T. Pateman, Language in mind and language in society. Oxford: Oxford Press, 1987. xiii+ 194 pp. f25.00. Reviewed by Robert D. Borsley, Department of Linguistics, University North Wales, Bangor, Gwynedd LL 57 2 DG, U.K.
University College of
This is a collection of papers by a philosopher which deserves to be widely read by linguists.’ It is possible that most linguists doubt whether there is anything of value to be found in the writings of philosophers on the study of language.2 It may be in fact that most linguists take the view that philosophers’ writings are at best harmless nitpicking and at worst ignorant nonsense. In any event, this book falls into neither of these categories. Pateman addresses real issues and he knows what he is talking about. Accepting this, one might still wonder whether linguists need to concern themselves with the philosophical issues that arise in the field. It seems to me that it is important not to overstate the case here. I can see no basis for the view of those who Pullum (1984) calls the ‘methodological moaners’ that linguistics is enmeshed in a thicket of philosophical problems which threaten its future. It also seems to me that Pullum is right that linguists’ time is more usefully employed in getting on with linguistics than in debating philosophical issues. Nevertheless, there are real philosophical issues which linguists ought to take some interest in, and clarity about these issues is at least important when it comes to presenting the findings of linguistic research to outsiders. Pateman suggests that the originality of the book is ‘in exploring the character and consequences of nativism from the standpoint of an interest in social theory, and in introducing arguments from the recent philosophical realist literature into debates about the nature of linguistics’ (p. 1). It seems to me that this is a generally accurate characterization. However, Pateman is not just concerned with nativism but with Chomskyan conceptions in a more general sense, and in particular with the view that mentally represented grammars are a legitimate focus for linguistic research. It is quite possible to hold this view without sharing Chomsky’s nativism. Hudson (1984) illustrates this point. When Pateman speaks of ‘recent philosophical realist literature’, he is referring especially to the work of Roy Bhaskar (1975, 1979). Bhaskar’s ideas are very congenial to any linguist of a broadly Chomskyan persuasion. It is somewhat surprising, therefore, that there is no reference to them in Newmeyer’s (1983) defence of Chomskyan conceptions. It is not at all surprising, on the other hand, that Bhaskar (1975: 245) refers approvingly to Chomsky’s characterization of linguistic research. Bhaskar’s ideas are especially prominent in chapter 1, ‘The philosophy of linguistics’, a general discussion of the scientific status of the discipline, and chapter 2, ‘A realist theory of linguistics’, a critique of the views of Lass (1980) on linguistic change. Bhaskar argues that the world is an open system in which a variety of mechanisms *
I am grateful
to John O’Neill for helpful comments though.
Grice have done valuable
work on language.
on an earlier version that
of this review.
interact to produce phenomena, and stresses that exceptionless regularities do not generally occur, except in an experimental context in which a closed system is produced where the number and kind of mechanisms at work is strictly controlled. These ideas are relevant in various ways for linguistics. For example. they permit a response to Sampson‘s (1980) argument that semantics cannot be a scientific discipline because speakers are essentially creative in relation to the production of meaning. From Bhaskar’s perspective, this is no problem since ‘there is no bar to the existence both of finite mechanisms of speech production and understanding, and of other mechanisms (free will, creativity; but also memory, psycho-physiological states, and processes) which interact with or override those mechanisms’ (p. 8). Bhaskar’s ideas are also relevant to Lass’s (1980) worries about the scientific status of historical linguistics, in particular his argument that ‘ease of articulation’ is not an explanatory concept since languages do not always change in the ways that a preference for case of articulation would suggest. A central claim of Bhaskar’s is that causal mechanisms may be operating without their operation showing up in observable events, Illustrating, Pateman points out that: ‘Alcohol tends to produce slurred speech, a claim which is not falsified by the fact that a drunken motorist may succeed in avoiding slurred speech when confronted by a police officer’ (p. 22) and addressing Lass’s point directly. he comments: ‘That a language does not change in the direction of ease of articulation . . . does not falsify the claim that there is a mechanism at work tending to produce that outcome. It may simply show that there are other, more powerful, mechanisms at work’ (p. 25). Bhaskar’s ideas also inform Pateman’s critique of various objections to mentally represented grammars. Pateman is especially concerned with the objections of those who draw their inspiration from the later Wittgenstein. He addresses these objections in chapter 3. ‘What is a language?, chapter 5. ‘Using and defending cognitive theory’, and chapter 6, ‘Wittgensteinians and Chomskyans’. He notes that ‘mentally represented grammars as Chomsky conceives them ,.. are not social objects. nor copies (internalizations. representations) of social objects. They are about as social, and no more social, than infectious diseases’ (p. 91). Wittgensteinians like ltkonen and Kripke feel that the Chomskyan position is undermined by Wittgenstein’s famous (or notorious) private language argument. which seeks to show that rules are necessarily social since one can only talk of rules where there is the possibility of a mistake and of correction by others. Pateman argues that ‘claims for the necessarily common knowledge character of linguistic rules are falsified by the form which it appears that any true theory of learning or acquisition must take ___’(p. 66). He develops this point with reference both to ‘ordinary’ language acquisition and to the creolization of pidgin languages and the development of signing systems by isolated deaf children of hearing parents who do not use sign language. Pateman surveys a number of other strands of Wittgensteinian criticism. He suggests that the Wittgensteinian literature ‘places unnecessary restrictions on the prescriptions were form of legitimate science’, and notes that ‘if Wittgensteinian adhered to. ___we would be obliged to give up on the effort to explain many cognitive
that ‘much Wittgensteinianism
is simply an expression
anti-scientific attitudes’ (p. 136). Pateman devotes considerable space to the defence of mentally represented grammars. He has much less to say in defence of nativism. His main point, in chapter 4, is that nativism is not incompatible with a ‘From nativism to sociolinguistics’, recognition of the social dimension of language. He notes that he provides no arguments there in support of nativism, and he offers little in the way of argument elsewhere. Hence, anyone looking for a real defence of nativism here will be disappointed. As was indicated earlier, another important feature of the book is a concern with the social dimension of language. In defending mentally represented grammars. which Pateman is defending the view that Chomsky now calls internalized languages, languages in one sense are linguistic facts which are not social Facts. In chapter 3, he argues that languages in another sense are social facts which are not linguistic Pacts. More precisely, he argues that they are intentional objects of mutual belief. He argues in support of this conception that it allows for the possibility that speakers who agree that they are speakers of English can disagree about what English is, and that it allows for speakers to acquire, add to, and change thair beliefs about what English is. He also argues that it makes sense of presciptivism and hypercorrection, and that it allows an approach to questions of linguistic standardization and linguistic hegemony. Perhaps the most interesting element in Pateman’s discussion of the social dimension of language is the idea that grammars are not only ‘mental representations to which speakers are liable’ but also ‘things on which or on the output of which speakers are capable of acting with reason’ (p. 14). Pateman develops this point in chapter 4, drawing on Henning Andersen’s (1973) work on adaptive rules. He characterizes an adaptive rule as ‘an ad hoc adjustment to the output of the speaker’s phonological system which allows the production of a non-stigmatized phonetic form’ (p. 95). When speakers develop adaptive rules, their speech output is a heterogeneous object. For hearers, however, including language learners, there is only a single speech output which is unlikely to contain any surface markers of its dual structure. Hence, the grammars which the output stimulates will be grammars which account for that output, and not an unmodified output which may no longer occur. The result is that ‘through time the content of mentally represented grammars comes to contain a content which was in origin quite clearly social or cultural in character’ (p. 98). Pateman also notes that from a nativist standpoint ‘some linguistic modifications will, ex hypothrsi, go more against the grain of natural tendencies that others. They will then require more reinforcement and monitoring if they are to be reproduced and they will tend to fail or suffer unintended modifications’. He suggests that this ‘may help to explain differences in the success of different elements in prescriptive grammars’ (p. 97).3 3 In connection with this, it is worth noting that Emonds (1985: 238) argues that ‘prestige usage in Modern English is not a version of natural language’, but the artificial
I have now indicated the main themes of Pateman’s book. There is, however, muc 1 else that is of interest here. Most noteworthy, perhaps. is the discussion in chapter 7 of David Lewis’s conventional account of language. Also of interest, however, is Pat<,. man’s critique in chapter 3 of the view, which he associates with Saussure and Labor, that languages are social facts which are also linguistic facts, and his critical discussion in the same chapter of Katz’s view that they are abstract, platonic objects. There are also some amusing remarks. For example, discussing the Wittgensteinian view that abilities are manifest in performances and not causally related to them, Pateman comments as follows: ‘suppose someone said, “We say someone is suffering from consumption when they grow pale, weak, thin, clear-headed, cough blood and die ~ and that’s what consumption is.” This attitude would rightly be regarded as antiscientific; had people thought to take it seriously, we should all be dead of tuberculosis’ (p. 137, n. 13). This is, then, an extremely interesting book, which deserves to be widely read. Pateman indicates in his preface that he intends to turn to work in aesthetics. It is to be hoped that he will not altogether abandon linguistics. As the discipline develops, the character of the philosophical issues that arise within it can be expected to change and there will continue to be a place for the commentary of philosophers as wellinformed as Pateman.
References Andersen, H., 1973. Abductive and deductive change, Language 40, 765-93. Bhaskar, R.. 1975. A realist theory of science. Leeds: Leeds Books. Bhaskar, Hudson,
R.. 1979. The possibility of naturalism. Brighton: R., 1984. Word grammar. Oxford: Blackwell.
1985. A unified theory
Lass. R., 1980. On explaining Newmeyer,
University of Chicago Press. Pullum, G.K.. 1984. The revenge Linguistic Sampson.
of a number
Dordrecht: Cambridge and
Press. Foris. University
G.. 1980. Making