[I can’t remember what I was responding to.]
There’s nothing “cartesian” about the language of cognitivism. Information processing is just a viewpoint on phenomena which doesn’t give a damn about ion flows or gene expression. It just posits that there’s something transforming what’s perceived into the actions, and whether it’s a set of cogs or a Turing machine isn’t particularly interesting. These guys need to go back to Neisser!
I imagine a load of these conceptual analysts (using a priori wisdom they received from where?!) pounding their fists on a table, some of them agreeing it is a table, some of them arguing that, no, that’s ridiculous, it’s a collection of atoms, electrons, and protons, … There are multiple levels of analysis, and somewhere those levels have to connect to what it feels like to be a person and how people communicate with each other about what they’re doing. I agree that sometimes the language that we use at the personal level gets applied, by analogy, to what the brain’s doing at the sub-personal level, but often that’s just to try to tell a story about what’s going on. For instance today [a fairly famous researcher] talking about a parietal area “caring” about something or other. It was just a cheap way to get an idea across instead of saying, “We were able to reject the null hypothesis that there is no difference in BOLD activation between the two conditions (with alpha = 0.05).”
Many of the theories used by fMRI folk seem not far from the folk psychological vernacular and thus are much in need of refinement to make them more consistent with what a charming Italian professor termed the “meat machine” is up to. That’s the point, to me, of fMRI et al: improving consistency between what the brain’s up to and our models of information processing.