I found two essays on a kind of response to the Gettier problem. One is a Philosophy Now article, and another is a blog post. On both sources, they argue that Smith's belief (on the original Gettier example) that "The man who will get the job has ten coins in his pocket," (or belief b) refers not just to any man who might get the job, but only to Jones. In the Philosophy Now article, the author claims that belief b is false because Jones is not the one who gets the job. In the blog post, the author has considered Donnellan's argument such that even though the description found in belief b, "The man who will get the job," is not accurate of Jones, Smith can still use it to successfully refer to Jones. As per Donnellan, although the definite description denotes (Smith, who has ten coins), people refer (to Jones, who has ten coins), however inaccurate their description is.

Either way, the general idea is that the belief found in a Gettier example isn't as it seems. The proposition in the belief (the belief that P) is actually distinct from the proposition in the knowledge (the knowledge that   P  Q) that the example tries to show to be absent. Are there any scholarly works that pursue and critique this kind of idea?


2 Answers 2


The approach that (at least) some Gettier cases are the result of misdescribing beliefs due to disregarding linguistic conventions (such as Grice implicatures) is still actively pursued and even rediscovered. See for example Jose-Mabaquiao, Resolving the Gettier Problem in the Smith Case: The Donnellan Linguistic Approach (2018), Ludlow-Segal, On a unitary semantical analysis for definite and indefinite descriptions (2004), Schmidt-Petri, Definite Descriptions and the Gettier Example (2002) and its discussion on Philosophy News in 2011:

"Christoph Schmidt-Petri in his “Definite Descriptions and the Gettier Example” provides a framework for the argument I outlined in an earlier post in terms of a distinction between the attributive use and the referential uses of definite descriptions drawing heavily from Donnellan’s analysis of definite descriptions... Schmidt-Petri has shown that whether we take Smith’s belief as referential (he believes A or B) or attributive (he believes C) in all three cases he does not know P2 because he has not satisfied the requirements of JTB not because JTB is inadequate in some way."

Yakubu in Truth Analysis of the Gettier Argument goes into the root cause of such misdescription, the mismatch between natural language and its treatment in formal semantics and epistemology. He also points out that analytic philosophy of language (especially on the side of pragmatics as advanced e.g. by Austin and Strawson) has been much more sensitive to the ambiguities involed.

"The methods Gettier used to construct his challenge,however, utilized certain principles of formal logic that are actually inappropriate for the natural language discourse of the Gettier cases. In that challenge to epistemology, Gettier also makes truth claims that would be considered controversial in analytic philosophy of language. The Gettier challenge has escaped scrutiny in these other relevant academic disciplines, however, because of its facade as an epistemological analysis.

However, as was suggested early on linguistic nuance may not resolve epistemic flaws of JTB, see e.g. Hooker's patch in In Defense of the Principle for Deducibility of Justification, p. 403ff to avoid the Donnellan's attributive/referential ambiguity:"argument would not have applied had the inference been [formulas] where the canonical sentences are of a lower predicate calculus with definite descriptions whose semantics does not allow for two Donnellan-type readings". In other words, even if we assume that colloquial interpretations of Gettier examples are afflicted by linguistic ambiguity one can cook up variations that would cause troubles for JTB. But it does suggest that Gettier examples may not have been as natural as they may seem to be initially.

  • Thank you! This ought to be noticed by epistemologists. Even if JTB still needs to be modified, I don't think they should accommodate Gettier cases that can be resolved through the analytic philosophy of language. Trying to solve problems that don't exist won't be good for them. But if the development of the philosophy of language turns out to solve all Gettier cases then sure, why not?
    – brendt
    Oct 16, 2021 at 3:28

There are Gettier cases that don't have this alleged problem. There is Gettier's second example, which involves disjunction, from the original paper. Smith has a justified false belief that Jones owns a Ford. Because p entails p or q, Smith deduces that Jones owns a Ford or Brown is in Barcelona. It turns out that Brown is in Barcelona, which Smith didn't know, so Smith's disjunctive belief is justified (because it was logically deduced from a justified belief) and true, but intuitively not knowledge.

It's easy to construct other cases. You see from the window of your apartment your friend Alice coming toward your building and you think "Alice is coming". But in fact that is just someone who really looks like Alice. A moment later, Alice comes in. Your belief was true and justified. But intuitively it was not knowledge.

There is also Russell's famous stopped clocked case, which has been retrospectively recognized as a Gettier case. You look at the clock and see that it says 3:15. You believe on this basis that it is 3:15. It really is 3:15. But the clock is not working -- it stopped exactly 24 hours ago.

There is by now a known recipe for constructing Gettier cases, due to the philosopher Linda Zagzebski (see https://plato.stanford.edu/entries/knowledge-analysis/#KnowAnal). The main ingredient is luck. Consequently, there has been a lot of work in epistemology on epistemic luck. See for instance https://iep.utm.edu/epi-luck/.

  • Thank you! But I already know that. Btw, linguistic approaches to solving Gettier problems seem to be done on a case-by-case basis. So it is unlike the other ways in which they modify JTB to solve everything at once. I am only looking for a solution similar to the one in my question. Those other cases you mentioned might have a different strategy on the linguistic approach. See Yakubu's paper on the different strategies to answering Gettier problems using the philosophy of language. (credits to @Coniford for the link to Yakubu's paper!)
    – brendt
    Oct 15, 2021 at 15:38
  • @brendt It is unclear to me how any linguistic approach might solve the cases I mentioned. The received wisdom in epistemology seems to be that it is not the way to go.
    – E...
    Oct 15, 2021 at 16:16
  • Exactly why yakubu wrote his paper. "Thus, natural language and classical logic have different standards for truth, a fact the Gettier cases do not seem to recognize. This appears to have slipped under the radar in the epistemological dis- course, and the Gettier truth claims have been taken for granted, even though epistemologists think the justification for the claims is dubious." That's a quote from his paper. See also Grice on the gap between classical logic and conversation.
    – brendt
    Oct 15, 2021 at 17:04
  • By the way the paper is free. He has examples there. Including Gettier's Case II, the first example in your answer.
    – brendt
    Oct 15, 2021 at 17:11
  • I think Russell's Stopped Clock example merely shows how difficult it can be to justify a belief. Beliefs are built upon other beliefs. Here the belief that the clock is accurately showing the time is not justified. I think it is a non-problem, holistically speaking but a problem when given complex situations of beliefs where sorting out the justifications is complicated. Similarly with the truth conditions. Oct 15, 2021 at 18:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .