Functionalism came into prominence as a possible solution to the mind-body problem in the 1960s. Philosophers who have defended the theory have included Hilary Putnam, David Lewis and Jerry Fodor.
Like the identity theory, functionalism is a form of materialism and a form of mentalism. Still, it is importantly different than the identity theory…
To see exactly how it differs, we need to look both at its negative thesis and its positive thesis.
[4.5.1.] Functionalism’s Negative Thesis: We Should Reject Type-Type Identity.
As an approach to understanding this thesis, we must first understand what is called the type/token distinction…
[126.96.36.199.] The Type/Token Distinction.
tokens: specific individual things
types: kinds of thing
· the shirt that I am wearing
· a specific sample of water
· my big toe
· my copy of Core Questions in Philosophy
· shirts (in general)
· water (in general)
· big toes (in general)
· Core Questions in Philosophy (considered independently of any specific instance of it)
This distinction is frequently applied to, and illustrated with examples from, language… For example, how many words are in the following list?:
The answer depends on whether you mean word tokens or word types. There are three word tokens in the list but only two word types. The first two items in the list are separate tokens of the same type.
This distinction can be applied to mental and neurological types/tokens as well...
· my belief that Atlanta is the capital of Georgia
· the pain that I now feel in my throat
· the firing of my C-fibers
· the belief that Atlanta is the capital of Georgia (considered apart from any particular person having it)
· throat pain (in general)
· C-fiber firing (in general)
[188.8.131.52.] Token Identity vs. Type Identity.
The Identity Theory accepts both of the following:
the token identity theory (df.): each token mental state is one and the same thing as some token physical state; e.g., the pain I’m feeling right now (a token) is identical to the current firing of my C-fibers (a token).
the type identity theory (df.): each type of mental state is one and the same thing as some type of physical state; e.g., pain in general (a type) is identical to the firing of C-fibers in general (a type).
Notice that the type identity theory implies the token identity theory, but not vice versa; for example:
· Suppose that the types pain and C-fiber firing are identical (this is what the type identity theory maintains)…
· This implies that every token of pain is identical to some token of C-fiber firing, and vice versa... you can’t have pain without C-fiber firing, and you can’t have C-fiber firing without pain.
To see this more clearly, compare it to this analogous case:
· The types water and H2O are identical…
· This implies that every token of water is identical to some token of H2, and vice versa; you can’t have water without H2O, and you can’t have H2O without water.
Functionalism accepts the token identity theory – so on this, functionalists and identity theorists disagree.
[other types yet to be discovered by the sciences]
[184.108.40.206.] The Mousetrap Analogy.
Every individual mousetrap (every token of the type mousetrap) is a physical object… but beyond that, there is no specific type of physical object that each mousetrap has to be.
These three token mental states are all of the same mental type—belief that 2+2=4—but are all of different physical state types.
According to functionalism, just as there is more than one physical way to build a mousetrap, there is also more than one physical way to build a creature with mental states: mental states have many different possible physical realizations. In other words, they have multiple realizability.
If this is right, type identity theory is false—and so is the Identity Theory.
The above example (involving aliens and robots) is a bit dramatic. Here’s a more realistic example:
These are three examples of the same mental type (pain) but not of the same physical type. But if the identity theory were true, then in order for humans and rabbits and dogs to experience the same type of pain, we would have to have exactly the same type of mental state. In this way, the type identity thesis, and therefore the Identity Theory, is “chauvinist” (in the words of Ned Block; see Sober p.289).
To sum up…
· Functionalism rejects the type identity theory (“type-type identity”), because that thesis implies that other creatures must be like us physically if they are to be like us mentally.
· But it accepts the token identity thesis (“token-token identity”).
[4.5.2.] Functionalism’s Positive Thesis.
According to functionalism, what makes a token mental state the mental state that it is, is not its physical characteristics (this is because functionalism rejects type-type identity).
Rather, what makes a token mental state a token of a specific mental type is the causal relationships in which that token takes part.
The defining feature of any type of mental state is not physical. Rather, it is the set of causal relations it bears to three things:
(i) environmental effects on the body
(ii) other types of mental states
(iii) bodily behavior
other token mental states ®
environmental stimuli ®
® other token mental states
® behavioral effects
injury to the body ®
® distress, annoyance
® vocalization (e.g. saying “ouch”)
® nursing of injured part of the body
Any token that stands in those causal relations with those specific stimuli, other mental states, and behavior is a pain. It does not matter what physical material that state occurs in.
Compare the case of a mousetrap:
construction by some human being ®
® previously free mouse becomes trapped
Since all three tokens of pain (Jim the human’s, Bongo the rabbit’s and Murphy the dog’s) can stand in those causal relations with stimuli, other mental states, and behavior, they can all be the same type of mental state—the three different tokens can all be of the type pain, despite the fact that they are of three different physical types.
[4.5.3.] The Inverted Color Spectrum
Here is a potential problem for functionalism: It seems possible for two different individuals, A and B, to have very different subjective experiences of color—that is, to have different color qualia when seeing the exact same object.
· A looks at a tomato and experiences what you and I experience as color red; B looks at it and experiences what you and I experience as green.
· Where A looks at the sky and sees what you and I see as blue, B sees looks at the sky and sees what you and I see as yellow; and so forth.
It’s not that B is color-blind; it’s simply that their color-experiences of the world are completely reversed.
Of course, there is no way for A and B to compare their qualia to see whether they really are different.
Now consider the following:
· the stimuli that causes their different color sensations would be the same... a tomato would reflect light of exactly the same wave-length at both A and B.
· their different experiences would cause would result in the same beliefs about objects and their colors; they would both believe that tomatoes are red, that the sky is blue, etc.
· there would be no behavioral indications that their qualia are different, since they can make all the same observational discriminations. When asked what color the tomato is, they will both say “red”; when asked to point out a red object, each would point to the tomato. So their behavior with regard to colors would be exactly the same.
The possibility of an inverted spectrum is a problem for functionalism because A’s color sensation when viewing a tomato would be of a different type than the sensation B has when viewing the same tomato, despite the fact that those mental states play the same causal roles!
Functionalism says that if two token mental states play the same causal roles, they are of the same mental type. But if an inverted spectrum situation is possible, then it is possible for two token mental states that play the exact same causal roles to be of two different mental types: A’s sensation while viewing the tomato is of the type sensation-of-red, and B’s sensation while viewing the tomato is of the type sensation-of-green.
So if an inverted color spectrum is possible, then functionalism has to be false.
[4.6.] Eliminative Materialism.
A fifth proposed “solution” to the mind-body problem, one not discussed in your textbook, is
eliminative materialism (df.): “the radical claim that our ordinary, common-sense understanding of the mind is deeply wrong and that some or all of the mental states posited by common-sense do not actually exist.”
· Philosophers who have defended this view include Paul and Patricia Churchland and (at one time, but no longer) Stephen Stich.
According to functionalism, type-type identity is false because types of mental state are multiply realizable: tokens of one mental type can be realized in different types of physical state.
According to eliminative materialism (EM), type-type identity is false because the mental state types that folk psychology talks about are not real. In other words, EM maintains that folk psychology is largely or entirely mistaken:
folk psychology (df.): the common-sense conceptual framework in which ordinary people (not just psychologists) explain and predict behavior; it includes concepts like belief, desire, joy, fear, pain, sensation, intelligence, etc.
According to EM, folk psychology is not simply incomplete, or partially incorrect. It is “a false and misleading conception of the causes of human behavior and the nature of cognitive activity.” As neuroscience progresses and we discover more and more about the real, neurological basis of behavior, folk psychology will eventually be eliminated.
· caloric: 18th and 19th century scientists believed in this substance, which they thought was a fluid held in bodies (like water in a sponge) that made those bodies hot; when they believed in caloric, they put forward a lot of theories about how is produced thermal expansion, melting and boiling.
· phlogiston: a spirit-like substance in wood and iron given off when they burn or rust—scientists no longer believe there is such a substance
· “the starry sphere of the heavens”: scientists at one time believed that the earth was at the center of the universe, which itself was a hollow sphere; they wondered whether it was made of crystal or something else.
These were never identified with anything else…
· Caloric was not identified with kinetic energy or anything else;
· Phlogiston wasn’t identified with anything; and
· The starry sphere wasn’t identified with anything.
According to EM, beliefs, desires, etc. are no more real than caloric, phlogiston, or the starry sphere, and eventually we will stop talking about them.
[6.] Freedom, Determinism and Causality.
[6.1] Two Conceptions of Human Nature.
Consider these two conceptions of human beings. Each of them seems plausible when considered on its own, but they seem to clash with one another:
“The Ordinary Conception”
Humans have free will; i.e., at least some of the time, we perform actions that are free. This involves two conditions:
(FW-1) Some of our actions are such that whether we perform them is under our control.
(FW-2) Some actions that we do perform could have been different, even given no difference in preceding conditions.
It seems like both (FW-1) and (FW-2) have to be true in order for humans to have free will. So in order to show that we do not have free will, it is sufficient to show either that (FW-1) is false or that (FW-2) is false.
“The Causal Conception”
Humans are part of a causal network. Our mental states, including our beliefs and desires, are the effects of a complex series of causes, including things in our environment and our individual genetic backgrounds. And those mental states are the causes of our behavior—we act the way we do because we have specific beliefs and desires that cause us to act that way. [Sober’s diagram, p.293:]
beliefs and desires
[6.2.] Arguments Against Free Will.
The clash between these two competing conceptions of human beings is revealed by the arguments that are used against (FW-1) and (FW-2):
The Distant Causation Argument (against FW-1)
1. All of a person’s actions are caused by her beliefs and desires.
2. All of a person’s beliefs and desires are caused by her environment and genetic make-up.
3. One’s environment and genetic make-up are not under her control.
4. Therefore, none of the beliefs and desires a person has is under her control.
5. Therefore, for any action a person performs, whether she performs it is not under her control; i.e., (FW-1) is false.
The “Could-Not-Have-Done-Otherwise” Argument (against FW-2)
1. All of a person’s actions are caused by her beliefs and desires.
2. If an action is caused by beliefs and desires, then, given those beliefs and desires, it is not possible that the action not occur.
3. Therefore, none of the actions that a person performs could have been different, given the beliefs and desires she actually had; i.e., (FW-2) is false.
It would not be a trivial matter simply to accept that one of these arguments is sound and that in fact humans lack free will. This is because, at least on a very plausible view of the matter, having free will is a necessary condition of moral responsibility, and therefore of deserving moral praise or moral blame for what one has done. It would have other very gloomy consequences, as well:
Most philosophers suppose that the concept of free will is very closely connected to the concept of moral responsibility. Acting with free will, on such views, is just to satisfy the metaphysical requirement on being responsible for one's action. … But the significance of free will is not exhausted by its connection to moral responsibility. Free will also appears to be a condition on desert for one’s accomplishments (why sustained effort and creative work are praiseworthy); on the autonomy and dignity of persons; and on the value we accord to love and friendship.
The second premise of the CNHDO argument (“If an action is caused by beliefs and desires, then, given those beliefs and desires, it is not possible that the action not occur.”) is related to an idea called determinism:
determinism (df.): the doctrine that for any point in time, given the way things are at that point in time, there is only one possible way that things could be at the very next point in time.
Determinism itself depends on the idea of causation. The idea itself is very widely used, but it is not very clear. Let’s try to shed a little light on the idea of causation by focusing on a specific example: Suppose that I strike a match, thereby causing it to ignite.
· Striking the match is not a necessary condition of the match igniting. There are other things which could result in the ignition of the match (e.g., a magnifying glass focusing strong sunlight on the match). So causes are often not necessary conditions of their effects; x can cause y without being a necessary condition of y.
· Striking the match is not a sufficient condition of the match igniting. Other conditions must be met (e.g., there must be oxygen present; the match head must be dry, not damp).
So a cause of an effect need not be either a necessary or a sufficient condition of that effect.
A further idea about causation: we need to distinguish between a cause of an effect and the whole cause of that same effect. My striking the match is a cause of its igniting, but not the whole cause (the entire set of causally relevant facts); when I strike the match, the whole cause of the match’s ignition includes my striking it, the presence of oxygen, the head being dry, and other facts relevant to the match igniting.
Now that we are a bit clearer on causation, let’s look at determinism itself…
Consider once again the match example. If determinism is true, then, given that all of those causally relevant facts are the way they are (i.e., given the whole cause), there is only one possible way that things can turn out, viz. the match ignites. Those facts don’t just make it probable that it will light; they guarantee that it will light.
According to determinism, every single physical event is like this: given all the causally relevant facts at one point in time, there is one and only one way that things can be at the very next point in time.
[6.3.2.] Determinism and Materialism.
Why think that premise 2 of the CNHDO argument (“If an action is caused by beliefs and desires, then, given those beliefs and desires, it is not possible that the action not occur”) is true? If you accept determinism and materialism, then it seems like you have to accept premise 2 as well.
If materialism is true, then there is nothing more to human beings than the elementary particles that make up everything else in the universe—we are made of matter, just like everything else.
If determinism is true for all physical systems, and if every human being is nothing but a physical system, then determinism must be true for human beings: all the processes in our bodies (and minds) are deterministic.
determinism is true (and the physical state of the universe at one point in time uniquely determines the physical state of the universe at the very next point in time)
materialism is true (and human beings, including their minds, are completely physical),
all human actions are physically determined and could not have been otherwise: i.e., premise 2 is true.
[6.3.3.] Is Determinism True?
Many physicists today believe that determinism is false, at least for actions that occur at the sub-atomic level.
According to one popular interpretation of Quantum Theory, the behavior of sub-atomic particles is non-deterministic. At the quantum level, the state of a physical system at one point in time does not guarantee what will happen at the very next point in time. It can make certain causal effects more probable, but it cannot guarantee them.
Let’s assume for now that these physicists are correct and that determinism is false about the sub-atomic world.
By itself, quantum indeterminacy does not show that premise 2 of the CNHDO Argument (“If an action is caused by beliefs and desires, then, given those beliefs and desires, it is not possible that the action not occur” is false. It is possible that determinism is false but that certain portions of the physical world (including human beliefs, desires, and behavior) are nonetheless deterministic.
So let’s make a further assumption: assume that quantum indeterminacy “percolates up” and affects the relationship between mental states and behavior, so that mental states do not completely determine the actions a person will perform.
If we make this assumption, then premise 2 is false and the CNHDO Argument fails.
But now consider what things are like if quantum indeterminacy does in fact percolate up: our mental states (including our beliefs and desires) do not totally determine how we behave. At least some of our actions could have been otherwise, but this is so because there is an element of chance where the causes of behavior are concerned. [See the “roulette wheel brain surgery” story on p.300 of Sober]
Indeterminism no more leads to free will than determinism did – in particular, if your behavior is partly the result of chance, then you still have no control over your behavior. Indeterminism might save (FW-2) that some actions that we do perform could have been different, even given no difference in preceding conditions), but it doesn’t seem to be able to save (FW-1) (that some of our actions are such that whether we perform them is under our control). Consider these two options:
quantum indeterminacy doesn’t
quantum indeterminacy does
S’s behavior is completely determined by prior conditions; S has no control over that behavior and could not have done otherwise
(FW-1) and (FW-2) are both false
S’s behavior is not completely determined by prior conditions; S could have done otherwise, but still has no control over her own behavior
(FW-2) is true, but (FW-1) is false
Under the ordinary conception of free will, there is a condition of free will that does not obtain, whether or not quantum indeterminacy percolates up. So it seems that IF we accept the ordinary conception of free will, WE DO NOT HAVE FREE WILL.
Stopping point for Friday June 25. For next time read Sober ch.25 and ch.26. This is the last reading assignment of the course.
 For much more information on functionalism, see Janet Levin, “Functionalism,” The Stanford Encyclopedia of Philosophy (Summer 2010 Edition), Edward N. Zalta (ed.), forthcoming URL = <http://plato.stanford.edu/archives/sum2010/entries/functionalism/>.
 William Ramsey, “Eliminative Materialism,” The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/fall2008/entries/materialism-eliminative/>.
 Churchland, p.43.
 Timothy O'Connor, “Free Will,” The Stanford Encyclopedia of Philosophy (Fall 2008 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/fall2008/entries/freewill/>.
 Sober defines determinism as: “the thesis that a complete description of the causal facts at one time uniquely determines what happens next” (531). Determinism is frequently defined, not in terms of descriptions of causal facts, but in terms of the causal facts themselves; Sober himself defines it that way later in the chapter: “the thesis that the facts at one time uniquely determine what comes next.” (305) For much more information about determinism see Carl Hoefer, “Causal Determinism,” The Stanford Encyclopedia of Philosophy (Spring 2010 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/spr2010/entries/determinism-causal/>.
This page last updated 6/25/2010.
Copyright © 2010 Robert Lane. All rights reserved.