Abstract
Against contextualism, Duncan Pritchard has argued that conversational pragmatics give rise to an argument in favour of invariantist neo-Mooreanism. More specifically, he argues that when we conjoin a Moorean view with a warranted assertability manoeuvre, we can satisfy our pre-theoretical intuitions (which are decidedly invariantist), whereas contextualists cannot. In the following paper, I challenge Pritchard’s argument and contend that he is too quick to declare victory for invariantism, for not only does the WAM he employs appear to be ad hoc vis-à-vis DeRose’s plausible criteria, but it also seems to have very implausible pragmatic implications when subjected to close scrutiny.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Briefly, Pritchard’s two-tier relevant alternatives account is as follows. Take an epistemic agent named Zula, who has normally functioning perceptual faculties in a cooperative (demon-free) external environment. If Zula sees what she takes to be a zebra at the zoo, and no error-possibilities such as the zoo using cleverly disguised mules are raised, then she knows she sees a zebra because her reflectively accessible perceptual belief is factive. This is “first-tier” perceptual knowledge—when no defeaters are in play and conditions are optimal, one’s bare perceptual belief is sufficient for knowledge. If, however, Zula sees what she takes to be a zebra at the zoo, and the error possibility that she is instead looking at a cleverly disguised mule is raised, it prima facie appears that she cannot know that she sees a zebra by virtue of this bare perceptual belief because its putative factivity is defeated by our introspective awareness of an indistinguishable error possibility. Despite this, Pritchard thinks that such a relevant error possibility defeats her zebra belief only if we assume a principle to the effect that epistemic support for a perceptual belief must come from a perceptual source, e.g. that Zula can perceptually discriminate a zebra from a disguised mule:
The perceptual principle
If S has perceptual knowledge that ɸ and S knows that another (known to be inconsistent) alternative ψ does not obtain, then S must be able to perceptually discriminate between the object at issue in ɸ and the object at issue in ψ.
Pritchard, however, denies the Perceptual Principle, and argues that if we hold the following alternate principle of epistemic support:
The favouring principle
If S (i) knows that ɸ, (ii) knows that ψ, and (iii) knows that ɸ entails ψ, then S has better evidence in support of her belief that ɸ than for believing not –ψ.
then Zula can know that she sees a zebra despite the raising of the cleverly disguised mule possibility. This is so because Zula can call upon non-perceptual background evidence which supports her zebra belief, i.e. she believes that in normal circumstances, disguising mules as zebras would be a waste of time, and that zookeepers are not normally interested in deceiving zoo crowds. These background beliefs, according to Pritchard, rationally support her belief that she sees a zebra rather than a mule: it is more rational, on his view, for Zula to infer that she sees a zebra, based on these sorts of background considerations, than for her to reject her zebra belief. Thus, Pritchard argues that if we accept the favouring principle rather than the perceptual principle, we can defeat perceptually indistinguishable defeaters to our perceptual beliefs (in normal conditions) and as a result, by the closure principle, know that we see a zebra instead of a mule. This type of knowledge which is attained through our perceptual belief that p plus rational epistemic support is what he calls “second-tier” perceptual knowledge. Note that the disjunctive case is an application of the favouring/discriminating epistemic support distinction. Thanks to an anonymous referee for pointing this out. For more, see Duncan Pritchard, Epistemological Disjunctivism, Oxford: Clarendon Press, 2012, pp. 1–35.
This problem, moreover, is not isolated to Pritchard’s two-tier RA approach, rather it generalizes to any invariantist plus WAM strategy.
For a more specific example, take DeRose’s “Bank Case” (1992). Non-sceptical invariantists use a WAM to argue that it is conversationally appropriate for John to assert that he does not know that the bank would not be open Sundays when the possibility of defeaters is raised by his wife Mary despite the fact that it is true that he knows. What is happening is that Mary mistakenly thinks that the standard for knowing whether the bank is open is much higher than John’s evidence would warrant; thus, her question generates an implicature to the effect that John only knows the bank is open if this (mistaken) standard is met. So, when John says he does not know, he is saying something conversationally appropriate (qua Mary’s implicature) yet false. So, it is not that the truth conditions are shifting, rather it is the assertability conditions which are varying. Contextualists mistake assertability conditions for truth conditions. See Keith DeRose, Contextualism and Knowledge Attributions, Philosophy and Phenomenological Research, 52 (4), 1992, pp. 913–929.
Pritchard 2012, p. 142.
Pritchard 2012, p. 145.
How exactly this goes depends on whether the contextualism in question is attributor based or not. But this is beyond the scope of the present discussion. Here, I will assume attributor contextualism because it is Pritchard’s target.
Some contextualists construe relevance in an externalist fashion, some in an internalist manner. Here, I shall not attempt to adjudicate the issue.
Duncan Pritchard, Neo-Mooreanism Versus Contextualism, Grazer Philosophische Studien 69 (1), 2005a, p. 6.
Duncan Pritchard, Neo-Mooreanism Versus Contextualism, p. 7.
It was pointed out to me by a referee that one might ask why we should accept DeRose’s constraints upon a WAM. In reply, I think that the expectation that a WAM be well motivated is independent of DeRose’s constraints, and thus does not essentially depend upon the soundness of his argument. This is why I mention above that it is orthodoxy to demand that a WAM be well motivated.
Keith DeRose, Contextualism, An Explanation And Defense, In J. Greco & E. Sosa (eds.), The Blackwell Guide to Epistemology, Blackwell Publishers, 1999, pp. 15–18.
Why? It is because if I know that my colleague is in her office, I should not respond with a logically weaker possibility claim, for this weaker claim generates the conversational implicature that I do not know that my colleague is in her office.
According to DeRose, “Assert the Stronger” is a sub-maxim of the Gricean Maxim of Quality. See DeRose 2002, p. 9.
DeRose 2002, p. 10.
Cf. DeRose 2002 p. 17: “What of the defense of the crazed theory about ‘bachelor’? As it stands, it does not appeal to the generation of a false implicature to explain away apparent falsehood. It’s rather an instance of what we may call a “bare warranted assertability manoeuvre” — a WAM that simply explains away the problematic intuitions of falsehood by claiming the assertions in question are unwarranted or explains away intuitions of truth by appeal to the warranted assertability of the relevant assertions, without further explaining why the true assertions are unwarranted or the false ones warranted.”
More specifically, DeRose argues that we cannot fit the “crazy” theory into any Gricean norm; so it follows that any rule we cook up to accommodate the WAM will be particular to assertions of the term “Bachelor” and nothing else. So, it is paradigmatically a special rule. See 2002, pp. 7–9.
An anonymous referee asked why we should hold Pritchard hostage to the DeRose account of WAM acceptability. My brief reply is that the DeRose account is one of the more detailed in the literature and thus, outlines a plausible set of desiderata. If one so wishes, they can construe my argument as a conditional claim: if something approximating the DeRose account of warranted assertability is correct, then Pritchard’s WAM is ad hoc.
See Paul Grice, Logic and Conversation, in Syntax and Semantics 3; Speech Acts, eds. Peter Cole and Jerry L. Morgan, New York: Academic Press, 1975, pp. 45–50.
Note that this line of argument is also suggested by Jonathan Ichikawa (2014) in his blog post ‘Pritchard on pragmatics of knowledge ascriptions’ dated August 30, 2014. Accessed from http://blog.jonathanichikawa.net/2014/08/pritchard-on-pragmatics-of-knowledge.html
In a similar manner, Baumann 2011 extends contextualism to mental knowledge attributions. And if this is so, he argues, the WAM is not a fully general reply to the contextualist, as WAMs do not plausibly extend to thoughts. Thanks to an anonymous referee for alerting me to this.
Where by “evidence class or type” I mean basic evidence classes like auditory or perceptual evidence.
The term “explicitly” here is very important; for I take the example to differ from the previous examples featuring Zula in the following way: in the present example, Dave does not ask whether Zula knows if the truck is a Dodge or a Ford, and I have stipulated that he is not of the sceptical disposition. In the previous examples, Zack the sceptic explicitly asks Zula how she knows a relevant alternative is ruled out. I take the difference in assertion to correspond to a difference in conversational context: whereas Zack’s questioning raises the epistemic standards of the conversation, Dave’s mention of the Dodge does not. And this makes the latter example ‘everyday’ in the sense that the epistemic standards are quite low. And because it looks as though the PDI is still going to make Zula’s assertion in the truck example conversationally inappropriate, Pritchard is faced with a further problem: we make many such everyday assertions, and it seems plausible to think that they are warranted. He cannot hold this if I am right here.
For a related discussion, see Adam Leite, Review of Duncan Pritchard, Epistemic Luck, Notre Dame Philosophical Reviews, 2006 (4).
References
Baumann, P. (2011). WAM’S: why worry? Philosophical Papers, 40(2), 155–177.
DeRose, K. (1992). Contextualism and knowledge attributions. Philosophy and Phenomenological Research, 52(4), 913–929.
DeRose, K. (1999). Contextualism: an explanation and defense. In J. Greco & E. Sosa (Eds.), The Blackwell Guide To Epistemology (pp. 187–205). Blackwell Publishers.
DeRose, K. (2002). Assertion, knowledge & context. Philosophical Review, 111(2), 167–203.
Grice, Paul. Logic and conversation. In Syntax and Semantics Vol. 3: Speech Acts, eds. Peter Cole and Jerry L Morgan, New York: Academic Press, 1975, pp. 41-58. 1989
Ichikawa, J. J. (2014). Pritchard on pragmatics of knowledge ascriptions. There is some truth in that, blog post, August 30th, 2014. Accessed from http://blog.jonathanichikawa.net/2014/08/pritchard-on-pragmatics-of-knowledge.html.
Leite, A. (2006). Review of Duncan Pritchard, Epistemic Luck. Notre Dame Philosophical Reviews (4).
Moore, G. E. (1959). A defence of common sense. Philosophical papers (pp. 32–59). London: Allen & Unwin.
Pritchard, D. (2005a). Neo-Mooreanism versus contextualism. Grazer Philosophische Studien, 69(1), 1–24.
Pritchard, D. (2005b). Neo-Mooreanism, contextualism, and the evidential basis of scepticism. Acta Analytica, 20(2), 3–25.
Pritchard, D. (2005c). Epistemic luck. Oxford: Clarendon.
Pritchard, D. (2007). How to be a neo-Moorean. In S. Goldberg (Ed.), Internalism and Externalism in Semantics and Epistemology (pp. 68–99). Oxford University Press.
Pritchard, D. (2012). Epistemological disjunctivism. Oxford: Clarendon.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Cockram, N. Pritchard, Revisionism and Warranted Assertability. Acta Anal 31, 439–454 (2016). https://doi.org/10.1007/s12136-016-0288-x
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12136-016-0288-x