Abstract
Can knowledge be justified true belief? It may seem unlikely. For aside from the most serious problem with it (the justification requirement), that knowledge (that p) even entails belief (that p) has often been doubted. Zeno Vendler (1972, Chapter V) has recently carried this doubt to the extreme, not just appealing to ordinary usage but bringing to bear a likely scientific account of the relevant linguistic facts. Vendler purports to show that there is a grammar theoretic defense of the thesis that “What is known may be something that cannot be believed or disbelieved at all” (pp. 90–91). So, even if “... it is nonsense to say one knows that p but does not believe it” (p. 90), it is not the nonsense of believing a contradiction. Rather, it is nonsense either way—to believe or disbelieve what is known nonsense because ungrammatical (when said) and ill-formed (when purportedly thought).
The first version of this chapter appeared in Philosophical Studies, 1977, 32, 203–209. Footnote no. 6 did not appear in the original version.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
Vendler introduced the idea of the object of verbs for speech and mental acts (and mental states) by way of the distinction between ‘perfect’ and ‘imperfect’ nominalizations. It’s the imperfect ones that lead to the distinction between those that are Wh-nominals or not. Perfect nomināls refer to, designate, and/or denote a particular event, process, or action. Earlier, Vendler (1967, p. 131) said perfect nominalizations contain dead verbs (tenses, auxiliaries, and adverbs are inappropriate), and imperfect nominalizations contain live verbs. There is much further grammatical work to be done here; e.g., integrating these distinctions with (Chomsky’s (1970) distinction between gerundive (live verb?) and derived (dead verb?) nominals.
So, in defining ‘F’, if the ‘completed’ wh-nominals are still thought of as themselves Wh-nominals, then F is more useful in paralleling concreteness. Being F (as defined) is apparently the same thing as Kiparsky and Kiparsky’s (1970) being a factive clause. For this reason, I disagree with McConnell-Ginet & Ginet (1976) when they say it’s “...likely that some of his arguments would not have survived the rigors of actual syntactic analysis as practiced by linguists (for example, the argument in [Chapter V] for the existence of two syntactically distinct ‘that’-complementizers)” (fn 2, p. 217). I believe actual grammatical analysis can support Vendler on ‘know’ and ‘believe’, such as by incorporating a concept of ‘completed’ Wh-nominals (certain factive clauses) in a fuller syntactic and semantic account. Further, the philosophical motive of finding a way to disagree with Vendler’s philosophical conclusion suggests, I believe, a grammatical hypotheses—viz., that in certain deep structures (or, alternatively, in semantic structure) there is required an ‘is true’ predicate that undergoes optional deletion (or, semantically, is mandatory in underlying semantic representations).
Kiparsky and Kiparsky (p. 163) mention that the idea of there being a syntactic structure underlying factive clauses might parallel the idea of verbs that take concrete objects (‘hit’) as vs. those that don’t (‘clarify’), paralleling the grammatical account of the seeing vs. calculating given above. An important trouble Kiparsky and Kiparsky note is that though ‘know’ is semantically factive (preserves presupposition through negation), it’s syntactically non-factive (fails syntactic tests they give). Since Rosenthal (1976, p. 248) says, “Vendler rightly stresses that the possibility of ‘wh’-nominalization is distinctive of so-called ‘factive verbs’ ...”, it might be thought Vendler himself characterizes Wh-nominals as factive clauses. However, nowhere in Res Co gitans does he use the word ‘factive’. Anyway, nothing Rosenthal goes on to say (pp. 248–249) casts any doubt on my analysis. His interesting footnote 9 (p. 249) might reveal where his evident belief originated that factivity was mentioned by name in Res Cogitans; i.e., Vendler probably referred to it this way in recent personal communications.
This analysis may be what Jones (1975) is trying to express (though I doubt it) in his Section IV, 232–233, prodded by Shoemaker (fn 3. p. 233). Some readers may think that I am inspired to pursue the approach I take by Jones. I definitely am not. (I first hinted at my way with Vendler’s analysis in the public discussion of an early version of his Chapter V (1968).) Even though Jones gives at least two good examples to support my analysis (p. 227 and p. 229) and he does hold that the that-p clauses in question might be ambiguous (p. 226), still he does not see the importance of the fact-proposition distinction and how to defend it as I suggest (cf. his flat dismissal of it at the top of p. 224 and bottom of p. 225). And, Jones is simply mistaken in his analysis of why Vendler’s arguments don’t show knowledge doesn’t entail belief (middle of p. 223). Finally, Jones’ suggestion (p. 226) about his (9) and (10), though as clever as his alternative explanation (top p. 227) fails to be, is misleading. When (10), ‘I know that what he said is true’, is ‘shorn of three words’ we do indeed get an ambiguous string ‘I know what he said’ (though Vendler claims it’s not). What this shows is that its ambiguity is (or could reasonably be explained as) one of deep structure (or else semantic structure) vs. surface structure (analogous to explaining ‘Flying planes can be dangerous’). The string could be derived from an underlying form in which no genuine Wh-nominals occur (reflected in (10)), or it could be derived from a form marking the ‘what’-clause as a genuine Wh-nominal (the obviously preferred reading). The short form of (9), ‘I believe what he said’, is not at all ambiguous in the same way (not permitting a Wh-nominal derivation).
Note that in Res Co gitans, Vendler has correctly given up a view he held earlier that there is not a single fact that a given statement or proposition might separately correspond to (or not) and, rather, that there are only ‘the facts’ in general (cf. Vendler, 1967, pp. 145–146).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1997 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Peterson, P.L. (1997). How to Infer Belief from Knowledge. In: Fact Proposition Event. Studies in Linguistics and Philosophy, vol 66. Springer, Dordrecht. https://doi.org/10.1007/978-94-015-8959-8_2
Download citation
DOI: https://doi.org/10.1007/978-94-015-8959-8_2
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4856-1
Online ISBN: 978-94-015-8959-8
eBook Packages: Springer Book Archive