Teaching Philosophy 10:3, September 1987 269

different periods it will be of great use to students of James.

T. L. S. Sprigge, Philosophy, University of Edinburgh, Edinburgh, EH89JX, Scotland

Language and Reality: An Introduction to the Philosophy of Language,

Michael Devitt and Kim Sterelny
A Bradford Book. MIT Press, 1987,286 pp .. $12.50 pbk.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - --- - - - -

CHRISTOPHER GAUKER

Devitt and Sterelny's Language and Reality is a staunchly partisan, issue-oriented introduction
to the philosophy of language. The centerpiece of their view is the causal theory of reference.
In addition to reference, the topics covered include the psychological reality of grammars,
the relation between thought and language, the explanatory value of the concept of truth,
and linguistic relativism.
More than a quarter of the book is devoted to the causal theory of reference. In essence,
the authors' theory is that for certain sorts of terms, especially names and natural kind terms,
a term comes to refer to a thing when that thing causes a perceptual act that leads to the
introduction of the term (55). Such a term may retain its reference when used by speakers
who themselves have not been perceptually affected by the referent provided these speakers
acquire the term in the right way from other speakers whose uses of the term refer. Other
sorts of terms, such as terms for artifact kinds, may acquire reference by being defined using
terms that already refer.
As Devitt and Sterelny are aware, their causal theory of reference is fraught with problems.
Thc most striking of these is what the authors call the "qua-problem." Of all the things
involved in the causation of the act of perception that leads to the introduction of a term,
which will be the term's referent? What makes it the case that "Nana" comes to be the name
of a cat and not the name of a head of a cat or the name of a time-slice of a cat? The authors
reply that the type of the referent is determined by a descriptive element in the thoughts that
lead to the introduction of the term (64-65). Thus the authors reject the "pure-causal theory"
in favor of what they call a "descriptive-causal theory." But as the authors acknowledge,
the qua-problem arises as well in accollnting for the reference of the relevant descriptive
elements. Sinee a solution along the same lines as before would beg the question, the authors
have to concede in the end that they really have no comprehensive theory of reference at
all (75).
Since the authors, by their own admission, cannot make the causal theory work, they owe
us at least some good reasons for thinking that it ought to work. The rationale they offer
proceeds in two stages. In the first stage they explain why truth-conditions are of central
interest in a theory of meaning (section 2.1), and they explain how truth-conditions depend
on reference relations (section 2.2). The second stage is to argue that one alternative theory
of reference, the description theory, has all sorts of problems that the causal theory avoids
(Chapters 3 through 5).
This rationale does not go to the heart of the issue. The issue is, what may we take for
granted in explaining what reference is? May we explain reference in terms of linguistic
meaning or in terms of truth-conditions, or must the order of explanation be the reverse of
this'? If we could give an independent account of meaning, then there might be various ways
to explain reference in terms of it. For instance, reference might be explained as a function
of meaning and the context of utterance. If the authors could persuade us that reference is
more fundamental than meaning, then they could make short shrift of the description theory
of reference since this takes for granted the meanings of descriptions. Even if we cannot
270 Teaching Philosophy 10:3, September 1987

give an independent account of meaning, maybe we can give an independent account of

truth-conditions. Then, as Davidson and McDowell maintain, reference might emerge as a
theoretical entity serving to unify otherwise disparate truth-conditions (see Donald Davidson,
"Reality without Reference," Dialectica 31 (1977) 247-58 and John McDowell, "Physicalism
and Primitive Denotation: Field on Tarski," Erkenntnis 13 (1978) 131-52). In the context
of their presentation of the causal theory of reference, the authors do not consider at all the
idea that meaning may be more fundamental than reference, and they simply dismiss the
views of Davidson and McDowell as "obscure and unconvincing" (36).
However much later, in Chapter 9, "Truth and Explanation," Devitt and Sterelny do take
up a closely related issue. Here they first briefly present the disquotational and prosentential
theories of truth. These theories purport to explain truth in terms of the use of "true" and
without appealing to reference relations. Second, the authors argue that truth, considered as
a function of reference relations, is irrelevant to the explanation of behavior. The upshot is
that there may be no work for their theory of reference to do. Undaunted, the authors reply
that the reference-theoretic account of truth belongs to "the theory of symbols" and not to
psychology; but they do not make very clear what distinction that label is supposed to draw.
Devitt and Sterelny have little to say about meaning at all, considered as something beyond
reference. They briefly discuss the need for the concept of sense (section 2.5) and what
sense is not (section 2.6); but as for a positive account, all they say is that the sense of a
term may be identified with the "causal network" that ties it to its referent (58). They make
little effort to show that this really solves the problems that lead to the postulation of senses.
In particular, they have chosen not to deal with the problem of referential opacity in any detail.
Throughout their book, Devitt and Sterelny stress that they seek a "naturalistic semantics"
(e.g. p. 9). One might have expected that a naturalistic semantics would locate the concepts
of meaning and reference in an account of the causes and effects of speech or in an account
of language as a medium of cooperation. But this is not what naturalism amounts to for
Devitt and Sterelny. At every turn they shy away from the question of how language works.
Their naturalism comes to nothing more than their commitment to the causal theory of
reference.
On the relation between thought and language, Devitt and Sterelny have a novel view.
On the one hand, they think there is something right about Grice' s program of explaining
meaning in terms of the thoughts words express. On the other hand, they argue that much
thought actually belongs to the thinker's public language. How can they have it both ways'?
Their answer is that words acquire and maintain their reference by way of certain language-
independent thoughts on the part of those whose rapport with the referent is suitably direct,
but that others may have the same thoughts by thinking in the vocabulary that in this way
acquires reference.
One of the virtues of Language and Reality is that it contains two chapters on linguistics.
One of these (Chapter 6) is a lovely introduction to transformational grammar. The other
(Chapter 8) is an extensive treatment of the question whether grammars are psychologically
real. The authors' position on this is that there's no good reason to think so. Their primary
strategy is to distinguish between several kinds of rule-following. They conclude that a
grammar need not be an object of the speaker's knowledge and (what is different) that a
grammar need not be represented, i.e., written out, in the speaker's brain. One difficulty
with their discussion is that they don't make very clear what linguistics is if not a branch
of psychology. What they say is that linguistics belongs to "the theory of symbols" (133).
But again, what that means is not adequately explained.
Chapters 10 through 15 examine the views of a host of authors with whom Devitt and
Sterelny disagree: Whorf, Dummett, Kuhn, the recent Putnam, Saussure, Wittgenstein,
Davidson and others. Here and there they find some kernel of truth, but by and large their
treatments are unsympathetic. The fault they find again and again is failure to accept a causal
theory of reference (182, 197,205,208,217,243).
Language and Reality is intelligent, honest, stimulating and highly readable. The authors
 Teaching Philosophy 10:3, September 1987 271
do not "talk down" to the reader, but at no point does the discussion become very technical
or arcane. In graduate and advanced undergraduate courses the chapters on the causal theory
of reference, together with the chapter on truth and explanation, would serve very well as
the sole text for a unit on that subject (following, perhaps, a unit on Frege and Tarski). The
chapter on transformational grammar might also stand on its own. The chapter on thought
and meaning might be useful in conjunction with H. P. Grice's classic paper "Meaning"
(Philosophical Review 66 (1957) 377-88). The chapter on the psychological reality of grammar
might be useful in conjunction with something by Chomsky.
Very extensive bibliographical notes follow each chapter. These include references to
both well-known and less well-known works in the analytic tradition, to philosophical works
outside the analytic tradition and also to works in psychology and linguistics.
Christopher Gauker, Philosophy, University of Wyoming, Laramie, Wyoming 82071 USA

-------- - - - - . - - -- - - - _. - - -- - - ----.--- --'-

On the Musically Beautiful: A Contribution Towards the Revision of the Aesthetics
of Music, Eduard Hanslick. Geoffrey Payzant, ed. & tr.
Hackett Press, Indianapolis, 1986, 152 pp. $16.50 c1; $5.95 pbk.
------------- ------- - ---

LYDIA GOEHR

For nearly one hundred years, English-speaking readers have had to resort to Gustav Cohen's
loose translation of Eduard Hanslick's essay on the musically beautiful. Now we have an
accurate translation, by Geoffrey Payzant. Apart from the translation, Payzant's carefully
researched bibliography, notes, and essay, 'Towards a Revised Reading of Hansl ick," provide
valuable insight into what many consider the canonical statement of musical formalism.
(Payzant's translation is taken from the 1891 edition of Yom Musikalisch-Schonen: ein
Beitrag zur Revision der Asthetik der Tonkunst. Cohen's translation was published in 1895
(Liberal Arts Press edition, 1957).
Hanslick provided only the beginnings of an aesthetics of music. His treatise, nonetheless,
has helped to shape our contemporary understanding of music; it has influenced numerous
aestheticians and musicians, not least Schoenberg and Stravinsky. It stands in contrast both
to the eighteenth-century theories of mimesis and affect and to the more mystical aspects
of the romantic aesthetic. Music traditionally was viewed as a language which could express
feelings, imitate nature, or somehow express the transcendental or inexpressible. Hanslick
countered with the view that music is to be understood only in the specifically musical terms
of tonally moving forms ltonend bewegte Formenj. The form and content of musie are sui
generis----constituted solely by the melodic, harmonic, rhythmic, and thematic shaping of
tonal sequences. Herein, alone, lies the musically beautiful.
Morris Weitz has suggested, in his introduction to Cohen's text, that Hanslick's brief
essay (83 pages in translation) is to musical theory what Hume's Enquiry Concerning Human
Understanding is to speculative philosophy. Certainly Gurney's treatise on formalism, The
Power of Sound, produced thirty years after Hanslick's essay is more systematic. Even so,
the latter has advantages: it makes the formalist position strikingly clear; it is stimulating in
its polemical and provocative style; last, but not least, it is short. It is undoubtedly essential
reading for any student of aesthetics and for anyone else seriously interested in understanding
the nature of music. Hanslick's text serves as one of the best introductions to the problems
of music's meaning and purpose, music's relation to other arts and to nature, musical form
and content, the objectivity of musico-aesthetic contemplation, jUdgment, and criticism,
and, finally the nature of composition and style.
Methodologically, Hanslick's purpose is three-fold. First, Hanslick proposes-in Kantian