Meaning without content: on the metasemantics of register

ABSTRACT

What, exactly, is the difference between words such as ‘dead’ and ‘deceased’? In this paper, I argue that such differences in register, or style, ought to be construed as genuine differences in non-truth-conditional meaning. I also show that register cannot plausibly accounted for in terms of either presupposition or conventional implicature. Register is, rather, an instance of what I call pure use-conditional meaning. In the case of register, a difference in meaning does not correspond to a difference in the contents speakers intend to convey.

KEYWORDS:

• Register
• style
• conventional implicature
• presupposition
• non-at-issue content
• use-conditional items

Acknowledgements

For comments on previous versions of this paper, I would like to thank several anonymous referees for several journals. Needless to say, I am solely responsible for all the remaining shortcomings.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 See Frege (1892, 31) and (1897, 150).

2 For recent discussions of that notion see Horn (2007, 2013), Picardi (2007), Sander (2019, 2021).

3 Cf. Potts (2007) who, after initially offering some propositional paraphrases of that kind, opts for specifying the semantic contribution of expressives by what he calls ‘expressive indices’.

4 One might claim that, by choosing a more formal register, you express respect to your addressee, but ‘expressing respect’ appears to be vastly different from expressing an emotive state.

5 The confounding of felicity conditions and communicated contents is one of Bach’s (2006, 27) ‘top 10 misconceptions about implicature’. Of course, there are cases where register serves communicative purposes. For instance, in the (non-fictional) book title ‘A Teeny-Tiny History of Philosophy’, the word ‘teeny-tiny’ is also, or primarily, a way of sneakily informing adults that the book is intended for an immature audience.

6 Or, more exactly, mean ⊦ p (see Grice 1989, 118), but here I shall ignore illocutionary force as an element of Gricean meaning.

7 While register has been an issue in sociolinguistics (see, e.g. Hudson 1996, 45–49) and also in literary theory, there are very few works that discuss register as a semantic phenomenon: Dummett (1981, 84–89; 1991, 121–122); Predelli (2013, ch. 5); Kortum (2013, 65–74). For works on adjacent or overlapping phenomena such as connotation, honorifics, and taboo words see, respectively, Allan (2007), McCready (2019), Pinker (2007, ch. 7). In the next section, I will say more on how ‘register’, as I understand that word, differs from these phenomena.

8 Words that convey essentially the same expressive content may differ in social acceptability and thus in register (‘damn’ vs. ‘fucking’). More generally, alleged differences in expressive strength (cf. Potts 2007, 177) perhaps ought to be construed as differences in offensiveness.

9 This is, of course, not the whole story. Speakers who are aware of dialectal variants (and who know that the same goes for their addressees) may decide to use a particular variant to identify themselves as members of a certain group via such a shibboleth. I am not sure how to characterize such forms of ‘communication’, but in any case, this does not appear to be an example of Gricean meaning. There is, after all, a natural correlation between your origin and the words you use. See Grice’s (1989, 218) discussion of similar cases.

10 One might even claim that register is a kind of honorification (McCready 2019, 6) or, alternatively, that honorifics form a subcase of register (Diaz-Legaspe, Liu, and Stainton 2020, 158).

11 Coprolalia and certain kinds of selective aphasia may be taken to show that swear words as well as slurs work quite differently from ordinary terms. For an overview, cf. Pinker (2007, 334–339).

12 Note that Katz’s claim is much more general than is suggested by his two examples. He also mentions obscene words and slurs as examples.

13 Couldn’t one argue that register is a case of meaning because a sentence such as ‘The bow-wow is snoring, but I am not talking to a child’ is odd? I don’t think so. Moore sentences are similarly odd, but that plainly does not yet show that the proposition that I believe that dogs bark is part of the linguistic meaning of ‘Dogs bark’.

14 In Frege’s original version of (Sub) ‘synonymous’ is to be replaced by ‘have the same Sinn’. Given Frege’s insistence that words such as ‘cur’ and ‘dog’ differ only in tone, but not in sense, his version faces quite a few counterexamples (see Textor 2011, 144–145; Sander 2019, 375) to which (Sub) seems immune.

15 For similar reasons (‘IQ test’), Bach (1999) has claimed that alleged conventional implicature devices such as ‘but’ are part of what is being said.

16 Of course, one might claim that (4) is true but misleading since the sentence suggests that Darwin himself used the word ‘bow-wow’, but I find it hard to convince myself that (4) is true in the first place.

17 It is perhaps worth noting that some of Dummett’s examples of ‘style’ do not only differ in register, but also in truth-conditional content. The descriptive conditions for being a statesman appear to be stricter than the conditions for being a politician.

18 Needless to say, nursery language is also appropriate when the speaker is a child, but in what follows I shall ignore that complication.

19 Frege (1892, 40) offers a similar argument with respect to the existential presuppositions carried by names.

20 See also Dummett’s (1981, 454) inferentialist treatment of ‘Boche’. On his view, that word licenses the inference of ‘A is German’ and ‘A is barbarous and more prone to cruelty than other Europeans’ from the premise ‘A is a Boche’. Since ‘barbarous’ and ‘cruel’ are arguably thick terms, I am not sure whether Dummett regards the second conclusion as partly expressive or just as an empirical claim.

21 More exactly, Kaplan (ms, 22) says that he has the ‘inclination’ to regard the argument as valid. Kaplan would probably resist being called an inferentialist, but combining a use-conditional account of items such as ‘that bastard’ with the claim that there is a logic of expressives seems to come close to a broadly inferentialist account.

22 I think much the same could be said about (5). We do not ordinarily deduce a partly expressive conclusion such as (5c) from a descriptive statement about our attitudes such as (5b).

23 Since I am not concerned with either conversational implicature or with pragmatic presupposition here, I shall drop the adjectives in what follows.

24 One might argue, moreover, that the distinction between these two phenomena is rather artificial. Compare Potts (2015, 192–193) and Sander (2022).

25 Schlenker himself notes that his ‘expressive presuppositions’ differ in some respects from ‘standard presupposition triggers’ (2007, 239).

26 It is also worth noting that presuppositions can be directly cancelled or suspended (‘Carla doesn’t regret learning cursive because in fact she never learned it’). In contrast, attempting to cancel commitments carried by register would result in rather weird utterances (‘The bow-wow isn’t snoring because I am not talking to a child’).

27 Moreover, Grice’s explanation of CIs in terms of ‘higher-order speech acts’ (1989, 362) seems rather implausible with respect to register.

28 For a brief discussion of how the relevance-theoretic notion of procedural meaning is related to Kaplan’s ‘semantics of use’, see Wilson (2016).

29 I have borrowed that term as well as the abbreviation ‘UCI’ from Gutzmann (2015, 10).

30 Bach and Harnish add a second option according to which a speaker just has to intend that ‘his utterance satisfy the social expectation that one express pleasure at seeing … someone’ (1979, 52).

31 I am using the term ‘phatic’ as introduced by Malinowski in his supplement to Ogden and Richard (1946, 315). Phatic speech is ‘a type of speech in which ties of union are created by a mere exchange of words.’ For a plausible account of greetings in terms of ‘social recognition’, more specifically in terms of either initiating an engagement or ‘welcoming or at least accepting the initiation of an engagement’ see Goffman (1963, 213).

32 Of course, greetings and farewells, like other items, come in various registers (‘Hi’ vs. ‘Good morning’ etc.).

33 In a recent paper, Beaver & Stanley discuss how the theory of meaning might have to change in order to be able to cope with all kinds of problematic language. Merely adding some new tools is the most conservative option they discuss (2018, 534).

Additional information

Funding

Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – 439302327.