GRICEIANA

Speranza

In The Lewd, the Rude and the Nasty, Pekka Väyrynen systematically applies the tools of philosophy of language and linguistics to the investigation of thick terms. The book is an outstanding example of methodological accuracy, interdisciplinary stance, detailed arguments and clearly articulated theses. The bulk of Väyrynen’s proposal is the following: The evaluation that thick terms convey is not built into the lexical meaning of these terms, it arises from pragmatic mechanisms, “as a function of our communicative and practical interests in discourses involving thick terms and concepts” (p. ix). The standard view on how thick terms and concepts are associated with evaluation is that they are inherently evaluative: Inherently Evaluative (IE) is the principle according to which “the meanings of thick terms and concepts somehow or other contain evaluation” (p. 9). Väyrynen’s thesis — which is per se one of the possible accounts of thick terms — has strong theoretical consequences, because it follows from his view that thick terms do not have the deep philosophical relevance that they are typically taken to have. Thick concepts and terms allegedly challenge the fact-value distinction, and — assuming that one endorses Inseparability, i.e. the idea that the evaluative and nonevaluative contents of thick terms and concepts cannot be disentangled (pp. 12 ff, 202 ff) — they also challenge noncognitivism and expressivism. However, Väyrynen argues, if one leaves aside IE, then thick concepts and terms do not play such a crucial role in evaluative thought and discourse. In order to support his pragmatic thesis, Väyrynen discusses the main possible accounts of thick terms. He does not just explore the landscape of the theories that philosophers have actually put forward, but investigates all the relevant theoretical options, by going through their advantages and shortcomings. His overall strategy is the fol- 296 Book Reviews lowing. By generalizing the principle of Grice’s Razor (Paul Grice, “Further notes on logic and conversation”, in Studies in the Way of Words, edited by Paul Grice, Cambridge, MA, 1989, 47-8), according to which other things being equal, it is preferable to postulate conversational implicatures rather than multiplying senses and semantic implicatures, Väyrynen claims that other things being equal, it is preferable to explain the evaluations associated with thick terms (T-evaluations) in terms of conversational implications rather than multiply senses and other semantic properties (p. 55). His challenge is — in a sense — to investigate and verify the “other things being equal” constraint. The book is structured as follows: in Chapters I and II Väyrynen provides the necessary framework to develop his arguments. In Chapter I he introduces crucial notions, such as the problematic distinction between thick and thin — namely between purely evaluative terms (like ‘good’ and ‘bad’) and terms that mix evaluation and description (like ‘lewd’, ‘courageous’, ‘brutal’, etc.). In Chapter II he discusses the notions of evaluation and meaning and assesses the definitory issue of what counts as thick concepts and terms. In particular, Väyrynen takes a term or concept T to be “evaluative in meaning if T-sentences of the form x is T entail, as a conceptual matter or in virtue of a semantic rule, that x is good in a certain way (…) or that x is bad in a certain way” (p. 34). Therefore, it is not enough for a term to convey evaluative content in order to count as “evaluative”, as in principle any term could come to carry some evaluation in context. Chapters III and IV are devoted to the discussion of semantic views, where Väyrynen argues against the idea that the T-evaluations are semantic entailments. In Chapter III, he presents phenomena such as Projection and non-Deniability of T-evaluations. The fact that the evaluation associated with thick terms projects out of semantic embeddings such as negation, antecedent of conditionals, question, modals, denials, etc., strongly suggests that T-evaluations are not truth-conditional components, nor they can be semantic entailments. In particular, if the unembedded occurrence of ‘lewd’ in (1) conveys an evaluative content that sounds like (7), all the following embedded occurrences of ‘lewd’ and B’s denial in (6) convey the same Tevaluation (please note that I changed the numbers in the examples): Book Reviews 297 (1) Madonna’s show is lewd. (p. 70) (2) Madonna’s show is not lewd. (p. 78) (3) Is Madonna’s show lewd? (p. 64) (4) Madonna’s show might be lewd. (p. 64) (5) If Madonna’s show is lewd, tabloid press will go nuts. (p. 64) (6) A: Madonna’s show is lewd. B: No, it isn’t. (…) (p. 74) (7) Overt displays of sexuality that transgress conventional boundaries are bad in a certain way. (p. 62) Väyrynen observes these patterns for utterances involving objectionable thick terms, thick terms that convey an evaluation that speakers are not willing to endorse (typically, ‘lewd’, ‘chaste’, etc.). However, because in principle any thick term could be seen as objectionable, Projection and non-Deniability data speak against a truthconditional analysis of thick terms in general (p. 56). Another argument against the semantic view, to which we will come back later on, is Defeasibility, according to which T-evaluations can be contextually suspended without infelicity or contradiction. According to Väyrynen, a lewd-objector who wants to reject the T-evaluation conveyed by (1), can do that by uttering (8) or (9): (8) Whether or not Madonna’s show is lewd, it’s not bad in any way distinctive of explicit sexual display. (p. 70) (9) Whether or not Madonna’s show involves explicit sexual display, it would be in no way bad for that. (p. 70) Väyrynen takes (8) as a felicitous and literal use of ‘lewd’ that fails to convey a negative evaluation. He denies that there is a contrast between (8) and (9), where (8) is strikingly less felicitous than (9). If Väyrynen is right about this case, then T-evaluations are in fact defeasible. 298 Book Reviews In Chapter IV, Väyrynen discusses some alternative explanations of the data presented in Chapter III (Projection, non-Deniability, Defeasibility) and argues for the superiority of his own treatment. So far, Väyrynen’s arguments only target a truth-conditional analysis of thick terms and leave open the possibility for T-evaluations to be analysed in terms of other semantic properties different from entailment (such as presuppositions). Chapters V and VI are dedicated to discard such further semantic approaches, present various pragmatic theories and argue in favour of Väyrynen’s own version of a pragmatic approach. Väyrynen characterizes T-evaluations as pragmatic implications of utterances featuring thick terms, typically not-at-issue: they usually are not relevant to the main point of the conversation at hand. Certain conversational moves can nevertheless make these not-at-issue contents at-issue and therefore T-evaluations can be challenged when they are not backgrounded (p. 127). In Chapter V, the author discusses some general psycho-social principles that are taken to explain how T-evaluations conversationally arise as pragmatic implications. It is crucial for Väyrynen to explain why T-evaluations are so systematically associated with thick terms, given that for him evaluations are not built into the meaning of thick terms. He does so by appealing to the principle of Parochiality, according to which “The application of a thick term or concept tends to derive its point or interest from the term’s or concept’s relation to the evaluative perspective reflected in its application” (p. 128). The idea is that because a community typically shares a background moral perspective or practice, a certain linguistic expression can get routinely associated with a not-at-issue content, which gets triggered in all ordinary contexts as part of the default interpretation of the linguistic expression. In Chapters VII and VIII, Väyrynen assesses some issues that scholars typically take to favour the IE approaches over pragmatic ones and argues that, on the contrary, they can be accounted for by his theory. In particular, Chapter VII is dedicated to the problem of extension. Philosophers have argued that it is a distinctive feature of the evaluative that linguistic meaning underdetermines extension; given that the extension of thick terms and concepts is underdetermined, then T-evaluations must be part of the meaning of thick terms. Väyrynen rejects the argument by acknowledging that Book Reviews 299 the extension of thick terms and concepts is underdetermined while arguing that this feature can be explained in terms of certain general features of gradable and context-dependent expressions, independently of whether the expressions at stake involve evaluative contents or not. In Chapter VIII he discusses other two allegedly pro-IE issues: Shapelessness and Inseparability. Shapelessness amounts to the idea that “The extensions of evaluative terms and concepts aren’t unified under independently intelligible nonevaluative relations of real similarity” (p. 190). One interpretation is that to be shapeless with respect to the nonevaluative dimension is what is distinctive of the evaluative; therefore the fact that thick terms and concepts are shapeless with respect to the nonevaluative dimension would reveal that they are inherently evaluative. Väyrynen’s strategy to reject this argument is again to show that shapelessness is not in fact distinctive of the evaluative: He provides many such examples of similarly shapeless non-evaluative psychological notions. Moreover, he considers Inseparability, the thesis according to which “Thick terms and concepts are or represent irreducible fusions of evaluation and nonevaluative description; these aspects cannot be “disentangled” from one another” (p. 204). Väyrynen rejects Inseparability by relying in part on Simon Blackburn’s point that the objectability of certain terms and concepts makes Inseparability implausible, as for Inseparability one could not conceptually separate evaluative and nonevaluative content (Simon Blackburn, Ruling Passions, Oxford, 1998). In Chapter IX Väyrynen employs an opposite strategy to the one adopted in Chapters VII and VIII: he presents an argument typically taken against IE and argues that many theories endorsing IE can actually account for it. His set of arguments against IE is therefore not superfluous. The issue at stake is Variability, according to which thick terms and concepts are contextually variable with respect to the valence of T-evaluations. Prototypical instances of Variability would be cases like the following: (10) The carnival was a lot of fun. But something was missing. It just wasn’t lewd. I hope it’ll be lewd next year. (p. 221) The term ‘lewd’, instead of carrying its typical negative evaluation, seems to convey a positive one. However, according to Väyrynen, (10)-like examples are problematic to IE only if they 300 Book Reviews are analysed as literal uses of thick terms. On the contrary, they do not challenge the idea that the relation between thick terms and evaluation is systematic and stable if they are analysed in terms of non-literal uses of language, “a way of mocking the sorts of prudish evaluations that lewd conveys as a matter of meaning” (p. 224). This kind of account, suggested by Blackburn (Simon Blackburn, Ruling Passions, Oxford, 1998, 103) has interesting relations with some research conducted by Jesse Harris and Christopher Potts on perspectival shift concerning expressives and appositives (Jesse Harris and Christopher Potts, “Perspective-shifting with appositives and expressives”, Linguistics and Philosophy 32 (6), 2009, 523-52) and to Bianchi’s analysis of appropriated uses of slurs in terms of irony and echo (Claudia Bianchi, “Slurs and appropriation: An echoic account”, Journal of Pragmatics 66, 2014, 35-44). The underlying idea in these apparently unrelated works is that evaluatives involve a perspective and this is typically the speaker’s one, but, by employing the relevant pragmatic mechanisms, one can succeed in shifting this perspective from the speaker’s one to another party’s one. However, regardless of whether Väyrynen favours the “non-literal” account, Variability does not suffice to knock down IE because it can be in principle explained in terms of pragmatic mechanisms. Therefore, because IE can account for Variability, other arguments against IE are needed to discard the approach. In Chapter X — the last — Väyrynen completes his deflationary project, by arguing that, in the light of his pragmatic analysis, thick concepts and terms actually lack strong philosophical implications with respect to issues such as the fact-value distinction and the cognitivism/non-cognitivism debate. Let us now consider the negative contribution of The Lewd, the Rude and the Nasty, namely the arguments against the claim that thick terms lexically encode evaluations. Väyrynen considers two main arguments: Variability and Defeasibility. As we have just seen, Variability is set aside because instances of Variability à la (10) can be analysed as non-literal uses of language. The strongest argument against IE is therefore Defeasibility, namely the idea that T-evaluations can be contextually suspended without infelicity or contradiction. All in all, a reader might harbour doubts about the fact that the main argument against IE is based on Defeasibility, which is supported in Book Reviews 301 the book by a quite small set of examples and which relies on the assumption that there is no significant contrast in felicity between (8) and (9), an intuition that is not so obvious. However, if one accepts the generalization of Grice’s razor (according to which, other things being equal, it is preferable to analyse T-evaluations in terms of pragmatic implications rather than appealing to semantic notions), for Väyrynen’s main thesis to go through, it would suffice to show that the pragmatic account can in principle account for all the relevant phenomena, without relying on the Defeasibility data. On the other hand, Väyrynen’s positive contribution — the deflationary thesis according to which T-evaluations arise from conversational mechanisms — would have in a way profited from a more detailed discussion of what pragmatic implications are taken to be. Väyrynen talks about “generalized pragmatic implications” that are different from the well-known gricean implicatures, and he appeals to general communicative and practical interests, but it is not entirely fleshed out how to fully characterize such implications and systematically distinguish them from generalized implicatures. The under-determination of the notion of ‘pragmatic implication’ might be partly due to the fact that for Väyrynen thick terms are not homogeneous enough to allow a uniform analysis and different mechanisms can be at play each time a T-evaluation arises. As to how to best characterize such mechanisms, a comparison with pragmatic and deflationary accounts of pejoratives might come in handy. In the literature on slurs, scholars developed various accounts to explain how certain terms can systematically be associated with evaluative contents, without these contents being lexically encoded. For instance, Bolinger (Renée Bolinger Jorgensen, “The pragmatics of slurs”, Noûs 50 (3), 2015) talks about “contrastive preferences ”, Nunberg (Geoffrey Nunberg, “The social life of slurs”, in New Work on Speech Act, edited by Daniel Fogal, Daniel Harris and Matt Moss, Oxford, Oxford University Press, forthcoming) relies on gricean conversational manner implicatures, Rappaport (Jesse Rappaport, “Communicating with slurs”, manuscript) appeals to the relevance-theoretic notion of ‘showing’ (as opposed to ‘meaning’). While appealing to different notions, all these approaches aim to characterize the evaluative content of pejoratives as stable and nevertheless non-encoded. I conclude by suggesting that these kinds of 302 Book Reviews proposals from the literature on pejoratives provide some interesting insights as to how to flesh out a pragmatic analysis of the evaluative content of thick terms. Väyrynen himself suggested a comparison between his view on thick terms and the analysis of slurs put forward in Bolinger’s proposal in a talk (Pekka Väyrynen, “Evaluatives and pejoratives”, Handout for Linguistics Seminars-Scuola Normale Superiore, Pisa, 2016). I shall add that also Nunberg’s and Rappaport’s proposals are very relevant in this respect.