Brian Leiter has finally said something about Evolutionary Psychology that is more than just quick verbal jab. He did so in the update to this post. Note that the update is many times longer than the original post. I like it when people back up their claims with arguments. Maybe Leiter will do it more often, now.
If you're familiar with Gould's critique of EP, then you probably don't need to read Leiter's post (though there are some interesting facts in it), because it's pretty much the same thing. Interestingly, Buller, whom Leiter has mentioned favorably in the past, spends some time in his book arguing that this line of attack (the "we can't just assume it's an adaptation" line) is a critique that is neither fair to EP, nor extremely damaging to it. It will be interesting to see whether Leiter, in the article on biology in law that he says he is writing, references Buller or some of the replies to this critique by Evolutionary Psychologists. For the most part, on the biological side of things, I'd prefer to take Buller's approach, which, by the way, also notes that EP is woefully poor biologically (and neuroscientifically, and methodologically, and cognitively, and so on).
UPDATE: I should have noted that Leiter's arguments are more than those of Gould. Tacked on are those of Elisabeth Lloyd. I mention this mostly to post a link to her very good paper on EP, which you should read.
6 comments:
Hey, could you elaborate on your opinion that we have a better/simpler theory of metaphor comprehension than that of Lakoff et al.? Sorry for posting out of context but I wanted to make sure you read this. If you've already written about the theory you allude to then please just post a pointer; I couldn't find it anywhere in your blog myself.
Sure, try these posts:
Metaphor I
Metaphor II
Metaphor III
Metaphor IV
Thanks. I am not in cognitive science, but I have poked around the literature regarding metaphor comprehension for a little while.
Based on the findings of Fauconnier et al., I think that the structural-mapping theory of metaphor-type comprehension alone is almost certainly insufficient. However, one might recourse to a categorization theory of comprehension of comprehsion here. There seems to be good emprical evidence to do so, since familiar metaphors are processed as quickly as literal counterparts, and pragmatics should often allow for inference to be made more quickly than by special methods of metaphor comprehension. Work in this direction has been done by Brian Bowdle (at Indiana U.) and his colleagues. I think he is coming out with a new paper about it.
One interesting point about this theory is it is based on familiarity and conventionalization of metaphors which are understood categorically. Since less familiar metaphors are not understood as quickly they maintain that they are not understood through categorization. The insight of Fauconnier's work here, I believe, is that blending theory demonstrates that there exist principled ways of creating these meanings without using categorization (and without mere structure-mapping, which is insufficient).
It seems to me that there are also some theoretical reasons for liking a blending view of metaphor comprehension as well. For example I think that it boasts a high degree of re-use of specific conceptual modules and hence incredibly efficient reasoning. Consider a pattern of inference in a one domain which we wish to map onto another domain. If we view the source domain AS the target domain then we can reason in the source domain, and with sound reason map the final inference directly onto the target domain without having to first verify the validity of the inference in the target domain (although this might be done in parallel). In this way we can make very fast inferences, and this method can been viewed as a kind of blending.
For similar theoretical reasons metaphoric inferences made by categorization might also be viewed as a kind of blending.
This is probably the wrong place for this (it certainly has nothing to do with Leiter or EP), but I'll just quickly say that I've never seen structure mapping as in any way inconsistent with Fauconnier's conceptual blending, and since his theory is so underspecified (we have no idea about the specifics, computational or otherwise, of any of the steps in the blending process), structure mapping seems to be the preferred approach. In fact, you could probably combine the two if you wanted to keep some of the blending metaphors (e.g., the spatial ones).
yeah, sorry about that! since you post anonymously i couldn't email. i would be interested in discussing this stuff further if you have the time. my email is amoneill at ucsd.edu
My email address for the blog is on the sidebar.
Post a Comment