Kasher, Asa, ed. (1998) Pragmatics: Critical Concepts,
Volume IV: Presupposition, Implicature and Indirect
Speech Acts. Routledge, vi+722 pp.
Reviewed by Aldo Sevi, Linguistics department, Tel Aviv
[The volumes that make up the collection Pragmatics:
Critical Concepts, edited by Asa Kasher, are being reviewed
separately. This is the first of those reviews to be posted.
The volume under review is the forth of a six volume, 14-
part, beautifully bound boxed set collection of
previously published papers with several postscripts.
Volume IV contains 27 papers (5 with postscripts), which
are divided into three parts (parts six, seven and eight
of the collection): Presupposition (6 papers),
Implicature (17 papers) and Indirect Speech Acts (4
papers). I'll first give very brief descriptions of the
contents of the articles, and then comment on the whole
volume as a unit.
Essays 48-53 on presupposition contain mostly classical
papers from the seventies, which focus on two major
issues: What sort of phenomenon is presupposition
(semantic or pragmatic), and how the presuppositions of
the sentence depend on the presuppositions of its parts
(the projection problem).
The opening piece of this section is a short paper by
Strawson called "Presupposition", where he defines the
notion of a sentence's presupposition as a statement
which its truth is a precondition of the sentence being
true or false.
The second paper, by Keenan, "Two Kinds of Presupposition
in Natural Language", suggests that there are two notions
of presupposition: logical and pragmatic. A logical
presupposition of a sentence a la Keenan is a statement
which is entailed both by the sentence and its negation
(This definition includes also non-restrictive relative
clauses which don't seem to be intuitively presupposed,
but Keenan isn't bothered by that). A pragmatic
presupposition a la Keenan is a culturally defined
condition on the context of utterance that must be
satisfied in order for the sentence to be felicitous
(for example the use of tu vs. vous in French).
The next two papers: "Pragmatic Presuppositions" by
Stalnaker and "Presupposition and Linguistic Context" by
Karttunen, see presupposition as a pragmatic phenomenon
(something which is "taken for granted", or
alternatively, already included in the context), and
suggest an elegant solution to the projection problem.
According to the Stalnaker-Karttunen approach a
presupposition of a clause within a complex sentence must
be already included, not necessarily in the context of
utterance of the whole sentence, but in the local context
of that clause. Consider for example the simple case of
conjunction, where the second conjunct has a
presupposition, as the case in (1) and (2):
(1) John used to smoke, and he stopped smoking.
(2) John took his doctor's advice, and he stopped
The first conjunct plays a role in determining whether a
presupposition of the second conjunct will be inherited
by the whole conjunction. This makes sense -- the second
conjunct is added to a context, which already includes
the first conjunct. Hence, what is relevant for
determining the whole conjunction's presupposition is the
local context of the second conjunct; this local context
is derived from the original context by adding to it the
information in the first conjunct. In example (1) above,
the whole sentence does not inherit the second conjunct's
presupposition that John used to smoke, because this
presupposition is already entailed by the first conjunct.
In example (2), the first conjunct does not entail the
second conjunct's presupposition, and in order for the
local context to satisfy it, it must be included in the
context of utterance of the whole sentence, and hence
presupposed by the whole sentence.
The fifth paper, "Pragmatics and Presupposition", by Katz
and Langendoen, rejects the Stalnaker-Karttunen pragmatic
approach, and argues for a purely semantic account of
The last paper in this section, "How Presuppositions are
Inherited: A Solution to the Projection Problem" by
Soames, discusses two influential approaches to the
projection problem: Gazdar's (1979) 'cancellation'
approach (a presupposition is cancelled if incompatible
with the conversational implicatures of the sentence),
and Stalnaker and Karttunen's approach (described above).
Soames gives counterexamples to both theories, and
suggests a synthesis: We cancel first, and then use the
other strategy. For a recent critique of Soames see
Kadmon (2001), ch.6.
I'm surprised not to see in this collection Heim's (1983)
influential paper on the projection problem.
Essays 54-70 on implicature open naturally with Grice's
seminal "Logic and Conversation", where the notion of
implicature is introduced, and their derivation is
explained by the famous Cooperative Principle and the
four super-maxims of Quantity, Quality, Relation and
Manner, which supposedly follow from it. This paper is
followed by Grice's less known "Further Notes on Logic
and Conversation" and by a "Retrospective Epilogue" from
The forth paper in this section is Kasher's
"Conversational Maxims and Rationality" where he argues
that the relation between the Cooperative Principle and
the maxims of conversation is problematic (generally the
participants in a conversation do not have a mutual
purpose). He suggests replacing Grice's Cooperative
Principle with a rationality principal (achieving ends
with the most efficient means at the lowest costs).
Kasher argues that the maxims follow from this principle.
The next paper, "The Universality of Conversational
Postulates" by Elinor Ochs Keenan, argues that the
Gricean theory is not universal. The author presents her
findings that the inhabitants of a small village in
Madagascar violate systematically the maxim of Quantity -
they give too less information. Kasher (in a postscipt
from 1982 to his paper) suggests that these findings
could be explained by an interaction between his
(universal) rationality principle and local cultural
Harnish, "Logical Form and Implicature", investigates
certain types of implications (for example, "Russell and
Whitehead wrote Principia" implies they wrote it
together). He claims that these kinds of implications
are not entailments, and uses a modified version of the
Gricean theory to explain them. According to Harnish, we
should distinguish between direct and indirect
conversational implicatures. Indirect implicatures
require a maxim to be flouted, direct implicatures
require that the (more important) maxims are intended to
be observed. The implications Harnish investigates are
claimed to be of the second type.
The contribution by Sadock, "On testing for
Conversational Implicature", investigates three criteria
for deciding whether a certain implication is an
entailment or a conversational implicature: cancelability
and non-detachabilty (which were proposed by Grice) and
reinforceability. He thinks that no criterion alone, and
no combinations of the criteria is a necessary property
of conversational implicatures.
McCawley, "Conversational Implicature and the Lexicon",
argues that differences in use between some lexical items
and complex expressions matching their suggested semantic
decomposition (i.e.; kill vs. cause to die) are due to
Grice's maxim of Quantity and hence don't provide
evidence against decomposition.
The next two papers are by Sperber and Wilson, "On
Grice's Theory of Conversation" and "Mutual Knowledge and
Relevance in Theories of Comprehension". The authors find
Grice's distinction between "what is said" and "what is
conversationally implicated" inadequate. They claim that
the proposition expressed by an utterance ("what is
said") is also derived using pragmatic processes. They
criticize Grice's analysis of irony and metaphor, and
call for a separate treatment of these within a theory of
rhetoric. The authors suggest a reduction of Grice's
maxims to a single principle of relevance: a rational
speaker will choose an utterance that will provide the
hearer with a maximum number of contextual implications
in a minimum processing effort. A feature of Sperber and
Wilson's theory which is significantly different from
Grice's is that the processing of an utterance involves a
construction of a context in which the effects of the
utterance are evaluated. The context is not given, but
enriched in such a way that facilitates the processing of
the utterance. For a recent critique of relevance theory
see Levinson (2000), ch.1.
Horn, "Toward a New Taxonomy for Pragmatic Inference: Q-
based and R-based Implicature", suggests two principles
(Q and R) which are intended to replace all the maxims of
conversation, except Quality. The Q-principle, hearer-
based, "Say as much as you can (given R)", and the R-
principle, Speaker-based, "Say no more than you must
(given Q)". These principles correspond to two competing
forces identified by Zipf (1949), Speaker's economy:
"(Given m meanings) a vocabulary of one word which will
refer to all the m distinct meanings", and Auditor's
economy: "a vocabulary of m different words with one
distinct meaning for each word".
Hintikka, "Logic of Conversation as a Logic of Dialogue",
shows how to reinterpret 3 of Grice's super-maxims
(Quantity, Quality and Relevance) in his game-theoretical
dynamic discourse framework. He argues that the maxims
apply primarily to answers to questions. For example,
Quantity is the requirement that the utterance must be a
full answer to the question.
Carston, "Implicature, Explicature, and Truth-theoretic
Semantics", introduces the notion of "explicature", the
proposition explicitly expressed by an utterance. Carston
argues that this is not the minimal proposition that we
get from the logical form of a sentence after
disambiguation and reference assignment, but something
which is derived by a pragmatic process. She considers
the problem of distinguishing explicatures from
implicatures in a relevance-theoretic framework. The
paper is followed by a postscript written for this
Fretheim, "The Effect of Intonation on a Type of Scalar
Implicature", argues that the 'at least' interpretation
of cardinals is possible in Norwegian only when the
phrase containing the cardinal refers to some entity that
is salient in the discourse. Salience is reflected in
intonation, which is an important information-structuring
device. In a postscript specially written for this
volume, the author admits that 'at least' interpretations
are sometimes natural without a special intonation
The next two papers are by Recanati, "Truth-Conditional
Pragmatics" and "Primary Pragmatic Processes", are
intimately related to Carston's contribution. Recanati
argues that many cases that were analyzed as implicatures
are pragmatic constituents of the proposition expressed.
He rejects the popular view in formal semantics that
"what is said" is derived from "sentence meaning" by
filling in empty slots (such as an appropriate domain of
quantification), on the grounds that there are cases who
cannot be explained in this way. However, it is not at
all clear that the cases he discusses in this context are
not just implicatures and are indeed a part of the
proposition expressed by the sentence. The author
proposes that there are pragmatic processes which operate
locally (below the sentence level) before the computation
of the proposition from word meanings.
The last paper in part 7 is "Minimization and
Conversational Inference" by Stephen Levinson. Levinson's
work is another reformulation of Grice's maxims. He uses
Horn's Q-principle (mentioned above), the I-principle
from an earlier work with Atlas (Atlas and Levinson
1981): "Say no more than your hearer needs, given Q", and
a Gricean maxim of relevance, and gives a resolution
mechanism for cases of principle-clash. Levinson argues
that conditions B and C of the Binding Theory (and
possibly other significant portions of the Government and
Binding Theory in syntax) fall out from the interaction
of his neo-Gricean principles.
Essays 71-74 constitute the part on indirect speech acts.
The opening piece, "indirect Speech Acts" by Searle,
introduces the notion of an indirect speech act (one
illocutionary act is performed indirectly by way of
performing other), and analyzes the phenomenon in terms
of the author's theory of speech acts (for example, a
question which is used indirectly as a request is about
the felicity conditions of the request), combined with a
Gricean theory of cooperative conversation.
The next paper, by Morgan, "Two Types of Convention in
Indirect Speech Acts", emphasizes the conventional nature
of sentences such as "Can you pass the salt?" Language
users have knowledge about conventions governing the use
of certain classes of expressions for certain purposes.
The third paper in section 8, "Short-circuited
Implicature: A Negative Contribution, by Horn and Bayer,
tries to explain the use of sentences such as "I don't
think he has come" to mean "I think he has not come", as
a conventionalized instance of an implicature.
In the last paper, "Indirect Acts and Illocutionary
Standardization", by Bach and Harnish, the authors reject
the conventionality approach, and suggest instead what
they call "standardization", which combines frequent
usage and Grice's Cooperative Principle.
The last two papers are followed by postscripts (one by
Horn, the other by Bach) written for this volume.
At a price of 650 British pounds (more than $900) for the
6 volumes (the volumes are not sold separately), I guess
the collection is aimed at the library market. I doubt if
it is intended for professionals; they probably already
have most of the articles (but it would be a very
aesthetic and convenient replacement for huge piles
of xeroxed material for anyone who can afford it). The
collection can be helpful to students -- it contains much
of the reading material required for beginning and
advanced courses in pragmatics (especially those with a
philosophical orientation), and it can save a great deal
of running around in the library.
While reading the book, I wondered if it forms a whole,
which is greater that the sum of the articles in it. As
Kasher's introductions are so uninformative - 6 pages of
editorial comment out of 722, which consist mainly from
sentences such as "The author is a leading philosopher of
language" or "[The] paper is another important
contribution to the linguistic literature..." -- the
answer is left to the reader to infer.
Although most of the papers certainly "speak for
themselves", I feel that there are three things missing
from Kasher's introductions. First, there is almost no
attempt to motivate why a certain paper was selected, why
it is important or was important in the context of its
original publication. Second, there is no information
about developments in the particular research directions
that the articles represent. Which directions turned out
to be fruitful? Which were abandoned? Which survive as
competing alternatives? As many articles in volume IV
clearly "talk" with each other, and as the editor
couldn't possibly include endless replies, objections and
counter-objections, the debates are arbitrarily cut at
some point, possibly giving the wrong impression that the
issue at hand is closed. In such cases a few lines by the
editor (even just a few suggestions for further reading)
would be very helpful. For example, the editor's choice
to close the part on presuppositions with Soames' paper
from 1982, without informing the readers about the more
recent work done in dynamic semantic theories inspired by
Stalnaker's approach, truly surprised me. Thirdly, hardly
any historical background is given. As the title,
"Pragmatics: Critical Concepts", implies, one of the aims
of the collection is to provide the reader with the means
to gain insight into the development of the key notions
in pragmatics. Starting the section on presuppositions
with Strawson and the section on implicatures with Grice
is natural, of course, and Kasher credits them
appropriately in his introductions (although early
versions of these notions appear already in Frege's
famous "On Sense and Reference"), but I think he should
have said something about the circumstances in which
these concepts entered into the linguistic discourse.
Some of the postscripts do part of the job I wished the
editor had done, but there are too few of them.
A note on the selection of papers -- these are mostly what
one expects to see, however I was surprised not to find
Lewis 1979 and Heim 1983 on presuppositions and Thomason
1990 on implicatures. I guess that adherents of Relevance
theory would miss a contribution by Blakemore.
The book has some truly irritating features. Nowhere in
the volume it is indicated where and when the articles
were first published. Michiel Leezenberg, the reviewer
of the first volume, informed me that a list of
bibliographical data appears at the start, but he has
noticed that references are lacking for items 71 through
86! It is extremely unfriendly to the reader not to
provide such information in the beginning of each
article, and it is scandalous to leave 16 papers without
publication acknowledgements whatsoever. Browsing through
the book is very inconvenient - the names of the articles
or the authors do not appear on the pages' headings! A
potential buyer of such an expensive collection deserves
more. One wonders, for example, who was the proofreader
who decided to change 'intensional logic' to 'intentional
logic' throughout Kasher's own contribution.
Summing up, despite my criticism, anyone who wishes to
access the most important ideas in the fields covered by
this volume, as they were originally presented, will
certainly benefit from the book.
Atlas, J., and Levinson, S., 1981, "It-clefts,
Informativeness and Logical Form". In P. Cole (ed.), Radical
Pragmatics. New York: Academic Press.
Gazdar, G., 1979, Pragmatics: Implicature, Presupposition
and Logical Form. New York: Academic Press.
Heim, I., 1983, "On the Projection Problem for
Presuppositions", in M. Barlow, D. Flickinger & M.
Wescoat (eds.), Proceedings of WCCFL 2, Stanford
Kadmon, N., 2001, Formal Pragmatics: Semantics,
Pragmatics, Presuppositions and Focus. Oxford: Blackwell.
Levinson, S., 2000, Presumptive Meanings: The Theory of
Generalized Conversational Implicature. Cambridge, Mass.:
The MIT Press.
Lewis, David K., 1979, 'Score keeping in a language
game', in: Journal of Philosophical Logic 8: 339-59.
Thomason, R. C., 1990, 'Accommodation, Meaning and
Implicature: Interdisciplinary Foundation of Pragmatics'.
In P. Cohen, J. Morgan and M. Pollack (eds.), Intentions
in Communication, 325-363. Cambridge, Mass.: The MIT Press.
Zipf, G. K., 1949, Human Behavior and the Principle of
Least Effort. Cambridge, Mass.: Addison-Wesley.
About the reviewer:
Aldo Sevi is a graduate student at Tel Aviv University.
He is writing a PhD dissertation on the projection problem
for conversational implicatures.