Embodiment critique in press in Cognition

Hmmm.. let's see how to be objective about this. This guy, Machery, wrote a paper critiquing the methods of what he calls neo-empiricists, or as I prefer, embodiment freaks, regarding how to deal with specific amodal hypotheses. In brief, there are three points:


First neo-empiricists' data do not rule out "the" amodal model, but merely a specific prediction of a few amodal hypotheses.

Second, the neo-empiricists' support for their claims can also be accounted for by a combination of Kosslyn-style imagery along with amodal networks. This is similar to Paivio's dual coding model.

Third, the scope of the amount of representations are amodal vs. perceptual symbol systems is a problem. Machery points out that abstract concepts such as number must be amodal representations, and thus evidence against wide-scope neo-empiricism, which claims that all representations are perceptual symbol systems. Since number is one example of an amodal representation then there is likely to be more and then the question becomes to what extent, according to Machery.

My reaction to this paper as an embodiment freak, could fill a response paper, but so far the reaction from others in my camp have decided that Machery's article is not worth the rebuttal. Here are a few reasons why I think his critique fails.

Regarding his first point, ruling out specific hypothesis in a lab setting is simply how science works. Neo-empiricists have not underestimated the ability of amodal models to be unfalsifiable. The approach to designing experiments to dissociated between amodal vs. embodied has evolved since the late 1990s, and Machery should realize that. As a whole there is a LOT of evidence that amodal theories cannot account for, and maybe he can find a few new models which can account for Barsalou's findings in (1999) and (2001), but what about Glenberg's, Kaschak's, Zwaan's, Rinck's, Spivey's, Altmann's, or Matlock's (or in fact my data)? Maybe we should ask amodal freaks to design experiments which will go down the line with each of these empirical results and rule them out as Machery has suggested neo-empiricists should do. This failure on his part to realize what he is proposing demonstrates to me his lack of understanding science.

Concerning his thoughts on imagery, I think he has an interesting point. I'll leave it at that and also say that I would rather develop a research program to look into these issues than his first proposal.

Lastly, I think the scope problem is also a good point. However, I wish he had done a little more research into where the other neo-empiricists besides Barsalou fall on his continuum (e.g., Glenberg). Actually Glenberg is mentioned in the abstract, but not again, nor is Glenberg and Kaschak (2002), which is a well-known staple of citing empirical support for embodiment theory. I myself am pushing to see if perceptual symbol systems are involved in abstract concepts, including number, and I disagree with his assumption that number concepts are amodal. So maybe I'm a wide-scope embodiment freak, but I'm curious to see how wide I can push the scope.

In conclusion I don't know if there is going to be a formal response paper anytime soon, but I can say that the embodiment people I know think that this paper is lacking a clear understanding of scientific methods.