Around a year ago, I wrote about a really
beautiful study by Machery and colleagues on cross-cultural similarities in epistemic intuitions. The study looked at intuitions about Gettier cases in four different cultures and found that the Gettier intuition was remarkably robust across demographic differences.
But of course, I could imagine some readers seeing this as relatively poor evidence of cross-cultural robustness. "After all," they might say, "the Gettier intuition is one of the most fundamental aspects of our practice of knowledge attribution. Even if this one intuition turns out to be widely shared, the more subtle and complex aspects of our epistemic intuitions might easily turn out to vary across cultures."
Now, a year later, we have more information about this question. The unstoppable team of Minsun Kim and Yuan Yuan has just completed a
new paper on the topic, and they provide evidence for a very surprising degree of cross-cultural robustness.
Kim and Yuan look at three different effects that have been obtained in studies on Western participants:
Friedman & Turri (2014) showed that people were less inclined to attribute knowledge in cases of probabilistic evidence than in cases of perceptual evidence, even when the probability of error was held constant.
Starmans & Friedman (2012) showed that even within Gettier cases, people are more inclined to attribute knowledge when the epistemic subject has what they call 'authentic' as opposed to 'apparent' evidence.
Buckwalter (2014) showed that even within Gettier cases, people are more inclined to attribute knowledge that something morally bad will take place than that something morally good will take place.
Kim and Yuan translated the materials used in these existing studies into Chinese and Korean and then ran the studies with participants in China and South Korea.
Strikingly, all three of the effects that had been found with Western participants also emerged with participants from these other cultures. In other words, even the most surprising and quirky effects on people's knowledge attributions -- including effects that were only discovered in the past few years -- seem to be remarkably robust across cultures.
Drawing on these results, Kim and Yuan argue that it is time for a pivot in our more metaphilosophical discussions. The time has come to start thinking more deeply about the philosophical implications of the rather extraordinary cross-cultural similarities we find in people's epistemic intuitions.
[If you are interested in learning more about what experimental philosophy has shown about demographic factors and philosophical intuitions, please take a look at this
attempt at a comprehensive list.]
This is a really interesting paper. Thanks to Kim and Yuan for their terrific work and to you, Josh, for sharing it here.
Let us assume these intuitions are universal. What could explain this? Kim and Yuan distinguish nativist and social explanations (i.e., culture). I lean towards nativism. Here’s why:
There is another aspect of these intuitions that wasn’t highlighted in the paper that is relevant: their remarkable sophistication. These intuitions track very subtle and non-obvious features: the difference between a Gettiered and non-Gettiered agent, probabilistic versus perceptual evidence, means versus foreseen but unintended effects (i.e., side-effects). The processes and principles that produce these intuitions thus reflect a very sophisticated psychological competence. The socio-cultural explanation thus has to propose that there are these highly abstract and sophisticated competences across cultural groups that arose more or less independently but, nonetheless, these competences ended up all being highly similar. To the extent this seems unlikely, this pushes us in the direction of nativism.
There are also Chomsky/John Mikhail-style arguments about learnability, poverty of the stimulus, etc. in the vicinity here. But I will stop here.
Posted by: Chandra Sripada | 08/01/2016 at 06:03 AM
Hi Chandra,
For what it's worth, I completely agree. My sense is that this conclusion actually has some interesting implications for questions about the role of knowledge in theory-of-mind. One possible view would be that the core of theory-of-mind is our capacity to attribute beliefs and desires. Then one might think that the capacity to attribute knowledge is just some little extra thing, of considerably less importance. However, if we suppose there is an innate basis even of these quite subtle patterns in knowledge attribution (as you argue here), it seems like we might have some reason to shift toward a view according to which the capacity to attribute knowledge is more central or fundamental to the way theory-of-mind works more broadly.
p.s. I am super impressed that Chandra has already read this whole paper, but please do feel free to write in with comments on the basic idea of this finding even if you haven't read the paper itself.
Posted by: Joshua Knobe | 08/01/2016 at 08:31 AM