Can a materialist conception of the mind really preserve free will?

[polytropos]

This post is directed at varieties of materialist philosophy of mind which hold that all mental activity in some way “supervenes” upon neural activity. What I am essentially arguing is that there is a lite interaction problem for materialists, the resolution of which necessitates the denial of free will.

Suppose I am sitting at my desk. I am thinking about a math problem. Eventually, I come up with the solution and reach for my pencil to write it down. Such an action seems voluntary. On a materialist account, can we maintain that it actually was voluntary?

I submit that we cannot.

We can begin our analysis at my action of reaching for the pencil, which is the simple part. My muscle movements were caused by the release of ions into my muscle cells, which was in turn caused by a series of action potentials stretching from my brain to my arm. What we are concerned with is what occurred in my brain.

I will define conceptual nexus to be the mental activity supervenient upon my brain. In other terms, as I am working through the conceptual content of the math problem, there is neural activity in my brain. This neural activity in some way represents the logical inferences that I am “making” in order to solve the problem.

What I take to be the materialist interaction problem is this: the conceptual nexus essentially constitutes our relevant conscious experiences. If the conceptual nexus “emerges” from the activity of the brain, then it is still wholly explained by the activity of the brain, and changes in the conceptual nexus must be accounted for in terms of specific neural activity. For instance, I decide that my current approach to the math problem is not promising, so I begin to think about another way of solving it. This seems conscious and voluntary, but must also be represented by neural activity. What seems to be the issue is that since the neural activity is primary, the conceptual nexus - ie. the conscious, voluntary experience - is not “acting” on the neural structure of the brain, which would seem to be necessary for genuine free will.

By materialist hypothesis, my mental activity is also fully represented by reductively neural activity. If physiological events require prior physiological events, then this would seem to remove free will from the materialist picture. The activity among my neurons must have been guided by the laws of physics; each event, by hypothesis, has an antecedent physiological cause, as opposed to an antecedent conceptual cause, since the conceptual nexus only supervenes on my neural activity and must be fully represented by neural activity. The logical inferences I make while solving my math problem, then must be reductively identified with some causal neural chain, however complex. The conceptual nexus does not cause my reasoning through the problem since it is conceptual and formally inert; the conceptual nexus represents my reasoning. The fact that the neural activity in some way represents mathematical logic cannot imply that I am actually making conceptual inferences, just that my conceptual inferences reflect reductively neural activity. We are left with an illusion of freedom, certainly, but not freedom. While the neural activity apparently leads me to believe that I freely chose to switch from one approach toward the problem to another, this was really the result of un-free neural activity.

Can the materialist evade this consequence by arguing that reality is indeterministic? It does not seem so. We still have a conceptual nexus supervenient upon particular brain states. Even with recourse to indeterminism, we are still left with the need for the logical inferences of the conceptual nexus to be represented by reductively neural activity. An appeal to stochastic laws could not save this picture. Even if we were to say that the decisions we make in solving a problem are not determined solely by prior physiological activity, indeterminism would just allow us to insist that the decisions are at the reductive level random, and this does not seem in any way to vindicate free will; rather than the solution deterministically emerging from a complex series of interactions between my neurons, some of the interactions were random, and the solution emerged anyway. This only seems to add a note of implausibility to the picture; that the “free” logical inferences I make in the course of solving a problem are in fact random does not seem to make them genuinely free.

As such, that mental activity is just supervenient upon neural activity seems to leave no room for free will.

If anyone disagrees with this characterization of materialist philosophy of mind, please say so. However, as I said, I do only mean for this to apply to varieties of materialism which hold that mental activity is supervenient upon neural activity; there are many varieties of materialism, and I do not expect this to apply to all of them (although more eliminative versions might deny free will outright anyway).

 

[Peter_Plato]

What has puzzled me is how consciousness enters into the picture. Perhaps it is simply an aspect of the "conceptual nexus” analogous to light from a candle? However, if intentional states are all reducible to neuronal activity then what would be the point of conscious awareness, or even the conceptual nexus itself? It would seem to be a waste of biological resources and much more efficient to simply bypass the conceptual nexus (and consciousness) altogether and have the response be automatic.

What comes to mind here is how elite athletes practice movements and responses until these become automatic and bypass conscious considerations altogether because efficiency is much higher without the need to pause and “think” before responding. It would seem to me that if evolution was strictly the means by which our brains developed consciousness would have been disposed of by natural selection as an impediment to survival.

I suspect a strong case can be made that consciousness itself is evidence that biological processes would not have brought about the human mind - at least not its capacity for self-awareness - strictly on evolutionary principles.

From there it might be demonstrated that free will is dependent upon consciousness since free will presupposes the “transcendence” of the causal order enabled by consciousness in order to foresee options and make choices from them. This ability to originate novel causal sequences would seem to depend upon rising above the causal chain of events that consciousness of them makes possible. Consciousness would therefore seem to be the mechanism by which free will is enabled.

Free will would seem a useless artifact or illusional to any eliminative materialist explanation because causation is strictly incompatible with free will, as you show.

Why would free will be important? It seems to me, anyway, that individual subjective identity as a feature distinct from the causal order would depend - no matter how illusory - upon the possibility of originating novel causality (free will.) In what sense could we said to have an identity (conscious awareness of self) without having the capacity to distinguish between what “I” do and what simply occurs (free will?)

 

[Rete]

I think a materialist who defends free will could get around your entire argument by arguing free will is nothing more than nondeterminism, and that nondeterminism is an inherent property of the universe.

How a neuron fires does have a great deal of determinism involved, but how the synaptic weights are adjusted depends quite a lot on experience and environment, which is nondeterministic for the most part.

His deciding to choose one action over another is no more remarkable than an electronic being observed in a particular orbit with a particular spin and not some other possible configuration.

 

[polytropos]

What has puzzled me is how consciousness enters into the picture. Perhaps it is simply an aspect of the "conceptual nexus” analogous to light from a candle? However, if intentional states are all reducible to neuronal activity then what would be the point of conscious awareness, or even the conceptual nexus itself? It would seem to be a waste of biological resources and much more efficient to simply bypass the conceptual nexus (and consciousness) altogether and have the response be automatic.

I agree that consciousness also poses issues for a materialist conception of mind. Here I have been assuming it as though it were not an issue.

The problem would seem to be how exactly consciousness supervenes on a pattern of neural activity. Why should a series of actions - however complex - lead to emergent consciousness, and if all mental activity is supervenient on the brain, then how can fundamentally subjective experiences be reduced to firing patterns?

It is interesting to think of it in terms of biological resources and efficiency. We would ordinarily think of there being some required energy to, say, project a movie onto a screen. But if the argument is that the projection of the movie onto the screen is wholly reducible to the effect of photons exciting certain electrons etc., and the qualitative character of the projection is just “in the head,” then we can’t exactly approach the issue of our qualitative experiences in the same way, since there is no longer any subject for the purportedly subjective experience.

It should be noted that evolution by natural selection is not perfect - it would be better if humans did not, for instance, have tail bones, and the reason we still have them is that they are vestigial. The same might be argued for consciousness - it evolved because it was contingently useful, even though to an “ideal” organism it would be extraneous. But that would just be a historical explanation of consciousness, which is impossible without a constitutive explanation of consciousness. We can only make inferences about how consciousness evolved if we have some idea of how it does work.

 

[polytropos]

I think a materialist who defends free will could get around your entire argument by arguing free will is nothing more than nondeterminism, and that nondeterminism is an inherent property of the universe.
…
His deciding to choose one action over another is no more remarkable than an electronic being observed in a particular orbit with a particular spin and not some other possible configuration.

The materialist could take this route, but IMO this just trivializes free will. If free will just is indeterminism, then free will is, as you say, “no more remarkable than an electron being observed in a particular orbit with a particular spin and not some other possible configuration.” ie. that is to say that humans are as “free” as the rest of the unconscious, purportedly non-mental world.

 

[Rete]

Except now you have to explain how free will differs substantially from nondeterminism. For the purpose of the materialist, nondeterminism works exactly like free will.

 

[Bagheera]

Observation #1: We do not know if there is a “free will”. It is merely a plausible assumption.

Observation #2: There are several definitions of “free will”. Which one do you use? Without giving your definition, the whole discussion “hangs in the air”.

Observation #3: The reductionist view – to reduce the thinking process to the interaction of the neurons does not take into consideration that the same “firings of the neurons” can be associated with different thoughts. There is no one-to-one correspondence between a certain thought and specific neural activity. It is much more complicated than that. Consider the computer analogy (which is not really precise), where the “program” is reflected by the on-off switches of the hardware. By merely observing the state of the hardware you cannot decipher what the program does. In other words, not even a simple computer program can be reduced to the on-off states of the hardware. The computer hardware is not a “closed” architecture, even though it is designed to be one. The original “bug” was an actual bug that short-circuited some wires.

The activity of the brain is even more complicated. The connections between the neurons is dynamic, not static. What the electro-chemical activities represent cannot be reduced to those activities. And the brain is definitely not a closed architecture, it receives a current of outside stimuli.

To ponder if we have “really free will” is a fruitless endeavor. We act as if we had one, and use it in our legal system.

 

[polytropos]

Observation #1: We do not know if there is a “free will”. It is merely a plausible assumption.

Right, my argument was not:

1. We have free will.
2. Materialist conceptions of mind commit us to denying free will.
3. Therefore, materialist conceptions of mind are false.

The argument was that materialist conceptions of mind are incompatible with free will.

Observation #2: There are several definitions of “free will”. Which one do you use? Without giving your definition, the whole discussion “hangs in the air”.

I am not committing to a definition of free will. Implied in my OP was the idea that, a necessary (though perhaps not sufficient) condition for free will would be that the conceptual content of the mental which supervenes on the brain would have to have causal relevance to efferent nervous activity. As I noted in my response to Rete, I believe that any lower bar for free will would be trivial.

As I’ve noted also, if one were to claim that indeterminism were a sufficient condition for free will, then obviously one could hold that we have free will. One would just have to trivially hold as well that the rest of reality has free will also.

Observation #3: The reductionist view – to reduce the thinking process to the interaction of the neurons does not take into consideration that the same “firings of the neurons” can be associated with different thoughts. There is no one-to-one correspondence between a certain thought and specific neural activity. It is much more complicated than that. Consider the computer analogy (which is not really precise), where the “program” is reflected by the on-off switches of the hardware. By merely observing the state of the hardware you cannot decipher what the program does. In other words, not even a simple computer program can be reduced to the on-off states of the hardware. The computer hardware is not a “closed” architecture, even though it is designed to be one. The original “bug” was an actual bug that short-circuited some wires.

Can you explain this objection in more detail? Say I write a program in a higher-level programming language and compile it into a binary string. The computer can run the program as a string of 0s and 1s.

Is this not reducing the activity of the computer to the lowest possible level of languages?

Not to mention, the computer, in doing so, does not a). act freely, b). have conscious experience, or c). “know” what it is doing.

What you have perhaps stumbled upon is an indeterminacy (not to be confused with indeterminism) argument for the immateriality of thought. The computer architecture does not have determinate meaning, ie. it “interprets” a program only metaphorically, and its physical (name removed by moderator)uts and outputs do not have an intrinsic meaning; this is true of all of physical reality. By contrast, our thoughts do have determinate meaning, and as such cannot be physical. Related to this, if we are to deny that our thoughts have determinate meaning, then we cannot formally object to such a proposal.

The activity of the brain is even more complicated. The connections between the neurons is dynamic, not static. What the electro-chemical activities represent cannot be reduced to those activities. And the brain is definitely not a closed architecture, it receives a current of outside stimuli.

It is not clear how non-static connections between neurons or outside stimuli are relevant to the point I’m making. I do not think one could commit to the view that outside stimuli or dynamic neural connections are necessarily associated with any purportedly “free” action (or how the consideration of them would uniquely allow brain activity to be free). One could, I suppose, take the view that when I sit down to do a math problem, even if (as is conceivable) I can temporarily shut myself out from all external stimuli, all of my neural activity is part of a causal chain stretching back to when I first sat down. But that would almost certainly abolish free will.

 

[polytropos]

Except now you have to explain how free will differs substantially from nondeterminism. For the purpose of the materialist, nondeterminism works exactly like free will.

I think the ability to act randomly is generally taken to be the lowest condition for free will. But such a conception, I think, faces issues. Suppose I build a robot that looks just like a human, and it “makes decisions” completely randomly. Even though its decisions might appear free, it is conducting them according to a purely random algorithm (granting that genuine randomness does not exactly exist in computing). This hardly seems to imply free will.

As I indicated in my response to Bagheera, in materialist terms I believe a necessary but perhaps not sufficient condition for free will would be for the conceptual content which supervenes on the brain to have a genuine causal relation to efferent nervous activity.

 

[tonyrey]

Free will is a fantasy if the self doesn’t exist. Hume’s view that we are “bundles of perceptions” is similar to materialist’s reduction of the mind to brain activity. In both cases persons are presumed to be responsible for their actions even though they are thought to be no more than legal fictions!

 

[Rete]

I think the ability to act randomly is generally taken to be the lowest condition for free will. But such a conception, I think, faces issues. Suppose I build a robot that looks just like a human, and it “makes decisions” completely randomly. Even though its decisions might appear free, it is conducting them according to a purely random algorithm (granting that genuine randomness does not exactly exist in computing). This hardly seems to imply free will.

As I indicated in my response to Bagheera, in materialist terms I believe a necessary but perhaps not sufficient condition for free will would be for the conceptual content which supervenes on the brain to have a genuine causal relation to efferent nervous activity.

It sounds like you are arguing that free will requires some kind of Platonistic metaphysics. That’s not really conducive either.

I think the problem you have here (in convincing a materialist) is that you cannot easily distinguish free will from nondeterminism within the system of philosophical materialism. You can talk about metaphysics all day, but he doesn’t care. Your free will and his nondeterminism are the same thing in a materialist universe.

You also have to overcome the problem from his point of view of why you need all of this other baggage to show nondeterminism. The simplest explanation to him is nondeterminism, which explains his experiences and observations perfectly well, and seemingly more simply than what you are trying to describe.

 

[SentinelofTruth]

Congratulations, polytropos . You have discovered the argument from free will, my favorite refutation of materialism.
It is indeed a refutation and not just a reductio ad absurdum, because free will is actually provable. Consider this, skeptics of the argument:

• 
  *]It follows from the definition of “conscious” that an conscious event cannot occur within one’s mind without one’s knowledge, and that one cannot experience a conscious event in one’s mind unless this event is in fact occuring.
  *]Clearly, free will is a conscious thing, whether or not it exists.
  *]Therefore, we can learn infallibly from our introspection whether we have free will.
  *]Introspection supports free will.
  *]Therefore it exists.
  
  polytropos is further correct in saying the objection of indeterminism does not help. Free will is not random. We can see this by asking the indeterminist materialism whether he believes electrons get to freely choose their path around the nucleus. His answer will be no. Therefore, randomness =/= free will.

 

[polytropos]

It sounds like you are arguing that free will requires some kind of Platonistic metaphysics. That’s not really conducive either.

Can you clarify where I invoke Platonistic metaphysics?

I think the problem you have here (in convincing a materialist) is that you cannot easily distinguish free will from nondeterminism within the system of philosophical materialism. You can talk about metaphysics all day, but he doesn’t care. Your free will and his nondeterminism are the same thing in a materialist universe.

Well, there are many kinds of materialists. This post is aimed at those who believe that we can still speak of the mind as “existing” in the respect that it supervenes on the causal structure of the brain. I do not think that such materialists would claim that free will cannot be distinguished from nondeterminism.

Then there are other types of materialists, like Alex Rosenberg or the Churchlands, who would might even deny free will while believing that reality is indeterministic (I am not sure about their particular positions - but I would not be surprised, given that eliminativists deny mind altogether and are definitely the sort of philosophers who would accept a stochastic mathematical model as pure ontology).

Indeterminism is sometimes invoked as though it clearly solves the issue of free will for materialism. My point in bringing it up is that it seems at the very least to be irrelevant. Whether reality is deterministic or not we could have the illusion of free will - and that seems to be all you can get with a materialist conception.

You also have to overcome the problem from his point of view of why you need all of this other baggage to show nondeterminism. The simplest explanation to him is nondeterminism, which explains his experiences and observations perfectly well, and seemingly more simply than what you are trying to describe.

I’m not really showing nondeterminism; I apologize if that was not clear. I gave a picture of why free will is incompatible with materialism, and then said that taking an indeterministic view of reality does not solve the problem posed.

 

[polytropos]

Congratulations, polytropos . You have discovered the argument from free will, my favorite refutation of materialism.

It may or may not be a refutation; since as I said in #8, I did not claim as a premise that we have free will. The general intention is to “raise the cost” of materialism.

It is indeed a refutation and not just a reductio ad absurdum, because free will is actually provable. Consider this, skeptics of the argument:

• 
  *]It follows from the definition of “conscious” that an conscious event cannot occur within one’s mind without one’s knowledge, and that one cannot experience a conscious event in one’s mind unless this event is in fact occuring.
  *]Clearly, free will is a conscious thing, whether or not it exists.
  *]Therefore, we can learn infallibly from our introspection whether we have free will.
  *]Introspection supports free will.
  *]Therefore it exists.

• 
  I will leave it to others to object to this. I would just say that the materialist theory I am referring to may not be committed to holding that conscious “events,” being supervenient upon neural structures, are really “events” in a sufficiently robust sense.

 

[SentinelofTruth]

polytropos:

It may or may not be a refutation; since as I said in #8, I did not claim as a premise that we have free will. The general intention is to “raise the cost” of materialism.

True, so in the literal sense you didn’t refute materialism because you didn’t make the assertion that free will exists, but if you had’ve asserted that (and given the argument for it that I supplied) then you would’ve refuted it altogether.

polytropos:

I would just say that the materialist theory I am referring to may not be committed to holding that conscious “events,” being supervenient upon neural structures, are really “events” in a sufficiently robust sense.

If they’re a change, then they’re an event.

 

[polytropos]

If they’re a change, then they’re an event.

But it would seem that a materialist might deny that there is genuine change in what supervenes on a neural structure. Perhaps it would be better for a materialist to respond, though - since, IMO, this would be why materialists are implicitly committed to the more eliminative, fringe positions (which, I believe, would tend to deny free will anyway). What I’ve described as the “conceptual nexus” has to be causal inert (lest the position fall into some sort of dualism), so whether it “exists” and “changes” in a relevant way, to me, seems unclear. But that is another topic.

 

[SentinelofTruth]

If it supervenes on something that changes, then it would seem that it must change.

 

[Bagheera]

polytropos:

The argument was that materialist conceptions of mind are incompatible with free will.
I am not committing to a definition of free will.

I don’t see how is it possible to have a conversation about undefined categories. Maybe you wish to say that the materialist view of the mind logically leads to a fully deterministic worldview? Or, putting it differently, since the thoughts are “reduced” to neural activities and since the neural activities are deterministic, therefore all our thoughts are “determined”?

Is this the essence of your post? I need to know before we can go on.

 

[tonyrey]

The law of conservation of energy precludes the existence of free will which implies the existence of a **non-natural **form of energy (rejected on principle by materialists).

 

[polytropos]

I don’t see how is it possible to have a conversation about undefined categories.

I think that the necessary condition I offered for free will is about all we need. My contention is that if one seeks a lower bar for free will, then free will will end up being trivial (ie. if free will just is indeterminism), while materialism is compatible with any definition of free will that meets my necessary condition. This is obviously a generalization, so if someone thinks they have a definition of free will that evades my claim, then they should provide it.

Maybe you wish to say that the materialist view of the mind logically leads to a fully deterministic worldview? Or, putting it differently, since the thoughts are “reduced” to neural activities and since the neural activities are deterministic, therefore all our thoughts are “determined”?

Is this the essence of your post? I need to know before we can go on.

No, the point is that whether the worldview is deterministic or indeterministic, materialism and free will are incompatible. Determinism/indeterminism were not really the bulk of the point I was making; toward the end I just pointed out that free will could not be salvaged by an appeal to indeterminism.