In the extensive discussion of the Gettier cases (see my earlier entry), little attention has been given to the notion of justification. If his critique is to have a point, Gettier needs the justified-true-belief story. But how are we to understand the account that he is attacking? When do we speak of justification, how does talk of justification enter into the context of human conversation? Very often, the issue is discussed as though matters of justification could be debated regardless of why the question arises, who raises it, in what circumstances. (Contextualists are an apparent exception – I hope to get back to them in a later entry.)
Gettier cites Ayer and Chisholm as expressing a similar point by speaking in terms of “having a right to be sure” and “having adequate evidence”, respectively. According to Gettier, these forms of expression can be treated as equivalent.
Consider an example. Martha said, “The water here is deep enough for diving in”. Was she justified in saying this? In one case, our concern is with deciding whether or not we can jump in from here. Here the question of justification is a question of reliability. Is what she’s saying true? This is important, since if there are underwater rocks, diving in would be very risky. We might ask her how she knows, though if she lives here and we know she often goes swimming, we may just take her word for it, provided, of course, we take her to be a reliable person. Either way we may tell one another, “Martha knows the water here is deep enough for diving in”.
Would this be a case of knowledge equalling justified true belief? Well, it would be strange here to speak about the latter as a condition of knowledge. We don’t infer Martha’s knowledge from the truth of what she is saying, rather we reason to the truth from the conviction that she knows. So this doesn’t seem to be a case of what Gettier is criticizing.
In another case, our concern is not with how things stand, but with Martha’s right to say what she said. This is the most likely to be the case when something goes wrong, but then that wouldn’t be the sort of case Gettier was trying to exclude – no “true belief” there. On the other hand we might think that what Martha said was true, but upbraid her for giving her assurance too easily. Suppose she had asked a local whether diving was safe there, and he had told her it was. Then it turned out he had been speaking about a different shore with the same name. Though as it happens both shores are safe. That’s seems more like a Gettier case. We might say she had an excuse for making the mistake – she had a “justification” in that sense. But that is hardly what Gettier was attacking: “knowledge = true belief with an excuse”. (Neither would we say, with Chisholm, that her evidence was adequate, nor with Ayer that she had a right to be sure.)
We haven’t succeeded, it seems, in finding a case in which we are tempted to say what Gettier criticizes us for saying.
In the Gettier debate, there is little consideration for how words are used in actual human conversation (how long will the anti-Wittgenstein, anti-Ordinary Language Philosophy backlash last?). Consequently, what are evidently descriptions of different contexts of use get treated as competing theories. Thus, according to Jonathan Jenkins Ichikawa & Matthias Steup,
Internalists about justification think that whether a belief is justified depends wholly on states in some sense internal to the subject… Externalists about justification think that factors external to the subject can be relevant for justification…
(”TheAnalysis of Knowledge", Stanford Encyclopedia of Philosophy (Spring 2014 Edition), Edward N. Zalta (ed.).)
Now, if what is at issue is a person’s responsibility for claiming to know something, it is natural to assume that what will be relevant above all are “states internal to” her, assuming that to mean: the information at her disposal, her previous knowledge and experience, her competence for forming certain types of judgment, etc., whereas when the reliability of the speaker’s claim is concerned, matters independent of the speaker will become important. Thus, rather than internalists and externalists addressing a single question (what might that question be, in a context of regular conversation?), it is plausible to suggest that what we have here are two fairly plausible answers to two different questions.
In fact, neglect of conversation seems almost to have become an ideology. In the article referred to above, speaking about the context-dependence of knowledge-claims, Jonathan Jenkins Ichikawa & Matthias Steup seem rather dismissive of the relevance of considerations of use for the analysis of knowledge. According to them, the “relationship between contextualism and the analysis of knowledge is not at all straightforward. Arguably, they have different subject matters (the former a word, and the latter a mental state)”, though they admit that “the methodology of theorizing about knowledge may be helpfully informed by semantic considerations about the language in which such theorizing takes place.” (Presumably, “semantic considerations” are taken to include observations concerning the way a word enters into conversation.)
Getting clear about the use of the word “know” (“semantics”), on their account, can thus be distinguished from study of the “mental state” of knowing. Knowing, as it were, is like a porcupine which can be observed, measured and dissected without our having to pay much attention to the way the word “porcupine” figures in human conversation. What is left open is how this language-independent observation of knowledge is supposed to be carried out. I’m afraid that in practice it will simply consist in the philosopher asking himself, independently of any context, “Would this be knowledge?”, “Would that be knowledge?” , as it were, feeling the taste of the word on his tongue. How reliable is that?