Monday, March 20, 2006

Notes on Psychological Essentialism

The basic tenet of psychological essentialism is the idea that key human cognitive processes, those which determine how we approach experience, reflect a basic belief that unobservable essences are causally responsible for the surface features we observe. As such, the world is divided up into essences from which preset associated properties can be inferred.

The concept seems to have been introduced by Douglas L. Medin (Psychological Essentialism, 1989) during his investigation into similarity-based and explanation-based categorisation processes. Medin argued that categorisation is theory-driven, whereby concepts must conform to an overall world-view, rather than just a delineation of phenomena by lists of observable attributes. He also suggested that psychological essentialism can function with "placeholders" in that one can believe that a given category possesses an essence without knowing what the essence is, thus supporting inductive generalisation under ignorance of 'true causes'. Seen this way psychological essentialism can be understood as a reasoning heuristic which, according to several studies, e.g., by Susan A. Gelman (Essentialism in Everyday Thought, 2005), is readily available to, and used by, preschool children and adults alike.

H. Clark Barrett (On the Functional Origins of Essentialism, 2001) argues that, scientifically speaking, it is not enough to see psychological essentialism as simply a useful strategy and argues that speaking of a propensity for humans to 'essentialize' entails making proposals about where such a propensity comes from and, in particular, the cognitive mechanisms by which it is subserved. He sees a process of natural selection, operating on competing representational and inferential systems, as a rich area for further investigation.

Gelman (2005) argues that the propensity for essentialization is influenced by the language that children hear. For example, she has some evidence that, to a child learning language, nouns imply that a category is relatively more stable and consistent over time and contexts than adjectives or verb phrases - e.g. "carrot-eater" was judged by a group of 5-7 year olds as being a more stable property of someone than if they are merely said to "eat carrots when they can."

Johannes Keller (In Genes We Trust, 2005) argues that, as well as epistemic motives, there are existential and ideological motives behind psychological essentialism which provide potential links to increased levels of prejudice and stereotyping.

I will look into some of this in more detail and I have made these notes simply to pull out some of the threads of what appears to be a burgeoning area of investigation. Despite the demise of essentialist metaphysics, it seems to me that essentialism nevertheless dominates laypeoples' beliefs and attitudes. Reading the likes of Keller there seem to be grounds for correlating essentialist beliefs and attitudes with some of the negative aspects of the present state of the world. This I need to do a lot more work on. As previously noted, I am currently most interested in seeing how far these contemporary studies of the ubiquity of essentialism underpin the understanding of perception held by Madhyamikan Buddhism and, consequently, how far Madhyamikan Buddhism offers an appropriate approach to moving away from its dominance, should that be desirable.

To wrap up these notes, I find it interesting that the propensity to essentialize may be best conceived as a successful evolutionary expedient. This fits well with a pragmatist perspective and curtails the scope of metaphysics (in its "How It Really Is" guise) somewhat. Also, it seems as though psychological essentialism could account for the ubiquity of folk theories of causation which I have touched upon recently and, perhaps, could be what was behind the beginning of philosophy when it first went behind appearances to unobservable true causes.

5 comments:

Matt Kundert said...

Some forty years, I think Paul Feyerabend suggested that as long as we have common speech, we are going to have the idea of "mind", the one that he and Sellars, and later Dennett and Rorty, were trying to explain the existence of. This line of thought is the one that Dennett and Rorty have taken in basically arguing that "mind" or "consciousness" (as everything else we talk about) is created by playing a particular language-game.

I think the same thing about essentialism. I'm not sure there's necessarily anything wrong with common sense, at least in relation to essentialism, as long as people beginning in common sense don't get all "philosophical" (which is what, I think, Wittgenstein complained about, with the transposing of words and such from one context to another). In this relation, I'm thinking of the beginning of chapter four of Rorty's CIS, where he talks about ironists and metaphysicians.

I'm not sure what "psychological essentialists" are arguing for (surely not that are brains are hard-wired for essentialism?), but I can certainly agree that our common speech is suffused with it. As Nietzsche said, "I'm afraid we won't be rid of God as long as we still have faith in grammar." Can we get rid of it? Maybe. But it'd definitely be a long, slow cultural process as the counter-intuitive suggestions of philosophers and the like trickle down and become the common sense.

Paul Turner said...

I'm still getting to grips with psychological essentialism myself but at this stage I think that at least some of them seem to be saying that the propensity to essentialize has emerged as a successful cognitive representational system from an evolutionary point of view. Thus, common sense follows from this propensity, whether it be biological or social. The link to Buddhism I have made is that the Buddha also saw that humans have a propensity to essentialize and reify perceptual content.

So my question is - How do we keep the evident evolutionary gains of this propensity whilst progressing beyond its negative side?* In terms of philosophy and metaphysics I think we have seen how essentialism has gradually fallen away (although Rorty still seems to be waging a war against it(?)) but in terms of everyday reasoning and attitudes, it is still dominant, as far as I can tell. People still essentialize themselves, objects and, perhaps worse, other people, whole cultures, ethnicities, religions and so on - "Oh, he's an X, therefore Y, Z," etc.

I agree it will be a slow change, indeed Buddhism declares this a degenerate age!

* I'm really also trying to work out if it is a desirable change, which is, of course, my starting assumption.

Matt Kundert said...

Yeah, I think the philosophical climate still holds a lot of essentialism. As long as Rorty is seen as a radical, you know they're out there. It's evolving, to be sure. Essentialism has had to evolve since Parmenides to stay alive, and it gets tricker and tricker every time.

With whether this "psychological essentialism," as an evolved tool of survival, is biological or linguistic, it has to be linguistic. If one of these guys argues that it is a biological propensity, I think you're seeing a resurrection of the reductionist physicalism (kinda', I think, like what the Churchlands have). If one of these guys argues that it is linguistic, but that language _requires_ us to be essentialistic, then I think you're seeing a form of Chomskyan philosophy of language (which is also no good). The only position I would see as fair to our pragmatist instincts is the kind Dennett, Rorty, and Putnam elaborate. Putnam, I believe, used the analogy of "the brain is hardware, the mind/culture is software" years ago and Rorty wrote a recent article with that as his title that goes on to argue some of these points. And, you'll certainly notice, that Pirsig uses the same analogy (independently conceived, I assume) with his static levels.

I guess with common sense, dropping essentialism is a matter of becoming commonsensically nominalist and historicist (as Rorty suggests can happen in CIS). An effect, possibly, is that people will act and think more fallibilisticly and experimentally, as John Dewey hoped. We can't, for instance, drop saying "He's an X" without dropping the whole idea of getting a handle on people to interact with them. What we can (and should) drop is the idea that people are _essentially_ anything, at least in a way that runs philosophically, like for large groups and cultures. Like the idea of human nature. Or the idea that certain religions are naturally X (and always will be X).

What I'm hesitant to say we should drop is that, for instance, saying I'm essentially a know-nothing hack. Its like a handle on a person that everything should be interpreted around. I remember Rorty saying that, while Heidegger should be seen as a philosopher who happened to be a Nazi, Dewey should be seen as essentially a democratic socialist (who happened to write in philosophy journals on occasion). It's like putting emphasis on what is most important to that person (or to understanding that person), what part of their web of beliefs and desires is most important. I think that might be an okay example of common sense that we shouldn't get rid of. It's just human nature and all that junk we should get rid of.

Paul Turner said...

"What we can (and should) drop is the idea that people are _essentially_ anything, at least in a way that runs philosophically, like for large groups and cultures. Like the idea of human nature. Or the idea that certain religions are naturally X (and always will be X)."

Yes, that's what I think is a negative side of essentializing.

And I agree that the word 'essential' should not be dropped from common discourse.

Psybertron said...

Interesting. I can tell you a little anecdote about "essentialism" from an different perspective ... same point though.

When we were developing a generic business data model, at one point we had the concepts of "essential" and non-essential or "incidental" characteristics of certain classes of object. Some philosophical smart-arse pointed out that essentialism was flawed (and I agreed), but unfortunately we lost a useful (pragmatic) feature of our data model.