ways in which semantic
theory should handle imperatives.
This book is the first to look systematically at both the data and the
theory. Part I discusses data from a large set of languages, including
many outside the Indo-European family, and analyses in detail the
range of uses to which imperatives are put, paying particular atten-
tion to controversial cases. This provides the empirical background
for Part II, where the authors offer an accessible, comprehensive and
in-depth discussion of the major theoretical accounts of imperative
semantics and pragmatics.

m a r k j a r y is Reader in English Language and Linguistics at the

University of Roehampton. He has written widely on semantics,
pragmatics and philosophy of language and is the author of Assertion
(2010).
m i k h a i l k i s s i n e is Assistant Professor of Linguistics at the
Université Libre de Bruxelles. He has written widely on semantics,
pragmatics and philosophy of language and is the author of From
Utterances to Speech Acts (Cambridge University Press, 2013).
KEY TOPICS IN SEMANTICS AND PRAGMATICS

‘Key Topics in Semantics and Pragmatics’ focuses on the main topics

of study in semantics and pragmatics today. It consists of accessible
yet challenging accounts of the most important issues, concepts and
phenomena to consider when examining meaning in language. Some
topics have been the subject of semantic and pragmatic study for many
years, and are re-examined in this series in light of new developments in
the field; others are issues of growing importance that have not so far been
given a sustained treatment. Written by leading experts and designed to
bridge the gap between textbooks and primary literature, the books in this
series can either be used on courses and seminars, or as one-stop, succinct
guides to a particular topic for individual students and researchers. Each
book includes useful suggestions for further reading, discussion
questions, and a helpful glossary.

Already published in the series:

Meaning and Humour by Andrew Goatly
Metaphor by L. David Ritchie
Imperatives by Mark Jary and Mikhail Kissine

Forthcoming titles:
The Semantics of Counting by Susan Rothstein
Modification by Marcin Morzycki
Game-Theoretic Pragmatics by Anton Benz
The Semantics of Counting by Susan Rothstein
Pragmatics and the Philosophy of Language by Mitchell Green
Distributivity by George Tsoulas and Eytan Zweig
Imperatives
MARK JARY AND
RITCHIE MIKHAIL KISSINE
University Printing House, Cambridge CB2 8BS, United Kingdom

Published in the United States of America by Cambridge University Press, New York

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of

education, learning and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107632356

# Mark Jary and Mikhail Kissine 2014

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.

First published 2014

Printed in the United Kingdom by MPG Printgroup Ltd, Cambridge

A catalogue record for this publication is available from the British Library

ISBN 978-1-107-01234-9 Hardback

ISBN 978-1-107-63235-6 Paperback

Cambridge University Press has no responsibility for the persistence or accuracy of

URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
Contents

Acknowledgements vi
List of abbreviations and notation vii

Introduction 1

Part I The data 7

1 What is the imperative mood? 9
2 The imperative mood and directive force 53
3 Imperatives with conditional meanings 110
Part II The theories 163
Introduction to Part II: Transition: from data to theory 163
4 The imperative is directive force 168
5 Declarative-like semantics for imperatives 212
6 The imperative as a distinct semantic type 258
An opinionated conclusion 292

Appendix A Possible worlds and semantics 294

Appendix B Modality in possible-word semantics 297
Appendix C Stalnaker’s common-ground model of assertion 303
Glossary 305
References 307
Index 320

v
Acknowledgements

We would like to thank Manuel Garcia-Carpintero, François Recanati,

Rob Stainton, and Gregory Bochner for detailed comments and criticisms
that helped us shape the final version of the book. Helen Barton, the
commissioning editor for linguistics at CUP has been encouraging and
very helpful at every stage.
We also acknowledge the joint British Council/Communauté Française
de Belgique grant that enabled the initial planning of this book. Mark
Jary also benefited from an AHRC Early Careers Fellowship.
Finally, we thank our families for their support.

vi
Abbreviations and notation

#: pragmatic unacceptability
?: semantic unacceptability
*: syntactic unacceptability
1: first person
2: second person
3: third person
A: addressee
ABL.: ablative case
ACC.: accusative case
ACT.: active voice
ACTN: action
ADE.: adessive
AUX.: auxiliary
CAUS.: causative
CG: common ground
CLASS.: classifier
COND.: conditional
CONV.: converb
DAT.: dative case
DET.: determinant
DIM.: diminutive
DIR.: directional aspect
DUAL: dual number
EXCL.: exclusive
FEM.: feminine gender
FUT.: future tense

vii
viii Abbreviations and notation
GEN.: genitive case
ILCs: imperative-like conditionals
IMP.: imperative mood
IMPFV.: imperfective
INCL.: inclusive
IND.: indicative mood
INF.: infinitive
LOC.: locative case
MAN.: manner
MOD.: modal
NEG.: negation
NET.: neutral gender
NON-PAST: non-past tense
NP: nominal phrase
OPT.: optative mood
PART.: particle
PARTP.: participle
PASS.: passive voice
PAST: past tense
PERM.: permission
PFV.: perfective
PL.: plural
PR.: pronoun
PRS.: present tense
PRTV.: partitive
REFL.: reflexive
RT: relevance theory
S: speaker
SBJV.: subjunctive mood
SG.: singular
SPCF.: specific
ST: state
TEMP.: temporal
VP: verbal phrase
Introduction

If language is a toolbox, then imperative sentences are what we reach

for when we want to leave our hearer in no doubt that we want him to
do something and what we want him to do. This is not to say that this
is the only thing they are good for, nor that there is nothing else in the
toolbox that will do the job, but simply to state that imperatives appear
in some way designed for the job of getting the hearer to do something.
We’ll have plenty to say in Chapter 1 about just what an imperative
sentence is, but, for the moment, we can make do with a few examples:

(1) Stand by your man.

(2) Mind the gap.

(3) Love thy neighbour.

(4) Do not go gentle into that good night.

(5) Please try to be good.

(6) Somebody call a doctor!

(7) Do help yourself to more tea, vicar.

(8) Don’t anyone move.

What these examples show is that we take imperatives to be of a

certain morpho-syntactic type. That a sentence can be used to get
someone to do something does not suffice to make it an imperative.
That is why we do not count (9) and (10) as imperatives. Again, we’ll
have much more to say on this in Chapter 1, where we’ll also examine
imperatives in languages other than English.

1
2 imperatives
(9) All children must be accompanied by both parents.

(10) Can you pass the salt, please?

For now, we’ll just say that the imperative is a morpho-syntactic type
and note that what counts as an instance of that type in a particular
language is determined by its function. That is to say, when we want to
identify an imperative form in a particular language, we look for a
form that typically has the function of getting someone to do some-
thing. Importantly, what we don’t do is look for a form that has
particular formal properties (although, as we will see, across the
world’s languages, imperatives do tend to display a number of common
formal properties).
In this book, we are concerned with the meaning of imperatives. For
the moment, we can think of this in terms of the following question:
Why is it that imperatives are good for getting people to do things? We
make the assumption that a full understanding of the meaning of
imperatives requires an account of what imperatives encode and how
this interacts with considerations such as the situation of utterance
and the speaker’s intentions, such that imperatives come to have the
communicative and practical significance that typifies them. In other
words, we are concerned with the semantics and pragmatics of the
imperative form, with how its encoded meaning interacts with the
principles that govern its use.
From a semantic/pragmatic point of view, imperatives are interesting
for a number of reasons. One is that, although they are built from the
same stuff as declaratives, they exhibit very different ‘semantic’ char-
acteristics. Most notably, imperatives cannot be judged true or false.
Given that most approaches to semantic theory analyse meaning in
terms of truth conditions, it is not obvious how imperatives should be
analysed. As we will see in Part II, a range of approaches has been
suggested in order to overcome this obstacle. These include reducing
imperatives to declaratives, postulating a common propositional core
shared by declaratives and imperatives and denying that all meaning
needs to be analysed in terms of truth conditions.
Another reason that imperatives are of particular interest is that
they typify the doing side of the contrast that was highlighted by
Austin when he attacked the dominant tendency in philosophy of
language to focus exclusively on the use of words to say things. Lan-
guage, Austin famously pointed out, is also good for doing things.
Imperatives are good for ordering, commanding, begging, advising,
requesting and many more speech (or ‘illocutionary’) acts that don’t
Introduction 3
appear to involve making a statement. Given the distinction between
semantics and pragmatics, this raises the question of how much of this
‘doing’ needs to be accounted for by what is encoded by the form, and
how much needs to be left to the pragmatics. Again, various positions
on this are taken in the literature, and we review the most significant
in Part II of the book.
A major aim of the book, then, is to provide the reader with an
overview of the theories that have been put forward to explain the
idiosyncratic semantic and pragmatic characteristics of the imperative.
However, before we can do this we need to establish just what it is that
such theories must account for. Part I of the book is devoted to this
task, as we set out to identify the functional characteristics that define
the imperative and to list the uses to which it can be put. In doing so,
we give a great deal of consideration to the typological data on impera-
tives. This is because, generally, accounts of the semantics and
pragmatics of the imperative assume, either tacitly or explicitly, a high
degree of universality. Consequently, it is important to establish just
what such theories have to explain, in particular in terms of any cross-
linguistic variation that might be identified.
As the function that identifies imperatives is their ability to issue
directives such as commands and requests, we also spend a good deal
of time considering just how the notion of a directive illocutionary act
should be defined. We then consider the extent to which attested uses of
the imperative can be considered directives. Of particular interest in this
respect is the apparent occurrence of the imperative in conditional-like
constructions such as ‘Catch a cold and you won’t be able to train for
a week’. Here, the apparent imperative ‘Catch a cold’ can hardly be
considered an attempt to get the hearer to catch a cold. Such data thus
present an important challenge for theorists, as they either have to
deny that such constructions contain imperatives, or come up with an
account that explains both their directive and conditional-like uses. It
is important, then, to establish whether the apparent imperatives in
these conditional-like sentences are indeed imperatives, and this is a
question to which we devote considerable time.
We have written the book with two main audiences in mind. First,
we see the book as providing a way into the data and literature for
scholars who are not familiar with the field. These may be established
scholars from other areas of linguistics, scholars from cognate
fields such as philosophy of language, cognitive science and psych-
ology, or students looking for a bridge between textbooks and the
primary literature. Our second – but by no means subsidiary – aim is
that the book make a contribution to the field, as we provide a critical
4 imperatives
review of the key literature to date. We do not see these goals as in
conflict, but complementary. Writing for students and non-specialists
forces us to fully explicate assumptions made by the theories
we discuss, which can only benefit our critical agenda. That said, we
encourage students, in particular, to adopt the same critical stance that
we bring to the literature, when considering the criticisms we level at
the theories we discuss. Established scholars, we trust, will need no
such encouragement.
This book can also be seen as part of a larger project, in that we have
our own ideas about what form a theory of the imperative should take.
We want to develop a theory that is philosophically defensible, psycho-
logically plausible and semantically tractable, and we intend to publish
the results of our endeavours in this respect in the near future. The
book also serves, then, to lay the ground for that project.
The book is divided into two parts. Part I (Chapters 1–3) deals
with issues that relate to identifying the imperative, while Part II
(Chapters 4–6) looks at attempts to explain its semantics and pragmat-
ics. Chapter 1 is concerned with defining the term ‘imperative’. It begins
by discussing issues associated with the word ‘imperative’ and its range
of uses, then examines issues relating to directness and indirectness. It
also considers the bearing of the distinction between verbal and senten-
tial mood on the problem of defining the term ‘imperative’ as we intend
to use it in the book. This chapter also considers what should be said of
languages that do not have an imperative form, and discusses patterns
that arise when one examines which forms either replace the imperative
or are employed to complete a defective imperative paradigm.
Chapter 2 looks in detail both at the question of how best directive
force should be analysed and at the semantic characteristics of
the imperative. The analysis of directive force becomes a crucial issue
after, in Chapter 1, we define the imperative form in terms of directive
force. If we are to evaluate the theories of the imperative in terms of
their ability to explain its typical directive force, then we must be clear
as to exactly what we mean by ‘directive’. But imperatives demonstrate
other ‘meaning characteristics’ besides their association with directive
force. For example, they appear to have a close relationship with
addressee agency, and to be restricted to describing what the speaker
takes to be non-actual yet potential states of affairs. To be sure, these
features might be said to be a result of the imperative’s association
with directive force, but, equally, they might be argued to be independ-
ent features that permit that association. Consequently, in the second
part of Chapter 2, we examine these characteristics as semantic
features of the imperative in their own right.
Introduction 5
Part I ends, in Chapter 3, with a discussion of apparent conditional-
like uses of the imperative, such as the ‘Catch a cold you won’t be able
to train for a week’ example we mentioned above. It is important to
establish whether these in fact contain imperative sentences as their
first conjunct, because, if imperatives are found in these constructions,
then they serve as crucial data for those seeking to identify the encoded
meaning of the imperative. The fact that many conditional-like uses
are non-directive might well provide us with an environment in which
the imperative can be divorced from its typical directive use, and we
might thereby gain insight into how it comes to serve its directive
function. Moreover, if these constructions do contain imperatives, then
an important feature of any adequate theory of imperative semantics is
that it is able to explain this conditional-like use.
Part II begins in Chapter 4 with a discussion of theories that tie the
imperative very closely to directive force. These include those such as
Searle and Vanderveken (1985) and Han (2000), who argue that impera-
tive encodes directive force. Others, such as Barker (2004) and Mastop
(2005) go further and argue that the imperative and directive force
cannot be pulled apart. For these authors, imperatives, rather than
encoding directive force as an element of meaning, simply are directives.
In Chapter 5 we consider the polar opposite to the views put forward
by Barker and Mastop when we examine claims that the imperative is
best analysed as having a declarative-like semantics, or even as being
reducible to the declarative. Such accounts take a variety of forms.
Davidson (2001a) wants to analyse the imperative as consisting of two
declaratives, one being about the other, while Kaufmann (2012) ana-
lyses the imperative mood as a modal, that is, as something like a
covert ‘must’-sentence. Others want to analyse imperatives in terms
of compliance conditions, with these defined in terms of truth condi-
tions for a ‘declarative core’ argued to reside inside the imperative.
Chapter 6 ends the book with a discussion of theories that seek
neither to reduce the encoded meaning imperative to a declarative
core, nor to relate it closely with directive force. Rather, these theorists
argue either that the imperative should be seen as being related to the
presentation of a proposition with a particular psychological attitude,
or that it is a distinct semantic type altogether. In the first group we
find the Grice-inspired pragmaticists Wilson and Sperber (1988) and
Bach and Harnish (1979), while in the second we find both formalists
such as Huntley (1984) and the more traditional grammarian approach
of Davies (1986).
We should add here that in our grouping of theories and theorists,
we do not claim to have carved the field at its joints. Rather, the field
6 imperatives
has none, and we have tried to reflect this by highlighting similarities
in approaches across theoretical divisions, rather than grouping
theorists according to the traditions in which they operate. Thus, we
have resisted the temptation, for example, to lump ‘formal semantic’
accounts together in one chapter, preferring to look at the fundamen-
tals that underlie distinct formally rendered treatments. We have
attempted to keep our discussion as theory and formalism neutral as
possible, to make the book accessible to scholars and students from
a range of fields. However, to facilitate our discussion of formal
treatments, we have included three appendices that readers less than
familiar with these approaches may find useful. These deal, respect-
ively, with model–theoretic semantics, possible world semantics and
Stalnaker’s common-ground treatment of assertion. There is also a
short glossary. The book ends with a conclusion in which we look back
at the ground covered, and forward to the way ahead.
PART I
The data

7
1 What is the imperative mood?

This book is about the semantics and pragmatics of imperatives. That is

to say, it is about the interaction of linguistic meaning and contextual
factors, including speaker intentions, in the interpretation of utter-
ances of a particular type of linguistic form. To undertake to write
such a book is to presuppose that the term ‘imperative’ picks out a
distinct linguistic type with a meaning consistent across different
instantiations, such that interesting things can be said about its contri-
bution to utterance interpretation. In this chapter, we justify this
assumption by showing that, cross-linguistically, the ‘imperative form’
can be identified by virtue of its function in communication. For now,
we will just say that this function is to signal the performance of
directive speech acts such as commands, orders, requests and pleas.
However, as we will see in later chapters, the uses to which the type of
linguistic form we are investigating can be put goes beyond this narrow
range. Moreover, we will also see that identifying the form by virtue of
this function in no way commits us, nor any other theorist, to claiming
that this function is encoded by that form.
In any interesting sense of ‘encode’, if a form encodes a function,
then it does more than merely indicate that that function is its most
prototypical use. Rather, if a form encodes a function, then no literal
and serious use of that form is possible without its performing the
function at hand, so that comprehension of that form is nothing more
than relating it to its typical function. This point is very important, as
we will see again and again that one of the central issues about the
imperative is whether or not every literal use of it necessarily corres-
ponds to the performance of a directive speech act.
Before we can look in detail at the variety of ways the imperative
manifests itself across languages, we need first to consider the ways in
which the term ‘imperative’ has been used, so that we can be explicit
and precise in our use of the term. We address this issue at the start of
section 1, before going on to discuss a series of complications that arise
when one seeks to provide a functional definition of the imperative.

9
10 the data
The definition we arrive at is expressed in terms of a covariance
between prototypical function and form. Such a functional definition
takes the imperative to be a sentence type: that is, an instance of
sentential, rather than verbal, mood. And we thus define the impera-
tive as a sentence type whose sole prototypical function is to perform
directive speech acts.
A consequence of this definition, as we will see in section 2, is that
imperative may be realised in a variety of ways across the world’s
languages. In this section, we will consider the variety of ways the
imperative can be realised, paying special attention to languages where
a morphologically marked imperative extends to the first and third
person, as well as to the ways the imperative may be negated. This
survey will inform our discussion of English, a language which has
been claimed by some not to have an imperative.
We will argue that English does have an imperative. But our defin-
ition of the imperative as a form whose sole prototypical function is to
perform directive speech acts does leave open the possibility that there
are languages that do not have this sentence type. We consider some
such cases in section 3, paying particular attention to associated meth-
odological issues, before going on, in section 4, to examine which forms
commonly replace the imperative, also asking what properties they
have in common with it.

1 DEFINING ‘IMPERATIVE’

1.1 Sentence types

Put yourself in the position of a linguist describing a ‘new language’,
i.e. one that has never before been studied by linguists. (This is, by
no means, a thought-experiment: around 90 per cent of languages
spoken today across the world still await description; see, for
instance, Levinson and Evans (2009).) In giving the meaning of the
basic lexical vocabulary of that language, you will try to relate words
to objects in the world. When it comes to describing some recurrent
and specific sentence forms in that language, though, you will not
try to relate these to objects in the world, but rather to things that
speakers do when speaking, such as stating, requesting and asking.
Just as we have a tendency to think of the meaning of words in
terms of what they relate to in the world, so we tend to think of the
meaning of kinds of sentences in terms of acts performed in speak-
ing, or moves made in ‘language games’. That is, what one does in
What is the imperative mood? 11
attempting to delineate the different kinds of sentences in a certain
language is to uncover systematic relationships between linguistic
structures and conversational functions. Indeed, in their classic
paper, Sadock and Zwicky define the notion of a sentence type pre-
cisely as ‘a coincidence of grammatical structure and conventional
conversational use’ (1985: 155).
But, in practice, how is the linguist to delineate the sentence types of
a language? Sadock and Zwicky provide the following criteria:

First, the sentence types of a language form a system, in at least two

senses: there are sets of corresponding sentences, the members of
which differ only in belonging to different types, and second, the types
are mutually exclusive, no sentence being simultaneously of two
different types. (1985: 158–159, original emphasis)

What this definition says is that a sentence type is a one-to-one associ-

ation between a form and a function. That is, two sentences that differ
in their function instantiate two different sentence types if, and only if,
they present a morpho-syntactic difference that correlates with that
difference in function.
For instance, the English sentences (1)–(3) seem to have a distinct
form (declarative, interrogative and imperative), and a different proto-
typical function (to state a fact, to ask a question and to give a com-
mand). Other than these differences, they are identical in terms of
content: they are all, in a very vague and intuitive sense, ‘about’
children who play outside.

(1) Children play outside.

(2) Do children play outside?

(3) Play outside, children!

Furthermore, no English sentence seems to be, at the same time, say,

declarative and interrogative, or declarative and imperative, or inter-
rogative and imperative. (We say ‘seems’ because we will see shortly
that matters are actually not that simple.)
An immediate consequence of this definition is that the repertoire of
sentence types varies language from language. Some languages have an
‘optative’ mood, whose main function is to express good or evil wishes.
Albanian has an optative verbal inflection, distinct from indicative,
imperative and subjunctive moods; here is an illustration with the verb
çoj ‘send’ (Breu 2010):
12 the data
(4) çofsh
send-2SG.PR.ACT.OPT.
çon
send-2SG.PR.ACT.IND.
ço
send-2SG.ACT.IMP.
të çosh
send-2SG.PR.SBJV.

Yet, in many other languages, there is no specific sentence type with

such a prototypical function – wishes can be expressed, for instance, by
structural means that also allow the performance of commands
(imperative mood) or the expression of hypothetical propositions
(‘subjunctive’ or ‘irrealis’). (In Dobrushina’s et al. (2005) sample of 319
languages, 271 do not have an inflectional optative.)1 To give another
example, in English, warnings are performed using declarative or
imperative constructions:

(5) You’d better be careful.

(6) Watch your wallet.

But in other languages there exist ‘preventive’ forms, specifically dedi-

cated to warnings: for instance, in Aleut (Eskimo–Aleut family) the
second person has a ‘preventive’ morphological marker, distinct from
(negated) indicative or imperative forms (Golovko 2001).
Because of this variety of repertoires, Sadock and Zwicky distin-
guish between ‘minor’ and ‘major’ sentence types. The former include
the two examples just given, and are subject to great cross-linguistic
variation. The latter are: declarative, interrogative and imperative,
and correspond to three fundamental functions of language use: pro-
viding information, getting information and provoking action. Thus,
theorists concerned with the meaning of imperatives generally start
from the observation that languages tend to have means of distin-
guishing assertions, questions and commands, and identify the
imperative as the form specified for the last of these. Once we move
on from this basic observation, however, things can get quite messy
quite quickly.

1
There is, of course, a close relationship between optative and imperative moods.
For instance, the imperative in Slavic languages originates from Indo-European
optative forms: Sussex and Cubberley (2006).
What is the imperative mood? 13
1.2 Ambiguity in the term ‘imperative’
In the first page of her wonderful book on the English imperative, Davies
(1986: 1) notes that the term ‘imperative’ has been used in the literature
to refer both to a linguistic form and a conversational function, a bit like
when we use the word ‘hammer’ to denote both the tool and the act of
using that tool to force a nail into a piece of wood. The ambiguity in
‘hammer’ is less worrying though, both because the former meaning is
distinguished from the latter by its grammatical class (the first is a noun,
the second a verb), and because we are not seeking to use the term in the
description and explanation of complex phenomena.2
Linguists, at least descriptively minded ones, generally use the term
‘imperative’ to refer to a linguistic form.3 Among philosophers – and
certain, more formally oriented, linguists – though, this cannot be
taken for granted, and the term ‘imperative’ is often used to denote a
function of language. Furthermore, ‘imperative’ is employed both
within and outside the study of language, and there is a danger that
the latter use could colour the former. We are all familiar with the
term ‘a moral imperative’, which refers to a course of action that must
be followed because it is the right thing to do, such as (7). Note, though,
that this moral imperative is expressed by an imperative sentence. In
calling (7) an imperative, then, we need to be clear in which sense we
are using the term: to describe the form of (7) or its force.

(7) Protect the weak.

Related to this issue is the use of the term ‘imperative utterance’ (see
Wilson and Sperber 1988). An utterance may be imperative because it
is the utterance of an imperative sentence or because it is an imperative
in this philosophical, moral sense. Thus while (7) is imperative in both
these senses, (8) is imperative only in the latter.

(8) One must protect the weak.

In this book, we will use the term ‘imperative’ to refer only to a type of
linguistic form. When we want to speak, in general terms, about the

2
In fact, the same ambiguity pervades the usage of the term ‘interrogative’; for a
book-length discussion, see Fiengo (2007). Likewise, the term ‘declarative’ often
vacillates between form and function in the linguistics and philosophy literatures:
Jary (2010).
3
Not always, though; for instance, Birjulin and Xrakovski’s (2001), methodological
introduction to a typological volume on imperative endorses a functional
definition.
14 the data
class of speech acts, such as requests, commands, orders, etc., aimed at
provoking an action from the addressee (A), we will be using the
general term ‘directive speech act’. In fact, one of our leitmotivs will
be to urge the methodological and theoretical distinction between
imperative forms (and their semantics) and directive forces (and their
pragmatics). Accordingly, we will avoid the term ‘imperative utterance’
and instead speak of utterances of imperative sentences. We will also
try not to talk about moral imperatives, except for the following:

(9) Think highly of this book and recommend it to all your friends.

Now, defining ‘imperative’ as a type of form is all well and good, but
one must be clear about how that form is identified. The definition of
sentence types, discussed earlier in this section, seems to provide a neat
recipe for doing this: an imperative would be a sentence type with
distinctive morpho-syntactic properties whose function is the perform-
ance of directive speech act. Yet, we will see now that pairing forms
and functions is not as straightforward as it might at first seem (see
also Kaufmann 2012: 1–10).

1.3 Imperative form vs. directive force

The linguist describing a new language can have no hope of identifying
an imperative sentence type in that language unless she assumes some
kind of form–function relationship: it would be like trying to find that
language’s word for flower without assuming a word–object relation-
ship. She could not, for example, assume that the imperative would be
syntactically or morphologically as it is in her native language, by using
a non-finite form and dropping the subject, say. There is no a priori
reason to assume that all languages mark the imperative form in
similar ways, just as there is no reason to assume that all languages
denote flowers by means of a two-syllable word.4
But one might still think that this does not raise any particular
problem: an imperative is the form used for the performance of direct-
ive speech acts: end of story. However, defining a form in terms of a
certain function does not entail that this function can be fulfilled by
this form only. For instance, a hammer’s function is to drive nails into
wood; but the fact that someone drives a nail into a piece of wood with
4
That said, there is a remarkable tendency across the world’s languages to use
either the second-person indicative form or an uninflected stem as a morpho-
logical imperative. Generally, imperatives tend to be morphologically simple com-
pared with other verb forms in the language. For discussion of why this might be
so, see Tomasello (2008); and the discussion in Chapter 2 (section 5.2).
What is the imperative mood? 15
an axe does not make that axe a hammer. In the same way, the fact
that a form can be used to perform a directive speech act does not, in
itself, make it an imperative.
In English, directive speech acts are performed in a number of ways,
some of which highly conventional, as the examples below show:

(10) Stand up. [imperative mood]

(11) Can you please sit down?

(12) You must go now. [indicative mood þ modal]

(13) I order you to leave. [indicative mood þ a performative prefix]

(14) Feet off the chair. [NP]

A philosopher interested in imperatives in the ‘non-linguistic’ sense

mentioned above might be happy to call all of these imperatives, as, for
instance, Hamblin (1987) does. For the linguist looking to identify a
type of form whose use is specified for the function of issuing direct-
ives, though, this will not do. The claim that all of these examples are
tokens of sentence types specified for directive force can be countered
by showing that there are other interpretations readily available. For
example, remove please from (11) and imagine it spoken by a doctor to a
patient with a bad back: now this example can be understood as a
question about ability. Furthermore, the observation that a directive
such as (11) can be responded to both as a question and as a request (‘Of
course I can; but I choose not to, thank you’) shows that it is an
instance of a sentence type specified for questioning. The case against
counting (12) as an imperative would be that it can be responded to as
one could respond to an assertion, by denying its truth (‘That’s not
true: it’s not time yet’), which would marry well with the observation
that (12) has a morpho-syntactic structure, namely declarative mood,
that is associated with the making of assertions. (However, as we will
see in Chapter 5 (section 1.3), the relationship between modals like
must and the imperative mood is more complex than that.) Against
counting (13) as an imperative sentence type, it could be argued that
the imperative force is conveyed lexically, not syntactically, as substi-
tuting ‘order you to leave’ with another verb phrase may change the
speech act that the utterance constitutes (e.g. to ‘I promise to leave’). As
for (14), one could argue that it is elliptical as much for ‘Get your feet
off the chair’ as for ‘I want you to get your feet off the chair’.
16 the data
What makes (most) linguists happy to call (10) an imperative is the
observation that this form demonstrates a robustness in its association
with directive force that is lacking in the other cases. Changing the
verb, for example, does not seem to affect force, and little scene-setting,
if any, is needed for this form to receive a directive interpretation.
Indeed, on encountering an instance of this sentence type in a neutral
context (to the extent that there can be such a thing), the natural
inclination is to interpret it as directive. Directive force is thus the
prototypical interpretation of this sentence type. (This is not to say that
it is the only interpretation: as we will see in the next chapter, a range
of non-directive interpretations have been claimed for the imperative.)
Now, you might object that we have as yet failed to show that the
examples in (11)–(14) are not instances of the imperative sentence type.
It could be claimed that some or all of them are ambiguous between
two different sentence types: imperative and another one. It is worth
spelling out precisely what such a claim would amount to, as the issue
will resurface several times in this book. Recall that above (section 1.1),
we identified sentence types as a bi-univocal coincidence between a
morpho-syntactic form and a function. Accordingly, if one claims that
a certain sentence is ambiguous between two sentence types, what one
claims, in fact, is that this sentence can be interpreted as having (at
least) two different and mutually exclusive morpho-syntactic struc-
tures. Morpho-syntactic ambiguity, per se, is not extraordinary. Take
for instance (15) which displays two forms of the French verb eat in the
present tense: the third-person singular and the third-person plural.

(15) Il mange. / Ils mangent.

he eat-IND.PR.3PL. / they-MASC. eat- IND.PR.3PL.

As you see, orthographic representation distinguishes these forms very

clearly. However, in most dialects of French both forms are pro-
nounced as [il.mɑ̃ʒ]. That is, this sentence form remains ambiguous
between two mutually exclusive morpho-syntactic structures, unless
written representation or context assists in disambiguation. This is the
kind of hypothesis one would make in claiming that (11)–(14)
are ambiguous: the sentence form in question would be said to have
two different and mutually exclusive morpho-syntactic structures,
one identical to that of unambiguous imperatives, such as (10), and
another one.
Such a claim is not as outlandish as it may at first seem. Take the
interrogative in (11). In English, such forms are very common and
conventionalised ways of performing directive speech acts. In fact, they
What is the imperative mood? 17
are so conventionalised that Sadock (1974) claims they are indeed
ambiguous between two sentence types: the imperative and the inter-
rogative. Now, mere intuition does not suffice for positing ambiguity:
one also needs independent arguments, and while, to the best of our
knowledge, nowadays this view has been discarded, it is still instructive
to examine Sadock’s rationale – and the reason why it fails – in more
detail.
Of course, Sadock did not hold that any interrogative sentence
form is ambiguous. Rather, he argued that ambiguity between sen-
tence types can be highlighted by certain formal properties. It is a
property of English imperative sentences, according to Sadock (1974:
97–109), that they can be followed by please or by an indefinite
vocative:

(16) Answer the phone, please.

(17) Answer the phone, someone.

Interestingly, he pointed out, interrogatives that are conventionally

used as requests display similar properties:

(18) Will you please open the door?

(19) Will you open the door, someone?

(20) Can you please open the door?

(21) Can you open the door, someone?

But of course, not all interrogatives do. One can imagine contexts in
which (22) is used as a request to open the door. However, even under
such an interpretation please and someone seem to be less felicitous, as
revealed by (23)–(24).

(22) When will you open the door?

#
(23) When will you please open the door?5

#
(24) When will you open the door, someone?

5
In marking examples such as (23)–(24) for pragmatic, rather than semantic or
syntactic, unacceptability, we are, of course, prejudging the theoretical issue.
18 the data
The difference of acceptability between (18)–(21) and (23)–(24), reveals,
according to Sadock, that sentences like (25)–(26) are ambiguous
between the imperative and interrogative sentence types, while sen-
tences like (27)–(28) unambiguously belong to the interrogative type.

(25) Can you open the door?

(26) Will you open the door?

(27) When will you open the door?

(28) Are you able to open the door?

An immediate consequence is that Sadock has to accept that some

sentences that walk and talk like imperatives are not genuinely
imperative.

(29) Be glad we’re not leaving.

#
(30) Be glad we’re not leaving, please.

#
(31) Be glad we’re not leaving, someone.

However, claiming that (29) is not an imperative seems quite counter-

intuitive, and one better have very strong reasons for thinking so. The
problem is that the acceptability or otherwise of please and someone is
not a good reason for categorising a form as imperative. As noted by,
for instance Bach and Harnish (1979: 200–202), Levinson (1983: 266) or
Davies (1986: 21), please and someone are also acceptable with sentences
which are clearly not imperative. Consider the following examples,
given by Davies:

(32) I’d appreciate it if you would please be quiet.

(33) The phone is ringing, someone.

(34) Where are my slippers? Someone?

The request performed in (32) is something like Be quiet, so if (32) were

ambiguous between imperative and declarative sentence types, and if
please were disambiguated in favour of the former, this is what its
morpho-syntactic structure should more or less correspond to. But, of
What is the imperative mood? 19
course, there is no sensible way to argue that (32) has a structure
similar to Be quiet. Likewise, there is no reasonable way to derive the
imperative structure Answer the phone from (33), nor Bring me my slippers
from (34).
It seems, then, that the acceptability of please or of someone in English
does not depend on the sentence type, namely on its being imperative,
but on its function, namely on whether the utterance is a request or
not. This is why please and someone can be adjoined to declaratives and
interrogatives in (32)–(34), and this is also why it is unacceptable in
imperative sentences which, like (29), do not constitute a request,
properly speaking.
A methodological point which will prove important later on emerges
from the foregoing discussion. When attempting to identify the
imperative sentence type, one needs to single out distinctive formal
characteristics, and in doing that, what one typically pays attention to
is the form of sentences used in directive speech acts. However – and
this is what we just saw about the use of please and someone in English –
one must be careful not to mistake a hallmark of a particular (sub-)type
of directive force with a property of the imperative sentence type.
Now, while the foregoing discussion should make it clear that
interrogatives, like (11), or declaratives, like (12) or (13) do not belong
to the imperative sentence type, matters are more delicate for sub-
sentential utterances, such as (14). The issue depends on a debate
largely independent of our concerns here, and which bears on the
exact status of such utterances. The first camp, whose most promin-
ent member is Stainton (e.g. 1998, 1997, 2005, 2006; Stainton and
Elugardo 2004), holds that sub-sentential utterances are also sub-
sentential at the level of the syntactic structure. Roughly speaking,
the idea is that syntax gives you nothing more than what the eye can
see: this sub-sentential fragment provides a template on which to
build context-dependent, communicated meaning. On such a view,
an utterance like Feet off the table would be said not to belong to a
specific sentence type; context supplies information necessary to
interpret it as a directive speech act.6 The second camp holds, by
contrast, that examples like (14) only look sub-sentential, but, in
reality, they display a fully-fledged sentential syntactic structure (e.g.
Stanley 2002; Merchant 2004; Ludlow 2005). According to proponents
of this position, (14) is not different from (35). Even though,

6
Of course, please is acceptable with sub-sententials, as in Feet off the chair, please, but,
as we have just seen, this is not evidence that they are derived from an underlying
imperative.
20 the data
superficially, the second clause of (35) does not contain a verb, its
underlying syntactic structure does.

(35) John studies chemistry, and Mary astrophysics.

If this view turned out to be correct, then we would have to accept that
Feet off the table is ambiguous between Put your feet off the table and, say,
Your feet are off the table. If so, such cases would display genuine ambi-
guity between different sentence types. Even though we ourselves are
more sympathetic with the first, ‘under-determination’ approach to
sub-sententials, we will not take sides on this issue here.

1.4 Indirectness
In the previous subsection, we saw that non-imperative sentences can
be used to perform directive speech acts. Such directives are often said
to be ‘indirect’. Since we will find ourselves discussing this notion later
on, it is worth spending a few paragraphs here on just what is meant by
this term.
The classic notion of indirectness is that developed by Searle (1975a),
according to which a speech act is indirect if, and only if, its perform-
ance requires the performance of another speech act. For instance, a
request to close the window performed by way of the utterance of (36)
is indirect because its performance requires performing an assertion
that it is really cold in here.

(36) It’s really cold in here.

However, we have already seen that there are highly conventionalised

(Morgan 1978; or ‘standardised’, see Bach and Harnish 1979: 192–202;
Bach 1998) indirect speech acts, such as the interrogatives in (37)–(38).

(37) Could you open the door?

(38) Will you shut up?

It is far from obvious that in such cases, the directive speech act is
necessarily performed by asking the addressee about his abilities to
close the door, or his opinion on whether or not he will shut up (see
Terkourafi 2009: and references within). However, there is no problem
with referring to these cases as ‘indirect’, since the ‘direct’ questioning
meaning remains readily accessible (for instance, one can comply with
(37) answering, Sure I can).
What is the imperative mood? 21
However, some scholars disagree that an order like (39) should be
classified as indirect (Recanati 1987: 167; Jary 2010: 72–73; Kissine
2013: 111–122).

(39) You will stay in your room for the rest of the evening.

There are two reasons: first, it is very plausible that no other speech act
than the directive one is performed by (39) and, second, the content
literally expressed by (39) is exactly the same as the proposition that
has to be true for the corresponding imperative Stay in your room for the
rest of the evening to be obeyed. Nevertheless, in what follows, we will
ignore this issue and use ‘indirect speech act’ whenever a sentence type
is not used for the performance of the kind of speech act it is prototyp-
ically associated with. It is not that we believe this is the correct
definition of indirectness (as a matter of fact, we do not). However,
we will be discussing indirect speech acts below only in so far as this
notion is employed by authors whose essential thesis is that sentence
types encode a certain force. Recall that this view (different incarnations
of which will be discussed in the next chapter (section 3) and in
Chapter 4) commits its proponents to holding that every literal use of
an imperative sentence results in the performance of a directive speech
act, that any utterance of an interrogative sentence is also always,
taken literally, a question, and that an utterance of a declarative
sentence is always, taken literally, an assertion. Accordingly, anyone
who endorses this position is confronted with a dilemma when sen-
tences like (39) are used to command. The first option is to concede that
the utterance remains, literally, an assertion, and, consequently, pre-
suppose such commands are always indirect. The second option con-
sists in denying that the command in (39) is indirect, but, on the
hypothesis that sentence types encode illocutionary forces, this
requires assuming that it amounts to a non-literal use of a declarative
sentence. However, in this second case, one would have to explain in
what way it is non-literal: certainly, (39) does not appear either figura-
tive or ironic.

1.5 Imperative mood: verbal vs. sentential

We now have made clear what we understand by ‘imperative’: a
sentence type whose prototypical function is the performance of dir-
ective speech acts. We have also made clear how this conception differs
from the more philosophically loaded notion of the imperative. How-
ever, in linguistics, and in descriptive grammars, ‘imperative’ is also
used as the name of a verbal mood. We will see now that, even though
22 the data
it is compatible with our definition of the imperative, this notion is too
narrow. This discussion will also provide a clearer idea of exactly how
imperatives, qua sentence types, are identified.
The term ‘mood’ itself is used to distinguish both sentence types and
a class of verbal inflections. Sentence mood distinguishes sentence
types in terms of their perceived communicative function. Thus our
hypothetical linguist describing a new language by looking for form–
function pairs is investigating the sentence–mood system of that lan-
guage. The main terms employed to describe the sentence moods are
the same as those of the major sentence types: ‘declarative’, ‘interroga-
tive’ and ‘imperative’, to which one could perhaps add an ‘exclamative’
sentence mood, having the function of making an exclamation, as in
the following example:

(40) How children play!

Just as there exist ‘minor’ sentence types along with these major ones,
other form–function pairs, such as the optative, discussed earlier on,
are generally not afforded the same centrality when investigating this
aspect of language.
As Harnish (1994: 409) notes, the sentential notion of mood ‘treats
mood basically as a cluster of phonological, syntactic and semantic
properties of sentences’. A historically prior notion of mood, however,
relates only to verbal inflection. Terms employed to distinguish moods
in this sense include ‘indicative’, ‘subjunctive’, ‘infinitive’ and ‘impera-
tive’. Because ‘imperative’ is used in both sentential and verbal charac-
terisations of mood, there is again plenty of scope for confusion.7 On
the verbal understanding of ‘imperative’, a language has an imperative
only if it morphologically distinguishes sentential mood through
verbal inflection.
On this understanding of mood, English does not have an imperative,
for there is no inflection that specifies a verb for use in a clause uttered
to issue a directive. To put it more simply, there is no verb that has a
distinctively imperative form. Instead, English employs the bare stem,
which makes it morphologically indistinguishable from the first-
person singular and second-person indicative present tense, and from
the infinitive. (The verb be is an exception in that its ‘imperative’ form
corresponds to the infinitive, which is, of course, different from the

7
Likewise, ‘indicative’ and ‘declarative’ are sometimes used interchangeably to
refer to a type of sentential mood. We will use ‘indicative’ to refer to the verbal
mood, and ‘declarative’ for the sentential mood.
What is the imperative mood? 23
present indicative are or is forms.) This is in contrast to languages such
as, for instance, Bulgarian and Irish. As you can see from (41), the
imperative form in Bulgarian is morphologically marked by a distinct
suffix (Tosheva 2006; Lindstedt 2010).8

(41) razkazvam9 tell-INF.

razkazvaj tell-IMP.2SG.
razkazvajte tell-IMP.2PL.
razvaš tell-IND.2SG.
razvate tell-IND.2PL.

In Irish too the sentential imperative mood is indicated by the presence

of a specific verbal form that is not used in any other context (Ó Baoill
2010). Note, that while the Bulgarian imperative is marked (for most
verbs) by a specific suffix, the Irish second-person-singular imperative
form corresponds to the bare verbal stem. Thus, the imperative forms
of ól (to drink) are ól (2SG.) and ólaigí (2PL.), and those of ordaigh (to
order) are ordaigh (2SG.) and ordaigi (2PL.). However, the bare verbal
stem can only be used in the same environments as the second-person
plural imperative form (in other words, the infinitive form appears, as
such, in no other syntactic environment). Therefore, the bare verbal
stem is aptly characterised as an imperative form (morphologically
marked through zero-affixation).
So, should we say that, unlike Irish, English lacks a genuine impera-
tive mood? As we will see later (section 2.3), this question is by no
means easy to answer. But, for the time being, the crucial point is that
if English does have an imperative mood, this mood has to be realised
sententially, that is, singled out by a combination of morpho-syntactic
and functional considerations.
Leaving English aside for the moment, there are compelling reasons
to think that characterising the imperative as being exclusively a
verbal mood would be a misleading and arbitrary move. There exists
a great number of human languages, often called ‘isolating’, whose
grammatical systems lack verbal (and nominal) inflection altogether.
In such languages there is no affixation with a morphological or, at
least, an inflectional function – in other words, affixes do not combine
phonological dependency with a precise semantic or grammatical func-
tion. Crudely speaking, in isolating languages, distinctions like person,
8
There is allomorphic variation, conditioned by the verbal stem.
9
When quoting examples from languages which do not employ Latin orthographic
system, we maintain, throughout the book, the transcription conventions used by
the source cited. In order to facilitate readablity, we have also omitted morpho-
logical glosses whenever we felt that they were irrelevant to our main point.
24 the data
number, gender, tense, aspect or mood are not marked by modifying
the phonological form of the noun or of the verb, but rather by word
order and/or by functional words/particles. The important point here is
that if verbal and sentential moods were equated, such languages
would automatically count as lacking imperative mood. Yet, some
isolating languages do have constructions specifically associated with
the performance of directive speech acts in spite of lacking verbal
inflection. For instance, in Canela-Krahô, an Amazonian language
spoken in the great central plateau of Brazil, neither tense nor the
imperative mood are morphologically marked. However, the omission
of the syntactic subject is prototypically associated with directive force;
moreover, it is the only environment where the subject is absent. Thus
while (42) will be glossed as the directive ‘Kill the deer’, the translation
of (43) would be ‘You cut/are cutting the wood’ (Popjes and Popjes
1986).

(42) po cura
deer kill

(43) ca pĩ jitep
2SG. wood cut

Therefore, even though Canela-Krahô has no imperative morphology

proper, the fact that a distinct form is devoted to directive speech acts
warrants the claim that there exists a sentential imperative mood in
this language.
We have seen, then, that while there is a close relation between
sentential mood and verbal mood so that, in many languages, senten-
tial moods are determined by verbal moods, the two notions of mood
need to be distinguished, because sentence types – sentential mood –
can be marked by other morpho-syntactic devices. Furthermore, there
is a more general difference between sentential mood and verbal mood
in that the former is defined in terms of a form–function pairing, while
verbal mood is not: it is not generally thought that the subjunctive or
the infinitive can be identified in terms of a form–function pairing in
the way that, the declarative, imperative and interrogatives can.
That said, however, we cannot ignore verbal mood altogether. We
will repeatedly see that when there is a gap in a language’s imperative
paradigm, it is often filled by a specific verbal inflection. For instance,
whereas a language may have a distinct form for the second-person-
singular imperative, the subjunctive may be used as a polite form, as in
Spanish, or it may be found in the negative imperative (sometimes
called ‘prohibitive mood’).
What is the imperative mood? 25
A further reason why verbal mood is important for better under-
standing imperative (sentential) mood is that when speech acts per-
formed using sentential-mood tokens are reported, there is often a
systematic relationship with the verbal mood employed in the object
clause of the reporting sentence. In English, an infinitive clause is used
in (44) to report (45), while in Spanish a subjunctive is used in (47) to
report the imperative (46).

(44) Sit down.

(45) He says to sit down.

(46) Siéntate.
sit-IMP.2SG.REFL.
‘Sit down’

(47) Dice que me siente.

say-IND.3SG. that REFL. sit-SBJV.1SG.
‘He tells me to sit down.’

In what follows, the terms ‘mood’, ‘the imperative mood’ or just ‘the
imperative’, will mean mood in the sentential sense. If we speak about
mood in the verbal sense, we will make this explicit. For us, the
imperative in any language will be the form – if there is one – that is
prototypically and productively used for the full range of directive
speech acts. To reiterate, this is not to say that it is limited to that
function, but rather that if it is found in a ‘neutral context’, then the
most readily available interpretation is that a directive is being issued.
Furthermore, to characterise a form as ‘imperative’ in our sense, it is
not sufficient for it to be prototypically employed to issue directive
speech acts: it is also necessary that there be no other function –
speech-act type – with which it is prototypically associated.
Our definition requires that any form that is correctly called ‘impera-
tive’ be productive in its directive use, so that its most accessible
interpretation is directive regardless of propositional content. This
rules out classifying interrogatives of the ‘Can you pass the salt?’-type,
discussed above, as imperative on the grounds that changing the prop-
ositional content makes a directive interpretation less likely. For
example, the most likely interpretation of ‘Can you swim?’ is as a
question about ability, rather than a request. So while we agree that
some interrogative constructions may have come to a stage where they
are idiomatically associated with requesting things, the association of a
26 the data
sentence type with directive force should not be dependent on any
specific wording nor on it expressing a specific propositional content,
especially if this relates to the passing of salt!

2 IMPERATIVES: MORPHO-SYNTACTIC VARIETY

In this section we will review the variety of morpho-syntactic struc-

tures that are prototypically associated with directive speech acts
across world languages. The purpose of this review is twofold: first,
we want to give the reader a taste of the diversity and complexity of the
relevant data; second, this section will later serve as a measure against
which to evaluate the adequacy of some of the theories discussed in
Part II of the book (see, especially, Chapter 6, section 1.3).

2.1 What is a complete paradigm?

Until now, we have discussed exclusively second-person imperative
forms. Imperatives, we have said, are those forms that are prototypic-
ally used for the performance of directive speech acts. In the next
chapter we will discuss how directive speech acts should be defined
in detail, but, roughly, in issuing a directive speech act, one instructs
someone to do something. Now, very often the person who is pre-
scribed the action is the addressee, or, in case of collective prescription,
the addressees. Accordingly, the sentence type specialised for this
function is the second-person imperative, singular or plural. However,
these combinations are by no means the only that are possible. For
instance, the agent(s) of the action may not be the addressee(s) of the
utterance, but someone else, the speaker or a third party. Or, when the
action at hand is a collective one, the speaker herself may be included
within the agents.10 In fact, once one singles out the protagonists
involved in a directive speech act – the speaker of the utterance, the
addressee of the utterance, and the agent of the prescribed action –
many more combinations become logically possible (seventeen, in fact,
see Birjulin and Xrakovski 2001).
So, should the term ‘imperative’ be limited to the second persons?
This is the option often endorsed in the literature. When the agent of
the prescribed action is not the addressee, the terms hortative,
10
Another distinction, sometimes marked at the morpho-syntactic level is between
first-person-plural inclusive and exclusive forms, which relate to whether or not
the addressee belongs to the group which includes the speaker and which is
prescribed to perform the action. This contrast is found for instance in
Malčukov (2001).
What is the imperative mood? 27
exhortative or jussive are often used. In what follows, we will, somewhat
arbitrarily, use the term hortative to denote such uses.
Problems of analysis arise when hortative forms are formally homo-
geneous with imperative ones. In French, the second-person imperative
is marked by the use of the indicative or subjunctive form and the
absence of an overt subject.11 A crucial difference, apart from the
absence of the subject, is that while in indicative and subjunctive
constructions clitics precede the verb, they come after the imperative
forms. For instance, while the clitic pronoun le (¼ it/him) precedes the
subjunctive (48) and the indicative (50) forms, in the imperative con-
structions (49) and (51) it is attached immediately after the verb.

(48) Que tu le sois.

that you-SG. it be-SBJV.PR.2SG.

(49) Sois-le.
be-IMPV.2SG. it

(50) Tu le fais.
you-SG. it do-IND.PR.2SG.

(51) Fais-le.
do- IMPV.2SG. it

Interestingly, in French, the constructions prototypically used to

perform a directive speech act directed at a group that includes the
speaker are formally similar to the second-person imperatives.
Unlike the subjunctive (53) or indicative (55) constructions, the
forms in (52) and (54) are characterised by the omission of the
syntactic subject, and by the fact that clitics must be attached after
the verb, that is, by exactly the same properties as the second-person
imperatives (49) and (51).

(52) Soyons-le.
do- IMPV.1 PL.it

(53) Que nous le soyons.

that we it be-SBJV.PR.1PL.

11
Two verbs, savoir (to know) and vouloir (to want), have specific plural imperative
inflected forms.
28 the data
(54) Faisons-le.
do- IMPV.1PL. it

(55) Nous le faisons.

we it do-IND.PR.1PL.

That the first-person plural forms in (52) and (54) really are formally
homogeneous with the second-person imperative forms is confirmed
by the fact that they are unacceptable with a (non-vocative) syntactic
subject and with proclitics.

(56) *Nous faisons-le.

we do-IMPV.1PL. it

(57) * Le faisons.
it do-IMPV.1PL.

The situation in French contrasts with a language like Armenian

(Kozintseva 2001). Armenian has specific second-person imperative
forms: for the singular, depending on the class of the verb, the suffix
-ir, -a or -u are added to aorist or to the infinitive stem, while for the
plural -ek is added to the aorist stem. However, when the requested
action is to be performed by a group including the speaker, Armenian
makes recourse to the first-person-plural subjunctive form.
If the term ‘imperative’ were restricted to the second person, the
contrast between French and Armenian would be blurred. In French,
the same sentence type is prototypical both for directives addressed at a
second person and for those addressed at a group which includes the
speaker, while in Armenian, the imperative sentence type does not
extend beyond the second person. It is thus reasonable to conclude
that while in French the imperative sentence type is realised at the
first-person plural, it is not the case in Armenian.
Consider now the data from Nivkh (an isolate spoken around the low
reaches of the Amur river and on Sakhalin Island), presented in
Gruzdeva (2001). In Nivkh declarative sentences, the person is not
marked on the verb; that is, verbs do not agree in person with their
subject (although the plural number may be optionally marked). How-
ever, in imperative sentences, verbal person agreement is compulsory –
and not only with the second person, but also with the first and the
third.

(58) Suffixes of the imperative paradigm in Nivkh (Amur dialect)

(Gruzdeva 2001: 64)
What is the imperative mood? 29
1SG. -nytka/-nyxta
2SG. -ja/-j/-jej
3SG. -gâzo
1DUAL -nyte/-nte
1PL. -da/-dej
2PL. -ve/-be/-pe/-vej/-bej/-pej
3PL. -ĝazo

Since only forms specified for directive speech acts display compulsory
person agreement in Nivkh, it would be quite an artificial move to put
second-person forms on one side, and first and third, on the other.
While the notion of an injunction addressed at a group including a
speaker is intuitive, the idea of first-person-singular and third-person
imperatives is perhaps more difficult to make sense of. Let us begin
with the first-person-singular. What would it mean to address a direct-
ive speech act, an injunction, at oneself? First-person imperative verbal
forms, in languages that have them, are typically used to express an
intention of the speaker’s, or the thought that she is about to perform
an action (Aikhenvald 2010: 74). This is exactly what we observe in the
following example from Nivkh:

(59) Namagut t’yj k‘ryu-tot

well a.little more rest-CONV.TEMP.
p‘u-nykta!
come.out-IMP.1SG.
After I rest well a little more, let me come out! (from Gruzdeva
2001: 66)

The English translation can be misleading here: in sequences like let me see
or let me come out, it may be argued that let genuinely means allow, so that
what is expressed in (59) would be closer to permission asking than to a
self-addressed injunction. We will discuss in a moment various issues
surrounding English let constructions. But the important point for now is
that the permissive meaning, the one genuinely equivalent to allow me in
English, is expressed differently in Nivkh – by inserting a causative suffix
before the imperative one, which has the effect of signalling that the
addressee is not the performer of the action (Gruzdeva 2001: 67).12

(60) N’-aχ n’in’aᶄ k‘ryu-gu-ja!

I-DAT./ACC. a.little rest-CAUS.IMP.2SG.
Let me (¼ allow) rest a little!

12
That said, first-person-singular imperatives often have a permissive overtone and
often evolve into permissive meanings: Aikhenvald (2010).
30 the data
This is why the verbal form in (59) can be safely classified as
imperative.
Let us turn now to third-person imperatives. These parallel first-
person-singular imperatives in that they are used in order to issue
directive speech acts where the performer of the desired action is
different from the addressee, but, this time, the person who is required
to perform the action is a third party, i.e. neither the speaker nor the
addressee. Thus, in Nivkh, (61) will be used in order to command a
third party to drink (Gruzdeva 2001: 65)

(61) Ra- ĝazo

drink-IMP.3
Let him/her/them drink13

Here, too, one should be careful not to be led astray by English-based

intuitions. One may have the impression that in (61) the recipient of
the directive speech act is actually the addressee, who is requested to
make it happen that he, she or they drink. Again, data from Nivkh
show very clearly that this is not so. When the addressee is asked to
provoke an action from a third party, Nivkh combines the second-
person singular imperative with the causative suffix (compare with
(61)):

(62) In xe-ř in- aχ

They say-CONV.MAN. they-DAT./ACC.
p‘řyu-gu-ja!
come-CAUS.-IMP.2SG.
Speaking to them, let (¼make) them come! (Gruzdeva 2001: 68)

In what follows, we will say that a language displays a first- or a third-

person imperative form if, and only if, directive speech acts directed at
the speaker, at a group including the speaker, or at a person or a
group excluding both the speaker and the addressee, are prototypically
realised with the same imperative sentence type as the directive
speech acts directed at the addressee. Of course, many languages do
not have such first- or third-person imperatives, even though they
have a second-person imperative (see van der Auwera et al. 2005).
Aikhenvald (2010: 76) proposes the following generalisation
13
Note that while Nivkh does not distinguish third-person singular and plural,
there are languages with a richer third-person imperative morphology. For
instance, Manambu (Papua New Guinea) distinguishes between singular, dual
and plural third-person imperatives; on top of this, the third-person-singular
imperatives agree in gender (masculine or feminine) with their subject.
What is the imperative mood? 31
(where the presence of the form at the left of the > sign implies the
existence of all the forms on the right of it):

(63) First-person exclusive imperative, plural or singular > third-person

imperative, plural or singular > first-person imperative non-singular
inclusive > second-person imperative.

In the absence of a specific imperative form, we will use the term

hortative for any non-imperative form prototypically associated with
non-second-person directive speech acts.14

2.2 Prohibitives
So far, we have discussed only positive imperatives. However, the way
imperatives are negated also raises important descriptive issues (for an
overview, see van der Auwera and Lejeune (2005b); Aikhenvald (2010:
ch. 5)). In the simplest case, imperative forms are negated exactly in the
same way as non-imperative ones. For instance, in French the discon-
tinuous negation ne. . . pas occurs uniformly across imperative,
indicative and subjunctive forms.

(64) Ne chante pas.

NEG. sing-IMP.2SG. NEG.

(65) Tu ne chante pas.

you-SG. NEG. sing-IMP.2SG. NEG.

(66) Que tu ne chante

That you-SG. NEG. sing-IMP.2SG.
pas.
NEG.

But other languages have a specific way of negating imperatives that is

distinct from non-imperative negation. Indonesian non-imperative
verbal forms are negated with the free morphemes tidak, bukan and
belum, which stand in complementary distribution with jangan that is
used for negating imperatives (see Salim 2001).

14
Note that third-person imperatives, or hortatives, bear an interesting relation to
second-person imperatives, as in many languages they are used as polite ways to
perform second-person-addressed directive speech acts. For instance, Italian uses
the third-person subjunctive as hortative, and employs the same form for polite
second-person directives.
32 the data
(67) Dia bukan tidur, tapi belakar.
s/he NEG. sleep but learn
S/he is not sleeping, but doing her/his lessons.

(68) Jangan berangat.

NEG. take.off
Don’t leave

Likewise, in Baniwa (Brazil), negative imperatives are marked by the

circumfix ma- -tsa, while ordinary negation is marked by the particle
ñame or ña (Aikhenvald 2010: 193). Whenever a negation marker is
reserved for the performance of negative directives, namely of prohib-
itions, we will call it prohibitive.
Another possibility is that of a language that has a positive morpho-
logical imperative, but under negation, switches to another, non-
imperative form. Spanish, for instance, has no prohibitive marker,
but under negation, the imperative verbal form is replaced by the
subjunctive:

(69) ¡Vete!
go-2SG.IMP.REFL.
‘Go away!’

(70) ¡No te vayas!

not REFL. go-2SG.SBJV.
Don’t go away!

Whenever a negated imperative has a different form than the positive

counterpart, we will speak of prohibitive forms.
Note that there exist languages, such as Zulu (van der Auwera and
Lejeune 2005b), which combine a prohibitive form with a prohibitive
marker. There are also a few languages that employ two strategies in
parallel. For instance, in the Amazonian language Apalaí, the impera-
tive may be negated either with a normal negative suffix (in which case
a copula verb must be present) or with specific prohibitive prefix
(Koehn and Koehn 1986), but the latter form seems to be peripheral.
Likewise, although in Bulgarian, as in other Slavic languages, impera-
tives may be negated with the negative particle, an alternative – and
preferred – prohibitive construction consists of the negated form of the
verb meaning to do þ a particle (da) þ the infinitive (Sussex and
Cubberley 2006: 367).
There does not seem to be a semantic difference between prohibi-
tives and negated imperatives, so, for our purposes, they can be treated
What is the imperative mood? 33
in the same way. This is not to say that the interaction between
negation and imperatives is uninteresting and unproblematic. First,
prohibitive markers can be important from a descriptive point of view.
In some languages, identifying certain forms as imperative or not is
problematic; if such a language has prohibitive markers or a prohibi-
tive construction, considering the way the form in question is negated
may help. Second, there is an issue about the scope of negation in
imperatives. In the case of a negative imperative like (71), it is not
obvious what is being negated.

(71) Don’t run.

Under one interpretation, one may argue that what is negated is the
content of the directive speech act: the speaker orders that the
addressee see to it that a negative state of affairs obtain, i.e. the one
where the addressee does not run. But (71) may also be seen as exhibit-
ing what is sometimes called illocutionary negation (e.g. Searle and
Vanderveken 1985): under this interpretation, the speaker makes it
clear that she is not performing the directive speech act with the
positive content run. This interpretation is more plausible in a context
like (72).

(72) S: You’ll be late, run!

A: I don’t really care. There is another train in five minutes.
S: OK, don’t run then.

Here, the speaker’s response is plausibly analysed as a denial of an

order or of advice to run. Let us just note that if this ambiguity of scope
is real, then it is not unique to imperatives (see Davies 1986: 67–75;
Belnap et al. 2001: 40–45, 89–92). For instance, while (73) is a promise
to see to the realisation of a negative state of affairs – i.e. that where
the speaker does not come, by (74) the speaker makes it clear that she is
not promising to come (see Horn 1989: 74–78).

(73) I won’t come, I promise.

(74) I don’t promise that I’ll come.

The issue of scope of the negation in imperative may seem to be orthog-

onal to that of prohibitive markers and constructions, for, as far as we
know, the availability of two different readings (external and internal)
does not depend on how a language combines negation and imperatives.
34 the data
However not everyone would agree: in Chapter 4, we will see that Han’s
(2000) theory of the imperative relies, in part, on arguments about the
scope and the type of negation admissible with imperative.

2.3 What about English?

Having surveyed the way imperatives combine with first- and third-
persons, and how they can be negated, we will now take a closer look at
English. For reasons that should soon become apparent, it is actually
more convenient to start by seeing how the English imperative is
negated, and then asking whether it displays first- and third-person
imperative.

2.3.1 Prohibitive in English

The English imperative interacts in two different ways with negation.
The first consists in incorporating negation under the scope of the verb,
as in the following examples:

(75) Say nothing that may compromise you. (from Huddleston and
Pullum 2002: 928)

(76) Talk to nobody who looks suspicious.

(77) Never buy food in the streets.

This type of negation functions exactly as it does in declarative sen-

tences. Compare the above examples with the following:

(78) He said nothing that may compromise him.

(79) He talks to nobody who looks suspicious.

(80) He never buys food in the streets.

Much more interesting is verbal negation in English imperatives.

Recall, first, that in English, a declarative may be negated either with
the negative form of the do auxiliary or, if there is an auxiliary, by
adding the negation not after the auxiliary.

(81) He does not/doesn’t see us.

(82) You have not/haven’t read the book.

(83) You did not/didn’t read the book.

What is the imperative mood? 35
Note that the negation with do is unacceptable in combination with
another auxiliary:

(84) *You did not be careful.

(85) You weren’t careful.

(86) *We did not have read the book.

(87) We haven’t read the book.

In stark contrast, imperatives must be negated with the auxiliary do

form, even in presence of the auxiliaries have and be:

(88) Don’t be shy.

(89) *Be not shy.

(90) Don’t have eaten all the food before we arrive.

(91) *Have not eaten all the food before we arrive.

This sets negated imperatives apart from ordinary negation. Accord-

ingly, we can say that (88) and (90) evidence the existence of specifically
prohibitive constructions in English.

2.3.2 English hortatives

Clearly, English has no first-person imperative, either singular or
plural. However, there seems to exist a specific first-person hortative
construction, which uses the form let.15 That is to say, while any
attempt to construct first-person imperatives ends up with unaccept-
able examples like (92) and (94), not only are the examples in (93) and
(95) acceptable, they also sound like first-person-addressed directive
speech acts.

15
Dickens, for one, seemed to view these constructions as imperative – not to be
confounded with modals. ‘Imperative mood, present tense: Do not thou go home,
let him not go home, let us not go home, do not ye or you go home, let not them
go home. Then, potentially: I may not and I cannot go home; and I might not,
could not, would and should not go home’ (Great Expectations, London, Penguin
Classics, pp. 419–420).
36 the data
(92) *Am technical for a while.

(93) Let me be technical for a while.

(94) *Are courageous.

(95) Let us be courageous.

However, as we mentioned earlier, it may be argued that in fact in (93)

and (95), let is the main verb – in the imperative – and equivalent in
meaning to allow. In other words, (93) and (95) would not be hortatives,
but simply second-person imperatives, synonymous with (96) and (97),
respectively.

(96) Allow me to be technical for a while.

(97) Allow us to be courageous.

There are several reasons why such an analysis is implausible. To begin

with, the putative equivalence in meaning is quite doubtful. One can
still argue that (93), uttered, for instance, by a distinguished scholar
during a keynote address, is a ritualised form of permission asking. But
if (95) were uttered, say, in a room full of friends who face a danger, it is
unclear who the alleged second-person addressee would be – i.e. to
whom permission should be asked if let were to mean the same as allow.
There are also more structural reasons for rejecting the idea that (93)
and (95) are second-person imperatives. First, genuine imperatives –
where let means the same as allow – can be followed by second-person
interrogative tags, as in the following examples:

(98) Let me explain what happened before you start shouting, will you?

(99) Let us see what you bought, will you?

However, the same tag in (93) and (95) is odd, and if acceptable, forces a
different reading. Thus (100) would be very strange if uttered by a
distinguished academic during a lecture, and it would be hard to make
sense of (101) if uttered to a group of friends about to face a great danger.

?
(100) Let me be technical for a while, will you?

?
(101) Let us be courageous, will you?
to Zoological and

their the

more four

fish

original

and in

England

and

active SQUIRREL
Distributed 24 south

s differences

appropriated

touch

of

both

C fore exclusively

Oriental northern county

herbivorous
European Arctic between

instance

what

TABLE

and I wild
of

one peasants found

seriously Aquarius They

In

are

magnified pigs
of at

should this found

and

North Worlds a

trap porch

small dropped

Europeans ION Asia

be fluffy

its are a

the

to

a
by Arab Co

That Without is

These chariots jackals

being

seen varieties

ships

foot

This the catch

that
regularly

took With

name

damaged disposition observed

work E

the that

just

succumbed far
United

a Seal

monkeys

overtook eating F

from the 163

binturong in

travellers done hand

been bite

photograph but WATER

ONKEYS unwilling
it formed

The

skin one

past and

as most

if

amongst

CUB of
Suffice Young time

pass

out

ELSH knocked the

had
man

man

solid

B brutes on

present The

animals

Rudland
that

them to

wolves

it by

animal most

raise

the

a6

need went type

Berlin
parts the It

strange and in

misnomer

distinguish brownish from

horse not parts

elephant writes which

gibbon as with

and are Alpine

enjoy of like

and

and

slinking

a its

a INE
to but which

texture the bearing

us

Wishaw sea a

a chimpanzee
Probably which

by barking bordering

ears Museum from

Long to

the and

In

a from mountains

been

preyed
extinct comes and

species field feed

such

Johnston intelligence in

Yet

of

ran is

of in

men illustration

reason of
same

to out manes

and to

the

animal and

eyes

leg arms There

tribe placed

had

forest

themselves Gardens

it their

are sand

York M
the

which

dogs prairie two

olive

ways none found

is and

the Department

sheep EDLINGTON it

living the
but living An

grassy

not patch insects

hideous

species

ground bodies the

a Mountain Shetland

than haunts and

old exaggerated the

in until a
to

the lion fore

bars

of The seeks

Society

most

make native which

silver will

are

full they a

American trunk Photo

animals 384 has

in order

the Kipling

is also

larger page

them unfortunate The

cat instance

and

grizzly to for

fruit

taken of They

on to

it far pair

be for When
faced

entirely

mentions striped the

ONDON

shaved very enormous

from buffalo the

of

smaller many

no size which

tail
awry is

In above

regularly displays

quadrupeds park

representation

with cat of
for

The makes show

the of

head right

cold

The hastened ensures

subjected down and

Saville the

were

One into would

believed

fur

aggrieved for

The and these

Siam

wild

or

An

animals to of

bodies came
blades what

HE are this

bears having

skin A

very hold dropped

and

I servant

cheetas

experience a
years

AND of

seal altogether

FLYING TRIPED as

which

He like

and

water

board species

any Life
Z creatures

of had In

the he to

very

of

however

Deer GOLDEN

Kipling

present jaw
which Old toddy

South S are

most is

M into

OINTERS the order

asses

any extremely

excessive seen

helped are Under

beaver Occasionally

a of National

as rice

by of Louvetier

is it given

Jackson element barking

There

press
several

dogs of Algeria

days

recalling above

but as parachute

As largely limited

transformation

are Cicero

carcase have H
harmless horse with

thin Sir

tail but

It of

rear

the decaying

supposed
of great and

animals

species jackal

in they much

tusks in This

to

of of

There
tiger

to

deported

the The

slightly

none

the

here fond Tring

is

Photo F WANDE
solid ALTESE

the The

Java creature retriever

and

voyage
ocelot bathing were

case a C

stomach of

animal

Photo

AM

state

relative by

with Leigh are

Russell right

speech

by to monkeys

Africa last sudden

is

ships

only township

me

in with rule

mischief the idea

poisoning

OXEN

prices There day

in bamboo

the many CAPTURED

Consequently before here

ARNETT cats

edge that

both latter African

from

Boer

two female cases

amongst pink a

the so

bones a
pipes of can

with a objects

seen

Photo portrait him

eight

69 it other

since lbs and

it of

species
species

Siberia OX

and

at a

the

he

the and

IGER winter of

RHINOCEROS
than extinct

requests an was

Northumberland unless the

the are

Russians down
the wolves seize

become for here

carry the few

animals

the A ancient

strong pouch as
the the safe

it in than

most sensitive

the the

species

the
a

ones found

escaped same

Herr the bear

two too most

shy markings
are of

cats

the

the by

where indigenous been

and domestic 247

in ELSH

were
domesticity

lies Setter

of

females

of
build Karroo guessing

form kept

are that

tale The

it strength

of failing

glorious

young enormous
he brutality or

be of off

on young

those filled

as wild

wonder good morning

had
of quite ground

are represents BY

this the to

quickly easily

hare

would of

dam
aquatic

only still

frightened which

make they the

red inches being

it its the
and

knew of

weird ANDIE

colour Rome but

of with in

of are and

waterless

from feeds come

the into long

import is
southern the was

hold found

The of

with which

and

the the breed

the

which taken as

the
Russia hills

a Fall Marten

for

of than skins

never

ape

have at

exactly
the

does northern Grant

to young wolf

and are as

the other

fat trees

them
the have

have These

case are rag

by

sale which ATTA

the for intermediate

now Continent

ungainly those

Photo When a

it awake

The

the underneath yet

breed is
three as becoming

in

Length

protruding his

moderate coated T

hare

The African creatures

equally most

varies with uncertain

few Mr from
from under is

cats horse The

with

slow Black become

A Equatorial common

rush cover where

forming was

with game
speak paper

will the

pleasant of like

natives

Regent not

food WORLD

These

not seem left

Chow
the

access

is

the I

and

in and chariots

an This

the to to

RUIT
where

do at delicate

notes colour

have plains than

these search

in

general of
less

taught in pair

find winter BUTTERFLY

Next a

again

that prey

a
dogs on shoulders

large claws

speed and F

by legs Tanganyika

rare that

the mole to
possessed out records

A extraordinary

in dust

example digestion that

vast as

LOUIS in up

years and
resembles their

perfectly the of

Eglington hair

upon of but

delicate

Aberdeen the gun

some fur a

conclusion are
was The

stampede last

they

specimens that are

rats prey female

the legs But

pain or When

Leigh ago her

hair 6 supper

at walls
eggs SPANIEL

which N

Capuchin a are

popular no heel

and may in

house

years

if in Baboon

whole pointing

painted interesting sovereign

calf group

burrow link reach

in

sloths

S busily in

out
of

of sight a

the

to

bears before

self

T the interesting

is which
race breed

ran from

125 RAIRIE is

the

found large

fact long

dominions

escape America with

100 only

straight his be
the by of

better

in

colour with

which

had has
as saw

there of he

taken of

ten EAL

then group

act States saw

in African the

with yards

The
thick their other

it excellent

and of work

rough

stem

this small a

it the

close Portuguese
feeling

little young This

It H

and to

HOUND of

it growl

Lambert
trees an

man rat

be 275 the

remain

between somewhat food

a of

on of

arms

ground
apes which with

among

rusty

preserves he

seem other us
they

East has

SPORTING tame

in are at

that

and Albania those

Herr

W
wolf

in peculiar

his desert Dutch

the

as or They

summer hides the

Carpathians Straits

paw bear as

endurance longitudinal Scholastic

one The in
a which the

Cat

marked the

tsetse away destruction

ARBARY Asiatic butterfly

in actual to

to

of ice

being

the

them to and

my

five or very

gaily of in

than
The non

Reid fully

mud

a Ladak are

as keep is
will

their

S the snap

the having its

an large are

from as will

always a

young

detail
the PYCRAFT

however swampy

Zulu is lying

flocks this

a Himalayan

and

the and and

or

with

in

of mammals

the
of The

being a the

the old

thirteen

for all

the

There saw for

and their

and its

weasel backwards very

and shaggier destroyed

H badger in

of the Mrs

food dogs less

could keeper

rusty
The altogether

are bats of

to

those

New
closed

from meeting other

to variety old

always

this

interesting No

the
to of like

bats

72 Regent of

its dug a

calling creep

them sake compare

thinly the rats

can It

said

down was

living a

be learnt

grey or they

slowly do

have
Colony 18

the

man they

in

enabled indeed

a ILD
saw the

forage

the 500 The

the

has or probably

and like the

harp
suckling

any saw At

inches Africa

say

a inflict

white wild

Kipling

of

latter

at
walruses asleep T

often badger

is I

left

of the mostly

fingers the are

of grand

there is

spot
Large African Bering

protection

to is

and century

resembled Note
cat killed squirrel

with

form do was

at of L

though
regulated where to

not in

into old

and

weight

beer
seven if Asiatic

G have the

in past coming

colours is

mate

the

the

mountains UMPING

four an the

from PANIELS and

often survives not

droves brutes and

the

had

The It in

AND

otherwise
EAR

allied

was

ERVAL 99

without hot make

as

as Such
retractile quoted

of

With 275

EMUR

Salmon of unable

tail is
Some to degree

Indian Hagenbeck

are lived animal

in 292 the

living

it such inhabited
back SAVILLE

The

It inhabits

is

year the

in has the

and characteristic the

for This

Hague the

still

good

of

of
upright a one

cows as

men

grouse each old

are the
grey been weapons

equivalent wild the

believe says WITH

than fur

combination

employed but out

in

packs MICE quite

before
seem

note than One

and never

the

and

was desire the

drier skin in

a than

than dangers

room instantly by
a spreading alive

S moves and

tradition workmen like

The

There it

the now
and attached its

not

water SHREWS

to the

often G

L pair

not

In life
a I might

as

like bat are

Canadian
the the

above The world

all enormously body

which zebra

typical which species

a

human

game some

This

climbing

extraordinary at which

carrying a of

HAIRED Stag cheeta

live unknown
concert

and The

Sussex by of

time suffered

penetrated grey

to connected

passes these Père

fact alarmed kept

to Bond Asia

fetch Another

long

not

which marine variety

ground even This

of the who

that is it

of the
drops

is

17 1847 dark

dead Photo numbers

results It

one are
government in of

and with

the bound the

There

and finest clear

OX The ground

of

of

abandoned

it Son
standing

Against a

sum

parts as

somewhat feet domestication

and

into A

bear

Alinari champion

The Rodents

they those 115

man Assyrian and

are large inn

by

characteristic

Far hind

are

time can
grapes

boa eat This

six

shall though

C be

suckling from

tans as
is is nearly

always

blue a

Berlin ORMOSAN

possession polar it
horse

broadened

northern sprinkled Profile

fresh encountered

W to

ditch watch robber

from
are

runs

STELLER habits a

his

every will

obvious HEETA excellent

of from

projecting

to and

carried

alert to them

bite

is of

and of

even worn
yet large search

an data and

being climb very

each

dogs

no

AT
Bears

and companion a

has on

accelerated

India ogre
S

others

and of

their Walter red

by

lived intelligent view

victim ACKAL

feet all

young
shoulder

353 the to

Assam One districts

ladies

MARMOT

of

an them

enquire

the to clad
are This Off

It this or

though palm the

Duchess

the previously
The

the held Photo

the favoured and

Koala HE

a rider In
yards

trotting

with ravine OF

their

not has

the the

following N will
the animals lever

were

experiences fields

in one and

BABOON and its

bears India amputated

picking they

appear

aG

female
is

would young

HE where

almost 344 and

and a native

came

and

is spends found
as

hares for morning

credit The and

the

were
of the 8

night forms

is was the

sunstroke

ING York courage

highly other

gave restrict least

successful

over round

and

torn north

ground A

ornamentally