AUTHOR: Vega Moreno, Rosa E.
TITLE: Creativity and Convention
SUBTITLE: The pragmatics of everyday figurative speech
SERIES TITLE: Pragmatics & Beyond New Series; 156
PUBLISHER: John Benjamins
Jörg Jost, University of Cologne, IDSL II
The main concern of Vega Moreno's book is to provide ''a pragmatic inferential
approach to the comprehension of everyday metaphorical and idiomatic speech.''
This approach in its inferential processing character does not vary from
ordinary literal utterances (pp. 1- 2). Vega Moreno's study is situated in the
framework of Relevance Theory (RT) (Sperber & Wilson  1995). She
consequently argues from the wider perspective of human (creative) cognition. RT
presupposes that the human mind is (biologically) predisposed to efficiently
process information. In doing so, the theory claims, the human mind is generally
searching for relevance. Sperber and Wilson (1995:49) have called this
automatism ''Cognitive Principle of Relevance''. It is this tendency of the human
mind which also makes manipulations and predictions about the mental states of
others possible. Thus, a speaker may indicate (by giving a certain stimulus) his
communicative intention (i.e. manipulation) while relying upon the hearer
processing the given information in the intended way (i.e. prediction). For
scholars within the tradition of RT, inferential communication is ''ostensive'',
because it involves in addition to the informative intention an extra
communicative intention (Sperber & Wilson 1995:50- 54, see also Sperber & Wilson
2008:610- 614) which has to be recognized for efficiently processing information
(Sperber & Wilson 1995:50). With her study Vega Moreno explicitly refers to this
basic RT idea, when she argues for ''the comprehension of linguistic utterances
in general, and of figurative utterances in particular'' as being a ''selective
process'' whose result is ''either creativity or conventionalization'', for both
involve ''the construction of new ad hoc concepts'' (pp. 2, 3).
The first chapter adumbrates the human creative cognition perspective and
outlines ''selectivity'' as an important property for information processing. More
precisely, creativity is described as the cognitive ability of ''constructing,
combining and modifying mental representations in thinking or in understanding
what others think'' (p. 5). It is this cognitive process and the minds'
selectivity that are of outstanding interest for the argumentation of Vega
Moreno (p. 6). She finds evidence in psychological research for how the human
mind selects information by creating ad hoc categories. As the name implies,
these categories are not permanently stored but instead they are built in order
to carry out a cognitive task online, e.g. a problem-solving task. RT claims
that the human mind possesses stable concepts and furthermore has the ability to
build new concepts ad hoc, for the latter can be described as subsets of the
former. In line with this assumption, for which RT provides some empirical
evidence, it is argued that selection ''results in a representation which denotes
a subset (or superset) of the general category'' when merely some features of a
(stable) concept are accessible while processing a certain task (p. 21). The
newly constructed concept is what is then called an ad hoc concept.
Chapter two explores the theoretical basis for Vega Moreno's study on the
pragmatics of everyday figurative speech: Sperber & Wilson's ( 1995)
Relevance Theory. Vega Moreno briefly outlines the connection between
communication and cognition. Starting from the very basic RT notion of the
Cognitive Principle of Relevance and people's - certainly limited - ability to
read other people's minds (see e.g. Wilson 2000, Papafragou 2002), Vega Moreno
traces the main line of argument of the RT comprehension procedure. Brought
about by evolution, this procedure ''together with the notion of optimal
relevance and the communicative principle of relevance, are the key components
of relevance- theoretic pragmatics'' (p. 35). RT suggests that implicatures
derive not consecutively but parallel to explicatures when interpreting an
utterance. The interpretation process is driven by relevance and, consequently,
is selective. According to this, Vega Moreno argues that understanding is a
creative process (p. 43). When interpreting an utterance, the hearer's access to
contextual assumptions a speaker makes is guided by the comprehension procedures
RT explains. Vega Moreno relates these procedures to the so-called spreading
activation models of memory. According to these, activating a concept initiates
the further activation of semantically related concepts (in a certain degree of
semantic depth) and so forth. From the lexical-pragmatic RT position, the number
of constructed and represented cognitive concepts is much larger than the number
of lexicalized linguistic concepts (p. 47). Factors influencing the pragmatic
process of understanding an utterance, i.e. of processing relevant information,
are the narrowing and the broadening of encoded concepts in contexts, which are
''two different instantiations of a single process of pragmatic fine-tuning of
the linguistically specified meaning of a word'' (p. 49). Accordingly, for RT
''people often construct ad hoc concepts during utterance interpretation by
broadening or narrowing the encoded concepts. Furthermore, [RT; J.J.] proposes
that it is this ad hoc concept constructed on-line [...], and not the concept
encoded by the word, which the hearer takes as a constituent of the explicature''
(p. 50). Under this perspective, the pragmatic fine-tuning process remains the
same, both when a literal intended utterance and a metaphorical intended
utterance are interpreted. This is the main theoretical assumption that Vega
Chapter three focuses on metaphors under two perspectives, since metaphors are
viewed as being both a form of creative language use and a touchstone for every
theory of meaning. Vega Moreno asks how existing theories of metaphors explain
metaphorical meaning construction. She critically takes into consideration
different approaches: the classical view of metaphor that understands metaphor
as a derivation from literalness and truthfulness; theories that understand
metaphorical interpretation as derived from an utterance's literalness; the
comparison view and matching models; and, discussed in more detail, the
Class-Inclusion Theory by Glucksberg & Keysar (1990), which is important for
Vega Moreno's further argumentation (more precisely: the property- attribution
hypothesis). She basically agrees with the interactive idea of this approach,
but criticizes that no explanation is given of ''what [in the metaphorical
meaning construction process; J.J.] determines the formation of [...] different
ad hoc categories [...] on each occasion'' (p. 73). On the background of several
empirical studies and in line with different theories she discusses the
importance of the emergence of features for metaphorical meaning construction.
In this context, both selection processes and attributing processes come to the
fore. Yet, it is not enough to know ''the selection of vehicle properties and
attribution of these properties'' for understanding how interpretation of
metaphors works. It is more important to ask how properties are transformed,
when a metaphorical expression is interpreted, ''into properties that can be
appropriately attributed'' (p. 82). For this widely unresolved problem (i.e. the
transformation problem) a RT approach of metaphor comprehension in accordance
with empirical psycholinguistic evidence may offer an adequate explanation.
Chapter four elaborates the announced RT approach of metaphor interpretation. RT
allows overcoming the literal priority as claimed by some non- cognitive
metaphor approaches. It does so by not clearly distinguishing between different
shapes of loose use of language and between loose use and literal use of
language. Viewed from a RT perspective, a speaker communicates both the
expressed proposition and a set of implications of that proposition (the encoded
concept). While selecting encyclopedic assumptions from the encoded concept, a
hearer ad hoc builds a new concept. ''It is this new, broader ad hoc concept
[...] constructed during the comprehension process, and not the encoded concept
that [...] is taken to be a constituent of the proposition expressed by the
utterance'' (p. 92). Whether the utterance is understood as literal or as a form
of loose use, depends, according to RT, on the hearer's selection of assumptions
included in that ad hoc concept; the selection follows a hierarchical order of
cognitive accessibility. The criterion guiding the interpretation process is
'relevance'. The RT pragmatic processes involved in interpreting utterances like
''John is a soldier.'' or ''I am afraid about the divorce. My husband's lawyer is a
shark.'' are above all inferential processes. Such transformation processes
involve the pragmatic adjustment of the explicit content of that utterance, of
contextual assumptions, and of implicatures as well as it involves the
lexical-pragmatic fine-tuning in understanding the utterance either literally or
figuratively (p. 95). The specification of metaphorical force, and more
generally the distinction of looseness/creativity and convention/literalness of
an utterance, is regarded as depending on the requirement of pragmatic
adjustment, i.e. as depending on ''the strength of the explicatures and
implicatures which the hearer takes the speaker to have intended to convey and
the amount of processing required to derive them'' (pp. 112, 113). In other
words: the more creative the metaphor the more adjustment is required. Given the
RT premise of our cognitive disposition being triggered by relevance, frequency
of use and familiarity of an expression reduce the cognitive effort of the
hearer searching for relevance when interpreting an utterance. This is what Vega
Moreno calls ''pragmatic routine'' (pp.116- 119).
Chapter five distinguishes the RT approach of metaphor from Class Inclusion
Theory (CIT) and Conceptual Metaphor Theory (CMT). RT and CIT share the
assumption that in a metaphorical expression, the metaphorical word's content is
part of the explicit content of that utterance (in the utterance ''My lawyer is a
shark'' the content of 'shark'). According to both theories, the ad hoc concept
SHARK*, which entails 'aggressiveness of animals and people which makes one fear
...', is part of the explicit content of the utterance but differs from the
concept SHARK encoded by the word 'shark'. The theories vary, however, in their
explanation of how the conceptual construction is processed while understanding
the utterance (pp. 121- 130). In contrast to both of these theories CMT does not
explain (and does not want to do so) online comprehension of metaphors but
primarily looks at 'why' we live in metaphors. For Vega Moreno it is of primary
importance that RT allows explanation of the comprehension processes of
communication – be they literal or figurative. She criticizes CMT because not
only the concept of 'conceptual metaphor' remains unclear but also its role in
structuring our thought and in comprehending metaphorical language use.
Chapter six and seven report mainly psycholinguistic research on comprehension
processes of idiomatic expressions. Idioms share the ''ability to move back and
forth between literalness and looseness, creativity and standardisation'' (p.
185) and, hence, are interesting phenomena for an RT approach. Vega Moreno first
discusses the analyzability and compositionality of idioms. She then points out
that it is necessary to examine how the transparency of idioms in online
processing (p. 174), i.e. the relation between the single words of an idiomatic
expression and the expression as a whole, relates to their role in online
language processing. Vega Moreno argues that RT offers an answer to these
research questions. The author claims that hearers process idioms like ''spill
the beans'' always the same, i.e. irrespective of their degree of literalness or
looseness. She further claims that comprehension processes of idioms do not
differ from those of non-idiomatic expressions.
Chapter eight finally summarizes first the argument of ordinary language use
floating between creativity and conventionality, and secondly that of pragmatic
routines guiding hearers through these free floating boundaries of utterances.
It is this ''psychology of routines'' (p. 221) that Vega Moreno asks for and that
she defends against the background of general psychological and psycholinguistic
evidence about utterance comprehension as gained by RT.
This is a valuable book for scholars interested in pragmatic processes of
understanding figurative speech and sharing cognitive pragmatic presumptions.
With 'creativity' and 'conventionality' the book highlights two 'aggregate
states' of language use resulting while interpreting everyday (figurative)
speech. Embedded in the RT framework Vega Moreno starts her argumentation from a
critique of traditional models of figurative and conventional language use on
the basis of the criterion of familiarity. Under this perspective meaning
consists in retrieving static patterns from memory, and not - as RT suggests -
in online processing linguistic expressions on the basis of both their encoded
literal meaning and of the context. Consequently, traditional models fail in
explaining idiom variants and possibly different meanings resulting from one and
the same linguistic expression (pp. 1- 3). This critique is the very starting
point of the book.
There are two main arguments developed which seem to me the most interesting and
fruitful contribution of the book to explaining pragmatic (inferential)
processes: First the assumption that not the encoded concept (the concept PIGSTY
encoded by the word _pigsty_) but the construction of a new and broader ad hoc
concept (MAX'S ROOM IS A PIGSTY*) is a constituent of the proposition expressed
with an utterance, i.e. of the explicature. When interpreting an utterance as
e.g. ''Max's room is a pigsty'' the hearer chooses the 'right' aspects of the
constructed ad hoc concept by processing them in order of their accessibility.
He thereby is guided by his permanent search for relevance during which he
always chooses the path of least cognitive effort. There is no doubt that this
argument offers a plausible explanation for the selection problem in
interpretation processes. However, it is not extensive enough for describing why
selection happens in a way. When Peter asks: ''Can we trust John to do as we tell
him and defend the interests of the Linguistics Department in the University
Council?'', and Mary answers: ''John is a soldier!'' (pp. 92- 93) a subset of
possible encyclopedic assumptions is activated by the hearer when interpreting
Mary's answer in the given context:
a. John is devoted to his duty.
b. John willingly follows orders.
c. John does not question authority.
d. John identifies with the goals of his team.
e. John is a patriot.
f. John earns a soldier's pay.
g. John is a member of the military. (cf. p. 93)
Vega Moreno explains that ''[t]he process stops when once the hearer has arrived
at a combination of explicit content, context and implicatures which satisfies
his expectations of relevance. In processing [the given example; J.J.], Peter
may satisfy his expectations of relevance by only the implications in [a- d]. It
follows that, contrary to the prediction of standard pragmatic models, he [...]
may never derive the implication in [g] or derive at a literal interpretation of
Mary's utterance'' (p. 93). This explanation seems to me very similar to Max
Black's ( 1962) ''filter''. Black compares a metaphorical expression to ''a
piece of heavily smoked glass on which certain lines have been left clear.''
(Black 1962: 41) When interpreting the metaphorical expression only some
attributes of – what Black calls – systems of associated commonplaces, i.e.
commonplaces we associate e.g. with soldier, are relevant for interpreting the
metaphorical expression, comparable with those attributes that lie on the clear
lines of the glass, whereas attributes not associated while interpreting the
expression lie behind the smoked glass. In my opinion, the problem of the RT
approach (and of Black's) is the unidimensionality of the selection process
proposed, which results in the uniqueness of the criteria of relevance (or in
Black's 'filter function') for explaining selection. The approach proposed by
Vega Moreno doesn't address that problem either. An alternative going beyond
this unidimensionality would probably be Levinson's (2000) approach of heuristic
interpretation. Levinson claims a set of hierarchically organized heuristics
when interpreting, and therewith gets over one single criterion responsible for
explaining the complex process of selection, as proposed by RT.
The second important argument of this book is what Vega Moreno calls
''inferential routes and pragmatic routines''. The frequent use of utterances and
their resulting familiarity allows minimizing the cognitive effort of a hearer,
when repeatedly processing an inferential path. ''Pragmatic routines are a kind
of cognitive procedure that might be expected to develop given the Cognitive
Principle of Relevance. Use of available pragmatic routines is encouraged by the
Communicative Principle of Relevance since they make an utterance particularly
easy to process in a way that is likely to satisfy expectations of relevance''
(p. 118). Satisfying expectations, not of relevance (as in RT) but of
coordination problems in communication, is tied in Lewis' ( 2002)
philosophical approach to convention, a primary category of linguistic units.
Vega Moreno argues for the processing of familiar (linguistic) stimuli to
develop routines for minimizing cognitive processing effort, and that this
mechanism does not only apply to utterance interpretation but is rather a
characteristic attribute of cognition. Apart from the cognitively interesting
point, the development of pragmatic routines sheds light on the balance of
conventionality and creativity of language use. This idea is also central to
Nelson Goodman's (1976) theory of symbols. According to Goodman, realism of
representation (contrary to invention) is a matter of a relationship between a
system of representation and a standard system. Realistic representation
(realism) is not a case of imitation or illusion, but of inculcation, which
means that ''realism is a matter of habit'' (Goodman 1976:38). Under this
perspective creativity and convention become floating boundaries which gradually
appear through concrete language use. What Vega Moreno argues for being a
cognitive procedure, sheds light on the symbolic procedures in a symbolic theory
framework. Even for those primarily interested in the latter processes, Vega
Moreno's ambitious book is an inspiring and absorbing reading.
Black, Max. ( 1962). ''Metaphor.'' In: _Models and Metaphors_. Ithaca,
London: Cornell University Press, 25- 47.
Goodman, Nelson. (1976). _Languages of Arts_. Indianapolis, Cambridge: Hackett.
Levinson, Stephen. (2000). _Presumptive Meanings. The Theory of Generalized
Conversational Implicature_. Cambridge, Massachusetts: MIT Press.
Lewis, David. ( 2002). _Convention_. Oxford: Blackwell.
Papafragou, Anna. (2002). ''Mindreading and Verbal Communication''. _Mind &
Language_, 17 (1/2), pp. 55–67.
Sperber, Dan & Wilson, Deirdre. ( 1995). _Relevance. Communication &
Cognition_. Oxford: Blackwell.
Sperber, Dan & Wilson, Deirdre. (2008). ''Relevance Theory''. In Horn, Laurence R.
& Ward, Gregory (eds.): _The Handbook of Pragmatics_. Oxford: Blackwell, 607- 632.
Wilson, Deirdre. (2000). ''Metarepresentation in linguistic communication''. In
Sperber, Dan (ed.): _Metarepresentations_. Oxford: Oxford University Press, 411-
ABOUT THE REVIEWER
Jörg Jost teaches at the University of Cologne, IDSL II. He holds a PhD in
Linguistics from RWTH Aachen University. His 2007 book (in German) is _Topos und
Metapher. Zur Pragmatik und Rhetorik des Verständlichmachens_ (_Topos and
Metaphor. The Pragmatics and the Rhetoric of Enhancing Comprehension_). Current
research interests in the field of pragmatics concern metaphors and
implicatures. In the field of applied linguistics he is interested in
communication in educational contexts and in the pragmatics of writing competencies.