LINGUIST List 14.1934

Tue Jul 15 2003

Review: Discourse Analysis/Pragmatics: Carston (2002)

Editor for this issue: Madhavi Jammalamadaka <>

What follows is a review or discussion note contributed to our Book Discussion Forum. We expect discussions to be informal and interactive; and the author of the book discussed is cordially invited to join in. If you are interested in leading a book discussion, look for books announced on LINGUIST as "available for review." Then contact Simin Karimi at


  1. Gloria Cappelli, Thoughts and Utterances: The Pragmatics of Explicit Communication

Message 1: Thoughts and Utterances: The Pragmatics of Explicit Communication

Date: Tue, 15 Jul 2003 15:31:53 +0000
From: Gloria Cappelli <>
Subject: Thoughts and Utterances: The Pragmatics of Explicit Communication

Carston, Robyn (2002) Thoughts and Utterances: The Pragmatics
of Explicit Communication, Blackwell Publishing.

Announced at

Gloria Cappelli, PhD student, Dept. of English Studies, 
University of Pisa, Italy.


The book is made up of four main chapters plus an 
introduction, two appendices, references and indexes. 

In the introduction, the fundamental assumptions on which 
the relevance-theoretic approach is grounded are sketched. 
The author illustrates the Relevance Theoretic notion of 
Pragmatics as a capacity of the mind, a real information-
processing system for ''interpreting human ostensive 
communicative behaviour'' (p.4) . In this sense, pragmatics 
stops being just an addition to natural language semantics 
and becomes a real object of study itself, although 
maintaining strict interaction with semantics. In this 
cognitive pragmatic approach, the semantics/pragmatics 
interface is connected to the distinction between two 
strictly interrelated cognitive processes, namely, semantic 
decoding and pragmatic inference.

In the 1st chapter, ''Pragmatics and Linguistic 
Underdeterminacy'', the author starts by illustrating the 
distinction between linguistic meaning, encoded by 
linguistic forms, and speaker's meaning. A comprehensive 
survey of the most studied linguistic phenomena showing 
saying/meaning divergences is provided. All the cases 
described support the hypothesis that it is necessary for a 
hearer to undertake a process of pragmatic inference which 
allows him/her to understand what the proposition expresses 
and what the speaker implicates. This hypothesis is further 
developed into the ''Underdeterminacy Thesis'': ''the 
linguistic semantics of the utterance, that is the meaning 
encoded in the linguistic expressions used, [...], 
underdetermines the proposition expressed (what is said)'' 
 The Underdeterminacy Thesis is illustrated in 
detail and the alternative hypotheses on linguistic 
ambiguities and reference assignment (especially those 
pertaining to the truth-conditional semantics and to the 
logico-philosophical tradition) are considered, compared 
and, when possible, integrated via a relevance-theoretic 
reinterpretation of compatible questions. Some other 
proposals are refuted in a lucid, articulate way. Many 
sources of linguistic underdeterminacy and several 
pragmatic strategies employed by the hearer in order to 
arrive at the correct interpretation are illustrated, such 
that underdeterminacy is finally defined as an essential 
feature of the relation between linguistic expressions and 
the thoughts they are used to express. The author proposes 
her strong ''essentialist view'', comparing her position with 
other weaker views on the question and refuting several 
philosophical principles allowing for the possibility of 
''eternal reference'' and ''eternal predication''. Carston 
illustrates the relevance-theoretic model for 
communication, underlying the primacy of the human capacity 
to infer the mental states of others and to attribute them 
through the pragmatic inferential capacity. She recognises 
the fundamental role of the Background (in a version 
similar to Searle's Background) and concludes that 
underdeterminacy is not only an inalienable characteristic 
of linguistically encoded meaning but of thought in 

The 2nd chapter, ''The Explicit/Implicit Distinction'', like 
the first chapter, deals with theoretical issues concerning 
the semantics/pragmatics and the explicit/implicit 
distinctions. The author analyses different ways in which 
the distinction between the proposition explicitly 
expressed by a speaker and the implicated propositions can 
be determined. Moreover, she defines which are the 
different levels which play a role in a cognitive pragmatic 
account of the representations and processes required in 
the interpretation of an utterance. 
 She comments on different views of the explicit/implicit 
distinction in verbal communication held by other linguists 
and philosophers and this critical survey provides strong 
support for her argumentation. The guiding thread of the 
discussion is Grice's pivotal work, and other theorists' 
views are considered insofar as they support, modify or 
react against his positions. 
 Carston starts from Grice's approach to the ''saying'' and 
''implicating'' distinction. First, she provides a survey of 
Grice's theory about ''what is said'', the complete 
propositional (truth-conditional) content of the basic 
speech-act of an utterance, and ''what is implicated'' and 
then she describes the relevance-theoretic 
explicature/implicature distinction, which is considered a 
distinction between two different ways of deriving 
communicated assumptions. 
 Explicatures are defined as ostensively communicated 
assumptions inferentially developed from one of the 
linguistically given logical forms encoded by the 
utterance. They are further specified to include in the 
account several types of ''subutterances'', although the 
central concepts of ostensive communication and pragmatic 
enrichment maintain their primacy. Explicatures are then 
made distinct from the ''propositions expressed'', which, 
within the relevance-theoretic approach, are connected with 
an idea of a ''representational level of explicit content 
that may not be communicated (speaker-meant)'' (p.133). 
Following the argumentation developed throughout the first 
two chapters, Carston starts casting some doubt on the 
necessity for different explicatory levels other than the 
encoded logical form, explicatures and other contextual 
enrichments. This strong conclusion will be drawn in the 
final chapter. 
 Implicatures are defined as ostensively communicated 
assumptions which are derived exclusively by pragmatic 
inference and they are further divided in implicated 
premises and implicated conclusions. In Carston's account, 
only Gricean ''conversational implicatures'' figure (and they 
are considered an aspect of the proposition expressed), 
while phenomena categorized as ''conventional implicatures'' 
are usually considered as procedural constraints on 
inferential processes whose conceptual content is supplied 
by pragmatic inference and are compared with the relevance-
theoretic notion of ''procedural meaning''. Following the 
underdeterminacy thesis, the author also rejects the 
distinction between particularized and generalized 
implicatures, in favour of a continuum going from 
occasional context-dependent cases to those cases which are 
generally present in the vast majority of contexts and for 
which very specific contextual circumstances are required 
in order to be blocked. 
 Carston illustrates the global comprehension process as 
postulated by the Relevance Theory and adds further 
specifications building on her underdeterminacy thesis. 
 Once an utterance is uttered, it carries its own 
presumption of relevance. The hearer can retrieve the 
intended meaning relying on the so-called ''relevance 
theoretic comprehension strategy''. The first step is the 
decodification of the ostensive stimulus, in order to 
derive the logical form, an incomplete conceptual 
representation with variables whose value must be fixed, 
and with slots to fill in. Tentative interpretive 
hypotheses concerning the intended referents are made on-
line and subsequently confirmed or rejected, according to 
whether they contribute or not to the interpretation which 
meets the expectation of relevance, that is, considering 
interpretations in order of accessibility and stopping when 
the expected level of relevance is reached. The logical 
form is so pragmatically developed through processes of 
disambiguation, saturation and free enrichment 
 The author's position in favour of the underdeterminacy 
thesis is finally outlined through a critical survey of 
several accounts provided by outstanding voices from the 
philosophical and linguistic field of research. 
Travi's, Recanati's and Bach's proposals concerning the 
comprehension process and more specifically the 
saying/implicating distinction are analysed and integrated 
in Carston's account or progressively discarded. The level 
of the Gricean ''what-is-said'', intended as a minimal 
proposition, is finally put aside in favour of a three-
level account of the representations and processes required 
in the interpretation of utterances (logical form, 
saturation and free enrichment), although the possibility 
for the necessity of a fourth representational level in a 
philosophical-semantic analysis is left open. 

In the 3rd and 4th chapters, ''The Pragmatics of 'And'-
Conjunction'' and ''The Pragmatics of Negation'', the 
underdeterminacy thesis is applied to the analysis of two 
of the most studied phenomena within the logico-semantic 
 In both cases, Carston demonstrates that the relations 
expressed by 'and' and the fixation of the negation scope 
are pragmatically determined. The semantics of these 
natural language operators only provides a minimal 
framework, which needs being pragmatically augmented and 
complemented by inferential mechanisms contributing to the 
explicit content of the utterance. 
 The account provided by the author is strongly cognitively 
oriented, and it is based on the hypothesis concerning the 
ways in which the human mind conceives of the world and 
organises and stores encyclopaedic knowledge. According to 
this perspective, some relations are typical of the human 
cognitive system (i.e. cause-consequence relation), while 
other ones are developed out of our experience of the world 
(i.e. temporal relations). 
 Chapters 3 and 4 open up the way for the possibility of 
pushing even further the underdeterminacy thesis and of 
positing the existence of a process of on-line concept 
construction, which is pursued in the last chapter.

The 5th chapter, ''The Pragmatics of On-line Concept 
Construction'', is probably the richest in proposals, and 
the one in which most questions are raised. It deals with 
lexical semantics and with those processes of loosening and 
broadening of the encoded conceptual content which have 
been traditionally considered as not contributing to the 
proposition expressed. Carston doesn't support this view 
and shows that the processes of narrowing and broadening 
of the conceptual content are both reflected in the 
proposition expressed and, therefore, they are seen as 
contributing to the explicit level of communication. 
 In the relevance-theoretic framework, ''atomic concepts'' are 
considered nodes in memory which make available three kinds 
of information: the logical entry, (consisting of a set of 
inference rules which capture certain analytic implications 
of the concept), the encyclopaedic knowledge (including 
many different types of knowledge, from scientific to 
cultural specific knowledge and beliefs), and the lexical 
properties (specifying the phonetic, phonological and 
syntactic properties of the linguistic form encoding the 
concept). In the process of retrieving the intended 
meaning, there are not two distinct processes at work, one 
of narrowing and one of loosening, but only one process 
with two possible outcomes, deriving from the act of 
''picking and choosing from among the elements of logical 
and encyclopaedic information that are made available by 
the encoded concept'' (p.334). The difference consists only 
in whether, in this process of on-line ad hoc concept 
construction, the logical properties are retained (like in 
the case of narrowing) or dropped (like in the case of 
loosening), but whatever the outcome is, it still 
represents a move away from strict literalness. The 
derivation of the intended meaning is seen as involving a 
mutual parallel adjustment of these processes until the 
addressee's expectation of relevance is met. 
 The analyses is applied to several kinds of figurative use, 
 such as metaphor and hyperbole, as well as to the semantics 
of certain lexical items showing polisemy or vagueness, for 
which the idea is proposed that, like all lexical items, 
they encode only a very abstract and general concept 
providing a basis for pragmatic processes of enrichment 
or, pushing this idea even further, that they function as 
pointers to a conceptual area. 
 The very last paragraph of the chapter summarises the 
account of explicit communication supported and depicted in 
the book: a radical version of the underdeterminacy thesis 
is supported which allows only for a three-level 
interpretive process, where the ostensive stimulus is 
decoded and the derived logical form is then enriched by 
pragmatically filling and adjusting ''the semantic 
scaffolding provided by the linguistic expression'' itself 
(p.366) , (explicature) and, possibly, through other 
totally pragmatically derived information (implicatures). 

The book also contains two appendixes, ''Appendix 1: 
Relevance Theory Glossary'' and ''Appendix 2: Gricean 
Conversational Principles'', and more than 30 pages of 
References and Indexes.


''Thoughts and Utterances'' is much more than the application 
of the Relevance Theory: it represents a huge step forward 
in the theory itself, strongly founded in pre-existing 
debates and analyses. The subtitle, ''The Pragmatics of 
Explicit Communication'', doesn't do it justice: the book is 
definitely about explicit communication as a whole, 
although the main goal of the author is to demonstrate the 
essential role of pragmatics in recovering utterance 
 What is immediately evident, from the very first pages, is 
the strongly cognitively oriented but interdisciplinary 
nature of the study: the author covers perspectives ranging 
from the philosophical to the logico-semantic and the 
neurological ones on such debated issues as the 
explicit/implicit and the semantics/pragmatics 
 Carston provides an incredibly extensive survey of several 
traditions of research which, although from different and 
often irreconcilable perspectives, have contributed to the 
lively debate about these fundamental questions in 
linguistics. Her synthesis is carried out critically and in 
order to provide further evidence to the account of 
explicit communication that she proposes and to support the 
well delineated divisions that she makes. She doesn't 
dismiss all contributions coming from different areas of 
study (as is often the case), but tries to reinterpret some 
of the suggestions according to her framework of analysis 
and to include them in her account. This approach gives her 
work a philosophical taste, and, at the same time, it 
provides the research with steady foundations in order to 
build further within the relevance-theoretical framework. 
 In her attempt to provide a comprehensive panorama of the 
debate, the author also presents Relevance Theory as 
developed by Sperber and Wilson (1986, 1995) as well as the 
successive contributions added by Blakemore (1992), 
Infantidou (2001), Papafragou (2000) and others. This 
aspect, together with the critical analysis mentioned above 
and with an in-depth analysis of Grice's work and of the 
development of the Gricean theory by Levinson, makes 
''Thoughts and Utterances'' also a useful, but not at all 
basic, handbook even for all those students and researchers 
who approach the processes guiding explicit communication 
for the first time. The book represents, indeed, a complete 
state of the art on a specific topic, although even a 
little background in linguistics and in philosophy of 
language is recommended, given the non-elementary nature of 
the book. 
 All the chapters are very rich in content: sometimes this 
richness makes the argumentation slightly difficult to 
follow, although everything is made clear in the end and 
all the different considerations find their exact place in 
the puzzle. The author was clearly aware of this 
argumentative complexity, since at the end of almost each 
chapter, and often even at the end of some paragraphs, she 
provides the reader with a brief summary of the points she 
made. These summarising lines are always extremely useful 
and help the reader follow the discussion.
 As far as the readability is concerned, it must be 
signalled that the notes at the end of each chapters are as 
important and full of useful information and input as the 
main text: they are often used as a ''box'' for interesting 
material which would have led the discussion too far from 
the actual topic of the single paragraphs. The same holds 
for the two appendices at the end of the book. The first 
one provides the reader with Gricean Conversational 
Principles which are constantly mentioned in the book but 
never explicitly reported. The second appendix deals with 
the relevance-theoretic terminology. It provides an 
essential dictionary containing all the keywords necessary 
to the understanding of the discussion, and the single 
lexical items are defined according to Carston's use of the 
term, especially when it doesn't perfectly correspond to 
some previous well-established usage. 
 Of course, the value of this book is not only in the 
critical state of the art that the author provides. Robyn 
Carston makes the Relevance theory progress providing 
researchers within this framework with new tools. She 
redesigns the model of explicit communication previously 
proposed according to her strong view of the 
Underderterminacy Thesis, ruling out the intermediate 
representational level of minimal proposition, which 
departs only minimally from encoded meaning but already has 
truth-conditions. In 'her linguistic' account (and she is 
very precise about this specification), there is only room 
for the logical form and for the enrichment it receives 
through different pragmatic inferential processes, driven 
from the linguistically encoded elements themselves 
(explicature) or totally pragmatically derived 
(implicature). The possibility for an other intermediate 
representational level to be useful is left open, in 
consideration of the fact that all philosophical accounts 
seem to include it. What changes, though, is the 
explicatory purpose. Linguistic pragmatics doesn't want to 
account for the truth conditions of propositions, and 
therefore it doesn't need that intermediate level.
Robyn Carston also attempts to make headway in lexical 
semantics. She tries to apply her theory also to word 
meaning, in an attempt to account for ambiguities, 
vagueness and polysemy. She hypothesises the online 
construction of ad hoc concepts (relative to a certain 
context), starting from a raw basis provided by a lexical 
item to the inferential processes. In order to support this 
explanation the author also provides a model for the 
meaning provided by concepts as summarized above. She 
leaves many questions open for further research, but the 
path she opens seems promising, since it seems applicable 
also to figurative uses of language.
 Robyn Carston's book is not an isolated voice. It perfectly 
collocates in contemporary cognitive-linguistic research 
and shows numerous points of contact with other linguistic 
trends of study. 
 A similar view of communication is proposed in Bertuccelli 
Papi (2000), a study of implicitness in text and discourse. 
Implicitness and explicitness are considered gradable 
entities and the former is seen as a cover term for several 
phenomena (inexplicit, implicated and subplicit) related to 
different types of knowledge which the hearer can select in 
order to set up the context of interpretation. According to 
Bertuccelli Papi, the input is filtered and the most 
relevant information is selected by an evaluative 
mechanism. As I already mentioned, in Carston's book, the 
role of the hearer and of his ''intention-reading'' effort is 
widely recognised. However, contrarily to what happens in 
Bertuccelli Papi (2000), where a fundamental role is 
explicitly attributed to attitudes, here, the importance of 
attitudes in the communicative process (as a fundamental 
source for the identification of implicit meanings) is 
neglected. It is possible that attitudes are implicitly 
included in the ''speaker's intentions'', or that they are 
considered part of the situational context or of the 
background information. 
 Carston's proposals and hypothesis seem to be in line with 
the most recent development of lexical semantics and of the 
cognitive-functional language acquisition studies as well. 
The online construction of ad hoc concepts from a steady 
core of meaning finds support in Croft and Cruse (to 
appear). In line with the recent development of Cognitive 
Psychology, they support the Dynamic Construal Approach, 
according to which there are no fixed concepts: concepts 
and structural relations between lexical items emerge in 
actual situations of use. They speak of ''meaning 
construal'', which seems to parallel Carston's on line 
concept construction. More point of contact could be 
outlined between these two approaches to the lexicon but 
this would lead us too far. I will only mention that also 
in Croft and Cruse, a very general steady element of 
meaning is hypothesised which receives a particular 
interpretation in a particular context according to several 
 Carston's communication process is essentially a mind-
reading pragmatic process: the hearer has to imagine in a 
certain situation what the communicative intention of the 
speaker might be and to interpret the ostensive stimulus 
accordingly, trying to retrieve the most relevant 
information. The most recent studies applying cognitive-
functional approaches to language acquisition, this 'mind-
reading perspective', or more precisely this 'intention 
reading', is considered as one of the essential features of 
the human mind involved in the development of language. 
Tomasello (2003) sees intention-reading as uniquely human 
and as one of the foundational processes for the emergence 
of language, both phylogenetically and ontogenetically.
''Thoughts and Utterances'' is a very rich study, which marks 
a decisive step forward in the scientific study of explicit 
communication thanks to the solid linguistic, philosophical 
and scientific theoretical background witnessed by a very 
large bibliography of hundreds of titles. It is a work that 
shows links to the most recent trends in linguistics 
research and in cognitive studies in general, which is 
enough to make of Robyn Carston's book a necessary reading 
for all the people interested in the actual functioning of 
ostensive linguistic interaction.


Bertuccelli Papi, M. (2000), Implicitness in Text and 
Discourse. Pisa: Edizioni ETS.

Blakemore, D. (1992), Understanding Utterances: an 
Introduction to Pragmatics. Oxford: Blackwell.

Croft, W. & Cruse, A. (to appear), Cognitive Linguistics. 
Cambridge: CUP.

Infantidou, E. (2001), Evidentials and Relevance, 
Amsterdam: John Benjamins.

Papafragou, A. (2000), Modality: Issues in the Semantics-
Pragmatics Interface. Oxford: Elsevier Science.

Sperber, D. & Wilson, D. (1986, 1995), Relevance: 
Communication and Cognition. Oxford: Blackwell; Cambridge, 
MA: Harvard University Press.

Tomasello, M. (2003), Constructing a Language: A usage-
based Theory of Language Acquisition. Cambridge, MA: 
Harvard University Press.


I am a PhD student in Linguistics at the department of 
English Studies of the University of Pisa, Italy. In the 
past, I studied the acquisition of Italian as a Second 
Language and I am now interested in English linguistics, 
especially in the lexical semantics of cognitive verbs from 
a cognitive functional perspective. 
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue