What We Talk about When We Talk about Truth: Dewey, Wittgenstein, and the Pragmatic Test

ABSTRACT

Pragmatic theories of truth need to pass the pragmatic test: they need to make a difference. Unfortunately, defenders of the pragmatic theory have rarely applied this test. I argue that a Deweyan pragmatic account of truth passes the test by identifying the political and epistemic dangers of certain types of social networks that create a durable consensus around false beliefs. To better understand Dewey’s account of truth I propose an excursion through Wittgenstein’s later views on knowledge and certainty.

KEYWORDS:

• Truth
• John Dewey
• Wittgenstein
• social networks
• pragmatic theory of truth
• warranted assertibility

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. This doesn’t necessarily mean that every theory of truth needs to be true by its own lights. But just as the verificationist theory of meaning was tripped up over its own meaningfulness, it’s natural to wonder how a given theory fares on the terms that it itself sets – and this question may be especially pressing for a pragmatic theory of truth. That’s because pragmatic theories are often read as offering a criterion of truth, either in addition to or in place of a definition, which immediately invites the question of how and whether it meets that criterion. The question may not arise quite so quickly or as forcefully for other theories, but for pragmatic theories it is difficult to avoid.

2. In anutshell the concern is this. If we start with Misak’s claim that one can never know that is true, and then add the plausible assumption that knowing implies knowing that is true, it follows by modus tollens that one doesn’t know : p p p p

• (1) It’s not the case that S knows that p is true. (Misak)

• (2) If S knows that p then S knows that p is true.

• (3) ∴ It’s not the case S knows that p.

(3) is a statement of philosophical skepticism much more radical than the sensible fallibilism usually associated with Peirce and pragmatism more generally. Avoiding the conclusion requires rejecting one of the premises and, since (2) follows from truth being a necessary condition for knowledge (Capps 2017, 144), this suggests that (1) is at fault. But (1) is a direct consequence of the Peircean account of truth. In anutshell the concern is this. If we start with Misaks claim that one can never know that is true, and then add the plausible assumption that knowing implies knowing that is true, it follows by modus tollens that one doesnt know :

3. This leaves open the possibility that some truth claims do make a claim about the future. The Deweyan point is that this is not always (and perhaps not even often) the case. If I point to a hot stove, most would agree that it is true that one should not touch it, based most likely on past experience. It’s not at all clear that the future enters into the equation at all: after all, it may be perfectly safe to touch the stove in an hour.

4. But might there also be corresponding practical disadvantages to the Deweyan account? Might having the courage of one’s convictions lead, say, to fanaticism, over-confidence, or avoidable mistakes when one claims to know a fact that later turns out to be false? These are risks, but I would argue that exiling the concept of knowledge poses an even greater risk: an account of truth that prevents us from knowing what is true shifts the odds in favor of nay-saying, denialism, and obstructionism. Put another way, while the Deweyan account allows for us to discover that we didn’t in fact know what we thought we did, the Peircean account does not allow us to to discover that we do know those things we thought we didn’t. Misak (2008) makes a similar point: that a commitment to the concept of truth does not lead to ‘zealotry’ – and that the advantages of recognizing the role of truth in politics and morals far outweigh the risks. The same point applies to the concept of knowledge as well.

5. Truth being subject-independent in this epistemological sense isn’t to deny that truth might be subject-dependent in another sense. The truth of ‘my hair is brown’ obviously depends on me in one sense – it’s my hair – but not in another: I can’t wish my gray away.

6. For example, in the opening pages of On Truth (1991). See also Le Morvan’s point that ‘though Ramsey often sounded … as if he countenanced the existence of propositions, we must remember that his talk of propositions (for instance, of propositions being true or false) was just a facon de parler, and that strictly speaking he … did not posit their existence’ (2004, 716).

7. As Sullivan and Johnston (2018) note, while Russell viewed propositions as truth-bearers in 1904 by 1910 he had switched over to judgments. While it bears repeating that Russell’s conception of both ‘propositions’ and ‘judgments’ is quite different than Dewey’s, it is notable that both eschewed propositions in favor of judgments. This indicates the state of philosophical flux during the first decades of the 20th century.

8. After noting that the idea of a proposition ‘seems never to have been made quite clear and unambiguous’ Tarski concludes that ‘for several reasons it appears most convenient to apply the term “true” to sentences, and we shall follow this course’ (1944, 342).

9. Like Tarski, Quine prefers to speak of sentences as truth-bearers: ‘the doctrine of propositions seems in a way futile on the face of it.…why not then just talk of sentences…and let the propositions go? The long and short of it is that propositions have been projected as shadows of sentences, if I may transplant a figure of Wittgenstein’s’ (1970, 10).

10. It’s worth noting that Wittgenstein and Dewey were aware of each other’s work to some degree. Dewey possessed a copy of Wittgenstein’s Tractatus and, according to O.K. Bouwsma, Wittgenstein may have heard Dewey lecture (Bouwsma 1986, 39). But Bouwsma also reports that when Wittgenstein learned that Dewey was still alive in 1949 he replied ‘ought not to be’ (29).

11. One example of Ramsey’s influence is in nudging Wittgenstein away from the correspondence theory he defended in the Tractatus and toward some kind of deflationary theory of truth. However, whether Ramsey ultimately defended a pragmatic theory of truth is a topic of some debate (Misak 2016) and, in any case, Wittgenstein was never comfortable describing his position as ‘pragmatic.’ Given that both Dewey and Wittgenstein had reservations about ‘theories’ of truth, a more fruitful point of contact is Wittgenstein’s more extensive comments on knowledge and certainty. For more on Wittgenstein’s account of truth see (Dolby 2017; Glock 2004; Horwich 2016) – all entitled ‘Wittgenstein on Truth.’

12. Wittgenstein puts the point succinctly:125. If a blind man were to ask me ‘Have you got two hands?’ I should not make sure by looking. If I were to have any doubt of it, then I don’t know why I should trust my eyes. For why shouldn’t I test my eyes by looking to find out whether I see my two hands? What is to be tested by what?

13. Wittgenstein’s example may be an oblique reference to his work as a porter and laboratory assistant at Guy’s Hospital in London during WWII. It’s also tempting to read this passage in light of Wittgenstein’s presence during Ramsey’s final illness, at the same hospital, where he kept a vigil until a few hours before Ramsey’s death (Misak 2016).

14. Here’s one. Imagine the hospital has lost power, the lights and computers are down, and patients need to be evacuated. You’re in charge of the ward. Medical transport arrives, you tell them to move the patient in Room 7, but they don’t see anyone (it’s dark, it’s chaotic, maybe the patient has been moved already). They ask again if there’s a patient in Room 7 and you say, ‘yes, I know there’s a sick man lying there.’ That sounds fine but note why: in this set of unusual circumstances you have information or evidence that others do not.

15. It’s easy to confuse this point: Wittgenstein might agree that ‘I know that a sick man is lying here’ isn’t exactly false, but then it isn’t exactly true either. It isn’t exactly true or false and this is because the claim makes no sense at all.

16. While Stroll’s cites two passages (OC §253 and §358–359) that suggest only that hinge propositions are beyond justification, another passage (‘205. If the true is what is grounded, then the ground is not true, not yet false.’) does support Stroll’s interpretation. My thanks to an anonymous reviewer for pointing this out.

17. Such situations will seem contrived, which is exactly the point: saying ‘it’s true that grass is green’ is not something we would ordinarily or typically say. Still, if one insists, the following dialogue might suffice:Aimee: I don’t like the color green or green things.Bruce: But you like the feel of grass on your bare feet.Aimee: OK. It’s true that grass is green.

18. Putting aside the mathematical model, it’s possible to see intuitively how this could happen. For example, a community might be trying to determine which of two treatment options is best. Perhaps the difference between them is slight: say, 92% effective vs. 95% effective. It’s possible, because of random variation, that the less effective treatment could have an early run of results that makes it appear more effective; likewise, the more effective treatment could have a run that makes it appear less effective. These early results could lock the community into a false conclusion that would be difficult or impossible to dislodge, depending on the structure of the network as well as its goals. Some networks are more interconnected than others; sometimes results are needed quickly. When speed is of the essence, and networks are interconnected, there’s a greater chance of error (Zollman 2007).

19. Of course, a lot depends on what, exactly, is meant by ‘indefeasible.’ As noted earlier, Misak proposes that an indefeasible belief is one that ‘would not be improved upon; it would forever meet the challenges of reasons, arguments, and evidence’ (Misak 2008, 95; see also, 2009, 32)) and argues that this is preferable to Peirce’s description of true beliefs as those that would be held at a hypothetical ‘end of inquiry.’ But, like any modal term, ‘indefeasible’ is open to multiple interpretations: indefeasible in the actual world? indefeasible in every possible world? just the nearby ones? Pragmatists would most likely be inclined toward a conception of indefeasibility in practice (or for all practical purposes) in which case it’s easy to imagine how a belief might survive past the point that anyone is inclined to challenge it, yet still be false. Better, I would suggest, to drop talk of indefeasibility or else recognize that it, like other modal terms, is fundamentally contextual. (For a related discussion of how ‘reliability’ is context-sensitive see Baumann 2016 (especially Ch. 2).)

20. It’s worth noting that O’Connor and Weatherall embrace a ‘broadly deflationary attitude’ toward truth that includes a ‘strong dose of pragmatism.’ But they differ from ‘traditional’ (i.e. Peircean) pragmatists who tie truth to the end of inquiry because ‘inquiry may well lead us astray’ (2019, 188–189). My claim is that a Deweyan account fits the bill.

21. I would like to thank two anonymous reviewers whose comments improved my argument – truly!