LINGUIST List 12.2616

Sun Oct 21 2001

Review: Nuyts, Epistemic Modality

Editor for this issue: Terence Langendoen <terrylinguistlist.org>

Directory

1. Chaoqun Xie, Review of Nuyts, Epistemic Modality, Language, & Conceptualization

Nuyts, Jan (2001) Epistemic Modality, Language and Conceptualization:
A Cognitive-Pragmatic Perspective. John Benjamins Publishing Company,
hardback ISBN: 1-55619-983-X, xix+428pp, $100.00, Human Cognitive
Processing 5.

Chaoqun Xie, Foreign Languages Institute, Fujian Teachers University,
Fuzhou, Fujian, China

Traditional linguistic research is still to a large extent characterized
by an inclination to limit the scope of attention to purely linguistic
aspects of language use, and to the organization of the linguistic
system per se. Yet in reality language operates as an integral subpart
of the human behavioral repertoire, serving a quite specific function
in it. And the present monograph under review, built upon Nuyts
(1992b), tries to approach language from a cognitive-pragmatic
perspective. It "offers an in depth investigation into the semantic
domain of epistemic modality" (p. 1). Most of all earlier empirical
studies of epistemic modality have focused on the structural rather
than on the semantic category, and there has hardly been systematic
concern with the full range of structural and functional dimensions
of epistemic expressions in general, let alone with the question how
these might correlate with the nature of epistemic modality as a
semantic category (cf. xvi). For Nuyts, epistemic modality "concerns
an estimation of the likelihood that (some aspect of) a certain state
of affairs is/has been/will be true (or false) in the context of the
possible world under consideration" (pp. 21-22; cf. Aijmer 1980,
1996; Caton 1966; Sweetser 1990). And the analysis starts from the
assumption that "epistemic modality is intertwined with the
qualificational dimension of polarity" (p. 79). The aim of this
monograph, as clearly summed up in the Conclusion, is twofold: one
is to perform a paradigmatic investigation of the functional and
structural properties, and their interrelations, of 4 major epistemic
modal expression types in three West Germanic languages; the other
is to use this investigation to add to our understanding of the
cognitive infrastructure for language use, and specifically to inquire
into the nature of and the relationship between linguistic structure and
processing and conceptual structure and processing (cf. p. 1; p. 5). This
book is divided into six chapters, plus the Preface, the Introduction,
and the Conclusion.

Chapter 1, "Preliminaries" (pp. 1- 54), presents the preliminaries to
the investigation. First of all, it introduces the cognitive-pragmatic
perspective on language. The adjective 'cognitive' relates to the
observation that language is a dimension of human mental activity.
On the other hand, the adjective 'pragmatic' relates to the observation
that, in the context of the human behavioral repertoire, language has a
specific role to play, viz. (primarily) to allow communication with
other members of the species (cf. Grice 1969; Nuyts 1993b; Searle
1972). The cognitive and the pragmatic or functional dimensions of
language, which are central to strong research traditions in the
language sciences, are not just two separate issues; instead, they "are
two faces of one phenomenon, which must be mutually interrelated
and interdependent" (p. 3). According to Nuyts, the cognitive-
pragmatic perspective assumes that an adequate account of language
in general, or of any linguistic phenomenon in particular, has to do
full justice to both dimensions simultaneously, in an integrative way
(cf. p. xv). And even as two faces of one phenomenon, the cognitive
and the functional dimension have a different status in the
investigation of language. The cognitive dimension is hidden in the
black box of the human mind, and can only be accessed through
observing what the system perceives and how it behaves. The
functional dimension of language, however, is part of the observable
behavior. Hence, "analyzing the functional dimension of language is
logically antecedent to analyzing the cognitive infrastructure for
language" (pp. 3-4). Nuyts argues that language research is bound to
be concerned with the linguistic and conceptual dimensions at the
same time and that language is a dynamic entity (cf. pp. 5-21), herein
lies the two key properties emanating from the cognitive-pragmatic
perspectives. A clear understanding will undoubtedly help to get a
better grasp of the empirical facts of language as such and to make
functionalist linguistics a viable partner in the interdisciplinary
enterprise of cognitive science. In the sections that follow, Nuyts
presents an overview of the empirical analysis of epistemic modality
that will be dwelled on in the subsequent chapters.

Chapter 2 to 4 are devoted to a close scrutiny of corpus data for the
four major expression types of epistemic modality in Dutch and
German, with reference also to English, and with a discussion of
issues concerning the linguistic characteristics of these expression
types brought up in the literature. Chapter 2, "Modal adverbs and
adjectives" (pp. 55-106), discusses the modal adverbs and adjectives.
The author starts by presenting a comparative analysis of the
epistemic modal sentence adverbs and the predicative epistemic
modal adjectives as in the following two sentences:
 (1) Probably they have run out of fuel.
 (2) It is probable that they have run out of fuel.
The author's opinion, in this regard, is that the epistemic adverbs and
adjectives are closed classes, i.e. they consist of a fairly stable and
delimited set of forms. The two classes overlap, yet are not identical
(p. 54). Besides, there turn out to be several clear differences (cross-
linguistically) in the semantic-syntactic behavior of these two form
types. For one, it appears that adjectives can be questioned, whereas
modal adverbs cannot. Thus, Nuyts is right in holding that (3) is
perfectly acceptable, whereas (4) seems impossible.
 (3) Is it probable that they have run out of fuel?
 (4) a. *Probably that they have run out of fuel?
 b. *Did/Have they probably run out of fuel?
A second clear difference the author notices is that there are no
negative modal sentence adverbs, while there are negative modal
adjective. So, (5) is correct, while (6) is not.
 (5) It is improbable that they have run out of fuel.
 (6) a. *Improbably they have run out or fuel.
 b. *They have improbably run out of fuel.
And the corpus analyses that focus on Dutch waarschijnlijk and
German wahrscheinlich 'probably/probable' used as an adverb and as
an adjective evidently show that "both languages share a clear overall
tendency (strongest in Dutch) for the adverb to be much more
frequent than the adjective, irrespective of the discourse or text type"
(p. 63).

Nuyts calls our attention to the point that these properties, and the
differences between the expression types more generally, cannot be
accounted for in terms of just a single underlying cause, but require
recourse to several factors, namely, evidentiality, performativity,
information structure and, discourse strategy (cf. pp. 32-45). In the
sections that follow, Nuyts discusses the behavior of the adverbs and
adjectives in terms of these four factors. For each factor, the author
shows first what the corpus data reveal regarding its role in the use of
the expressions before pointing out how it contributes to
understanding some of the behavioral properties of the expression s
noted in the literature. Nuyts claims that by far, information structure
serves as the most important factor differentiating between the use of
the modal adverbs and adjectives (cf. p. 79). Also, the adjective can
be used if the epistemic qualification acquires strong focality in the
discourse context, which can find expression in the corpus (cf. 79-
100). Nuyts concludes in Chapter 2 that the adverb and adjective
"probably do involve the same semantic category of epistemic
modality, but with a different degree of likelihood" (106) and that
"the functional differences between the epistemic adverbs and
adjectives have nothing to do with (differences in) the semantic
category of epistemic modality as such" (106).

Chapter 3, "Mental state predicates" (pp. 107-170), expounds the
mental state predicates, or 'propositional attitude predicates', as
they are sometimes called, of the type illustrated in (7).
 (7) I think they have run out of fuel.
For the sake of their complex semantic structure and their mysterious
linguistic behavior, the mental state predicates are notoriously
difficult, and in-depth literature to unravel the properties of this
expression type is scarce. In this chapter, Nuyts, using the functional
analysis of the adverbs and adjectives as his stronghold, carries out
linguistic investigations of the use of the mental state predicates. The
author asserts that, apart from the difference in word use and
corresponding syntactic possibilities, these predicates differ from the
adverbs and adjectives in at least two respects, viz.: (i) unlike the
latter they both a qualificational and a non-qualificational meaning;
and (ii) the nature of the qualificational meaning, and the structure of
the lexical class in this respect, is quite different from that of the
adverbs and adjectives (cf. pp. 122). In the sections that follows, the
author, by drawing on corpus observations and on the special
behavioral properties observed for the latter form types, the author
discusses the status of the Dutch and German mental state predicates
in terms of the four factors introduced before, concluding that the
most important element distinguishing the mental state predicates
from the adverbs and adjectives is the omnipresence of the evidential
component.

Chapter 4, "Modal Auxiliaries" (pp. 172-233), is devoted to the
modal auxiliaries, as exemplified in (8).
 (8) They may have run out of fuel.
Despite the strong interests in them the modals remains a highly
controversial linguistic phenomenon. The reason for this is that "they
show an enormously complex structural (syntactic and morphological)
and semantic behavior" (p. 171). Besides, this behavior "turns out be
extremely variable between individual forms and semantic groups of
forms, in any one language as well as across languages" (p. 171).
First, the precise delimitation of the category of the modal auxiliaries
is a controversial matter. Opinions vary with regard to whether class
membership should be determined primarily in terms of the grammatical
or rather of the semantic properties of the forms. It is argued
by the author that the epistemic modals are only a subset of the
modals and that in the languages considered in this book, the
epistemic modals are a rather well circumscribed and closed category,
like the adverbs and adjectives, and unlike the mental state predicates.
The data in the corpus show that Dutch and English only have clearly
epistemic modals for expressing positions on the epistemic scale
rather far on the positive side, viz. zullen and will, and close to the
neutral point, viz. kunnen, and could or may. German has modals not
only for those positions, viz. werden, and koennen or moegen,
respectively, but also for a moderately positive position on the scale,
viz. duerfte. English and Dutch also allow expression of the latter by
means of combinations such as may well and kan goed, however,
which, are strongly idiomaticized (cf. pp. 172-174). For the
grammatical status of the modals, Nuyts points out that the modal
auxiliaries hold a strange balance between a lexical and grammatical
status: they figure as separate verbal morphemes, yet they also have
properties of grammatical markers (cf. 176-178). After discussing the
general characteristics of the modals, the author, recurring to data in
the corpus, explores at great length the multiple meanings of the
modals, with kunnen in Dutch and koennen in German as the object of
comparison. And corpus observations reveal that the epistemic
reading of these modals is far from well established. These modals
are quite unlike their English 'counterpart' may, well established as
an epistemic expression (pp. 187-195). Again, in the following
sections, the modals are examined in terms of the four functional
factors. The author concludes that 'there is not a single factor in
terms of which the modal can be said to have a special profile: it is
essentially neutral in terms of all of them' (p. 227). The postscript at
the end of this chapter discusses the diachrony of the Dutch modals,
pointing out that 'the Dutch modals in general have acquired their
epistemic meaning only relatively recently, much more recently than
the English modal' (p. 233).

Chapter 5, "Information Structure: An Experiment" (pp. 235-260),
aims to "see whether these correlations between (non-)focality and
form types can be further corroborated" (p. 235) and "also to widen
up the perspective towards other epistemic expression types not
considered in the corpus study" (p. 235). By adopting the method of
controlled data elicitation, this experiment, performed in a spoken
and a written version, has turned out that the corpus findings regarding
the information structure of epistemic expression types are
largely supported, or at least not contradictory, by the present
experimental findings: the adverbs never occur in focus; the
adjectives, predominantly used focally, are the most import[ant?]
expression type in focal conditions; the mental state predicates can be
but rarely are used in focus, and if they are, only in the
complementing form; and the modal auxiliaries can be used in focus,
but it is undecidable whether this observation concerns epistemic
uses of the modals, or rather dynamic ones. Nuyts concludes in this
chapter that this experiment confirms the tendency apparent in the
corpus investigation that speakers tend to avoid using epistemic
expressions under conditions of focality (p. 259), and a persistent
tendency for the epistemic qualification to assume a non-focal
position in linguistic expression can be observed (cf. p. 263).

Chapter 6, "The cognitive structure of epistemic modality" (pp. 261-
366), is the longest and, to the present reviewer's mind, the most
important one of this book. It is in this chapter that Nuyts, by
expounding the cognitive structure of epistemic modality, endeavors
to solve the puzzling problem of how to interpret and understand
human cognition and the role of language in it. After presenting some
general reflections on the syntax of epistemic expressions (pp. 261-
272), Nuyts goes on to briefly outline the 'Functional Procedural
Grammar' concept of language production in Section 6.2. Functional
Procedural Grammar tries to blend the basics of traditional
functionalist modeling of language structure with a general dynamic
procedural concept of cognition more typical of language psychology.
Functional Procedural Grammar attempts to offer a global
encompassing overview of the process of language production, from
the level of conceptual representation onwards (cf. pp. 272-279). And,
so far as qualificational categories are concerned, the notion of
'hierarchical' or 'layered' representation serves as a key source of
inspiration as well as object of critical reflection. For layered
representation, there have been too substantial hypotheses regarding
how to integrate a hierarchical system of layers of qualifications in a
functional grammar model, namely, in the context of Van Valin's
Role and Reference Grammar (RRG; cf. pp. 280-283) and of Dik's
Functional Grammar (FG; cf. pp. 283-287). Regarding epistemic
modality, the FG and RRG proposals differ in at least three respects:
(inter)subjectivity, polarity, and time (cf. pp. 306-315). Nuyts argues
that like RRG, Functional Grammar primarily deals with grammatical
expressions of qualifications and with lexical expressions in optional
constituents. Unlike in RRG, however, layered representation,
organized at two main levels, the lowered, 'representational' level
and the higher, 'interpersonal' level (cf. p. 284), is introduced at
only one level, viz. in the 'prediction', which is the basic
representation in the grammar. Nuyts claims that, on the one hand, the
validity of the principle of layered representation to explain the
behavior of qualifications in language adds a perspective 'in breadth'
to the 'in-depth' investigation of epistemic modality by situating his
findings in a wider context and evaluating their significance. And on
the other hand, the findings may help us better understand "the position
and format of layering in the cognitive systems for language use"
(p. 286).

In Section 6.4, Nuyts dwells on the matter of the nature of human
conceptualization in general, arguing that conceptualization is non-
linguistic, and maybe even non-propositional and that epistemic
modality is not just a linguistic phenomenon but a basic conceptual
category (cf. 287-304). Section 6.5 mainly addresses the global
division of labor between the linguistic and the conceptual levels,
focusing on the nature of the linguistic and the mapping relations
between the two levels. Section 6.6 explores the principles of
layering in conceptualization, concluding that even though what can
be conceptualized must be innately determined, what is actually
conceptualized depends on the needs and interactions of the organism
in/with its physical environment, and that, the functionality of the
conceptual system is quite different from that of the linguistic system
(cf. pp. 334-366).

Throughout this monograph, Nuyts tries to demonstrate the great
significance of adopting a methodologically traditional functional
linguistic approach to the study of epistemic modality, thereby going
beyond the narrow limits of language per se into its wider cognitive
context. Although the models and methods proposed in this book
need to be further tested and validated, the present volume, in the
present reviewer's opinion, serves as a valuable comprehensive and
in most respects convincing paradigmatic analysis of epistemic
modality, making a no-small contribution to helping us better
understand linguistic structure and processing per se and more
important, leading us to think more and know more about human
conceptualizations. Needless to say this is only a superficial sketch of
this rich and varied book, which is well-structured and written in an
accessibly academic style. This book is cogent, coherent, and
profound, meeting high standards both from a theoretical and from a
descriptive point of view. This brilliant book would be of great
interest to students of linguistics.

REFERENCES
Aijmer, K. (1980). Evidence and the declarative sentence. Almqvist
& Wilsell International, Stockholm.

Aijmer, K. (1996). "Swedish modal particles in a contrastiv
perspective". Language Sciences 18: 393-427.

Caton, C. (1966). "On the general structure of the epistemic
qualification of things said in English". Foundations of Language 2,
37-66.

Grice, H.P. (1969). "Utterer's meaning and intentions." Philosophical
Review 78: 147-177.

Nuyts, Jan (1992b). Aspects of a cognitive-pragmatic theory of
language. Amsterdam: Benjamins.

Nuyts, Jan (1993b). "On determining the functions of language."
Semiotica 94: 201-232.

Searle, John (1972). "Chomsky's revolution in linguistics." The New
York Review of Books, June 29: 16-24.

Sweetser, Eve (1990). From etymology to pragmatics: metaphorical
and cultural aspects of semantic structure. Cambridge: Cambridge
University Press.


ABOUT THE REVIEWER:
Chaoqun Xie is a lecturer with Foreign Languages Institute, Fujian
Teachers University in Fuzhou, Fujian Province, China. His main
areas of research interests include cognitive linguistics, pragmatics,
translation and communication. He is particularly interested in
relevance theory and politeness theory from the cognitive-pragmatic
perspective and is seeking cooperation with researchers of the same
interest.