Scientific Identities

The problems we've seen with behaviorism prompted philosophers in 1950s and 1960s to move to new kinds of materialist theories, which get called Causal Theories of Mind.

We will look at two Causal Theories of Mind:

  • the identity theory, which says that mental states are identical to states of the brain. This view goes by a variety of names. Some philosophers call it "Central State Materialism" (because it says that mental states are identical to states of your central nervous system). Others call it "type physicalism" (because it says that mental states are identical to certain types of physical state).

  • functionalism, which we will explain later

What these theories have in common is that, unlike behaviorism, they claim that our mental states as real inner states that cause our behavior. Behaving in certain ways is not all there is to being in pain; it just gives us evidence that someone is in pain, in the same way that smoke gives us evidence that there's a fire.

It's important to distinguish between two different kinds of claim that one thing is identical to something else:

conceptual analyses

Consider claims like these:
What it means to die is to have the biological processes of your body stop

What it means to be a square is to be a four-sided closed figure, all of whose sides are of equal length, and all of whose angles are right angles.

Anyone who grasps the concepts DEATH and SQUARE can tell whether or not these analyses are true, just by thinking and reasoning carefully. As philosophers put it, we can know a priori whether or not these analyses are true.

scientific identities

These are not knowable a priori, and they don't purport to tell us what the concepts in question mean. Here are some examples:

  • The temperature of a gas is the mean kinetic energy of its molecules
  • Water is H2O
  • Cows are animals with the genetic code GACCTAGCTA...
  • Lightning is a certain sort of electric discharge

As it turns out, water and H2O are one and the same chemical substance. This identity is a posteriori. We can only know that it's true by relying on our senses, and doing empirical research.

When the identity theorists say things like:

  • Being in pain is having your C-fibers fire

they are not trying to tell us what the concept PAIN means. One can understand the concept of PAIN perfectly well without knowing that pain and C-fiber firing are the same thing. Rather, the identity theorists are proposing a scientific identity. They are claiming that, just as science taught us that water is H2O, and that lightning is a certain sort of electric discharge, so too has it taught us that mental states are certain kinds of brain states.

At one point in his paper, Smart does offer an account of what some sensation-concepts, like "having an orange after-image," mean. He argues that the meaning of these concepts is topic-neutral. That is, it leaves it an open question whether the sensations in question are identical to processes in the brain, or to procesess taking place in Cartesian souls, or to some other sort of process. So, in Smart's view, the meaning of sensation-concepts leaves it open whether or not sensations are identical to brain-processes. It takes scientific investigation to show that sensations are in fact brain processes.

What Reason Do We Have to Believe the Identity Theory?

The natural sciences are having increasing success in explaining our behavior as the result of physio-chemical mechanisms. These scientific explanations of our behavior appeal only to the presence of certain processes in our brain (and certain neural and muscular connections between our brain and our limbs). In addition, we are getting better and better explanations of phenomena like perception, memory, and reasoning in terms of processes in our brain. So it looks like we have two choices:
  • Either we say that our mental states are perfectly correlated with certain brain processes (but still something over and above those brain processes)
  • Or we say that our mental states are identical to those brain processes
Smart argues that considerations of simplicity support the second account. (See the end of his paper, where he compares the choice between these two accounts to the choice between creationist and orthodox geology. The orthodox geology is more plausible because it is simpler. The creationist geology postulates too many brute and inexplicable facts. For just the same reasons, Smart thinks, we should prefer the Identity Theory over the view that our mental states are perfectly correlated with our brain processes.)

Leibniz's Law and its Limits

Leibniz's Law says that if X and Y are one and the same thing, then X and Y have to have all the same properties. So if X and Y have different properties, then they cannot be one and the same thing. Here is an example. We look out the window and see Superman flying by. We wonder who Superman is. Now as Superman flies by outside the window, we see our friend Jimmy Olson standing beside us. We can reason like this:
Superman is flying outside the window.
Jimmy Olson is not flying outside the window, he's standing right beside me.
So Superman has a property that Jimmy Olson doesn't have.
So Superman is not Jimmy Olson.
In general, this argumentative strategy is perfectly legitimate. However, we have to be careful. We can't apply it everywhere. For instance, we can't argue like this:
Lois believes that Superman is strong.
Lois does not believe that the mild-mannered reporter she works with is strong.
So Superman is not the mild-mannered reporter that Lois works with.
For Superman is the mild-mannered reporter that Lois works with. She just doesn't know that.

Let's introduce some technical apparatus that philosophers use in discussing these cases.

The extension of a term is the set of things which it picks out in the world, as the world actually is. So, for instance, the extension of the name "Jim Pryor" is a certain person. The extension of the description "the teacher of Phil 156 at Harvard in Spring 2002" is also a certain person, in fact the same person. So the name "Jim Pryor" and the description "the teacher of Phil 156 at Harvard in Spring 2002" have the same extension. If the world had gone differently, however, and Professor Siegel were teaching Phil 156 this term, then the description "the teacher of Phil 156 at Harvard in Spring 2002" would have a different extension. In that case, its extension would have included Professor Siegel, and not Jim Pryor. So although "Jim Pryor" and "the teacher of Phil 156 at Harvard in Spring 2002" have the same extension with respect to the way things actually are, they have different extensions with respect to other, counterfactual situations.

General terms like "cordate" also have extensions. The extension of "cordate" is the set of all the things which actually are cordates. (A cordate is a creature with a heart.) So, since you are a cordate, you are included in the extension of "cordate." In other possible situations, different cordates might existed, so with respect to those possible situations, the extension of "cordate" would be different.

Comment: It turns out that, in the actual world, "cordate" has the same extension as "renate." (A renate is a creature with a kidney.) Everything which is in fact a cordate is also a renate, and vice versa. But if evolution had gone differently, there might have been a species of creature which had hearts but no kidneys. These would be cordates, but not renates. So with respect to possible situations in which such creatures exist, "cordate" and "renate" have different extensions.

Some general terms have the same extension with respect to every possible situation. For instance, "closed figure with three sides" and "closed figure with three angles." Not only is it actually the case that all closed figures with three sides are also closed figures with three angles, there is no possible situation in which there is a closed figure with three sides which is not also a closed figure with three angles, and vice versa.

Next let's talk about different kinds of sentence fragments. A sentence fragment is a sentence with a hole in it. For instance:

____ is strong.
is a sentence fragment with a hole in the noun phrase. If we plug that hole with the name "Superman," we get the complete sentence:
Superman is strong.
A sentence fragment "___ is such-and-such" counts as extensional just in case filling in the blank with different terms having the same extension will always result in sentences with the same truth-value. (Such sentence fragments are also called "transparent.")
"___ is at least 30 years old" is an extensional sentence fragment. For example, substituting "Jim Pryor" and "the teacher of Phil 156 at Harvard in Spring 2002" into the fragment will result in sentences having the same truth-value.
A sentence fragment "___ is such-and-such" counts as intensional just in case it's not extensional. That is, just in case filling in the blank with different terms having the same extension can sometimes result in sentences with different truth-values. (Notice the spelling! It's "intensional" with an "s". Such sentence fragments are also called "opaque.")
For instance, "It is necessary that ___ is a human being" is an intensional sentence fragment, because substituting "Jim Pryor" and "the teacher of Phil 156 at Harvard in Spring 2002" into the fragment will result in sentences having different truth-values. (It is necessary that Jim Pryor is a human being. It is not necessary that the teacher of Phil 156 be a human being. We might have hired a robot, or an intelligent chimp, to teach the class.)

Likewise, "Sue believes that ___ is at least 30 years old" is also an intensional sentence fragment. For Sue might believe that Jim Pryor is 32 years old, but think that the teacher of Phil 156 is younger than 30. (She may falsely believe that someone else is teaching Phil 156.)

Examples like this last one show that attributions of beliefs are intensional. When you say "Lois believes that ___ is strong," putting different noun phrases in the hole can result in sentences with different truth-values, even if the words you put in the hole all have the same extension.

The same is true for attributions of any other representational or intentional state. Lois hopes that Superman will kiss her, but she does not hope that the mild-mannered reporter she works with will kiss her. She doubts whether Superman is in the office, but she does not doubt whether the reporter she works with is in the office. And so on.

Note: Notice the different spelling between "intentional" and "intensional." It's our mental states themselves that are intentional. That means they are about or represent things. It's the language we use to describe our mental states that's intensional. That means that its truth depends on more than just the words' extensions. It also depends on what the relevant subjects know about those extensions. These phenomena are related, but they're not exactly the same.

What is the upshot of all of this for Leibniz's Law? It means we can't use it in connection with claims about what people believe, or disbelieve, or doubt, or hope, or know is the case. For any sentence fragment like "I believe that ____ is strong" or "I hope that ____ will kiss me" or "I doubt whether ____ is here" will be intensional. So even when we do have two expressions "A" and "B" that refer to the same man, we can't be guaranteed that plugging "A" and "B" into the holes in these fragments will result in sentences with the same truth-value. We do get different truth-values when we plug in "Superman" and "the mild-mannered reporter that Lois works with," but that does not entitle us to conclude that Superman and the mild-mannered reporter are different people.

Leibniz's Law and Scientific Identities

Now, what is the upshot of all of that for the identity theory?

When we're talking about proposed scientific identities like "lightning is electrical discharge" or "pain is C-fiber firing," they will only be knowable by doing empirical investigation. One can possess the relevant concepts yet fail to know that these identities are true.

As a result of this, there may be some sentence fragments that come out true when you plug "lightning" in the hole, but false when you plug in "electrical discharge." And there may be some that come out true when you plug in "pain" but false when you plug in "C-fiber firing."

For example, we cannot prove that pain and C-fiber firing are distinct by this sort of argument:

I can tell that I have pain just by attending to how I feel.
I cannot tell that I have firing C-fibers just by attending to how I feel.
So pain is not identical to C-fiber firing.
That is not a legitimate use of Leibniz's Law. "I can tell that I have ____ just by attending to how I feel" is an intensional fragment. The mere fact that we get a different truth-value when we plug in "pain" than we get when we plug in "C-fiber firing" does not entitle us to conclude that pain and C-fiber firing are different states.

Smart thinks that such illegitimate appeals to Leibniz's Law lie behind several common objections to identity theory. Here are two of the ones he discusses:

Objection 1: Sensations and brain processes differ in epistemic respects. For instance, you can know lots about your sensations while being ignorant about your brain processes.

Smart replies: This does not support the conclusion that sensations and brain processes are distinct, any more than analogous reasoning supports the conclusion that lightning and electrical discharges are distinct. Whenever we're dealing with a posteriori scientific identities, the phenomena being identified will differ in epistemic respects.
Objection 7: I can imagine having sensations without having brain processes.
Smart replies: This does not show that sensations and brain processes are distinct, any more than the fact that I can imagine lightning being some non-electrical, purely optical phenomenon shows that lightning has to be distinct from electrical discharge.

 


URL:  http://www.courses.fas.harvard.edu
Last Modified: 
Copyright © The President and Fellows of Harvard College