Kasher, Asa, ed. (1998) Pragmatics: Critical Concepts,
Volume V: Communication, Talk in Interaction, Discourse.
Routledge, vi+490pp, hardback ISBN 0-415-16938-0,
Routledge Critical Concepts series.
Reviewed by Maite Taboada, Department of Linguistics, Simon
This is the fourth of a projected set of six reviews of this
anthology, one for each volume. Previous reviews can be found at:
http://linguistlist.org/issues/12/12-2673.html (Volume I)
http://linguistlist.org/issues/12/12-2154.html (Volume IV)
http://linguistlist.org/issues/12/12-2221.html (Volume VI)
For the original announcement of the six volume set, see
Pragmatics: Critical Concepts is a set of six volumes that
cover most of the areas in pragmatics. Volume six contains
three parts, each one including a number of articles, on
Communication (eight articles), Talk in Interaction (four
articles) and Discourse (six articles). The articles
have all appeared before in print. (No publication
information is provided in this volume; see the review of
Volume I for the location of this information.) Some of them
have been updated with postscripts. The first part of this
review provides brief summaries of each of the articles in
the book. The second part is an evaluation of the volume
and, to a certain extent, of the collection.
Part Nine: Communication
75. Meaning, Communication and Representation. John R.
Searle. In this paper, Searle furthers previous
explorations, by himself and by Grice, on what it is for a
speaker to mean something by an utterance. Searle believes
that intentionality is crucial in the analysis of meaning.
The speaker, in uttering something, intends to produce some
effect on the hearer, i.e., the speaker intends to
communicate. Communication is, then, a consequence of
Unlike Grice, who thinks that speakers have the intention
that hearers believe their utterances, Searle asserts that
the only intention is that the hearer understand the
utterance, i.e., an illocutionary act. The problem arises
when a speaker produces an utterance, means it, and yet
does not intend to produce understanding in the hearer. One
such case is talking to somebody that we know is not paying
attention, or talking to oneself, or even writing a diary.
Searle solves the problem by postulating that those are
defective instances of speech acts, with no illocutionary
effect. (Or maybe the effect in writing a diary is to
produce understanding in one's future self.)
76. Reflections on Language. Noam Chomsky. In a response to
Searle's (1972) criticism of his own views, Chomsky defends
the validity of studying language structure without making
reference to function, a position that Searle winnows as
"pointless and perverse." Chomsky objects to Searle's
identification of language and communication, pointing out
too that there are situations where speaker meaning does
not offer an insight into the literal meaning of what that
speaker says. (No consideration is given to indirect speech
Chomsky challenges Searle, and the group of "communication
theorists," to show that communication is an essential
function of language. If we classify talking to oneself as
communication (as Searle does), then we must conclude that
language is also a vehicle for expressing thought. This
again raises the question of which function is primary,
communication or the expression of thought. Chomsky's view
is that communication is one function, by no means the
Finally, and to address an objection by Strawson (1970),
Chomsky disregards any importance attached to the public
nature of language rules. We share rules of language, he
says, as we share other cognitive abilities, for instance,
the organization of a visual space, which also plays a role
in successful communication. Chomsky seems to overlook here
that we cannot choose not to see the world in terms of
objects, but we can choose to bend, or even break, some of
the language rules we share with others.
77. Language Without Communication: A Case Study. Marion
Blank, Myron Gessner and Anita Esposito. The authors
reflect on the relationship of meaning and communication
through the case study of a boy who can formulate correct,
age-appropriate, utterances, but who fails to engage in
communicative behaviour with anyone except his parents.
Language development includes both a syntax-semantic and a
pragmatic aspect, the authors argue. Those two seem to go
alongside each other in normal children, but may not evolve
in parallel in children with language disorders, which
leads the authors to speculate that the two aspects might
function independently of each other.
78. Overcoming Inadequacies in the 'Message-model' of
Linguistic Communication. Adrian Akmajian, Richard A.
Demers and Robert M. Harnish. The 'Message Model' in the
title refers to Shannon's (1948) and Shannon and Weaver's
(1949) theory of communication. The theory is based on the
concept of encoding and decoding of a message. The speaker
is the transmitter, who has some message in mind. The
speaker encodes that message and sends it across a channel.
The message is then decoded by the receiver. Akmajian and
colleagues expose a number of problems in this model. First
of all, it identifies the message with the literal meaning
of the words sent or received, thus preventing the
explanation of non-literal uses of the language, such as
irony or metaphor. Secondly, it assumes that both encoding
and decoding of words take place in a linear and sequential
manner. Although they do not propose an alternative, they
provide evidence against a simplistic view of
communication, one that many researchers do not hold any
79. Precis of _Relevance: Communication and Cognition_. Dan
Sperber and Deirdre Wilson. This is a summary of the 1986
book in which the authors explain their view of
communication, a view based on Grice's (1975) maxims,
reduced to one principle of relevance: that communicated
intention comes with a guarantee of relevance. Every
communicative act carries a presumption of optimal
relevance, which involves a guarantee that (i) the
communicative act is relevant enough to make it worthwhile
for the hearer to process it, and (ii) the stimulus is the
most relevant one available to the speaker. When processing
utterances that are assumed to be relevant, the hearer
interprets each utterance relative to the context, built
for each utterance out of the logical form, implicatures
Communication can be achieved, it has been argued, by
encoding and decoding messages (Shannon and Weaver's 1949
view, discussed earlier), or by providing evidence for the
inferences to be drawn about the speaker's communicative
intention. The two types of processes have traditionally
been seen as exclusive of each other; one could conceive of
communication as one or the other. Sperber and Wilson
propose that both can be combined, especially in verbal
The final part of the paper is devoted to explaining how
relevance accounts for poetic effects. Given a principle of
relevance that is mindful of least effort, an utterance
that has more implicatures (in the form of repetition,
metaphor, irony, etc.), and therefore requires extra
effort, will be justified because of the extra effect.
80. Language and Communication. Michael Dummett. Dummett
starts out this article by questioning what is the function
of language. It is natural to think, he argues, that
language has two primary functions: instrument of
communication and vehicle of thought. But the question
arises as to whether language is necessary for thoughts, or
merely an epiphenomenon. He then abandons this distinction
because the crucial question is what it is that we do with
language. He then proposes a philosophy of language based
on language use, a la Wittgenstein, that is, a description
of the activity of using language.
81. A Pragmatic Model for the Dynamics of Communication.
Jef Verschueren. The author proposes a general framework of
the discussion of pragmatics, here understood as a theory
of linguistic adaptation or adaptability. Communicative
dynamics comes in to explain that process of adaptation.
Communicating, according to Verschueren, consists in making
communicative choices. Those are to be understood as
choices of variability (range of possible choices),
negotiability (choices are not mechanical, or based on
fixed rules), and adaptability (speakers and hearers make
choices which satisfy their communicative needs). These
three notions apply at the micro-level of small-scale
interaction contexts, but also at the macro-level of
synchronic and diachronic processes in society. The paper
concentrates especially on the temporal dimension inherent
in communication---a process that takes place over time.
The paper ends with an example, which tries to illustrate
that any encounter can be analyzed from a dynamic point of
view. The illustration is, however, limited, as the author
himself acknowledges. It is not clear how the analysis
could be generalized to other interactions.
82. Communication and Strategic Inference. Prashant Parikh.
This paper presents the Strategic Discourse Model, a model
of communication based on situation theory and game theory.
The author tries to solve the problem of how to get from
meaning to content. Situation theory (Barwise and Perry
1983) provides an account of context, and game theory (as
developed by von Neumann and others) accounts for the
behaviour of rational agents in context. The model lays out
the tools for a systematic account of Gricean
communication, including mutual knowledge and non-natural
Part Ten: Talk In Interaction
83. A Simplest Systematics for the Organization of Turn-
taking for Conversation. Harvey Sacks, Emanuel A. Schegloff
and Gail Jefferson. The paper details the deceivingly
simple rules for turn allocation in conversation. It first
defines the unit necessary for the analysis of
conversation, the Turn-Constructional Unit (TCU). Each
speaker is entitled to at least one TCU, after each one
comes the point where turn can be allocated to any of the
participants, including the speaker herself---this is the
Transitional Relevant Point (TRP). The rules account for
conversations of any length, and with a number of
participants (although there is a bias towards the two-
person conversation, with conversations of four or more
typically breaking up in smaller groups).
The paper also introduces the notion of an adjacency pair,
a first-part and second-part pair, in which the first part
invites a preferred second part. Examples are sequences
such as 'question-answer', 'greeting-greeting' (Schegloff
and Sacks 1973).
Conversation is defined as an interactionally and locally
managed system. One effect of the application of the rules
is that, for spontaneous conversation, they introduce a
motivation for careful listening: the listener must pay
attention in order to know when the next TRP will occur.
84. The Preference for Self-correction in the Organization
of Repair in Conversation. Emanuel A. Schegloff, Gail
Jefferson and Harvey Sacks. Repair in conversation includes
substitution of words (corrections), search for words, or
clarifications. The authors work within the framework
described in the previous paper, to show that there is, in
conversation, a preference for the speaker to repair his or
her own utterances (self-repair), rather than have them
repaired by the interlocutor (other-repair). They also
distinguish between repair initiation, and repair outcome,
the latter being potentially unsuccessful. The paper
details, with numerous real-life examples, the possible
placements of self- and other-repair, and concludes that
speakers tend to self-repair. Other-repair may also be used
in particular contexts, especially adult-child
interactions. In those contexts, other-repair seems to act
as a device for dealing with a child that is learning how
to operate in a system where self-repair is the preferred
85. Response Cries. Erving Goffman. "Utterances are not
housed in paragraphs, but in turns at talk," Goffman states
at the beginning of this paper, emphasizing that utterances
are to be understood within the turn-taking system. He is
interested here in three types of utterances that appear to
produce communicative effects, but do not enter in the
dialogue. The first type is self-talk, which breaches
social rules because it does not qualify as communication.
Goffman proposes a rule, "no talking to oneself in public."
He discusses situations where one-sided talk is acceptable,
such as talking on a public phone, or addressing a pet.
This leads to the second type of utterances discussed,
interjections emitted when one stumbles, or is surprised.
These are not instances of ordinary talk either, because
they are not part of a conversational encounter, or of the
summoning of one. They are, however, not totally
involuntary, and are subject to self-monitoring. When in a
nursery, for instance, one might tone down an expression of
surprise elsewhere expressed as a curse. Finally, he
discusses response cries, vocalizations such as "oops" and
"phew" that constitute ritualized acts in which we align
ourselves with the circumstances of the world around us.
But we also use them even when we are removed from the
situation itself, as when a friend reports on something
unpleasant or painful that happened to him or her, and we
respond "ouch!". In this case, the response cry is part of
the interaction, not removed from it.
Goffman concludes that all three instances of expression
are conventionalized and specifically designed to be used
outside regular conversation. They can, however, enter in
conversation, but always understood as originating from a
ritual function. He challenges linguists to include social
situations, not simply talk, in their study of
86. Collectivities in Action: Establishing the Relevance of
Conjoined Participation in Conversation. Gene H. Lerner.
This paper explores multi-party conversation, especially
the type of interactions that occur when a group
participates in the conversation as a whole. For instance,
the audience during a talk and the students in a classroom
sometimes behave as one participant, for purposes of taking
the floor (choral response) or producing backchannel cues
(laughter, applause, booing). Lerner studies all the
possibilities for addressing a group, and how the group can
respond (in unison, through a speaker person, by conferring
among themselves first). He observes how group interactions
take turns according to the rules laid out by Sacks et al.
(in the article discussed above).
Part Eleven: Discourse
87. The Pragmatics of Discourse. Teun A. van Dijk.
Postscript (1995): The New Pragmatics. Van Dijk defines the
pragmatics of discourse as "the systematic relations
between structures of text and context," which takes into
account not only how certain discourse structures are
determined by the situation, but also how discourse
structures, when introduced in a conversation, become part
of the context itself. In this chapter, he focuses on
sequencing of sentences and speech acts, and the contextual
conditions under which connectives (such as "and", "so",
"because", "since", "therefore") can join two speech acts
in one sentence. Typically, sentence boundaries represent
the boundaries between speech acts. Sometimes, two speech
acts can be accomplished in one sentence. For instance,
"I'll give you the money, but you don't deserve it"
contains both a promise and an assertion. A change in
sentence is, however, required when there is a change in
illocutionary force ("#It's cold in here and please shut
Van Dijk proposes pragmatics as the study of how we
accomplish certain social acts, and he expands on that in
the Postscript, produced for this volume a few years after
the original paper. There, he reviews the definition of
pragmatics and, especially, how the label has been used
throughout the years. He discusses how pragmatics has
overlapped with other disciplines such as sociolinguistics,
the ethnography of communication, discourse and
conversation analysis, and, much less, psycholinguistics.
He dislikes that fuzziness of boundaries, and proposes a
restriction of the term, to cover exclusively the study of
action (parallel to the way syntax studies form and
semantics is concerned with meaning). Nevertheless,
pragmatics always needs to be discourse-based; one should
not confine the study of speech acts to the sentence.
88. Pragmatics and the Description of Discourse. Charles J.
Fillmore. The paper constitutes another suggestion for an
approach to the analysis of discourse. Fillmore considers
that the concern of those studying discourse should not be
either predictive or prescriptive, but rather descriptive
of the competence of speakers in judging the
appropriateness of utterances in context. Fillmore
describes a type of narrative, in which neither author nor
addressee are represented, and where the point of view can
be associated with a character in the narrative. He shows
how discourse type (also known as genre) places constraints
on other syntactic and semantic choices, such as use of
pronouns, personal names, type of verbs, etc.
The article concludes that face-to-face conversation is
the most basic type of discourse. By describing the
syntactic, semantic and pragmatic characteristics of that
type of discourse, we can further our understanding of how
other types of discourse differ from it. However, we can
discover the pragmatic conditions of that basic type of
discourse by looking at other discourses that deviate from
89. Generative Discourse Analysis in America. Susumo Kuno.
The author describes the research in generative syntax in
America at the time he was writing. He characterizes it as
being either purely syntactic or primarily non-syntactic,
with a lack of crossover between the two. He reviews some
of the research up to that time that calls upon discourse
factors to explain syntactic phenomena. In particular, Kuno
summarizes the explanatory power of functional sentence
perspective (FSP)---first outlined by Mathesius (1939), and
developed by the Prague School---in sentence-level
phenomena, such as gapping, pronominalization, and adverb
90. Speech Acts, Discourse Structure and Pragmatic
Connectives. Eddy Roulet. This paper presents a summary of
Roulet's model of discourse, a hierarchical model according
to which each conversation can be analyzed into exchanges,
exchanges into moves, and moves into a master act, possibly
with embedded exchanges, moves or acts. The relationships
between the master act and the elements subordinated to it
are indicated by pragmatic connectives. The model is very
close to Sinclair and Coulthard's (1975) analyses, first
applied to classroom interaction. His interactive relations
are also close to rhetorical relations (Mann and Thompson
1988), which have also been studied in relation to
discourse markers (Knott and Dale 1994).
The analysis of speech acts in discourse, Roulet points
out, has been illustrated in a few studies. He wants to
take it one step further, and apply it monologic discourse,
but from a point of view where the monologue is considered
as part of an ongoing dialogue (Bakhtin 1981), i.e., as a
move that constitutes part of an exchange, whether present
or implicit. For instance, a newspaper editorial can be
seen as a reactive move to some current issue, but also as
an initiating move, inviting a response from the readers.
91. Discourse Analysis: A Part of the Study of Linguistic
Competence. Ellen F. Prince. "'Discourse analysis' is
without doubt one of the most widely used and loosely
defined terms in the entire field of linguistics." With
this statement, Prince starts a discussion of the field of
discourse analysis, and its scope. She believes that
discourse analysis should be concerned, not with the
utterance of words in context, but with the influence of
context in linguistic form. And linguistic form refers to
the competence of a speaker in his or her language
(pragmatic linguistic competence), independently of other
social or cultural norms.
Prince studies some discourse functions in different
languages (English and Yiddish, mainly), to show that the
choice of linguistic expression in those has to be part of
the linguistic competence of the speaker, something that is
learnt with the language. For instance, she shows how, in
the same context, English uses an it-cleft, whereas Yiddish
uses a different construction. Thus, there is nothing
iconic or cognitive determining the use of the it-cleft; it
is rather a language-specific discourse constraint. The
article concludes with a call to include this type of
discourse analysis---the study of speakers' knowledge of
how to use linguistic forms in discourse---as part of any
theory of linguistic competence.
92. On the Informativeness Requirement. Rachel Giora.
Postscript (1995). Giora explains the informativeness
constraint, which states that (narrative) texts proceed
from a generalization to the most informative content, the
Discourse Topic. This is based on Grice's (1975) maxims,
and on Relevance theory (Sperber and Wilson 1986), with
relevance being viewed here as informativeness. The author
shows, with a number of experiments, that readers prefer
texts where the most informative message occupies
paragraph-final position. She compares this progression to
the organization of categories (e.g. Rosch 1973). A text is
organized like a category, going from the most typical
member (the least informative message) to the least typical
member (the most informative message). In the Postscript,
Giora discusses her work after the paper, and how it has
expanded the informativeness requirement to cover non-
narrative texts, jokes, and irony.
In my evaluation, I will try not to repeat what was said
by previous reviewers of the collection, although I do
agree with most of the comments they made. The evaluation is
mainly concerned with two areas: the format of the volume
and the series, and the choice and content of the articles
in this volume.
First of all, and with regard to format, the series is
obviously conceived to be purchased as a set. The
individual volumes do not have information on previous
place and date of publication for each of the papers, and
the introductions to the sections often make reference to
other volumes in the series for further detail (for
example, for biographical information about the authors).
The individual volume felt like a bit of an orphan by
itself, since it contained no general preface to the
series, or an explanation of its place and relation to
[It was the review editors' decision to split up the
anthology among five different reviewers, so that their
experience will not be shared by many readers. However,
anyone who borrows a single volume from a library will
have something like their experience. --Eds.]
Within the volume itself, it is difficult to browse the
articles, because the running header is the same for each
part (i.e., the only headers are "Communication", "Talk in
Interaction", "Discourse"). It would have been really
useful to add an index to each volume, although I
acknowledge this would have been a colossal task, on top of
what must have been an enormous effort of selection and
compilation of the articles.
The editor provides very little information on the choice
of particular papers for the collection, except to say that
they are "major contributions" (we assume that is the
reason they were chosen for the collection). The
introductions are uninformative about the choice of
articles, and are uneven with respect to the biographical
information provided for each author. The articles in the
Communication section do not seem to relate nicely to each
other, except for the debate between Searle and Chomsky.
Dummett's article was probably chosen because it mentions
language and communication. The article, however, is not
the best example of a discussion of communication; it is a
proposal for what philosophy of language should concentrate
on. There is no unifying definition of communication, and
the differing definitions are not presented in a way so
that they can be contrasted or compared. One also finds a
gap in the lack of discussion of animal communication.
Part Ten, Talk in Interaction, on the other hand, is more
coherent because the four papers describe work within the
same framework, of conversation analysis and interactional
The papers in the last part, Discourse Analysis, do
represent a good cross-section of the work being done on
discourse analysis (albeit, of course, not an exhaustive
compilation). It is interesting to note how all the papers
have some programmatic aspect: they all attempt to define
and characterize discourse analysis. The authors propose
objects of study and methodologies for the area. This is
probably a reflection of the wide scope of the field, and
also a result of the vagueness---polysemy at best---of the
term "discourse." It is also a clear indication that there
is no consensus, among linguists, about what it is that the
study of discourse needs to include, and its relation to
pragmatics (the section on discourse analysis is
a very small part of a six-volume series on pragmatics).
In all, the collection is welcome from an aesthetic and
practical point of view, since it provides a very nice
binding for important articles in the field. Lengthier
introductions and an index would have given it added value.
Bakhtin, M. (1981) Discourse in the Novel. In Michael
Holquist (ed.) The Dialogic Imagination: Four Essays by M. M.
Bakhtin. Austin: University of Texas Press. Translated by
Caryl Emerson and Michael Holquist. 259-422.
Barwise, J. and J. Perry (1983) Situations and Attitudes.
Cambridge: MIT Press.
Grice, P. H. (1975) Logic and Conversation. In P. Cole and J.
Morgan (eds.) Syntax and Semantics 3: Speech Acts. New York:
Knott, A. and R. Dale (1994) Using Linguistic Phenomena to
Motivate a Set of Coherence Relations. Discourse Processes 18
Mann, W. and S. Thompson (1988) Rhetorical Structure Theory:
Toward a Functional Theory of Text Organization. Text 8 (3):
Mathesius, V. (1939) On Information-Bearing Structure of the
Sentence. Harvard Studies in Syntax and Semantics 1.
Rosch, E. (1973) On the internal structure of perceptual and
semantic categories. In T. E. Moore (ed.) Cognitive
Development and the Acquisition of Language. Hillsdale, NJ:
Schegloff, E. and Sacks, H. (1973) Opening Up Closings.
Semiotica 8: 289-327.
Searle, J. (1972) Chomsky's Revolution in Linguistics. New
York Review of Books. June 29.
Shannon, C. (1948) A mathematical theory of
Communication. Bell System Technical Journal, vol. 27, pp.
379-423 and 623-656.
Shannon, C. and W. Weaver (1949) The Mathematical Theory
of Communication. Urbana, Illinois: University of Illinois
Sinclair, J. and M. Coulthard (1975) Towards an Analysis of
Discourse: The English Used by Teachers and Pupils. London:
Sperber, D. and D. Wilson (1986) Relevance: Communication and
Cognition. Blackwell: Oxford.
Strawson, P. (1970) Meaning and Truth. Inaugural Lecture,
University of Oxford, November 5, 1969. London: OUP.
ABOUT THE REVIEWER:
Maite Taboada is an Assistant Professor in the Department
of Linguistics at Simon Fraser University. She works on
discourse and computational linguistics.