Wednesday, March 24, 2010

Some notes on Churchland's How do Neurons Know?

Yes, of course this is in conflict with not just the AFR, but with a lot of what we think we know about ourselves. And yep, I think the whole mess is self-referentially incoherent. Showing that it is, however, is more of a chore than you realize.

On the one hand, the Churchlands are committed to listening to brain science and seeing what those disciplines have to tell us about what our cognitive enterprise is like. Interestingly enough you have a lot their fello naturalists, like Jerry Fodor, who aren't looking to brain science to provide much info on what our cognitive life is like. The reason is that there seems to be too big of a disconnect between our mental life as we know it through introspection and what science is going to come up with. There are people who say, yes, in some sense it is all material, but you shouldn't be too eager to throw over huge elements of what Sellars called the Manifest Image (as opposed to the Scientific Image), because to do that is going to undercut what we know about our mental life. So I guess there's a kind of promissory note put out there that the disconnect will go away someday, but an unwillingness to turn neuroscience into philosophy.

The Churchlands are the deadly enemies of people like Fodor; they think we have to start taking neuroscience seriously now, even if it means that we start thinking that we are dead wrong about having things like propositional attitidues like belief. So they are classed as eliminative materialists, eliminativists in the sense that they say that we have to be prepared to throw over our common-sense understanding of our cognitive life. It is typical to couch their position as the claim that there are no beliefs, but actually what they suggest is that as science develops on cognition, we may have to be prepared to drop the notion of belief. We may have better terms to use once brain science gets where it will eventually go. The problem that they face is precisely what you mentioned: that when they talk about all of this they use the langauge of common sense psychology. But it's not as if they're going to hit their heads and say "I never thought of that" if you point out to them that they are using prescientific language. I mean, what other language would you suggest they use? We don't have the scientific image mapped. So they may say "if you think about it" but they may think that, if they knew enough future neuroscience, they wouldn't talk like that. They routinely hit self-referential objections with a charge of begging the question.

My contention is that the disconnect between naturalistically acceptable talk about cognition and the common-sense talk (I did the mathematical question, and concluded such-and-such) is a logical disconnect, and that any reconciliation is bound to be based on some sort of confusion of categories. For starters, surely neurons don't know, even if things composed of neurons do know. Talk about neurons making predictions seems puzzling as well. What does that mean? If I say, "I predict the Lakers will win the NBA Finals this year", that doesn't make sense unless you attribute to me all kinds of states that are naturalistically illegal.

In fact, I have identified four dimensions of our mental lives which seem to be automatically excluded from base-level naturalistic accounts: intentionality, normativity, subjectivity and purpose. The divide between something that these terms can apply to, and that which these terms cannot apply, doesn't work. And being told "Aha, look at all the nifty stuff science is discovering about the brain! Surely this great divide will someday be crossed if we keep going on brain science!" doesn't wash. It is as if a lot of nifty maps of one side of the Grand Canyon can tell us how to get across the Grand Canyon. Bridging attempts, such as functionalism, appear to work, in my view, because we take out eyes off the ball and miss the category mistakes.

But this has been battled out numerous times on Dangerous Idea 2. Haven't been over there much lately, because I haven't done a lot of AFR stuff of late. There are also some very nice BDK/Doctor Logic vs. Darek Barefoot debates that were done around late 2007, when I was writing the Blackwell Companion entry on the Argument from Reason.


Anonymous said...

I'd dispute that showing that it's incoherent is a chore. In fact, that part seems pretty easy. Getting people who disagree to admit it's incoherent? That's the hard part.

But getting a solipsist to admit solipsism is wrong can be a chore too.

Blue Devil Knight said...

Victor asks:
Talk about neurons making predictions seems puzzling as well. What does that mean?

There are algorithms from control theory that predict where a missile will go (for example). Neural networks can implement such algorithms.

Neuroscientists tend to use the word 'prediction' in this engineering sense.

For example, say we find a neural network in the brain, part of our visual tracking system that lets us move our eyes to follow an object (this is a lot like the missle tracking problem). This network implements one of these algorithms from control theory perfectly. That would be an instance of a population of neurons making a prediction.

For those interested in the science....

On the motor side of things, where this kind of stuff gets the most traction, this article reviews one popular framework. If you follow it up in a citation index you'll find a lot more details.

The eye movement system is probably the sin qua non of this view. Older paper looking at smooth pursuit eye movements, and prediction, here. This is one of the more important model systems for the study of voluntary motor control.

On the sensory side, you can check out a paper examining the hypothesis that even neurons in the retina predict the state of the world:
Meister. While Meister introduces some wrenches into the machine, they do describe the math and analysis of the hypothesis, this empirical hypothesis.

J said...

Aha, look at all the nifty stuff science is discovering about the brain! Surely this great divide will someday be crossed if we keep going on brain science!" doesn't wash.

One might say that shows an "idealist" bias of a sort; those who define the Mind as immaterial typically reject any scientific accounts which hint "Mind" ( qualia, intention, language, etc) depends upon or relates to, or is synonymous with brain neurology. In a sense the dualist metaphysician wants to say, no observable or empirical evidence will suffice--.

Consider some of the new "neural" gear which interfaces directly with the brain . Merely by thinking of, say, turning on the light a handicapped person with an interface implanted (neuro-kinetics, I believe....more material here: can...turn on the light. The synaptic firings are translated into code which then issues a command to a computer, etc.

One might argue there's a philosophical implication to that technology/research. A person's intended thought ("I want to turn on the light") has been realized, in the forms of an electronic command. E-compatibilism...

that said, I am not completely down with the Churchlands. They do tend to a certain reductionism (as did their vati Quine): even if mind's dependent upon brain and neurology, that doesn't thereby imply humans are rats, or a return to behaviorism. And they have a certain Toffleresque like futurist hype--sound nearly...Ayn Randian...

Edwardtbabinski said...


Where do "intentionality, normativity, subjectivity and purpose" go when you're unconscious?

Also, wouldn't you agree that there are different types of consciousness, lying along a spectrum.

What about distinguishing among different types of "intentionality, normativity, subjectivity and purpose?"

And in split-brain patients the silent half responds to questions in a rational fasthion, i.e., when both eyes are shown different images and asked different questions in different ears about each different image, the silent half of the split-brain responds by pointing to the answer to the question that it hears, while the other half of the brain responds either by pointing or speaking to the question it hears. This happens at the same time.

Sometimes the silent part of the brain can speak one word answers. And I think I read about one very unusual case in which the silent part of the brain could even respond using whole sentences.

And sometimes in split-brain patients the silent part of the brain displays a different "intention" than the talking half of the brain, such as one hand trying to pull pants down while the other hand is trying to pull them up while getting dressed for work, or one hand is closing a door when the other hand is trying to open it.

In fact, if you split the hemispheres of a child's brain and subjected one ear and one eye to nothing but one type of metaphysical belief system, and the other half of the child's split brain to another type of metaphysical belief system, provided such an experiment could be carried out, though I admit being in the same body would make it difficult to complete split the environmental input since even full body hugs from people are part of that body's environmental input, but say if the one metaphysical side was hugged only by Christians and the other only by atheists, something along the lines of the division in input could be acheived. And if the child only experienced one parent out of one eye and ear, the atheist parent, and only experience the other parent out of the other eye and ear, the Christian parent, then something along those lines would also be acheived, but the child's movements and line of vision and hearing would constantly have to be enforced, and the child's ability to coordinate it's own movements, left and right sides of body would be impaired in the long run.

But IF such an experiment were carried out, you might have an "atheist/christian" in the same body, would you not?

And if you split the brain very very early, both hemispheres might learn to speak in infancy and grow more equally specialized in other respects as well.

I say this because there are cases in which some children ARE born with half a brain or have to have half removed due to cancer, and some of them do survive, at least one case I've read about, and the brain half that was left did function in a less specializes manner than that lobe normally does when two halves are present.

What do you suppose it means when you can cut the two halves of the brain apart and have two different halves able to simultaneously answer two different questions?

In the experiment I mentioned the two halves might even argue with one another's metaphysical beliefs.

William said...

If multiple personality disorder really exists (and there are those who think it sometimes is an artifact of the evaluator) then we have plenty of examples of split consciousness without doing callosal section.

But does this really help us with the subjective/objective problem other than demonstrating that the brain and mind must be connected in multiple ways and physical locations?

Blue Devil Knight said...

I agree split brains multiply the problem of consciousness in some ways. However, if we have two conscious hemispheres in split brain patients (and I'm frankly not convinced this is the case), this would at least tell us what neuropsychological features are not necessary for consciousness (e.g., anything that is strongly localized to one hemisphere, such as language, would not be strictly necessary for consciousness).

While data from split brain patients leave open the problem of consciousness, that's not to say they support dualism, especially any more than ordinary consciousness supports dualism. It doesn't seem to help the dualist much: do we end up with two selves? You can literally get a split brain patient to fight with himself, have each hand trying to accomplish a task incompatible with the other (e.g., put the key on the table versus put the key in your pocket).

Victor Reppert said...

I think the question had to do with whether the split brain cases actually hurt the dualist.