In the lead up to our upcoming conference, we thought it would be fun to read some relevant papers by our keynote speakers. First up is a paper by Gary Hatfield, written in collaboration with psychologist William Epstein:
Epstein, William and Gary Hatfield (1994) “Gestalt Psychology and the Philosophy of Mind,” Philosophical Psychology 7(2): 163–181.
Epstein and Hatfield examine the tension between the gestalt psychologists’ apparent physiological reductionism and their explicit realism about phenomenal qualities. They also shed light on some of the puzzling comments we saw in previous readings from Köhler and Koffka about the place of value and meaning in the natural world.
As always, please contact one of the organizers if you have trouble accessing the reading.
I have to admit, I find it quite difficult to find anything to disagree with in Epstein & Hatfield. Their exegesis of the gestalt psychologists seems spot on, and I’m extremely sympathetic to the morals they draw for contemporary cognitive science (morals which, 20 years later, still have force).
As someone sympathetic to the gestalt project, I was somewhat comforted that E&H consider the notion of a “physiognomic character” to be underdeveloped. If one endorses the goal to find a place for “value” in a “world of facts,” then the gestaltist appeal to “physiognomic character” can seem like sleight of hand — they introduce a technical term that appears to merely rename, rather than explain primitive perceptions of significance. I do think there is an evolutionary story to tell about the perceivable characteristics that evoke such primitive perceptions of value, but neither the gestaltists nor Gibson have done so in sufficient detail (moreover, I don’t think either of them has a nuanced enough understanding of evolutionary explanation to see the real challenges here).
The characterization of Gibson’s discussion of ecological physics as “an exercise in phenomenology” was also instructive, although here I’m curious to hear other list members’ reactions. For me, the tension between gestalt psychology and phenomenology/enactivism has turned most heavily on whether one must maintain a Koffka-style distinction between behavioral (phenomenal) and geographic (physical) environment to explain perception or not. Gibson explicitly rejects the distinction, yet E&H take him to simply restate it in new terms. Given that Merleau-Ponty also seems to reject this distinction and (like Gibson) take the gestaltists to have missed the most radical implications of their own view, I wonder if these remarks by E&H cast those claims in a new light? Perhaps not, as E&H don’t so much argue for the distinction as assume it, and the justification for their reading of Gibson is not so much textual support, but the principle of charity — perhaps then they simply lack the imagination to read Gibson’s more radical claims at face value? I’m afraid I find myself sympathetic to E&H here.
A final thought concerns the morals for contemporary cognitive science. E&H stress that one needs “commensurate . . . conceptual vocabulary” for phenomenal and neurophysiological descriptions, since only then may one look for systematic correspondence between the two. But how strict a constraint is imposed by this mandate? While I agree with it, it’s not clear to me that it constrains cognitive science much at all. Wearing the hat of a traditional philosopher of mind, one might worry that terminology about the mental is inherently intentional, semantic, or otherwise normative, while typical neurophysiological terminology is not. Yet the understanding of mental terms like “belief” in these heavily semantic / normative ways is the result of philosophical analysis, not any implication inherent in the terms per se, nor forced upon the philosophically naive cognitive scientist.
My own view is that the real moral for contemporary cognitive science following from the gestalt tradition is not that special vocabulary must be chosen to ensure commensurability between phenomenological and physiological domains, but rather that whatever vocabulary is used to describe either domain, it must be considered strictly in terms of its structural features while asking how the two relate. I take it this is how Köhler attempts to naturalize value, by searching for a norm-neutral structural feature that characterizes it.
I found it very helpful to read this immediately after finishing Ben’s piece on MP, and in the middle of wondering about what exactly is at issue between MP and gestalt psychology. The summaries of GP positions in the paper suggested several points of contrast between MP and GP to me, though puzzlingly none of them seem to be quite the ones MP focuses on.
One potential tension concerns GP’s commitment to reductive explanation of forms in terms of physical/mathematical structures. It’s still not obvious to me that there’s too much to worry about here (for reasons I’ve tried to give in my comments on the other MP posts) – whilst it’s crucial for MP that formal structures *as they figure in our experience* can’t be exhaustively or faithfully characterised in objective vocabulary, what I’ve read so far hasn’t convinced me that GP would have anything to lose in agreeing with this. So, on p.169 when E&H say that GP ‘adopted the position of methodological naturalism in delimiting the domain of psychology’, it looks possible to me to accept this as a criterion for psychological explanation, but to reject it when it comes to Phenomenology (or philosophy of mind). Of course, this places restrictions on how we can understand the relation between philosophy and psychology. But I nonetheless think a satisfying story about their interrelations can be provided on this picture.
A more worrying source of tension, I think, arises from the specific way in which (on E&H’s reading) GP wishes to pursue its reductionism. According to the material summarised here, reduction will proceed by identifying internal structures that are isomorphic to the external structures to which the relevant cognitive episode is attuned. I think it’s crucial for MP that the forms/structures that constitute cognition are relational (at least in the foundational cases of perception and thought). Whilst, for the reasons hinted at above, I think MP should be fine with the possibility of providing physical/objective descriptions of forms/structures, all his paradigmatic examples will encompass both organismic and environmental parts.
I wondered whether the same considerations are at the bottom of the disagreement between Gibson and GP suggested on p.175. Is the reason that E&H take GP to be committed to rejecting Gibson’s ‘resonance’ view of perception that the latter view construes perception as fundamentally relational in a way that E&H’s reading of GP does not?
Two additional ways in which I think MP (for better or worse) might go beyond GP as characterised by E&H:
– On p.172 E&H suggest that GP’s claim that perceived meaning is more primitive than sensory qualities ‘parallels’ GP’s claim that whole-perception is more primitive than part-perception. For MP in the Phenomenology of Perception these claims are more intimately related than this might suggest: I read him as arguing that perceived meaning arises *due to* the gestalt (figure/ground) structure of perception and the way in which that structure is constituted and taken up by embodied organisms.
– E&H note that GP doesn’t have much to say about how perceived value or significance can arise as a result of learning, rather than as a result of innate sensibility. I wonder if MP has helpful things to add here, at least insofar as he commits himself to the claim that perception’s structural properties are grounded in the body schema, and that the body schema is both plastic and culturally malleable.