The Courage to Conjoin
What renders atheism incompatible with a coherent account of morality, when it is incompatible, is physicalism (or what is sometimes described as reductive materialism). If it is true that the universe consists entirely and without remainder of particles and energy, then all human action must be within the domain of caused events, free will does not exist, and moral reasoning is futile if not illusory (as are other kinds of reasoning).
This is a stupefyingly widespread view that flows from an elementary error in thinking.
Suppose you know that there is free will or that moral reasoning is not futile. Next, suppose you find that the universe is made out of only whatever the universe is made out of. What do you infer? You infer that free will and moral reasoning, which occur inside the universe (or as aspects of the universe), whatever they may be, are made possible because of whatever it is the universe is made out of. And there you are.
Here is what you do not do. You do not start with a mystifying conditional like “If the universe is only physical (or whatever), then there is no free will,” because how do you know that? You don’t. But you may think you do and so you get caught in a retarded ponens/tollens showdown: the universe is physical, ergo no free will, or… free will, so the universe is not physical. But, again, through what method of divination do we validate this conditional? None. Because we already know it is false.
Here are two things you know: free will exists (it is obvious: go ahead, touch your nose) and the universe is made of whatever it is made of (obvious, if anything is). Therefore, you know the conjunction of those two things. Therefore, you know that the crazy proposition that says that one of them must be false isn’t true! There’s no need to get hung up on an arbitrary conjecture about the trascendental conditions for the very possibility of the existence of something when things you already know rule it out. P & Q implies ~ (P —> ~Q). Logic: try it!
If we find out tomorrow that the universe is made of jello, all we will have learned about morality is that it, like everything else, is ultimately jello-dependent.
July 18th, 2007 16:35
Well, more strictly, we know that the experience of free will exists, phenomenologically.
It could just as easily be so that being told to touch your nose (or to try to as a thought experiment) causes responses in a way that’s strictly deterministic (with no actual freedom of will in the sense of non-determinism).
It’s long been my position that it doesn’t matter*, since we’re incapable of seriously thinking that there’s no free will, because of the ubiquitous (even if possibly false) experience of having it.
* In any non-metaphysical sense.
July 18th, 2007 17:21
My guess is that Ponnuru would deny that we know that free will exists.
The common fallacy is that physical determinism implies a lack of meaningful free will. That’s the mistake. It’s a confusion of levels.
As you demonstrated with the nose-touching, we do control our behavior (to some extent) with our ideas. It’s not futile or illusory. And, it doesn’t require magic.
July 18th, 2007 17:28
Deliberation exists. I prefer to stay as far away as possible from the term “freewill” simply because it comes with the aforementioned metaphysical (or, at least mistaken metaphysical) baggage and because it does no more actual conceptual work than the concept of deliberation.
July 18th, 2007 17:29
“a retarded ponens/tollens showdown..”
Great stuff.
July 18th, 2007 18:40
[...] Will Wilkinson explains why it is silly to claim that materialism is incompatible with free will or morality. [...]
July 18th, 2007 19:39
Well said! Are you sure you’re not aching to return to academic philosophy?
July 18th, 2007 20:06
I thought the central fallacy of incompatibilism was actually the fallacy of composition. By the same logic you can say “if the universe is made of atoms and energy then humans are invisible”.
July 18th, 2007 22:43
[...] line of argument always sounded dumb to me, and Will Wilkinson explains why it is in quite an economical way. Here’s a snippet: Suppose you know that there is free [...]
July 18th, 2007 23:54
I would say the problem is with the notion of “free will.” It’s never been entirely clear to me what this term is supposed to mean.
If it means something very simple and prosaic (i.e. if Will’s “I can touch my nose” serves as good evidence of free will) then it’s not the sort of thing anyone should ever disagree over.
If it means something ambitious (like “humans are in principle not reducible to a mechanism” or, more concretely/specifically, “it will never be possible to create a machine that can predict my actions”) then it seems to me very very unlikely to be true.
July 19th, 2007 01:25
I’ve always felt that materialism and free will were compatible, and I really like Will’s post.
However, the problem is that the implication (materialism -> not free will) is very appealing. I accept materialism, and I accept free will, so I feel like the above implication is false. But to be metaphysically happy, I’d love to read a clear explanation as to why the implication is false. It might be as simple as mk says, just showing that free will isn’t even really a coherant concept. Or it might involve some very deep, mindbending philosophy =).
July 19th, 2007 05:33
I think the point here is that the philosophy involved needn’t be “complex” or “mindbending” — it need merely consist in not making a (on consideration obvious) error. There are dozens of ways of accepting materialism and still saying no to determinism.
You can attack it as Matt does as an unjustified composition, or as composition’s big sister sorites. You can attack the concept of free will which is, as Sigivald notes, more phenomenon than noumenon, and is in any case somewhat silly. In the opposite direction, you can attack the notion that material components are necessarily deterministic in the first place. You can spin any of half a dozen different compatibilist stories — that free will implies not an absolute independence from material but merely an inability to predict action as a practical matter; that determinism is underdefined and when defined coherently excludes or includes both human action and free will; that the distinction between free will and determinism is not in fact of any practical import. Blah blah blah.
The point, as Will I think correctly points out, is not that a particular one of these stories is correct. The point is that some story along these general lines must be; for most purposes we may take that as read and move on. Determinism, like an other skeptical proposition, is subject to its own variant of “here is a hand” — we are more certain of free will than we are of any particular story of the universe’s (or free will’s) composition, of the definitions of these comically fuzzy terms, of the nature of the phenomenon of “certainty that free will exists”, or what-have-you. To bootstrap from premises you are unsure of to a conclusion contradicting something you are sure of is the opposite of good philosophy. It is college dorm room too-late-at-night bull session philosophy.
July 19th, 2007 06:12
I don’t think you’ve made a very good argument here Will, it’s not at all as obvious as you would like it that free will exists.
Free will is by definition super-natural i.e. beyond, outside, or above the universe. The universe being everything which is within the realm of the laws of physics and accessible to science. Only if I were able to relive the exact same moment in time and make two different decisions would the touch your nose thing prove anything to me.
July 19th, 2007 08:29
“Here are two things you know: free will exists (it is obvious: go ahead, touch your nose) …”
Not at all obvious to the scientists actually studying the phenomenon, however. (Start with Libet, e.g., and work your way up to contemporary neuroscientists.)
The current and widespread undestanding in the field is that “free will” is, at best, a very convincing illusion.
July 19th, 2007 08:42
free will exists (it is obvious: go ahead, touch your nose)
What the hell?! If I find myself doing what Will Wilkinson tells me to, that’s good evidence that I have no free will whatsoever!
July 19th, 2007 08:44
Free will is by definition super-natural i.e. beyond, outside, or above the universe.
If that is your definition of “free will,” (or rather, part of your definition, since that still doesn’t explain what the nature of “free will” is by your lights) then it follows, doesn’t it, that you and I, who exist inside the universe, will never experience it. And thus, whatever this “free will” is, we will never experience it and its existence or nonexistence is of no concern.
July 19th, 2007 09:21
[...] Thursday, July 19th, 2007 in Politics Will Wilkinson, former philosophy student (natch), gets appropriately huffy. [...]
July 19th, 2007 09:26
I endorse your compatibilist conclusion Will, but your argument for it is fallacious.
Here are two things you know: free will exists (it is obvious: go ahead, touch your nose) and the universe is made of whatever it is made of (obvious, if anything is). Therefore, you know the conjunction of those two things. Therefore, you know that the crazy proposition that says that one of them must be false isn’t true!
The problem with your argument: “The universe is made of whatever it is made of” is not the thesis of materialism, and is not the antecedent of the conditional premise of the argument. So showing that it is compatible with free will does not show that materialism is compatible with free will. (Since it’s a tautology, it better be compatible with free will, since otherwise contradictions are true!)
Suppose I have a lump of metal, and a nearby Geiger counter, which is not detecting any radiation. I reason: “If the lump were made of uranium, my Geiger counter would be clicking. It is not clicking. Therefore, the lump is not uranium.”
One cannot refute the major premise of this argument by pointing out the compatibility of “The lump is made of whatever it is made of” with “The Geiger counter is not clicking”, because “The lump is not made of whatever it is made of” is not the antecedent of the conditional.
July 19th, 2007 09:28
Oops, the end of that last sentence should read
because “The lump is made of whatever it is made of” is not the antecedent of the conditional.
July 19th, 2007 09:31
“Here are two things you know: free will exists (it is obvious: go ahead, touch your nose)”
The problem here is that this is the proof that free will does not exist.
For some number of people, because the universe is a mechanistic place, they can not touch their nose. The mechanisms don’t work. For all people, a similar easily phrased task can not be done. For some, it might be as mundane as “Touch your toes.” for others, it might be “Jump to the moon”.
You may argue that people have the will to attempt those things, so they have free will. That’s fine. But there are things that people do NOT have the will to even attempt. Our will is constrained by fear, despair, and even revulsion.
You may argue that we have free will because we may desire to do things even if we can not bring ourselves to attempt them. Even that stretch falls short. What we desire is constrained and easily manipulated by biochemistry and physiology. An alcoholic may wish he did not desire another drink, but he is out of luck. He has no choice in the matter. He is just an extreme example though. We are all programmed through our experiences to desire things in such a way that no act of will can alter.
Ironically, the very lack of free will is what ensures the efficacy of moral reasoning. Morality is an entirely mechanistic process. It is simply what is “best” for “us”. The definitions of those two words in quotes are all that is important. By putting our virtual mothers and babies on train tracks we more accurately resolve what is meant by “us” and “best”.
Even the illusion of free will itself is a part of our morality. It is a useful device for the curtailment of harmful actions. It is best for us to believe we have free will. It removes restraint in the punishments used to dissuade harmful actions agains us as a whole. By holding an individual responsible for their actions, they can be punished in isolation, which is much easier than redressing entire socialogical environments which bring about harm. People today can no more disregard free will than people could disregard God 800 years ago. Society would collapse if that happened. Because morality will not destroy its society, we will not accept the absence of free will until we can survive doing so.
July 19th, 2007 09:34
Brock, “Matter” is one of the things the universe could be made of. I don’t happen to know what the universe is made of, which is why I made the argument general. But I know that whetever it is made of, matter included, it is evidently compatible with free will.
July 19th, 2007 09:52
Ponnuru’s point is simplistic and cheap and is only a means to an end (cultural conservatism), but this response isn’t particularly deep either. Free will, a completely undefined concept, exists? And that’s that? And the proof is our intuition?
It’s not as if National Review types deserve (or are interested in) anything better, but this is a pretty fascinating debate simplified awkwardly.
July 19th, 2007 09:54
[...] suggesting that some forms of atheism make free will and moral reasoning absurd. Will Wilkinson responded by essentially saying that this is a psuedo-problem (link via Unqualified [...]
July 19th, 2007 10:04
Will, taking the argument to an epistemological level doesn’t work. Compare:
(1) I know that the Peano axioms are all true.
(2) I don’t know whether the Goldbach conjecture is true.
(3) Therefore, the Goldbach conjecture is compatible with the Peano axioms.
July 19th, 2007 10:04
Free will is problematic under an omniscient God, as well. If all are actions are known in advance, were we free to choose them.
July 19th, 2007 10:38
The argument Ponnuru makes drives me crazy. It ranks right up there with Dostoevsky’s ‘if there is no god, everything is permitted,’ and with Descartes’ ‘if there is no god, there is no knowledge.’
All 3 are skyhook arguments. They start with the mistaken premise that some (spooky, wildly improbable) thing X is required for some other thing. This premise is almost always false. In Ponnuru’s argument: contra-causal authorship of actions is required for the non-futility of moral reasoning (which in turn is required for free will, which in turn is required for moral responsibility). James above gives a definition which I suppose Ponnuru would accept. (In Dostoevsky’s: a guarantee of eternal moral scale-balancing is required for moral imperatives to have meaning. I won’t get into Descartes’) Then, they give that spooky thing a really nice name–free will, in this case–and hang absolutely everything on it. Remember that Ponnuru makes his argument in the context of, well, trying to argue that theism is necessary for morality to have meaning, so the Dostoevsky example is apt.
Ponnuru’s argument is really something like this:
1. There is moral responsibility only if moral reasoning is non-futile. (If there is moral responsibility, morality has a meaning, etc., etc.)
2. Moral reasoning is non-futile only if it is causally relevant to the production of intentional action.
3. The causal relevance of moral reasoning (and other mental operations) to intentional actions is called free will.
We’re ok so far. The definition in 3 needs a hell of a lot of refinement, but it works for now. Here’s where Ponnuru screws up:
4. A thing or event is causally relevant only if it is not itself caused. (This is contra-causal, prime mover free will.)
5. If determinism is true, then every event is caused.
And then everything follows:
6. If determinism is true, then nothing (except perhaps atoms or sub-atomic particles) is causally relevant (from 4&5)
7. Moral reasoning is futile (or illusory). (from 2&6)
8. Free will does not exist. (from 3&6)
9. Moral responsibility does not exist (from 1&7)
The problem, of course, is with 4. It confuses levels, and elides a distinction between causal relevance and causal efficacy. If determinism is true, then only atoms or sub-atomic particles (or strings or whatever) are causally efficacious. It does not follow that physical objects and events that can be spoken of only at a higher level are not causally relevant. One billiard ball’s striking another is causally relevant to the second ball’s moving, even if the first ball’s movement is itself caused and causally non-efficacious. For if the first billiard ball had not struck the second, the second would not have moved. (In the possible world where, …) (The paper to read here is Frank Jackson and Phillip Pettit ‘Causation in the Philosophy of Mind.’ PPR 50 (1990), repr. in their anthology with Michael Smith.)
Eliding the distinction between causal relevance and causal efficacy is eliding the distinction between determinism and fatalism. If fatalism is true, then there is no free will and moral reasoning is futile. But fatalism is the idea that whatever is going to happen will happen regardless of what we do. This is false.
What happens clearly depends on what we do. And what we do clearly depends on what information we are able to acquire and how that information stands for us as constituting or comprising reasons for action. (That is, what we do depends on our reasonings, moral or otherwise.)
The fact that there are three cows loose on Connecticut Ave. is a good reason for me to avoid Connecticut Ave. The traffic will be horrible. But for the farmer who has lost his cows, it is a very good reason to go to Connecticut Ave. to collect them. Similarly, the fact that some action is torture is sufficient, over-riding reason for me not to perform or condone the action. For others, the fact does not have this standing (though it should: they are bad people for not being sufficiently sensitive to the reason-giving status of the fact that an act is torture). Reason-responsiveness and the capacity to learn (that is, a less than one probability that tomorrow’s responses to information are not the same as today’s responses) are all we need for free will (the causal relevance of mental operations to intentional actions) and moral responsibility. And these are both clearly compatible with determinism.
July 19th, 2007 10:40
err… the capacity to learn (that is, a less than one probability that tomorrow’s responses to information are the same as today’s responses)…
i snuck a ‘not’ in there by accident.
July 19th, 2007 10:53
Listen to McNeel, people!
July 19th, 2007 11:30
Suppose you know that Ramesh Ponnuru is an idiot… (That would save several steps in the argument.)
July 19th, 2007 11:46
By the way, Douglas Hofstadter’s latest book “I Am a Strange Loop” does a good job at describing the different levels and why the low-level doesn’t determine the high-level in an interesting way. He’s more interested in refuting Searle’s anti-AI “arguments”, but the points are relevant to anti-free-will arguments as well.
I didn’t find very much new in the book, but he’s a master of analogies and word-play, so it was still a fun read. And, I’m sure it will be even better for people not already familiar with the arguments or the stories he relates.
Dennett’s “Freedom Evolves” is a good read as well.
July 19th, 2007 12:01
McNeel, you make good arguments, but the truth of determinism isn’t completely unproblematic to morality. There is the issue, after all, of how one feels guilt without the ability to have done otherwise. And there is the issue of punishing criminals who could not have done otherwise.
It’s not coincidental that these issues center around our psychological responses to determinism - that is, after all, the hardest part.
While we’re suggesting books: “Free Will and Illusion” by Saul Smilansky.
July 19th, 2007 12:14
“…The fact that there are three cows loose on Connecticut Ave. is a good reason for me to avoid Connecticut Ave. The traffic will be horrible. ..”
And yet, many will be stuck in traffic on Connecticut Avenue despite knowing about the cows and knowing what the consequences would be.
July 19th, 2007 12:17
cg,
What do you mean by “The ability to have done otherwise?”
How would your decision being undetermined by your prior state (ie somewhat random) make you more responsible for it?
July 19th, 2007 12:31
CG & Gil,
I think the question here is the relevant sense of “able”. If this is supposed to mean, “defies the laws of nature,” then of course we aren’t ever able to do other than what we did do. But then it is mysterious what this has to do with agency as we notmally understand it.
The right idea of “able” (or “power” or “can”) is the lack of certain relevant constraints, like coercion, brainwashing, intoxication, etc. These sorts of things can cripple the free exercise of agency. But that’s rather different from the fundamental nature of the universe ruling it out.
July 19th, 2007 13:04
McNeel (and Will, I guess),
With respect, it is you, and not Ramesh, who are confused. Ponnuru does not (and does not need to) deny that reasons and reasoned behavior can exist in a deterministic universe. What he needs is that morally blameworthy or praiseworthy actions cannot exist.
You are the one guilty of tricky elision. We think of morally blameworthy behavior as behavior that is wrong but also that one might have chosen otherwise. It is this latter aspect of freedom that is neither ruled in nor ruled out by your appeal to reasons as underlying behavior. It is quite possible that mental phenomena are, as you suggest, capable of determining real world events, of influencing them, even if in fact our “reasonable” actions are not free in the sense that we could do otherwise.
The causal relevance of mental states, the reality of reasons, in other words, is not in fact sufficient to give rise to blameworthy or praiseworthy behavior in the relevant sense here. Thus, the capacity to learn and reason-responsiveness are not sufficient conditions for making moral arguments that matter, for making praise and blame coherent concepts on a proper view of the world.
It may well be that the closest the world comes to the freedom necessary for a meaningful “morality” is the reason-responsive behavior that you describe. But that is qualitatively different from the notion of freedom that underlies moral praise and blame.
You may be suggesting that it’s possible to imagine freedom of the morally relevant sort as a kind of emergent property of sufficiently developed beings–so that today’s smartest computers, say, don’t have it, but tomorrow’s might. But that’s a contestible argument about reasons and freedom–the connection is not necessary as your post implies. I myself am inclined to that view; but I don’t think that our observation of reason-responsive behavior coupled with our observation of the capacity to learn is enough to rule out the idea that freedom of the will is in fact illusory.
And there are those who think that “emergent properties” talk is mumbo-jumbo (these folks tend to be the materialist reductionists that Ponnuru is addressing.)
You might also read the original Ponnuru post to see that he does not believe that after God, all things are possible. Perhaps you won’t find him such an annoying thinker after all.
July 19th, 2007 14:01
Gil, Will: Oh, absolute undetermined agency (the ability to have done otherwise) would not confer more serious responsibility on people than compatibilist agency; moreover, since incompatibilist agency isn’t even theoretically coherent, it’s not worth debating whether such an agency is possible. (But try telling that to Ponnuru.)
My point, however, was one’s emotions and intuitions strongly and repeatedly assert that it feels unfair to be punished and held accountable for actions over which we don’t have absolute, undetermined agency - even when we have compatibilist agency. Real, serious suffering is being dealt to people for actions they did do, but which they couldn’t have not done; the crime, guilt, and suffering of those who we punish had to happen.
Does that not make you feel a bit queasy? If that is our neat philosophical resolution to the issue of free will, I’m still unsatisfied.
July 19th, 2007 14:06
The right idea of “able” (or “power” or “can”) is the lack of certain relevant constraints, like coercion, brainwashing, intoxication, etc.
Good, but this becomes more subtle the more we begin to scientifically understand the causes of our thoughts.
Coercion we understand: we are pitting a person’s survival instinct/desire to avoid physical pain against their desire to not do something. Intoxication might quiet someone’s normal moral reasoning pathways.
Think of the brain as a machine that follows reasoning rules and then causes action. Now suppose someone does something bad, like kill someone. What is the cause of the action? The proximate cause is the perpetrator’s reasoning and decision to kill. But the brain is not a closed system, so the causes, if we trace them back, will come from a lot of other places: maybe it was hot out that day. The victim made the perpetrator feel very badly about something. The perpetrator had elevated stress levels due to lack of sleep or a stressful work situation. The lack of sleep may be because the perpetrator was not “taking care of himself.” And that failure to take care of himself has causes.
Ultimately what we find is that all humans are not prime-mover causal agents, but rather receptacles of the entire universe around them. We can make a model that focuses on the most “important” (most causally determinative) events and thoughts, and in that way we can distribute “raw causal blame” for an action across hundreds of antecedent events.
Now, “raw causal blame” isn’t “moral blame” because moral blame depends on things like being expected to foresee the effect of your action. But I think we need to reform our concept of “moral blame” anyway, on consequentialist grounds.
Ultimately, we should not make a priori commitments about whether the perpetrator or the “antecedent actors” should get punished for a murder. Ultimately the goal should be to reduce murders.
So a consequentialist legal system after 100 years of scientific advancement might look like this:
1) Suppose there’s been a murder, and we know the perpetrator.
2) Suppose according to an accurate model, we also know 100 antecedent events that substantially caused the murder (made it 90% likely according to the model, say). Perhaps “the perp’s parents giving birth to the perpetrator” is one of the events.
3) We would like to trace the causality back forever, but science can only tell us so much, so in reality we just get the 100 antecedent actions.
4) Some of the antecedent actions may have been intended to cause harm, and some may not. Presumably almost none were intended to cause a murder.
5) Penalize each actor (including the murderer) in whatever proportion minimizes the likelihood of future murders. Thus, you will probably want to penalize the murderer very highly, so that game-theoretically “murder” is generally a very unappealing option.
5.1) Penalize people who unintentionally contributed to the murder within the framework of 5), that is, in whatever proportion minimizes the likelihood of future murders.
In other words, the legal system should rely on a broader scientific model that predicts human behavior. Since this model will always be incomplete, we can’t trace actions back to the Big Bang, instead we will always have to penalize a set of “proximate” causes.
Right now, science isn’t very good at modeling human behavior, so we can’t trace the network of causes back much past the perpetrator’s own reasoning. In special cases like coercion we can (we have a pretty good model of how coercion affects human reasoning.) But science will evolve, and I think the consequentialist approach provides a much richer framework for assigning blame and achieving the real goals of minimizing crime.
In short,
1) Trace the causes of actions using a scientific model,
2) Consider all of the “perpetrators” of these antecedent actions to be candidates for punishment,
3) Assess punishments using a set of rules,
4) The set of rules is endogenous to the scientific model in 1), because of deterrence effects.
I have a feeling that comment needed some more editing than I gave it
July 19th, 2007 14:08
‘The right idea of “able” (or “power” or “can”) is the lack of certain relevant constraints, like coercion, brainwashing, intoxication, etc. These sorts of things can cripple the free exercise of agency. But that’s rather different from the fundamental nature of the universe ruling it out.’
How can you ever rule out coercion and brainwashing? Continued existance is submission to coercive necessities of survival. All of observed experience is brainwashing. Just because we can conceive of much more horrible results of coercion and brainwashing does not mean that it hasn’t happened to us.
July 19th, 2007 14:55
“retarded ponens/tollens showdown”: definitely the phrase of the month, if not the year
Quite right that if you know that p and q (where p is [that free will exists] and q is [that physicalism] in whatever iteration of physicalism you choose), it takes a certain willful stupefaction to worry about ~q. If there’s free will and physicalism is true, then free will facts are supervenient on/dependent on/reducible to (again, take your pick) physical facts. No great mystery.
I’m not convinced that we know free will exists. I’m sure you’ve seen this before, but here goes:
A) There is free will.
B) If the world is deterministic there isn’t free will.
C) If the world is indeterministic there isn’t free will.
D) The world is either deterministic or indeterministic.
Which one’s wrong.
July 19th, 2007 15:04
Either B & C or B or C. Either the whole issue of in/determinism is irrelvant, or whichever one is true of the world is a necessary condition for free will, and probably everything else.
July 19th, 2007 15:18
Here are two things you know: free will exists (it is obvious: go ahead, touch your nose) and the universe is made of whatever it is made of (obvious, if anything is). Therefore, you know the conjunction of those two things. Therefore, you know that the crazy proposition that says that one of them must be false isn’t true
This is silly. No one has ever denied the tautology that “the universe is made of what it is made of.” What Ponnuru is saying is this: “There exists something outside the universe and its collection of particles, and this something is what makes rationality and free will possible.” You can’t possibly disprove that argument merely by making the absurdly obvious point that the universe is made of whatever it is made of. That doesn’t even lay a finger on what Ponnuru was saying.
July 19th, 2007 15:25
cg,
No, it doesn’t make me feel queasy.
We don’t punish atoms, we punish people. And people are patterns of symbols within brains that make choices. They are (usually) responsible for those choices because of the patterns that they are.
It just doesn’t matter whether the underlying atoms (both within the brain, and everywhere else in the universe that could affect it) could theoretically be predicted to lead to those patterns. It’s the theories and reasons that we care about, and that’s what causes the behavior in the only morally relevant sense. So, everything is ok. You should be satisfied.
This notion of “Couldn’t have done otherwise” is a confusing irrelevance.
July 19th, 2007 15:44
What? “People are patterns of symbols within brains that make choices”? There is a remarkable level of abstraction at work here.
Imagine yourself suffering the excruciating punishments which are more or less deemed defensible by good, tolerant liberals: a lifetime in prison, say. Imagine suffering those punishments for something that, because of amazingly distant and inscrutable workings in the past (the original composition of the universe, even), had to happen. You came into existence, and you suffered this profound punishment, and that was your fate.
It’s a miserable tragedy when it comes as a result of earthquakes and hurricanes; isn’t it something even worse when it comes as a result of men?
mk’s hyper-consequentialism is the most palatable offering yet, but it’s also a fantasy.
July 19th, 2007 15:57
What is it about morality, exactly, that is supposed to become impossible if free will doesn’t exist?
Consider that “social system which works well” or “economic system which works well” have some general similarities to “good moral system” in many people’s minds. (Objectivists, as I understand it, profess to nearly equate them, but I don’t think you need to go that far to admit that there’s some resemblance.) Even in a perfectly deterministic world it’s perfectly reasonable for borders between markets and command economies to end up with different outcomes on the two sides. Similarly, it seems to me, even in a deterministic world immoral actions can lead to bad consequences. Perhaps determinism confounds some people’s intuitions about punishment or revenge or other aspects of justice. But I think even if one had to back off from concerns about whether revenge is emotionally satisfying or not, and could only concentrate on good or bad consequences, it wouldn’t obviously doom the philosophical project of morality.
(I’m no philosopher, though.)
July 19th, 2007 16:12
“What is it about morality, exactly, that is supposed to become impossible if free will doesn’t exist?”
Free will allows us congratulatory self-importance for our moral behaviour. If it isn’t a free willed act, working hard to feed your kids makes you a chump.
July 19th, 2007 16:24
But, cg…
“You came into existence, and you suffered this profound punishment, and that was your fate.
It’s a miserable tragedy when it comes as a result of earthquakes and hurricanes; isn’t it something even worse when it comes as a result of men?”
Why is it worse? Why do you think the behavior of men is different from that of earthquakes and hurricanes? Aren’t they equally inevitable? Isn’t it a waste of time for you to even pretend to be thinking about this?
If not, why not?
July 19th, 2007 16:37
Why is it worse? Why do you think the behavior of men is different from that of earthquakes and hurricanes? Aren’t they equally inevitable?
Since human beings are not causally closed, the ultimate events that cause human actions are non-human. (Goes back to the big bang).
But, I contend that those causes are irrelevant to us because inanimate objects do not respond to incentives.
Humans do respond to incentives. Thus, even though they are not the prime causers, you can still improve society by setting up the best possible incentive system, because that will affect the likelihood of future good/bad events happening.
(There are lots of other ways to improve human society besides “set up a better incentive system”, but this is one way.)
The best possible system of justice is that which minimizes the likelihood of crimes happening. (Slightly less roughly, “minimizes the aggregate loss due to crime”, or something)
July 19th, 2007 16:50
Ok, let’s start with the easy one first:
Njorl (12.14): Yes, and we call those people imprudent. Or tourists.
Bill and cg make a reasonable point about ‘could have done otherwise,’ but I am not convinced. Let me explain why. But first, Bill, I never said I found Ponnuru an annoying thinker. What I said was that the argument he made drives me crazy. He was not the first to make it, and he won’t be the last. The argument is what concerns me. I have no opinion one way or the other of Ponnuru as a thinker, for I seldom read what he writes. Ponnuru does, however, claim that physicalism makes morality incoherent; which is a skyhook analogous to Dostoevsky’s argument, but with a different hook and thus a different bogeyman.
Bill’s claim, if I’ve read him right is:
An action is blameworthy (praiseworthy) only if the action is wrong (right) and if the agent could have done otherwise.
The first point in response is this: only agents are praise- or blameworthy. Actions can be right or wrong, good or bad, prudent or imprudent. We blame/praise agents for actions. Praise and blame attach to agents in virtue of their actions, because actions are manifestations of the agents’ dispositions/character.
So your claim could now be: An agent is blameworthy only if her action is wrong and she could have done otherwise.
This seems right, provided we locate the proper sense of CDO. Surely there are cases (which cg points to) where an agent could not have done otherwise, and we exempt the agents from blame owing to extenuating circumstances or the fact that there was nothing they could have done.
Here’s why I think reason-responsiveness and capacity to learn give us the right sense of CDO.
An agent’s reason-responsiveness is the agent’s disposition to make some fact a reason for action. Dispositions can be thought of as probabilities. So, an agent’s disposition is the probability that some fact will constitute or comprise a reason for action for some agent. An agent is more or less reason-responsive depending on how high or low the probability is that she will make some fact a reason for action. Probabilities can, of course, shift according to circumstance. I am more likely to say something unkind to someone who’s being an ass if I have not had sufficient sleep, or if the person has a history of being an ass (and I have had to keep my mouth shut several times before). What I end up saying may or may not be excusable, depending on the severity of the circumstances (how much of an ass, how little sleep, how unkind were my words, etc.). If the probability that I murder the ass is so easily affected, however, there is something wrong with me. I should be said to have an inferior set of dispositions, I would be a bad person.
The capacity to learn is the capacity for these probabilities to change over time depending on the information and feedback the agent receives. Treating agency, along with consciousness, as a process, we have to realize that it is a non-ergodic process. The structures of individual agency change constantly, there is no guarantee that any state will recur, and there is no clear connection between our beginning state and our present state. The non-ergodicity of the process and structures of agency negate the possibility of any sort of reductive determinism (by which I mean genetic or cultural determinism). An agent’s dispositions/character/structures of agency are the product of her unique learning history (including her actions and the feedback received) as that has acted on and transformed an original, underdetermined, set of genetic predispositions.
In a sense, one’s present set of dispositions is noncaused. This does not mean that there are no sufficient causes for the present state of the structures of agency. Rather, it means that there is nothing (or no set of things) we could isolate by means of subjunctive conditionals (counterfactuals) as the necessary cause(s) of one’s present state. Not, that is, without jeopardizing personal identity. That is, we might say ‘if only I had been born to rich parents (or, French), then I would be a much better person.’ Chances are (depending on remoteness of the possible world in which the antecedent of the conditional is true) that the referent of ‘I’ in the consequent of the conditional could not in any sense be said to be you.
A problem arises when we ask ‘are you responsible for your character/dispositions?’ Trying to answer this question threatens an infinite regress which seems to threaten the very idea of responsibility. mk intimates something along these lines with the argument that responsibility has to be distributed. Now, in situations where very specific quirks of personality can be attributed to very specific past events, this might work. (Think of, say, Bradley Whitford’s character in The West Wing, and the way music would send him into a panic, and how this was a result of PTSD following the attempted assassination.) Failing this, the point is rather that you are your character and dispositions, at least for the purposes of moral evaluation and navigation of the social world. What you are depends on what context we are considering (a doctor treats you as a machine that is either functioning correctly or incorrectly, and tries to figure out how to fix the machine; the government treats you as a vampire bat treats its prey, wanting nothing else than dollars and votes; etc.). And for the purposes of moral praise and blame, you are your character and dispositions and the actions that are manifestations thereof.
What has this to do with CDO? Well, it is likely that ‘could he have done otherwise,’ cannot mean ‘was it possible, at the very moment of action, that he could have done B instead of A?’ Such is impossible in a deterministic world, and in an indeterministic world this possibility leaves us with uncaused actions, which agents cannot be held accountable for. Instead, ‘could he have done otherwise,’ means (something like) ‘in light of his dispositions, is there a nearby possible world in which he did otherwise?’ Is it reasonable to believe that, had something been minutely different, the agent would have done something different.
Suppose (and now I get to cg’s point), a man is driving a car down a residential street and runs over a child. The child ran into the street at the last minute. Could the man have done otherwise?
Well, let’s consider some possibilities:
1. If the man had taken a different route, he would not have run over the child.
To make the antecedent true, we have to know why the man would have taken a different route. If the route was his normal one, the only reason for him not to have taken it (in the absence of, say, construction) would be that he knew he would end up running over the child. The man is not omniscient, and there is no way for him (or anyone else) to know the child would run out at that time. This scenario does not interest us.
2. If the man had a faster reaction time, say a reaction time the speed of a cockroach (1/30th of a second) instead of his slow human reaction time (1/3rd of a second), he would not have run over the child.
This clearly does not interest us. Unless, of course, the man knows he has an unusually slow reaction time, in which case it is incumbent on him to drive more slowly than would other drivers, or to avoid driving altogether, as he poses an unacceptable risk to other motorists and pedestrians. Further, there is very little anyone can do to tighten up his reaction time.
3. If the man had been driving slower, he would not have run over the child.
How responsible the man is depends, in this case, on how fast he was driving. If he was doing 50 in a 25, he is surely culpable for having run over the child. Had he been driving the speed limit, we would think him less culpable. If he was driving the speed limit and knew (as it was his normal route) that children often played in or near the street during that time of day, it may have been incumbent on him to have been driving below the speed limit. If he knows the street, it is reasonable to expect him to take precautions. Does he always speed on residential streets, thus inviting a tragedy? Or was his wife whose water had just broken in the car, and this was the only reason he was driving so fast?
4. If the man had been paying attention, he would not have run over the child.
Was the man lighting a cigarette, or fiddling with the radio, or eating, or talking on his cell phone? The less he was attending to driving, the more culpable he is. Is he the kind of person who does not pay much attention to driving, or was this a fluke owing to the fact that he had just spilled hot coffee in his lap?
The point here is that we ask about people’s dispositions, wanting to know whether they are ticking time bombs or whether their actions were fluky results of extenuating circumstances. These questions are independent of determinism, and questions about what count as extenuating circumstances, thus exempting or excusing individuals to what extent from praise or blame, are political questions instead of metaphysical.
Further, if the man had been paying attention, driving (below) the speed limit, and there really was nothing he (or anyone else) could have done to prevent his running over the child, he will (as cg notes) feel guilty. The feeling of guilt does not signal some intuition about in/determinism. Rather, it signals that we acknowledge that our actions have consequences, and that our responsibility extends as far out as our causal relevance. The man feels guilt because he has, even non-intentionally, put a harm on the scales (they need not be moral scales; most scales aren’t). Even Austin’s man who shoots his neighbor’s goat by accident knows that he owes the neighbor a new goat. Focusing on punishment and retribution (how much of a harm should we visit on the agent), rather than on compensation and restitution (how much of a benefit is the patient owed) distracts our moral attention. Both punishment and restitution are means to the end of balancing the scales, but one repays a harm with a harm ignoring the harmed, while one restores a harm with a benefit with our attention where it should be (and who really gives a damn about the harmer, unless he is so reckless that he constitutes a permanent unacceptable danger to others).
One last point about guilt is that guilt is often associated with a searching for what one could have done differently. Guilt prompts us to the examination of possible worlds I explored above. This is a good thing, for this examination can lead to learning, and can alter one’s dispositions, making one less likely to, say, speed in the future, and thus reducing the possibility that any more children become road pancakes.
Will, sorry to have posted such a long thing on your blog. I’ll start my own soon enough, so I don’t take up your space.
July 19th, 2007 18:16
“causes responses in a way that’s strictly deterministic (with no actual freedom of will in the sense of non-determinism).”
My take is that at very short scales of time and distance, it’s deterministic. But in the timescales on which humans operate, the mind has plenty of opportunities to weigh the various deterministic impulses that pop up, and different activation patterns will come into play to reinforce or suppress things, so that by the time something happens outside the brain it represents the volitional free-will choice of the person.
It’s not like a single neurotransmitter in a single synapse that activates a neuron is all it takes to produce a behavior. There are plenty of yea and nay votes in the brain before something happens.
July 19th, 2007 19:04
I’d love to see Will and Ramesh debate this stuff. (AFF event?) My head might explode, but at least my final moments would be intellectually stimulating.
July 19th, 2007 22:45
McNeel,
I think you are having trouble distinguishing arguments about what you think is the fact of the matter about the universe (a descriptive account of human agency and, if it exists, moral agency (praiseworthiness/blameworthiness)) and what Ponnuru is trying to do, which is contrast some people’s vision of the way things are with traditional formulations of moral agency.
You say that his argument doesn’t work because your descriptive view is correct. But your descriptive view, which may well be correct (more or less) does not save morality in the old fashioned sense. That may be no loss, and obviously one ought not to lament what reality will not abide–if we have to abandon old notions of morality then so be it.
Ponnuru does not claim that no kind of moral reasoning is possible without God, for one thing. He posits that a certain kind of strict reductionism (which you do not appear to hold) is incompatible with a certain traditional view of moral agency (which you also do not seem to share). Your answer is somehow that Ponnuru is wrong because he misunderstands that some meaningful account of some sort of morality is compatible with your not-entirely deterministic account of morality.
Everything you have written can be true and it would leave Ponnuru’s argument untouched.
Meanwhile, you have done much more than Will, who seemed to think that you could prove the mutual coherence of morality and reductionist physicalism by proving that one can touch one’s nose and that the universe, after all, only contains what it contains. This was enough for Matt Yglesias and Andrew Sullivan to swoon, which is disappointing enough. It’s not quite as bad as Hitchens trying to prove that atheism and morality are consonant because atheists can behave morally, but it’s not much better. (Mind you, atheism and morality may well be consonant, but the fact that there are moral atheists wouldn’t prove it, any more than the fact that lots of scientists believe in God proves that God and science are consonant.)
Will, after all your study of philosophy, do you really think that “touch your nose” constitutes a proof of free will? Or that the claim that “the universe is whatever it is” is either equivalent to or (worse yet) a proof of reductionist materialism? (Recognize that reductionist materialism is not generally regarded as pleonastic.)
Cheers.
July 19th, 2007 22:54
mk and William Newman make similar points, worth addressing. If you’re view of the universe is correct, then yes, you can posit “better and worse” outcomes for people, and if you want to call that morality then feel free. Just recognize that it’s at best a shadow of the conception of morality that has generally dominated western discourse. (Again, if it’s the best account, then so be it–it’s just not the same account).
cg made the point well with natural disasters: people are better off as a matter of states of affairs when they are not covered in lava, but when a volcano erupts we don’t say “Bad Volcano!” If it turns out that the best accounts of human agency suggest that, though suggestible and able to learn, we are more like the volcano than like the free-willed beings we have traditionally imagined ourselves to be, then we will have moved away from what Ponnuru, and most people who currently employ moral vocabularies, mean by the term “morality.”
It doesn’t mean that we will have lost all evaluative capacity–we can say whether a tree is healthy or not, and that a storm caused a fire that made a forest far less healthy, but we would not call the conditions that made the tree healthy morally praiseworthy, nor the storm that destroyed the forest wicked.
And with that run-on, I’m sure I have overstayed my welcome.
July 19th, 2007 23:27
Gil: Now it seems you’re falling onto fatalism. Our actions are inevitable, yes, but they are still a result of our rational decisions; you seem to be saying we shouldn’t even consider how to achieve the best possible outcome, and I know you don’t believe that.
I think my point still stands: there is something fucked up about a world in which people are forced to suffer for things they couldn’t have avoided. What’s fucked up about this? It is making us agents of (what you seem to admit is, in the case of hurricanes) an unfair system: a system in which people are born, are causally lead to a certain predetermined outcome, and suffer greatly as a result of their original, unchosen conditions.
There is another, seemingly more fair, choice - we can not punish criminals! But that’s not really a choice, and that’s the issue.
July 20th, 2007 00:11
cg,
Right, I don’t believe it. I was putting on my cg hat to point out the contradiction your statement revealed.
You know that it’s worse when people behave unjustly, because you expect them to choose responsibly. You know that people, unlike earthquakes and hurricanes, can think about, and choose “How to achieve the best possible outcome.”
And that’s what matters, right?
July 20th, 2007 08:30
Bill,
I get iffy around claims to represent ‘traditional morality.’ I don’t know whether this means the peculiar institution of rules and interdicts, or whether we’re talking about sets of virtues, or what. In any case, claims about what ‘traditional morality’ is are never more than second-order beliefs. And we are not very good at forming second-order beliefs.
My argument was intended to sketch a conception of agency which is consonant with our first-order moral/ethical beliefs and practices. And my claim was that at this level, our traditional beliefs and practices are very much independent of determinism. The truth of determinism (and I don’t know if determinism is true) may shatter some second-order moral beliefs, but it does not touch moral practice and first-order moral beliefs. If this is right, then Ponnuru is wrong to say that physicalism or reductionism is incompatible with actual moral practice. (I do think eliminativism does, but that’s a different kettle of fish.)
The only place I diverge greatly from traditional morality is in rejecting the revenge impulse to re-balance the moral scales downward. I prefer a compensatory view of justice which would balance the scales by leveling up as best as possible the state of the victim. But this is at best a side issue.
In any case, I think you’re right about over-staying, but I would be glad to continue the discussion elsewhere. Via electronic mail, I can be reached at kyle (my last name is the handle I’m using) the mail service run by google.
July 20th, 2007 11:26
Anyone else amused that the proof of free will is to follow a command?
July 20th, 2007 12:08
It seems ironic, but it’s not the following of the command; it’s the deliberation about whether or not to follow it, and then acting on that decision.
July 20th, 2007 12:50
McNeel,
All that I meant by traditional morality above was what we have traditionally thought we meant when employing terms like right, wrong, ought, etc.. To wit, we traditionally had a notion of a robustly free agent who could have done otherwise and whose action therefore merits praise or blame on a moral basis (as opposed to for its technical skill, say).
I was in no way referring to the content of the moral scheme, just to its presumptions about human freedom. My very simple point is that if human freedom to do otherwise is in fact not clear, or if freedom amounts only to the capacity to learn and the ability to respond to reasons then that is not likely (certainly not necessarily) a sufficient freedom to sustain the old meanings behind obligation, rightness, or wrongness.
Cheers.
July 20th, 2007 12:56
Don’t anyone worry about overstaying your welcome. I love it!
July 20th, 2007 13:18
Re: a consequentialist moral/justice system..
The scenario I posed (use science to model/predict human behavior, use this model to assess punishment to maximize expected utility) is indeed fanciful by today’s standards.
But we can use the same principles to set up a consequentialist justice system, even today. If we don’t know most of the causes, oh well, that means that the punishment is mostly focused on the perpetrator (that’s already the case today). It would still change the game to focus explicitly on minimizing the number of crimes*, rather than on some more abstract notion of morality.
Then, as scientific capacity to accurately model human behavior ramps up, our justice system won’t be faced with a “human deliberation is a result of external causes” crisis– instead, we’ll have a mechanism in place to assign blame and punishment to the external causes (the human ones, anyway).
(* “Minimizing the # crimes” is too specific. You have to take into account the reduction in utility from the penalties themselves– otherwise simply executing all criminals might seem the best policy)
July 20th, 2007 13:33
Will,
I hate to rain on your parade because I fear doing so will feed the trolls (Sorry, but I’ve just called anyone who carries on to the effect that freedom, goodness, etc… requires God a troll), but there’s something epistemically questionable about what you’re suggesting.
Let’s suppose you know we have freedom on the basis of e. You then learn that the universe is made of what it’s made of. However, we’re assuming this isn’t a tautology b/c otherwise you knew it all along. Rather, we learn what the universe is made of. Surely what you learn, when combined with e could in principle lead to lowering the probability that you had free will even if you initially knew you had free will on the basis of e alone.
To deny this is a possibility out of hand, it seems you’d have to deny that you could have knowledge on the basis of uncertain evidence. Surely, however, we don’t have certain evidence that we’re free, in which case even if you denied that we could have knowledge on less than certain grounds, we’d not know we’re free.
July 20th, 2007 16:15
“There is another, seemingly more fair, choice - we can not punish criminals! But that’s not really a choice, and that’s the issue.”
Even if it were possible to perfectly rehabilitate wrongdoers without punishing them, and to perfectly ascertain the causitive factors that led to the wrong doing, society could still conceivably be served by punishing the wrongdoer for their actions. Punishment would still act as a deterrent. Otherwise, the anti-social man of the far future could think, “OK, I’ll kill my wife, take the rehabilitation pill and marry my mistress. That’s a good deal for me!”
Until immorality vanishes, acting as if there is free will even if we do not accept it’s existance is the preferable course.
July 20th, 2007 16:59
Njorl:
Until immorality vanishes, acting as if there is free will even if we do not accept it’s existance is the preferable course.
I would put it differently. There are no uncaused decisions, which depending on your definition may mean there’s no free will. One of the “causes” (or “influences”) of a decision is the surrounding incentive structure. So, we should set up the incentive structure to minimize people doing bad things. As you say, punishment acts as a deterrent. It does this, in fact, because there is no free will.
We tend to see coercion as a “subversion of free will” because we clearly observe a causal mechanism influencing the decision. But in fact there is always some causal mechanism that influences the decision, it’s just usually much murkier to us than coercion.
Society’s role should be to embrace this (let’s be happy that human actions are not uncaused) and design incentive structures to minimize bad things happening. This social engineering literally is morality, it is not that we should pine for the days when humans will never have incentive to do bad things to each other.
July 22nd, 2007 13:14
Isn’t Ponnuru making a go at a reductio here? He seems to be arguing that there is free will and thus moral responsibility (this, as Will says, seems obvious). But some of the people who accept free will and morality are physicalists. This is a problem, Ponnuru says, because physicalism (as Ponnuru understands it) entails a repudiation of the notion of free will (as Ponnuru understands it). The problem with this doesn’t seem to me to be a logical one, as Will suggests, but a conceptual one (i.e., with Ponnuru’s understanding of free will, causality, and the relationship between the two).
August 20th, 2007 10:52
Taking Ockham’s Razor straight to Ramesh’s jugular…OW.