New Study: No Difference in Gettier Intuition Across Cultures

Within the more metaphilosophically-oriented literature on experimental philosophy, there has been a great deal of discussion of the philosophical implications of cross-cultural differences in intuitions about Gettier cases. This work has been extremely impressive from a purely philosophical perspective, but at times, I worry that it has not been sufficiently closely connected to the actual empirical work in this area. In particular, much of it starts off from the assumption that people of different cultures differ in their intuitions about Gettier cases, but it turns out that the majority of the empirical studies actually find  that Gettier intuitions do not depend on culture in this way (see here, here and here). So it sometimes seems that people are investigating the philosophical implications of an effect that doesn't actually exist. 

Happily, Noûs has just published a truly amazing study on this topic by a team of experimental philosophers (Machery, Stich, Rose, Chatterjee, Karasawa, Struchiner, Sirker, Usui & Hashimoto), and I think this new study gives us a much better understanding of the relevant empirical facts. The researchers presented two different Gettier cases to participants in the United States, Brazil, India and Japan, yielding a total sample size of 521 participants. The study is extraordinarily impressive from a methodological perspective and very much worth reading in full, but the basic result can be expressed pretty simply in the following figure:

Overall, the study finds a robust tendency, found across all four cultures, to conclude that people do not have knowledge in Gettier cases.

Of course, this finding does not mean that philosophers were mistaken to think that there was something of deep metaphilosophical importance about looking at Gettier intuitions in different cultures. On the contrary, the result obtained here is a truly fascinating one, which surely has rich metaphilosophical implications. The key point is just that the metaphilosophical question we need to be asking is the opposite one from the one people have been discussing thus far. The question worth asking is not 'What are the metaphilosophical implications of cross-cultural differences in Gettier intuitions?' but rather 'What are the metaphilosophical implications of the extraordinary cross-cultural similarity in Gettier intuitions?' This latter question has not yet been sufficiently explored, but it opens up a whole new range of exciting issues that I hope philosophers will begin exploring over these next few years.

[The full paper is available to subscribers at Noûs, but please do feel free to write in with comments even if you have not yet read the paper itself.]

Comments

[Joachim Horvath]

Hi Josh, thanks for drawing everyone's attention to this newly emerging pro-Gettier consensus in the empirical literature. However, it may seem puzzling to some of the metaphilosophically-oriented people that the cross-cultural similarity "opens up a new range of exciting issues" in metaphilosophy. For wasn't this always the tacit consensus among (analytic) epistemologists anyway? So why doesn't this simply restore the status quo before 2001 (although now with a lot of extra empirical support!), and thus reduce rather than open up the range of related metaphilosophical issues?

[Joshua Knobe]

Hi Joachim,

Thanks for writing. Perhaps I'm mistaken, but my sense is that 20th century epistemologists simply weren't thinking about this issue either way. Clearly, there is something quite striking and surprising about the fact that this very subtle sort of effect (which even expert epistemologists have been unable to fully characterize) shows this shocking degree of cultural universality. Of course, it is possible that further inquiry will show that there is nothing of metaphilosophical importance to be said here, but I would put my bets on the other side.

[Clayton]

"Clearly, there is something quite striking and surprising about the fact that this very subtle sort of effect (which even expert epistemologists have been unable to fully characterize) shows this shocking degree of cultural universality."

I think I probably agree with Joachim on this point. I don't see what's striking or surprising about agreement, per se. I also don't see what's striking, surprising, or interesting about disagreement, per se. Perhaps with certain background assumptions in play, some agreements are interesting (and some disagreements are interesting) but I don't yet see what the rich implications of this study are supposed to be.

[Joshua Knobe]

Hi Clayton,

This is definitely a very reasonable perspective, but let me say just a little bit about why one might reach the opposite conclusion.

First, note that this empirical result raises new questions in cognitive science. When we find that people across all of these different cultures are arriving at the same judgments in Gettier cases, we are immediately faced with a question about *why* their judgments come out so similarly. Presumably, continued work on this cross-cultural similarity will tell us something about the underlying cognitive processes that generate people's knowledge attributions.

Now consider the metaphilosophical implications in light of that point. Of course, it is possible that the mere fact that people across all of these different cultures arrive at similar judgments has some metaphilosophical importance in and of itself, but in my view, that is unlikely to be the most fruitful place to look for metaphilosophical implications. A more fruitful approach would be a more indirect one, where we use results like this to learn something about the cognitive processes that generate knowledge attributions and then begin exploring the metaphilosophical implications of facts about those processes.

Needless to say, these implications will depend on specific empirical facts about the relevant cognitive processes, as well as on more substantive metaphilosophical claims about the relevance of facts about underlying cognitive processes for questions about the degree to which our judgments are warranted. So what I have said thus far does not itself argue for any particular conclusion. Still, perhaps it provides at least some sense of why this sort of issue would be worthy of investigation.