More Free Will Than You Really Need

Since I've been busy, and the free will issue is hot right now, here's a repeat post from my free will series:

(Originally posted April 3, 2010)

Free will and quantum mechanics

I never used to spend much time thinking about the problem of free will. If I can think, "I'm going to move my hand now," and then move my hand - well, that seems like enough free will for me.

Recently, though, I started reading up on the free will debate. Here are two of the arguments philosophers use. (The first argument is courtesy of the Stanford Encyclopedia of Philosophy, the second is cribbed from Norman Swartz.)

Argument #1 – There is No Moral Responsibility if Determinism is True
Premise 1.	No one has power over the facts of the past and the laws of nature.
Premise 2.	No one has power over the fact that the facts of the past and the laws of nature entail every fact of the future (i.e., determinism is true).
Premise 3.	Therefore, no one has power over the facts of the future.
Thus:	If determinism is true, it appears that no person has any power to alter how her own future will unfold, and therefore no moral responsibility for it.

Argument #2 - Causal Determinism is a Necessary Condition for Moral Responsibility Premise 1: Unless there are extenuating circumstances, persons are (to be) held morally responsible for their actions. Premise 2: Being unable reasonably to have foreseen the consequences of their actions is one such extenuating circumstance. (Recall that young children who cannot reasonably foresee the consequences of their actions are not to be held morally responsible for the consequences.) Premise 3: In order to be able to anticipate or foresee the likely (or even the remotely likely) consequences of one's actions, the world must not be random, i.e. the world must be fairly regular (or causally determined). Thus: Moral responsibility requires that there be causal determinism.

So the first argument (known as the consequence argument) seems to show that if the universe is deterministic, then there is no free will (because everything in the future is determined by events that happened long ago), and hence no moral responsibility. But the second argument (I don't know if it has a name) says that if there is no determinism, then there can't be any moral responsibility, either. Taken together, the two seem to imply that there can be no moral responsibility. I don't think this conclusion is correct, and neither does Swartz. You can read his notes to see how he resolves the problem; here, I want to consider a different aspect.

The current best models of how the universe works are based on quantum mechanics, and quantum mechanics is not deterministic. Or perhaps I should say quantum mechanics is partially indeterministic. That is, the quantum state of a closed system determines the quantum state of the system at a later time,  but the quantum state of the system doesn't determine all of the interesting aspects of the system. In fact, for any quantum system, there will always be questions I can ask about the system that don't have a definite answer - even if the quantum state of the system is exactly known. However, quantum mechanics can give us the probabilities of the various possible answers to any question, and so the future is partially determined: determined up to the range of possible values allowed by quantum mechanics.

How does quantum mechanics relate to the two arguments outlined above? The consequence argument is right out: Premise 2 is simply false if quantum mechanics gives a true picture of the world. What about Argument #2? According to Premise 3 of this argument, the world must be "fairly regular," which, according to quantum mechanics, it is. But the indeterminacy of quantum mechanics provides some elbow room (in Daniel Dennett's phrase), so that the future is not completely determined by the past. Quantum mechanics thus seems to walk a fine line between a clockwork universe in which every action of every person is, in principle, completely predictable, and a chaotically random universe in which nothing is predictable and so there can be no moral responsibility.

Quantum mechanics may not be the explanation of free will, but it does seem to allow for free will.

Free Will Rides Again

The blogosphere has been full of free will discussions once again. Julian Baggini reviewed several recent books, including Sam Harris's Free Will. The Chronicle of Higher Education ran a series of short articles on free will: start here and follow the links on the page for the others. As if that weren't already a feast, Russell Blackford has been reviewing each of the CHE articles on Talking Philosophy.

And last, Jerry Coyne responds to Blackford's critique on his blog. There's some good discussion in the comments, about evenly split between eliminativists and compatibilists.

Level With Me Again

This post is a continuation of the previous post, and is part of my series on free will.


 Now suppose instead of a rover, we are talking about an animal - say, a bear. The bear has to decide whether to go left or right around the rock. It has various goals in mind: eat, find a mate, find a safe place to sleep. And it has memories of places it has found food or mates before, and some sort of mental map that relates those places to the choice to go left or right. The bear, we could say, is "running a survival program" that makes decisions on the basis of its goals and memories. (Again, I am not suggesting that the bear's brain works like a computer. I am using the computer program as an analogy to gain understanding of the levels of description.) Does the flow of electrons and neurotransmitters determine the bear's behavior, or does the bear's "survival program" determine the flow of electrons and neurotransmitters? What does your intuition say?

In the case of the rover, the program and the computer architecture were clearly designed to produce the particular decision-making behavior that we see. But in the bear's case, the hardware and software were designed, too: by the processes of evolution. Moreover, they were specifically designed to allow the bear to realize its long-term goals of survival and reproduction.

A human has these abilities to an even greater extent: a human can imagine different possible futures and make a choice between them. This, it seems to me, is the essence of free will. And it doesn't really matter whether the underlying micro-physics that runs the decision-making "program" is deterministic or indeterministic. Either way, it is me - my goals, desires, memories, plans - that (probabilistically or not) enact the choice.

Level With Me

This post is part of my series on free will.


Let's return to the idea of levels of description. An action can be described on many different levels: the electron/molecular level, the neural level, and the mental level of thoughts and desires. Very possibly there are intermediate levels of description in terms of various brain subsystems, but I don't think those levels are currently well enough understood to be very useful.


At the level of electrons, atoms, and molecules, the description ought to be quantum mechanical, and thus indeterministic.


At the level of individual neurons, the description might be effectively deterministic, or it might not. It all depends on whether the neurons act as quantum amplifiers (like the Geiger-counter-bomb) or quantum dampers (like a computer). It seems conceivable to me (having very little knowledge of neuroscience) that at times, a single ion tunneling across the cell wall could be the difference between a neuron firing or not. This would make the neuron a quantum amplifier: the randomness of the tunneling event could get amplified and end up determining whether or not you perform some action. But maybe not - maybe the neuron is more like a computer transistor, which is designed so that a few electrons more or less don't make a difference to the outcome.


What about the level of mental events? Here the description is so loose, I would find it difficult to call it deterministic. Suppose I am considering which university to attend. I have reasons in favor or against both of my top choices: one is closer, the other gave me more aid, one is small and intimate, the other is large and has lots of opportunities.... But someone else with the same list of reasons could end up with the opposite decision, without being irrational about it. I suppose we could put weights on all the reasons and devise a formula that would determine the result - but how to decide on the weights and the formula? And would it work again if I had to make the decision again? So perhaps at this level Ekstrom's "caused but not determined" makes sense.


In my opinion, many of the difficulties involved with free will are the result of confusing different levels of description. Here is Ekstrom, for example (Free Will, pp. 195-196):

The idea that we can direct our behavior by our thoughts ... is welcome, but it is only superficially comforting. It comforts until we think about the possibility that even our thoughts are driven to be what they are by previous neurophysiological events (between which there are deterministic causal links), a chain going backward through events in our childhood brains and to events prior to our birth.

Notice how she slides from the mental level, to the neural level, to the micro-physical level of "events prior to our birth" without batting an eye. (Clearly there cannot be a neural level before there are neurons, so she must be thinking here of the electronic/atomic level.)

Sometimes the argument is phrased in terms of ultimate responsibility: You cannot have ultimate responsibility for something that you do not have control over. You do not have control over the events of the distant past that are the causes of your behavior today (if determinism is true). Therefore, you do not have ultimate responsibility for your actions.

But recall the Mars rover that had to turn left or right when confronted with a large rock. It would be very strange to say the computer program that made the rover turn left doesn't have responsibility for the decision to turn left because, at the electronic level, everything is determined by the laws of physics. That seems to get the causality exactly backwards. We would rather say that the computer program, together with the computer hardware, caused the electrons to flow in such a way as to make the rover turn left. So here is one case where the higher level is the cause of what's happening at the lower level, rather than the other way around. Or perhaps this is using "cause" in a different sense - another thing I find missing in the philosophers I've read is a careful analysis of causation.

To be continued....

Command or Description?

This post is part of my series on free will.

Early scientists like Isaac Newton believed they were discovering the principles by which God governed his creation. Such "laws of nature" were absolute; they admitted no exceptions and could not be broken. A similar vein of thought can be found in Einstein: "I want to know God's thoughts; the rest are details," and in modern theoretical physics, where some speak of an ultimate theory, a Theory of Everything (TOE).

According to another vein of thought, the laws of nature are descriptions, not commands. They are a way of organizing a large body of observations according to the regularities found among them. They are mathematical models that capture some aspect of reality. This view acknowledges that such laws are always provisional and approximate. There will always be some realms - some scales of size or energy - in which the known laws have not been tested, and in which they may well fail to be exactly true.

Many philosophers of free will seem to adhere to the seventeenth-century view of the laws of nature. This is especially evident in their discussions of determinism. There are, it is assumed, some ultimate Laws that, God-like, determine everything that will ever happen.

An exception is the philosopher Norman Swartz, who argues that if we take seriously the view of laws as descriptions, not commands, then the problem of free will  does not even arise. I am not completely persuaded by his argument: read it yourself and see what you think. But I agree that thinking of laws as descriptions is an important step in the right direction.

These issues leap to the foreground in discussions of quantum mechanics. If we take the mathematical formalism of quantum mechanics to be an absolute command that admits no exceptions, then we are driven to a metaphysically absurd interpretation (the Many-Worlds Interpretation). But if we admit that quantum mechanics is merely the most accurate description we can give of certain systems, then such absurdities aren't necessary.

I see the significance of this view for free will in the possibility that there is more than one description of a certain event that is (approximately) a true description. The existence of a (valid) description at the level of electrons does not rule out the existence of a (valid) description at the level of mental events.

I will return to this point next time.

Not All Things Considered

Si, abbiamo un anima. Ma e fatta di tanti piccoli robot.

Yes, we have a soul. But it's made of lots of tiny robots.

-- Giulio Giorelli, quoted in Dennett, Freedom Evolves, p. 1

 This post is part of my already-much-longer-than-I-expected-and-still-growing series on free will.


What I would like to see from these free will philosophers is some sort of model that shows how will works and why or in what sense it can be considered free. (I was hoping for this sort of model from Freedom Evolves, but was disappointed.) What follows is an off-the-top-of-my-head attempt that's just to illustrate the sort of model I have in mind.

No one has ever announced that because determinism is true thermostats do not control temperature.

-- Robert Nozick, quoted in Dennett, Elbow Room, p. 51

Let's start with a thermostat. As Nozick says, we have no problem with the idea that the thermostat controls temperature. But there is also the purely physical level of description, in which the parts of the thermostat are obeying physical laws, without any concern for what they are controlling or if they are controlling it. So already with this simple system, we can talk about it on two level, though, obviously, there is no question of any sort of free will involved.

Bumping it up a notch, let's consider the Mars rover that I discussed earlier. Let's focus on a single line of code that determines whether,  when confronted with a large rock in its path,  the rover will turn left or right. That line might look something like this (vastly oversimplified, of course):

(A)   If X, turn left, else, turn right.

 Here X is a variable that can only take on the values 0 or 1. If X is 1, the machine turns left, otherwise it turns right. X depends on some list of inputs: what the rock looks like in the video input, what the angle of tilt of the ground is on either side, whether there seems to be clearer path on one side, etc. Some considerations might favor left and some might favor right, but all of them must be boiled down and weighted so that a clear, and deterministic, decision is made.

We can, of course, talk about all this on the level of electrons, voltages, and circuits. The electrons are just following the laws of physics, without any sort of decision in mind. Yet, the physical system has been carefully set up (by the computer engineers and the programmer) so that the laws of physics result in a decision about something important to the rover's goals. Actually, it is the goals of the mission scientists in this case - but the rover may be thought of as analogous to a simple organism, "designed" and "programmed" by evolution to achieve its goals of survival and reproduction.


Now bump it up another notch. Suppose there is some sort of monitoring segment of the rover's program. Let's call it the monitor for short. It keeps track of X, as well as many other important variables in the program. After the variable X is evaluated, but before statement (A) executes, the monitor looks at the result and has the opportunity to override it. Why might it want to do this? Well, the monitor has access to a wider range of information than the inputs to X. It might be monitoring the level of charge in the batteries, the distance left to go the next goal, and so forth: the long term goals, as opposed to the local situation that is dealt with by the inputs to X. For instance, if the battery is low and there is more sun on one side of the rock than the other, then the need to maintain power might outweigh the local considerations of topography that X takes into account.

The monitor, I propose, plays a similar role to Ekstrom's evaluative faculty. It is responsible for taking into account the overall state of the organism, as well as both short-term and long-term goals. And - I suggest - it gives the organism something like free will: the ability to pause and consider alternative possibilities and their consequences before proceeding.

Why would you want such a faculty? Why not, for instance, just shovel those additional considerations into the determination of the variable X? From a programming point of view, there may be other reasons to want such a monitoring program, and it might just make more sense to include this override capability in the monitor rather than in X. Flipping the switch on our intuition - thinking about an evolutionary sequence instead of a human design - it might be that the evaluative faculty represents an evolutionary step that overlays the earlier, more mechanical, system. "If it ain't broke, don't fix it" probably goes for evolution, too: if some system is working well as it is, then, rather than tinkering with that system, it might be preferable to add a new system on top of the old one. (Of course, there is nothing more or less "preferable" to Evolution herself - she merely allows what works to succeed and what fails to fail. What I mean is that a mutation that changes the old system might be disastrous, while a mutation that adds a little bit of monitoring might enable enhanced survival without messing up the old system.)

But why not have the monitor do all the work? Why not monitor all the variables of the program, and all the conditions of the environment, and consider all the various permutations of options and outcomes? As Dennett points out in Freedom Evolves, there simply isn't enough time to consider all things. If you tried to consider all your options - cut your fingernails, cut your hair, walk the dog, eat breakfast, jump out the window, nail your hand to the table, eat the curtains, eat the dog, walk the cockroach, ... - you would never get out of bed in the morning. We are overwhelmed at every instant by input - sights, sounds, smells - one of the most important tasks is to ignore stuff: to filter out the unimportant, in order to focus on the important. The other important task is to filter the possible outputs - choose, from the infinite range of possible actions, the one to do next. A monitor that tried to take everything into account, to consider every possible course of action, would be useless. It would never get anything done, and the organism would die.

I don't know if I've managed to capture some hint of the nature of free will in this very simple model, but just for kicks let's see what it implies about Ekstrom's theory of free will. Suppose I put such an evaluative faculty, such a monitor, into my Mars rover. Would I want to make it deterministic or probabilistic? From the point of view of control, determinism is the clear winner. As the programmer or human monitor of the rover, I want to know what it's going to do in a certain situation and why it's doing it - what portion of the program sent it down that path. From the point of view of an organism - I'm not sure. It might be helpful in certain situations to have a random component to one's actions - in flight from a predator, for instance. But in considering the issue of free will, well, a random component might make my actions less predictable, but would it make them more free? Wouldn't I rather make the optimal decision based on the (necessarily limited) inputs I have available, instead of flipping a coin to determine my actions? Personally, I think I would prefer that my actions be determined by my deliberation process, not just probabilistically caused by it.

When Ekstrom faces up to what is the good of an indeterministic component to the evaluative faculty, all she can come up with is, "Well, we need it to avoid determinism, because determinism is unthinkable." But if all it does is to lose us some amount of control over our actions, maybe we don't want it after all. Maybe it's time for another look at determinism.

Indeterminism Where?

This post is part of my series on free will.


One of the great things about Ekstrom's Free Will is the way she anticipates objections. There were many times as I was reading the book that I started to think, "But what about...?" and almost immediately found her posing the same question and answering it. At the end of Chapter 4, she gives a whole list of potential objections to her approach. She doesn't pitch herself softballs, either - she does a great job of putting the opposing case as strongly as possible. For the most part, I think she succeeds in answering these objections.


The most difficult issue she faces is the problem of control. The indeterminism in a libertarian account has to occur somewhere - but where should it go? If there is indeterminism in the process of deliberation, so that it is a matter of chance whether some consideration occurs to the deliberator, then the outcome of the deliberation seems to be a matter of pure luck, rather than something that is under the deliberator's control (p. 121). Likewise, if there is randomness after the deliberation is complete, then we have the same problem. So the only logical place for indeterminacy to occur is during the deliberation process itself. This all seems correct to me.


But then, given the same mental state before the deliberation begins and the same external inputs, the deliberator could come to a different conclusion. This, of course, is what the libertarian wants: the possibility of different outcomes. But if the outcome is not determined by the state of the deliberator or the external inputs, then it seems to be still out of the deliberator's control. Ekstrom sees this, and effectively punts: you have to have indeterminism somewhere, she says, or revert back to determinism - which she has already rejected.

Ekstrom makes no attempt to explain why there should be indeterminism at this point, apart from a brief comment that it might arise via amplification of quantum randomness (p. 124, citing Robert Kane). She just insists that we must have indeterminism to have free will, and it must occur in this way if it is to make any sense.

Kane's approach has the opposite problem. He locates the indeterminism in quantum events occuring in the brain, so we know why (in the reductionist sense) the indeterminism is there. But this puts it too early in the causal chain - random quantum events in the brain are outside the control of the agent, and so their outcomes are purely a matter of luck. (Dennett argues this point against Kane in Freedom Evolves.)

So it seems that indeterministic accounts of free will have a serious problem: We can get indeterminism from quantum physics, but we can't get it where we need it for free will. 

My Gut Says, "Meh..."

This post is part of my series on free will.

Last time, we saw how Laura Ekstrom gives a libertarian account of free will.There are a lot of things I like about Ekstrom's account. As an indeterminist myself from the point of view of physics, I am partial to an indeterministic account of free will.I also like Ekstrom's clear definition of a "self"  - something lacking in other accounts I've read. And I like how Ekstrom doesn't shy away from the consequences of her approach. If her account conflicts with how we view moral responsibility, well, then maybe our view of moral responsibility is wrong.

But I have some problems with her approach. I'll start out with a couple of general complaints. In contrast to Daniel Dennett, she seems unaware of current research in psychology and neuroscience. I think her account of the decision-making process, for example, could have benefited from a more science-based approach. And I wish she had made some attempt to deal with the level problem: how a description at the level of reasons and preferences interacts with a description at the level of neurons and electrons. Her account takes only the higher-level processes into consideration. But determinism, if there is such a thing, would occur at some lower level.

To expand on the last point, consider this (Free Will, p.111):

Why did the free agent decide in that way? Because of reasons x,y, z, and so on. Why did those reasons lead him to decide as he did? The determinist would answer: Because of a deterministic causal law linking such reasons to such a decision. But the proposed account answers: Because the agent exercised his evaluative faculty in a particular way. Why? For reasons that inclined but did not necessitate a particular outcome to his deliberation process.

But a deterministic account need not link reasons to particular decisions. A deterministic account could operate entirely at the level of (say) individual neurons: If neuron A fires when neuron B is in state X, then neuron B will fire.... This account doesn't deal with the level of reasons and decisions at all.

It seems it might be possible for indeterminism to reign at the level of reasons and decisions, even if determinism reigns at some lower level of description. This is what a compatibilist would argue, and Ekstrom doesn't seem to recognize even the possibility of such an account.

(Contrast Robert Kane's libertarian account of free will: he traces the indeterminacy down to the level of individual quantum events occurring in the brain.)

Another problem arises from the idea that an act is only free to the extent that it is undetermined by the reasons that occur in the deliberation process. (Free Will, p.125):

We sometimes speak of a range of freedom of action, some acts being fully free and others less so. The probabilistic model gives one way of making sense of degrees of freedom. Perhaps the most free acts derive from preferences whose probability of occurring was raised by the occurrence of certain previous considerations to values within a range of, say, 0.2-0.8, whereas the act would be less free when resulting from a preference at either end of the spectrum, that is, in cases where the considerations made the probability of the preference's occurence near 0.9 or 0.1.

So, it seems that if I really, really, really, really wanted to kill someone, then my action wasn't a free act - and I shouldn't be held morally responsible for it. This seems odd. It also brings to mind Dennett's comment on Martin Luther: "Whatever Luther was doing, he was not trying to duck responsibility." (Elbow Room, p.133)

When we have very strong feelings about something, we identify with the feeling - it seems to be a part of us, to express something about our innermost self. Ekstrom's account has it the other way around - our innermost self is expressed only in those decisions where we don't feel strongly either way, where our gut says, "Meh...."

Ekstrom's Solution

This post is part of my series on free will.

 Ekstrom's next step is to tackle where in the decision-making process the indeterminacy should arise. She describes the decision-making process as period of critical evaluation of various factors leading up to a preference - a settled decision about what to desire - or an intention - a settled decision about what to do.

If there were indeterminacy between the intention to act and the act itself, it would be a very odd thing indeed. For then we would sometimes find ourselves having decided to do something but then failing to do it. It seems to be the case that every time I intend to reach for my pen I succeed in doing so. And if I did fail sometimes, then it seems my failure wasn't a free action on my part, because I had intended to succeed. So indeterminacy between intention and action doesn't seem to provide free will of the sort we want.

Likewise, if the indeterminacy arises between my forming a judgment about what to do and the actual intention to perform the act, then I don't seem to be in control of the outcome.

So the indeterminacy must be pushed back even further, into the deliberation process itself.

But what is the self that is doing the deliberating? According to Ekstrom, the self (or agent) is an evaluating and choosing faculty by which the agent creates preferences and acceptances, together with the preferences and acceptances themselves.

Ekstrom's solution, then, is to locate the indeterminacy in the deliberative process by which the self - the evaluative faculty - causes the formation of an intention to act. In her words (Free Will, p.115), she endorses:

Type 3d Theory - An action is free only if it results, by a normal causal process, from a pertinent intention ... that is caused by the agent, where this latter term ... is reducible to event-causal terms.

I take "event-causal" to refer to normal neuro-physiological processes. Even though it is the agent that is doing the causing, this is not an "agent-causal account", because Ekstrom insists that this causation is normal physical causation, not some mysterious power that pertains to intelligent agents alone.

But it is crucial for Ekstrom that the type of causation that the agent/self engages in is indeterministic, as she has already rejected any sort of deterministic account.

Next time I will try to poke some holes in this account.

Caused But Not Determined

This post is part of my series on free will.

To Ekstrom, here's how things look. I don't want my actions to be uncaused, because I want to be in control of them: I want them to be caused by me. But I don't want them to be deterministically caused, for the reasons outlined earlier. So Ekstrom endorses what she calls a Type 3 theory: free actions are indeterministically caused, in some manner involving the self.

Ekstrom needs to explain two things: what is indeterministic causation, and what is a self. She starts with indeterministic causation.

Not all causes (according to Ekstrom) are deterministic causes. She gives three examples of indeterministic causes.

1. Contact with an infected person may be the cause of my getting a disease. This is so even though such contact does not guarantee that I get the disease - it only increases the probability of my getting it.
2. A child's falling and scraping his knee may cause him to cry - even though he might not always cry from a fall.
3. A Geiger counter exposed to a radioactive source and connected to a bomb may cause the bomb to explode. But if the counter is set up so that it only activates the bomb if a certain number of clicks occur in a given time interval, then it is undetermined whether the bomb will be set off.

I have some difficulty with these examples. In the first two cases, it may be that the outcome (getting the disease or crying) is a perfectly deterministic result of several causes. Being exposed to the disease is only one part of the overall cause, the state of my own health might provide the remaining necessary and sufficient conditions for my getting the disease. Likewise, the child's state of mind when he falls might determine whether or not he cries. Only in the third case is there true indeterminacy: the Geiger counter acts as an amplifier of quantum indeterminacy.

In fact, the only true indeterminacy that I know of is quantum indeterminacy. All other types of indeterminacy come from lack of information about some part of the system. (In quantum systems, even with the maximum possible information about the state of the system there is still indeterminacy.) It seems to me that if Ekstrom wants to succeed in this approach, she will have to track the indeterminacy down to some underlying quantum event (as Robert Kane does), or else show that there is some alternative source of true indeterminacy. At this point, however, she simply notes that others have given accounts of probabilistic causation and moves on.

So I will move on, too.

Varieties of Libertarianism

This post is part of my series on free will.

Ekstrom's Chapter 4 is titled "Varieties of Libertarianism." This is not, of course, libertarianism in the political sense, but in the philosophical sense. Ekstrom identifies three types of libertarian free will. She dismisses the first two, so I will not say much about them.

Type 1 theories - Free actions are uncaused.

The main problem with this approach is that, if my action is not caused by anything, then it is not caused by me, in particular. That seems to mean that it is out of my control.

Type 2 (agent-causal) theories - Free actions are not caused by any sequence of physical events, but by the agent who is acting. Here an agent is taken to be something that is not reducible to a physical description (in terms of systems of neurons, etc), but is a substance or entity in its own right. Chisholm says an agent is a "prime mover unmoved."

All this (it seems to me) has an uncomfortably religious ring to it. The agent is a sort of immaterial spirit that, in some unspecified way, is able to act on the material body to bring about actions. I gather from Ekstrom that these theories have, in fact, come in for some "heckling" from other philosophers. She points out three difficulties for agent-causal approaches:

1. How can there be two different types of causation - agent causation and ordinary physical causation - interacting in a single human being?
2. If an agent has no physical structure, then how can we understand the changes in the agent that bring about a free choice at a particular time? 
3. What evidence is there of a separate, non-physical, sort of causation?

On a proper libertarian account, in Ekstrom's view, free actions should be caused, and specifically caused by an agent: I want to be in control of my actions. But these actions should not be deterministically caused, for then there is no room for freedom of will. Her challenge then is to explain what an agent (or self) is, and to do so in such a way that the agent is capable of causing actions in a non-deterministic way. Thus she introduces her preferred type of libertarianism:

Type 3 theories - A free action is indeterministically (or probabilistically) caused by a previous event involving an agent.

I will tackle Type 3 theories in the next post.

A Libertarian's View

This post is part of my series on free will.

Having solved the problem of free will, I would like to test and broaden my understanding of the issue. I am reading Free Will, by Laura Waddell Ekstrom, and I will be blogging about the issues she raises.

Ekstrom is an incompatibilist. That is, she believes that free will is incompatible with a deterministic universe. After some preliminary remarks in Chapter 1, she sets out several arguments to this effect in Chapter 2. She starts with the Consequence Argument, which I have discussed before. She then presents several more rigorous versions of this argument.

For example, there is the question of whether someone "could have done otherwise." (This is due to Van Inwagen.) Suppose that at time t I did not raise my hand. If the universe is deterministic, then the state of the universe at some remote past time, together with the laws of nature, imply the fact that I did not raise my hand at time t. Yet I claim that I could have raised my hand at time t had I chosen to do so. But I can't claim that I could have done anything about the state of the universe at some time before I was born. Nor can I claim to be able to change the laws of nature. (No magic allowed!) Either I am wrong in thinking I could have done otherwise, or the universe is not deterministic.

Now, I have already explained that (according to quantum mechanics) the universe is not deterministic. So these arguments don't bother me all that much. Yet I have difficulty seeing how an indeterministic universe helps with free will. Well, this is just what Ekstrom is going to show - so she says. We'll have to wait and see.

There are several ways of responding to the "could not have done otherwise" argument. (One of the great strengths of Ekstrom's book is the way she consistently presents responses and counter-arguments to her points - very impressive in a book only 236 pages long!) One way out is to assert that I could have done something else if something had been different. If what had been different? Well, there are several routes one can take.

One route is to say I could have done something different if my mental state had been different. I could have raised my hand at time t if I had felt like doing so, for instance.

Another route is to say I could have done otherwise if the whole past of the universe had been different - so that I ended up in the state (at time t) of desiring to raise my hand.

Another route is to say (bizarrely, to my mind) that I could have done otherwise if the laws of nature had been different.

Or, one could argue (as Dennett does) that the ability to do otherwise is highly overrated. Dennett cites Martin Luther, who famously said, "Here I stand, I can do no other." Perhaps Luther was exaggerating, but, "Whatever Luther was doing, he was not trying to duck responsibility." (Elbow Room, p.133)

Ekstrom points out difficulties with each of these possible compatibilist replies. She then turns to consideration of libertarian accounts of free will, which I will discuss next time.

Free Will Index

An index to my series on free will:

1. Free will and quantum mechanics
2. The Consequence Argument
3. Some Uninteresting Questions
4. Sphexishness and Other Concerns About Predictability
5. Electrons R Us!
6. Quantum Mechanics Revisited
7. The Problem of Free Will - Solved!
8. A Libertarian's View  
9. Varieties of Libertarianism
10. Caused But Not Determined
11.  Ekstrom's Solution
12. My Gut Says, "Meh..." 
13. Indeterminism Where?
14. Not All Things Considered 
15. Command or Description?  
16. Level With Me
17. Level With Me Again

I make no claim to originality in any of this. My views have been largely shaped by Daniel Dennett's books, but my take on the importance of quantum mechanics for free will is somewhat different from his. A nice collection of articles about free will is linked at this FRDB thread. See especially the articles there from the Stanford Encyclopedia of Philosophy, and this set of articles from the Routledge Encyclopedia of Philosophy.

The Problem of Free Will - Solved!

I decide to pick up my pen - I reach for my pen - and my hand moves in accordance with my will. The connection between my conscious thought and my actions is an empirical fact. If this is free will, there are few who would deny we have free will. And this, I think, is the only sort of free will that really matters.

(Daniel M. Wegner, in The Illusion of Conscious Will, argues that the connection is not nearly so close as our intuition suggests. He points to phenomena such as automatic writing, hypnosis, and multiple personalities that indicate it's possible for there to be action without the feeling of will, and will without a corresponding action. This may well be true, but the "normal" connection of will and action persists nonetheless.)

To summarize what I've been writing about:

• - We need not fear determinism, for the world is not deterministic.
• - We need not fear the fixity of the future; that is the fatalist fallacy.
• - We need not worry that strict microscopic laws of physics allow our every action to be predicted, because chaos and quantum mechanics will not allow such predictability.
• - We need not worry that the electrons are in control, because the electrons are us.

So it seems there's nothing much left to worry about. As Daniel Dennett puts it in Elbow Room (p.169):

When we look closely at the sources of our suspicion and dread, we find again and again that they are not indisputable axioms or overwhelmingly well-supported, empirical discoveries, but unfocused images, hastily glanced at - like the shadows on the bedroom wall that take on an apparent robustness and menace precisely because we do not look at them closely.

There are other varieties of free will discussed by philosophers. Some would prefer an immaterial spirit that somehow affects our material bodies. Others attempt to create free will out of the very quantum jumps that (perhaps) leave our actions unpredictable. 

But the plain ability to decide to reach for my pencil, and then to do so, seems to me a variety of free will "worth wanting," in Dennett's phrase. If I can make my body do what my thoughts choose to do, what more could I want? To defy the laws of physics? Wishing for a supernatural miracle, or a natural, quantum-mechanical miracle, is as pointless as wishing I could leap tall buildings in a single bound.

I do not need such miracles to be content. I can glory in the notion that all of the human race's complex thought is a result of purely physical processes, and that those processes are the result of millions of years of evolution. From simple one-celled organisms whose only need was to find the next organic molecule to devour, through uncountable small changes in the ability to respond to the environment, and hence the ability to survive, came the amazing information-processing creatures we call humans. That's a heritage we can be proud of.

Quantum Mechanics Revisited

Looking back over my free will posts, I see that sometimes I seem to be saying that quantum mechanics is important for free will, and sometimes I seem to be saying the opposite. Let me see if I can clear things up.

I have addressed several different issues regarding the free will debate, and quantum mechanics relates to each issue differently.

Determinism and the Consequence Argument: To the extent to which the Consequence Argument depends on the premise of deterministic laws of nature, quantum mechanics is certainly relevant. Since quantum mechanics is an indeterministic theory, if quantum mechanics is a true description of the world, then the Consequence Argument fails. However, we have seen that the fear behind the Consequence Argument is not really the fear of determinism, but of lack of control: If the electrons are in control, then I am not in control. The problem then is not deterministic laws of nature, but any sort of microscopic laws at all.

Predictability: Suppose someone were to write a computer program that accurately predicts everything I will do and say tomorrow. That would be a huge blow to free will (though perhaps not a fatal one). How can I be anything more than a machine if a mere machine can duplicate my actions? Here, I claim quantum mechanics is relevant. Accurate prediction of a classically chaotic system will run up against the quantum limit in a very short time, forcing us to adopt a quantum description. But a quantum description will never produce perfectly accurate predictions: first, because the computer cannot completely replicate my quantum state (thanks to the no-cloning theorem), and second, because quantum predictions are only probabilistic (even if the exact quantum state were known).

Control: Here it gets a bit tricky. We have seen that for me to be in control of my actions, we need a fair amount of determinism in the world. I have been arguing that quantum mechanics has enough indeterminism to make effective prediction impossible. Does quantum mechanics leave us enough determinism for moral responsibility? I suspect it does, but this point clearly needs more investigation.

For now, let me just say that I think the way to proceed is by distinguishing the lower-level and higher-level laws of nature, as I began to do in the previous post.

Electrons R Us!

An imaginary scenario of NASA headquarters as a new robotic Mars rover begins its mission on the surface of that planet: Evelyn, the software engineer who wrote the code that allows the rover to pick its way across the Martian surface without running into rocks and cliffs, has invited her philosopher friend, Pete, to watch the rover's progress. Evelyn exults as the rover successfully negotiates one obstacle after another, and brags to Pete about how well her control program is working. Pete, however, complains, "It's not your program that controlling the rover; it's just electrons in those electronic circuits following the laws of physics!"

At this point, I can only imagine Evelyn standing there agape, thinking of the years she spent developing that software, unable to come up with any kind of non-insulting reply.

Pete is guilty of level confusion: his objection amounts to the claim that if a physical description of the rover at the level of electrons is correct, then any other description of the rover's behavior must be incorrect. I think the free will deniers are guilty of the same thing.

(Note: I am not making the claim that human brains are like computers in any particular way. I am only proposing the rover story as an analogy to help us understand different levels of description.)

Let's recall Argument #2 from this post:

Argument #2 - Causal Determinism is a Necessary Condition for Moral Responsibility Premise 1: Unless there are extenuating circumstances, persons are (to be) held morally responsible for their actions. Premise 2: Being unable reasonably to have foreseen the consequences of their actions is one such extenuating circumstance. (Recall that young children who cannot reasonably foresee the consequences of their actions are not to be held morally responsible for the consequences.) Premise 3: In order to be able to anticipate or foresee the likely (or even the remotely likely) consequences of one's actions, the world must not be random, i.e. the world must be fairly regular (or causally determined). Thus: Moral responsibility requires that there be causal determinism.

This argument, I think, is sound. In order for there to be moral responsibility, I want my actions to be causally determined. Determined by what? By me, of course! But what am I? I am my thoughts, feelings, desires, beliefs, and so forth. But if we are right in assuming that the mind is a purely physical entity (no magic allowed!), then thoughts, desires, etc., are all the results of some sort of electro-chemical activity in my brain.. ("We are electrons, and electrons are us.") And if my actions are to be determined by my thoughts and desires, then they ought to be determined by those patterns of electrical activity that are the physical counterpart of those thoughts and desires.

So if free-will deniers claim that we can't do other than what we did because "it's all just electrons" following the mindless laws of physics, they are committing the same sort of level confusion as Pete.

There can be a low-level description of an action that consists of electrons following physical laws. And there can be a high-level description of the same action that consists of mental states like thoughts and desires. And there need not be any contradiction between the two descriptions, any more than there is a contradiction between the electron-level description and the program-level description of the rover. At the physical level, all we see is electrons affecting the motion of other electrons. But at the mental level, we have desires affecting thoughts, thoughts influencing desires, and, crucially, a mind capable of reflecting on its own mental state.

Free will is something that exists at the mental level; it is a characteristic of the relationship between thoughts and actions. To try to look for it at the level of electrons is to commit a category error - just as it would be to look for a FOR loop inside the rover. ("All I see is electrons!")

Sphexishness and Other Concerns About Predictability

Philosophers of free will give the issue of predictability short shrift, in my opinion. Typically they are worried about whether, given the past state of the universe and the laws of physics, the future of the universe is completely determined in principle. They aren't particularly worried about whether it is predictable in practice. I think they should be.

Consider the curious habits of the Sphex wasp. She lays her eggs underground, then flies off to get a cricket, which she paralyzes and places in the burrow for the hatchlings to eat when they emerge. In doing so, she follows a strict routine: after dragging her prey to the burrow, she goes inside to check on the eggs, then comes back out and finishes the task of dragging the cricket inside. Now, if some devious human moves the cricket away from the entrance to the burrow while she is inside checking on the eggs, she comes out, drags the cricket to the entrance again, and once again goes inside the burrow to check on the eggs.

If again the cricket is removed a few inches while the wasp is inside, once again she will move the cricket up to the threshold and re-enter the burrow for a final check. The wasp never thinks of pulling the cricket straight in. On one occasion this procedure was repeated forty times, always with the same result.

(Wooldridge, quoted in Dennett,  Elbow Room, p. 11.)

The first thing to note about the strange behavior of Sphex is that this is not a case of an agent doing the exact same thing under the exact same microscopic conditions. It doesn't matter if the experimenter moves the cricket three inches or four, the "same behavior" is repeated. And by "same behavior", we don't mean the microscopically exact same behavior, rather we mean the same general class of behavior. This is, I think, a much more interesting question than the question of whether we would do the exact same thing under the exact same conditions. Is there some general class of conditions under which I, as a human, would necessarily produce the same general response? If so, this is a more serious threat to free will than the possibility that I would do the exact same thing under the exact same conditions.

What is so striking about Sphex is the mindless predictability of her behavior. She apparently never stops to think, "Hey, didn't I just do this? Do I really need to do it again?" She is apparently pre-programmed to execute a series of steps. If those steps are interrupted, the series must be re-started at the point of interruption.

Are humans at all sphexish? Are there circumstances under which we will mindlessly repeat a certain behavior? We seem to be exempt from such behavior because of our ability to reflect on our own actions. Yet, if we behave according to (reasonably) deterministic laws of nature, there ought to be some (perhaps very small, yet not microscopically exact) set of conditions under which we would produce nearly the same behavior.

We can imagine someone, a burglar, say, who somehow gets hold of a complete physical description of  someone - down the quantum state of her very molecules. The burglar uses this information to predict when the victim will be out of her house, and then goes in and robs her.

This seems a reasonable concern, if the universe runs according to strict physical laws. But the laws of physics have been invoked to conjure up the spectre of sphexishness, and now the laws of physics rush in to the rescue. In a sufficiently complex, deterministic system, something called "chaos" comes into play. In a chaotic system, the set of conditions needed to produce with certainty a particular outcome is exponentially small. If we stipulate that humans are sufficiently complex, then the set of conditions needed to guarantee a particular human behavior will be smaller than the quantum limit allows, even for very short times. In other words, it will be impossible, without violating the laws of quantum mechanics, to predict the future behavior of a human from physical principles for more than some very short time.

If we move to a quantum mechanical description of our human (supposing such a thing to be possible, which is doubtful), one thing improves: quantum systems are never chaotic, as long as the Hamiltonian (the function that describes how the system changes with time) is completely known. So at first glance we seem to be better off with a quantum description, in terms of being able to predict outcomes. However, there are several problems with the quantum approach.

First problem: the Hamiltonian is never completely known as long as the system interacts with its environment. So, we could apply quantum mechanics to a person who was completely isolated from everything (including air, water, food, etc.)  - an unattractive prospect - or we could try to include the entire environment in our quantum description, greatly magnifying the difficulty of the task.

Second problem: Even with an exact quantum description (of, say, the person plus her environment) we will only get probabilistic results. That is, we can say she will have eggs for breakfast with 38% probability, jam and toast with 12% probability, etc. This is no more useful - and no more a violation of free will - than taking a survey of her eating habits.

Finally, I should point out that the problem I glossed over, of obtaining an exact quantum description of someone (to say nothing of her environment) is not just difficult, it is physically impossible. This is because of the no-cloning theorem I mentioned last time: it's impossible to take an unknown quantum system and make an exact copy of it.

So it seems that, from the point of view of predictability, free will has nothing to fear from the fundamental laws of physics. That is not to say that there couldn't be some higher-level laws that provide more predictability. After all, the same quantum considerations apply to Sphex, who remains quite predictable. (Not to mention a rock - also, presumably, a quantum system - whose behavior is extremely predictable.) So there is no way at this point to rule out the possibility of a nefarious burglar who exploits some as-yet-undiscovered psychological principles to generate useful predictions. But from what we have seen, there is no reason to think that the existence of such laws follows necessarily from the physical description of humans.

Next time: Electrons R Us.

Some Uninteresting Questions

As promised last time, I'm going to tackle some uninteresting questions about free will. What's interesting about these uninteresting questions is the amount of time and effort philosophers have spent on them. Well, I suppose they have to earn their salaries somehow.

Are Zombies Possible?

I'm not talking about the arm-ripping, brain-eating type of zombie, I'm talking about the philosophical zombie. In philosophy, a zombie is something that looks, acts, and talks in every way like a human, but there's no one home. These creepy entities have no thoughts, feelings, desires, or sense of self. If you think zombies could exist, then it's possible that everyone you know (including your spouse, parents, and children) is a zombie. As a physicist, I feel justified in ignoring these fantastical creatures: by assumption, there is no detectable difference between a zombie and a human. So what's the use in talking about them?


Would You Do It All Again?

One way of phrasing the free will issue is to ask, "If you were in the exact same situation, would you do the exact same thing, or could you have made a different choice?" This is often illustrated with Austin's Putt. Philosopher J. L. Austin wrote

Consider the case where I miss a very short putt and kick myself because I could have holed it. It is not that I should have holed it if I had tried: I did try, and missed. It is not that I should have holed it if conditions had been different: that might be so, but I am talking about conditions as they precisely were, and asserting that I could have holed it. 

The problem with this sort of discussion is in the phrases "exactly the same" or "conditions as they precisely were." Presumably we are talking about conditions being microscopically the same: same position of the ball, of every blade of grass and every air molecule. Further, I must be in the exact same state myself: every atom and electron in my brain and my body must be in the exact same state.

If that is what is meant, then I think this is a profoundly uninteresting question.First of all, there is clearly no way to test this question on the macroscopic level. Even if we could get the ball in the right place and the external conditions precisely the same (which we can't, of course) my own internal state will be different on a second go-round, for I will have a memory of the first try. Secondly, we already know the answer: from a physical point of view, having the exact same (quantum) state of the system is no guarantee of the same outcome. Finally, it is not at all clear what this has to do with free will. Let's humor the questioner and suppose we were able to do the experiment: make ten absolutely identical copies of me and my environment, and observe what the ten copies do. Suppose we found that I did the exact same thing each time. What have we learned? Only that identical copies do identical things - not particularly surprising or interesting. Suppose, on the other hand, that the copies do different things. On the physical level, this is possible if some quantum randomness is getting amplified into behavioral differences, as already noted. This would indeed be an interesting result - but it's not clear whether we've learned anything about free will. We have merely slipped back to the question of whether the underlying physics is deterministic or indeterministic. And, as I argued in the previous post, (in)determinism isn't really a crucial issue for free will.

(As an aside, I note that our inability to perform the experiment is not just a matter of technological sophistication: it is actually impossible in principle to put someone or some thing into the exact same state if the world runs according to quantum mechanical principles. This is the famous no-cloning theorem, to which I will return in the next post.)

What About The Nefarious Neurosurgeon?

Dennett, paraphrasing Frankfurt, presents this scenario (Elbow Room, p. 132):

Jones hates Smith and decides, in full possession of his faculties, to murder him. Meanwhile Black, the nefarious neurosurgeion, who also wants Smith dead, has implanted something in Jones's brain so that just in case Jones changes his mind (and chickens out), Black, by pushing his special button, can put Jones back on his murderous track. In the event, Black doesn't have to intervene; Jones does the deed all on his own.

Is Jones responsible for the murder, Frankfurter asks, even though he could not have done otherwise?

My question: why would anyone waste time worrying about such ridiculous stuff?

Next time: All sphexish things considered.

Free will and quantum mechanics

I never used to spend much time thinking about the problem of free will. If I can think, "I'm going to move my hand now," and then move my hand - well, that seems like enough free will for me.

Recently, though, I started reading up on the free will debate. Here are two of the arguments philosophers use. (The first argument is courtesy of the Stanford Encyclopedia of Philosophy, the second is cribbed from Norman Swartz.)

Argument #1 – There is No Moral Responsibility if Determinism is True
Premise 1.	No one has power over the facts of the past and the laws of nature.
Premise 2.	No one has power over the fact that the facts of the past and the laws of nature entail every fact of the future (i.e., determinism is true).
Premise 3.	Therefore, no one has power over the facts of the future.
Thus:	If determinism is true, it appears that no person has any power to alter how her own future will unfold, and therefore no moral responsibility for it.

Argument #2 - Causal Determinism is a Necessary Condition for Moral Responsibility Premise 1: Unless there are extenuating circumstances, persons are (to be) held morally responsible for their actions. Premise 2: Being unable reasonably to have foreseen the consequences of their actions is one such extenuating circumstance. (Recall that young children who cannot reasonably foresee the consequences of their actions are not to be held morally responsible for the consequences.) Premise 3: In order to be able to anticipate or foresee the likely (or even the remotely likely) consequences of one's actions, the world must not be random, i.e. the world must be fairly regular (or causally determined). Thus: Moral responsibility requires that there be causal determinism.

So the first argument (known as the consequence argument) seems to show that if the universe is deterministic, then there is no free will (because everything in the future is determined by events that happened long ago), and hence no moral responsibility. But the second argument (I don't know if it has a name) says that if there is no determinism, then there can't be any moral responsibility, either. Taken together, the two seem to imply that there can be no moral responsibility. I don't think this conclusion is correct, and neither does Swartz. You can read his notes to see how he resolves the problem; here, I want to consider a different aspect.

The current best models of how the universe works are based on quantum mechanics, and quantum mechanics is not deterministic. Or perhaps I should say quantum mechanics is partially indeterministic. That is, the quantum state of a closed system determines the quantum state of the system at a later time,  but the quantum state of the system doesn't determine all of the interesting aspects of the system. In fact, for any quantum system, there will always be questions I can ask about the system that don't have a definite answer - even if the quantum state of the system is exactly known. However, quantum mechanics can give us the probabilities of the various possible answers to any question, and so the future is partially determined: determined up to the range of possible values allowed by quantum mechanics.

How does quantum mechanics relate to the two arguments outlined above? The consequence argument is right out: Premise 2 is simply false if quantum mechanics gives a true picture of the world. What about Argument #2? According to Premise 3 of this argument, the world must be "fairly regular," which, according to quantum mechanics, it is. But the indeterminacy of quantum mechanics provides some elbow room (in Daniel Dennett's phrase), so that the future is not completely determined by the past. Quantum mechanics thus seems to walk a fine line between a clockwork universe in which every action of every person is, in principle, completely predictable, and a chaotically random universe in which nothing is predictable and so there can be no moral responsibility.

Quantum mechanics may not be the explanation of free will, but it does seem to allow for free will.