Quine on explication

ABSTRACT

The main goal of this paper is to work out Quine's account of explication. Quine does not provide a general account but considers a paradigmatic example which does not fit other examples he claims to be explications. Besides working out Quine's account of explication and explaining this tension, I show how it connects to other notions such as paraphrase and ontological commitment. Furthermore, I relate Quinean explication to Carnap's conception and argue that Quinean explication is much narrower because its main purpose is to be a criterion of theory choice.

KEYWORDS:

• Quine
• explication
• Carnap
• paraphrase

Acknowledgments

I would like to thank Chris Daly, David Liggins, Benjamin Marschall, Thomas Uebel, the audience of the 4th annual TiLPS History of Analytic Philosophy Workshop (including my commentator Christian J. Feldbacher-Escamilla), and an anonymous referee for helpful comments and discussions.

Disclosure statement

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

Notes

1 One reason might be that Carnap underwent his semantic turn after Tarski taught him how to explicate ‘truth’ semantically (Carnap 1963a, § 10; see also Leitgeb and Carus 2021, § 6.1 and Supplement F); Carnap might have modeled M&N-explication along the lines of Tarski's explication of truth: Tarski proposes his Convention T (Tarski 1933, 187f.) which any adequate definition of truth has to satisfy (see Tarski 1933, 187); Carnap also refers to Tarski's definition of ‘true’ (see his example 2 in Carnap 1962, 5).

2 We can assume, though, that Quine knew LFoP; he refers to its first edition (Quine 1951, 23, n. 4).

3 Gustafsson's interpretation (of the footnote) is that Quine ‘suggests that with regard to explication, he and Carnap endorse basically the same methodological viewpoint’ (Gustafsson 2014, 508).

4 Gustafsson claims to have shown ‘that at least some such instances of paraphrasing (such as Russell's way with singular descriptions) are counted as explications by Quine’ (Gustafsson 2014, 523). Even though I cannot find an explicit argument, he at least recognizes that Quine claims that his treatment of singular descriptions is an instance of explication. Gustafsson also recognizes the tension (Gustafsson 2014, 518) to the paradigmatic account to explicate nouns (Gustafsson 2014, 522), but still claims that singular descriptions are an explication example.

5 For the rest of this paper, ‘explication’ without qualification means Quinean explication. Thanks to an anonymous referee for making me be explicit here.

6 This contrasts Quinean and Carnapian explication. Carnapian explication is iterative, i.e. some explicata might, in turn, serve as explicanda. This is impossible for Quinean explication, because his explicata are formulated in canonical notation and, therefore, are neither expressions of natural language nor defective.

7 See Hylton (2007, ch. 10, § I) for a more detailed description.

8 See Chihara (1968; 1973, §III.3, 1974) for discussion.

9 An anonymous referee objected that ‘ordered pair’ is not defective. However, Quine does classify ‘ordered pair’ as defective noun; see the beginning of § 53 of W&O where he briefly discusses Pierce's account. From today's perspective, this is difficult to understand as we have settled on what here turns out to be the explicatum, i.e. our current understanding is the end-product of an explication (not necessarily a Quinean, though).

10 Note that, as Gustafsson (2006, 61, n. 2) points out, this account is not entirely accurate as Wiener was working within Russellian type theory.

11 The proposals diverge if $x\ne \mathrm{\varnothing }$ since $\left\{\right\{x\},\{x,y\left\}\right\}=\left\{\right\{x\},\{y,\mathrm{\varnothing }\left\}\right\}$ iff $x=\mathrm{\varnothing }$.

12 An example of a ‘don't-care’ for Wiener's proposal is that $\mathrm{\varnothing }\in \bigcup 〈x,y〉$ ($=\{x,y,\mathrm{\varnothing }\}$) for any ordered pair $〈x,y〉$ whereas this is not the case for Kuratowski's proposal.

13 Cf. what Quine says regarding ideal objects (see Section 7.1): ‘The appeal to ideal objects in mechanics occurs regularly through universal conditionals: thus (x)(if x is a mass point then…). The nonexistence of ideal objects consequently does not falsify mechanics; it leaves such sentences vacuously true for lack of counter-instances’ (Quine 1960, § 51, p. 249). Quine goes on to reject this option.

14 To be more precise, Gustafsson (2006, 58, 2014, 520) argues that we eliminate ontological commitment because the explication puts us into a position to adopt an equivalent theory not ontologically committed to the explicandum.

15 This point is commonly missed; e.g. it is completely absent in Hylton's exposition (Hylton 2007, ch. 9, §III).

16 As noted by Szabó, Quine ‘is often rather elusive on what he means by paraphrase’ (Szabó 2003, 21, n. 20). Szabó claims to find six uses in Quine's (1948). However, the word ‘paraphrase’ only occurs twice in Quine (1948) so that we have to look elsewhere to substantiate Szabó's observation. Furthermore, Gibson (1982, § 3.5.1.2) comes close to a similar point as mine when he distinguishes between explication and what he calls ‘simple paraphrase’.

17 Quine presents this argument at least twice: § 33 (p. 159) to dismiss the synonymy claim for paraphrase and in § 53 (p. 259) to dismiss the synonymy claim for explication. As the discussion of the text makes obvious, the argument reduces to the claim that synonymy cannot be an adequacy criterion for (Use 3) because it inhibits disambiguation.

18 Gustafsson (2006) misses this because he misses the connection to paraphrase. I also disagree with his assessment that Quine's points are ‘applicable only to forms of dualism that do not assume disembodied minds’ (Gustafsson 2006, 65; cf. Gustafsson 2014, 521).

19 Gustafsson (2014, 522) agrees that the following examples (Sections 7.1–7.2) are not explications.

20 Contrary to what Gustafsson claims, viz., that Carnap's and Quine's conceptions are both ‘broad and loose enough to fit many of the same particular cases’ (Gustafsson 2014, 509).

21 That is also what Quine (1998b, 430f.) suggests: if physics is not capable of explicating ‘telepathic effects’, assuming they have been properly established, then we reject that version of physics and need to work on a new one.

22 Explication is also not part of W&O's ch. V entitled ‘Regimentation’.

23 This is why explication is part of W&O's ch. VII entitled ‘Ontic Decision’; for one's overall ontic decision to be adequate, it has to explicate everything useful.