Talk:Primates-May 2018/@comment-35453425-20181213061153/@comment-35476160-20181213163758

Every time I've skimmed this paper I just don't really seem to be able to get on board with this paradigm. I think the intro lays out really nicely how face processing is largely experience driven. Apart from pareidolia stuff, I don't love how they split up species pairs here. I also would have liked to see conditions in which they are matching individuals within species rather than just matching identical images or matching between species images. To me, it seems like matching identical images or between species doesn't require second-order face processing strategies/ strategies where experience/familiarity matters more (whereas matching different images of the same individual within species would do this more). That way you would likely see much better performance for familiar species and drop off with unfamiliar species that they did not see here. In other words, perceptual differences seems to be the low hanging fruit of the way they set up these tests. If you aren't required to use more 'precise' second-order processing techniques to discriminate between species (as you would between individuals), then yes, perceptual similiartity is going to be the way to go. But I'm not sure I would say its being 'prioritized' over mechanisms that you see in paradigms that do elicit stronger familiartiy effects. I'd love to hear others' thoughts though because maybe I missed something in here?

As far as this relates to pariedolia, I think it'll be interesting to see how potential experience will play a role, because we don't really know if they do experience the perception of illusory faces or how 'familiar' they are with them. As with less familiar or less socialy relevant faces, processing will likely be reflected as more featureal based, but with more experience/sensitivity to it I would expect to see more configural. If the monkeys have little experience recognizing illusory faces, I'd expect pariedolia images to be completely featureally processed (i.e. they won't see them as a face at all). If their sensitivity to face-like configurations has given them experience in misperceiving faces then they should at least process paridolia images somewhat (probably not completely) configurally. That's why I think a form of a rating scale for them would be really insightful into whether or not/how much they experience the illusion!