Libertarianism and Laws

I've been surprised to find several writers say that Libertarianism is committed to the view that undetermined free actions are not covered by laws of nature. In some cases, the writer is a Libertarian; in others, the writer thinks that this view is, unfortunately for Libertarians, something to which they are committed. I haven't found much of an argument for thinking that Libertarianism entails any such commitment.

All of the writers I have in mind take Libertarianism to be at least metaphysically possible. Thus, they think there is some possible solution to the problem of luck. Perhaps the idea is that a solution to that problem requires that undetermined free actions be lawless. But why think that?

Any ideas? Also, has anyone else seen much discussion of this question?

Comments

[David Hodgson]

I am one person who says that free actions are not determined by the engagement of laws of nature with prior conditions. That is not saying that free actions are not covered by laws of nature, just that laws of nature do not uniquely determine the outcomes in the case of free actions. I would not have thought that position required much argument to support it, but I'm not sure if that is what Randolph is asking about.

[David Hodgson]

Two additional comments

First, I did not intend to beg the question about compatibilism. I was answering a question about what a Libertarian is committed to.

Second, I do argue that there is something about free actions that contributes to outcomes, with which laws do not engage, namely the person's response to whole feature-rich gestalts of conscious experience. I do not say this is something to which Libertarians are committed, but I do say it makes Libertarianism more plausible. I argue at length for this position in a number of articles since 2000 that are on my website, most succinctly in the 2007 TLS article.

[R. Clarke]

Here's an example of what I have in mind:

"First, libertarians must reject the possibility of an all-encompassing psychology. Human behavior would be governed by the laws of such a sceince, and libertarians deny that human behavior is controlled by any laws. ...Libertarians must also reject the possibility of an all-encompassing _physics_. ...According to libertarians, if physicists turned their measuring instruments on the sub-atomic particles composing a free person, formerly observed patterns would break down."

Sider, "Free Will and Determinism," in RIDDLES OF EXISTENCE (p. 121)

[Kip]

The people that you mention appear to regard laws as meaning: "laws that specify a single, unique future."

Indeterminate laws are still laws, I suppose. Consider the "law" that says:

The future can be A, B or C, but not D. Which one will it be? Heck if I know.

I suppose that is still a law. It rules out D, for example. But as the set including "A, B, C..." continues to grow, it stops looking like a law, and starts looking like chaos.

[Andrei Buckareff]

Of course, if one is a Humean about laws of nature I suppose there is not much of a problem. My guess is that Sider has governing conceptions of laws (like Armstrong's account) in mind. I have never thought of libertarianism as implying anomolous decision-making. Even a governing conception of laws can be characterized probabilistically.

For what it's worth, I share your puzzlement, Randy. I would think that removing laws from the domain of free decision-making only amplifies the problem of luck.

[Mark Smeltzer]

Randy,

Some Libertarians are dualists, right? As such, they may believe that the free-action-producing-substance is governed by laws, but that those laws are not physical laws.

These folks might feel comfortable talking about the laws of nature as the conjunction of the physical and the non-physical laws, but when talking with people who equate the laws of nature and the physical laws, they might say that freedom enables agents to act contrary to the laws of nature.

Van Inwagen, on the other hand, would have reason to assent to Sider's remarks, but maintain the idea that freedom operates within the bounds of options that are metaphysical open, according to the laws of nature.

[R. Clarke]

The view I have in mind takes Libertarianism to preclude even non-deterministic (e.g., probabilistic) laws governing free actions. That's what I find puzzling about it.

[Jeremy Dickinson]

Prof Clarke,

I've read the Sider piece you've quoted, and like you I found myself puzzled by the commitments he's tagged onto libertarianism. Straightforwardly, it seems to rule out event-causal libertarian views like yours and Kane's. Perhaps he is assuming that there is something metaphysically impossible about event-causal libertarianism for reasons similar to the ones offered by van Inwagen (and many others): undetermined events are random events, and random events are not freedom conferring events. Hence, there can be no solution to the Luck Problem. (This would hold for both the agent-causal and the event-causal theorist.) And, of course, Sider doesn't go mysterion like PvI; he goes compatibilist. So the upshot here is that I think your (reasonable) assumption that Sider thinks libertarianism to be metaphysically possible is probably not right.

Any thoughts? Thanks for the post.

[Chris Franklin]

I have been wondering about this myself I have been wondering about this myself Randy. In reading physicists or philosophers of physics (David Albert comes to mind, but I think Jeff Barrett also says something to this effect in his review of Hodgson (1999)), they will often say that quantum mechanics is no friend to free will since QM still requires that events be law governed. Now perhaps QM is no friend to free will but I assume it is not for *this* reason.

But more to your question, Chisholm claims in ‘Human Freedom and the Self’ that there can be no science of man by which he means free actions cannot be explained by subsuming them under laws. He assumes that laws relate events as cause to effect and as he argued above that free actions cannot be caused by events, it follows that free actions cannot be law governed, at least in sense that they cannot be explained by laws.

Doesn’t Pereboom argue in chapter 3 of LWFW that agent-causal libertarianism is incompatible with all events being law governed? I cannot remember if this is the exact conclusion he draws, but he definitely worries about the compatibility of agent-causal libertarianism with a view of the world as wholly law governed. He gives the “wild coincidence argument.” Although this argument is much cleaner against someone like Kant who thought the laws are deterministic, Pereboom thinks it can apply to those who think the laws are merely statistical. It goes something like this: if every event is governed by a statistical law, then certain objective probabilities will attach to the choices we make. But, if we are agent-causes, then we could falsify the laws that assign these probabilities: we could cause statistical deviations from the laws. Consequently, it would be a wild coincidence, given that we have agent-causal powers, that we never in fact act in ways that cause such deviations. Obviously this is a crude presentation of Pereboom’s argument, but I think this is the main idea.

Both of these arguments have agent-causal libertarianism in view, rather than event-causal libertarianism and so neither of them are intended to show that libertarianism per se entails that not every event is law governed, but perhaps they still explain why people sometimes attribute this view to libertarians.

[Ben]

Here's a reason to think event-causal libertarians cannot hold that actions can come under laws in any sense. Laws are either deterministic or indeterministic. Obviously the EC Libertarian doesn't want deterministic laws (a point driving van Inwagen's consequence argument), but nor should they want indeterministic laws, the point of the Mind argument. I think van Inwagen is right to suggest that the Mind argument is (in his words) "unanswerable". So what's causing the problem for the EC libertarian? It seems that any view that takes laws of nature not to be determined by the actions of agents is going to be the wrong conception EC libertarians want insofar as they're libertarians. If it is an Armstrong conception, the laws will govern them; if it's a Humean conception, the laws are just regularities. In either case, the agents aren't doing anything--they're just along for the ride really.

Why think the libertarian (NB, not EC), needs such a strong sense of sourcehood? I think it's because without being a source of control such that the laws don't just "make" you do what you're doing, they don't have a distinctive position from compatibilism; their view is trivialized without a strong kind of origination. If the EC libertarian is saying that moral responsibility is compatible with laws of nature in Armstrong's sense or Hume-style conceptions, what makes their view interestingly different? What is the agent doing? I can't see what she's doing other than instantiating a law of nature, and surely that's not libertarian agency worth the name.

[R. Clarke]

Chris, it does seem that Pereboom's empirical argument against free will includes the claim that free decisions would not be covered by even probabilistic laws concerning which events follow which others. For the argument (I think) is that we don't find deviations from expected patterns (distributions) that would almost certainly exist if we had free will.

[Bob Doyle]

The Laws of Nature are all fundamentally probabilistic when viewed from the atomic or molecular level.

But the number of molecules in most every object of interest to human beings is so large that the statistical fluctuations average out and we have an "adequate" determinism.

Newton's Laws of Motion are accurate enough to send men to the moon and back, but their apparent strict determinism is only an approximation in the limit of large numbers of atoms.

Consequently there is some microscopic randomness in all processes and events.

The standard argument against free will is that if we are not determined, then we are indetermined and randomness can not be what we want from freedom.

But many thinkers have suggested a third option, a two-stage model with both some randomness and some "adequate" determinism.

Think of it as some "free" and some "will."

We might randomly generate alternative possibilities for action, e.g., by accessing a random possibility generator somewhere in our unconscious. Our evaluation and decision process then selects from those possibilities with an adequately determined process that can reflect our character, our values, and our current needs and desires.

The Laws of Nature that Randy and Sider are discussing do not preclude this model of free will.

Of course it is not the metaphysical freedom that some dualists have looked for. It may be too determined for many Libertarians.

But many Compatibilists might be accept compatibility with an "adequate" determinism that allows us to be free and creative enough to bring new information into the universe that was not determined from a time before we were born (PvI consequence argument).

For more on adequate determinism, see
informationphilosopher.com/freedom/adequate_determinism.html

[Eddy Nahmias]

I think most scientists who think about free will, certainly the many who think a scientific approach to human cognition and volition inevitably conflicts with free will, think in the way Randy is asking about. As such, they are not properly concerned with determinism as philosophers understand it. They are really concerned with the idea that human behavior is lawlike (covered by laws), and if some of these laws are indeterministic, that won't change anything. They tend to think free will requires dualism. Here's just one quote of many I could offer that suggest this interpretation, Benjamin Libet writing:

"But we have not answered the question of whether our consciously willed acts are fully determined by natural laws that govern the activities of nerve cells in the brain, or whether conscious decisions can proceed to some degree independently of natural determinism.... Quantum mechanics forces us to deal with probabilities rather than with certainties of events.... [but] they might nevertheless be in accord with natural laws and therefore determined." (1999, p. 55)

I have a related question. Does anyone think that standard incompatibilist arguments (e.g., Consequence argument) can be formulated using "near-enough determinism" or what Bob calls "adequate determinism" (I put the "determinism" inside quote marks too, since these are simply NOT equivalent to determinism as philosophers define it)? Or do they then become skeptical arguments, like G. Strawson's and Pereboom's?

[Mark Smeltzer]

Eddy,

I think it is possible. The consequence argument is a species an argument for agential impotency. These arguments tell a story that, for various reasons, agents are impotent actors and that they are never/rarely the causes of the effects we observe.

Arguments for impotency work in this way:

1. If nothing in event P's causal history was caused by agent S, then P cannot be an attributed to S.
2. Due to agential impotency argument Q, agent S is impotent to cause any events.
3. Therefore, skipping a few easy premises to connect (2) to (1), P cannot be attributed to S.

If all events can be reduced to the governance of elementary particles according to natural laws, whether probabilistic or deterministic, then a reductionist argument could be made that yields agential impotency (viz., agents are reducible to elementary particles, and those particles are not free).

This is more or less the same thing that's happening in the consequence argument. The consequence argument tells a reductionist story that, in a deterministic world, agents are impotent actors due to a reductionist account of all events to the deterministic laws.

In these kinds of stories, the natural question is, "Where did the agent go?" So, another way of framing it is that these kinds of stories attempt to subtract agency from the world while still explaining the events in the world.

[Bob Doyle]

Hi Eddy,

As you know, I am not one of those scientists who think science conflicts with free will.

However, neither do I find recent neuroscientific evidence for free will the least bit convincing. We are not going to see any specific spot in the brain that is the locus of freedom.

I am really hopeful that philosophers will think this problem through again, and am trying to become one (joined the APA, etc.) to help.

I strongly second your suggestion that we rethink the standard arguments (PvI's Consequence and Mind arguments, G. Strawson's Basic Argument, etc.) from the standpoint of the merely "adequate determinism" the world offers us.

Although "adequate determinism" implies that - logically speaking - indeterminism is "true," it does not mean that randomness is rampant. On the contrary, it plays no part in the behavior of most macroscopic objects.

But randomness plays a vital role in creativity - both in nature and in our minds. Without it, nothing new, no new information can be created. Determinism is "information-preserving." If you know the past, you know the future, etc.

I have recently had an exchange on this subject with Galen Strawson, who as you know strongly opposes the idea of a causa sui.

I have asked Galen for permission to post the exchange the Garden and he says that that will be OK.

[Eddy Nahmias]

Just to be clear, I was not really trying to suggest anything here. I do think any indeterminism suggested by quantum physics does not help secure free will. But I also think that it is unlikely that such indeterminism (if it exists at micro level) plays no role at the macro level. That's why I think we should avoid any talk of "adequate determinism." If indeterminism is true even once in a while, determinism is false!

Let's clearly distinguish determinism from other potential threats to free will, such as reductionism (perhaps as Mark's reformulated argument uses), or the thesis that all events including human actions are covered under natural laws (as Randy asked about), or fatalism, etc.

If "incompatibilist" arguments like the Consequence argument work using "mostly determinism" [sic], then why are they incompatibilist arguments properly understood, rather than arguments for the incompatibility of free will and X, where X is some thesis *other than* determinism?

Having said all this, I would speculate that before quantum theory, indeterminism was likely understood by most to mean "not covered by laws of nature," and perhaps before Kane, most people thought that libertarianism required indeterminism of *this* sort. Does that sound plausible?

[Barry Loewer]

Readers of this blog might be interested in a paper I wrote years ago -"Freedom From Physics" in Philosophical Topics 1997 -concerning whether quantum theory provides any support for libertarian free will. The short answer is that it doesn’t. One of the arguments in the paper is a probabilistic variant of the Taylor- van Inwagen argument which may be the sort of argument concerning “near determinism” that Eddy is asking about. By the way, it is not correct that quantum indeterminism "plays no role" at the macro level. It is everywhere.
Regarding the issue of whether Humeanism about laws (e.g. David Lewis’ BSA) provides a better climate for libertarian free will note that the incompatabilist arguments (van Inwagen’s and my probabilisitic variant) don't assume non-Humeanism about laws.

[Bob Doyle]

Hi Eddy,

I agree that if just one single event is determined by chance, then indeterminism would be true, and therefore it is "true."

Some philosophers have said that real chance would undermine the very possibility of certain knowledge.

Some go to the extreme of saying that chance makes the state of the world totally independent of any earlier states, which is nonsense, but it shows how anxious they are about chance.

Regarding indeterminism before quantum theory, the debate was about the nature of probability, whether chance was the result of human ignorance, and thus merely epistemic, or ontological.

Probability was introduced as an acceptable euphemism for chance in the eighteenth century, by Pierre-Simon Laplace, for example, who called his work the "calculus of probabilities."

The root meaning of probability was originally "approbation." Something probable was deserving of belief. This connotation has been lost in modern times as the understanding of mathematical probabilities has become widespread. Although some epistemologists still connect "degrees of belief" with epistemic probabilities.

Chance, on the other hand, was associated with gambling and other disreputable ideas. Something chancy was risky and evoked disapprobation. Some even equated belief in chance with atheism.

The debate before quantum theory was whether the laws of nature were probabilistic, not that indeterminism was "not covered by the laws of nature."

But note that even if the laws are probabilistic (indeed, Ludwig Boltzmann proved this for some laws with his statistical mechanical explanation of the second law of thermodynamics), most 18th and 19th-century philosophers thought that the "normal distribution" of random events (according to a mathematical "law of errors") showed that underlying the randomness was a universal deterministic law.

As to the role of indeterminism at the macro level, I can assure you and Barry that quantum phenomena are physically irrelevant to the motion of the planets, for example. Though to be sure, quantum mechanics is completely consistent with classical mechanics in the limit of large numbers of atoms (Bohr's correspondence principle).

Most molecular biologists think that brain structures are too large to be affected by quantum uncertainty.

[Barry Loewer]

Bob,

Thanks for the assurance. I assure you that I am not worried that the planets are likely to jump out of their orbits. Nonetheless, your claim that quantum mechanics is completely consistent with classical mechanics when applied to many particle systems is mistaken. Think of a macroscopic device measuring the x spin of a y spin electron. If classical mechanics completely and correctly describes the situation the outcome of the measurement (where the pointer ends up pointing) deterministically follows from the system's initial state. But this is not so. According to QM the final state of the device can only be assigned probabilities. Exactly how qm probabilities are to be understood depends on the interpretation of QM but as long as quantum theory is understood as applicable to the system of the measuring device and electron this conclusion follows. And measurements are not unusual in this respect. There are many processes in nature in which large systems become entangled with small systems and so evolve indeterministically.

[Brian Parks]

Eddy,

One comment,

"But I also think that it is unlikely that such indeterminism (if it exists at micro level) plays no role at the macro level. That's why I think we should avoid any talk of “adequate determinism.” If indeterminism is true even once in a while, determinism is false!"

If indeterminism is true even once in a while, then yes, Laplacian determinism—the thesis that every property of the universe at every instant was set from the first instant—is false. But there are varieties of determinism other than Laplacian determinism that threaten free will. To accurately evaluate the compatibilist position, we need to agree on a definition of “determinism” that makes space for those varieties.

Let me be more specific.

Suppose I make some choice. The incompatibilist will claim that my choice is not free unless it is open to alternate possibilities. Clearly, it does not matter to the incompatibilist *when* in time my choice gets closed off to alternate possibilities. If it gets closed off at the big bang, or if it gets closed off one femtosecond prior to its actually being made, the effect is the same—the choice is not free.

In place of the Laplacian definition, then, a better and more inclusive definition of “determinism” would be:

For any choice C made at time (t), determinism holds with respect to C if and only if there was some time (t - dt) at which the conditions of the universe and the laws of physics necessitated the occurrence of C, let (dt) be any number greater than zero.

So, in other words, if you make a choice, determinism holds with respect to your choice if there was some small time prior to your choice, say 1E-100000000000000000000000000000 seconds, at which the conditions of the universe and the laws of physics necessitated the occurrence of your choice.

This kind of determinism is almost certainly true, even on the assumption that quantum mechanical events are indeterministic. To show that it is not true, it would not be enough to show that a quantum mechanical event somehow leads to my choice, or causes my choice, or factors into my choice. One would have to show that a quantum mechanical event literally *is* my choice. Otherwise, it will be possible to identify a time after the quantum mechanical event has occurred but before the choice has been made, and therefore determinism (appropriately defined) will hold with respect to the choice.

This is the main problem with Kane’s attempt to use chaos theory to explain how microscopic indeterminacies can trickle up to the macroscopic level. Yes, if you tweak the values of the initial conditions of a non-linear system like the brain by an infinitesimally small amount, the effect of that tweak will become infinitely large over time. But notice the key qualification—over time. If you take the time term down to zero, as you essentially have to do when talking about choice, the effect goes away.

[Eddy Nahmias]

People should read Barry's paper, which is very interesting. I cite it and develop a variation of the argument in chap 1 of my dissertation (see below). The goal of these arguments is to show that, if the Consequence argument succeeds, then a revised version of the argument undermines Kane-style libertarianism. (Jason Turner, now at Leeds, has another argument along these lines.) To be clear, I do not endorse the argument since I think FW and MR are compatible with determinism and the probabilistic indeterminism suggested by quantum mechanics. My point here was to try to push the debate towards the forced choice of compatibilism vs. no free will.

Brian, I see your point. But here I've just been emphasizing that the kind of determinism you describe (or that Mark describes or that Bob describes) is different in important ways from the determinism defined by, say, van Inwagen. And it's not obvious how the "determinism" you describe can be used in an argument for incompatibilism (or whether it should count as an argument for *incompatibilism* rather than an argument for the incompatibility of FW and some other thesis X).

Here's my old argument, which comes after a presentation of a parallel constructed version of the Consequence argument.
[L = laws of nature; P = Past]
(1) If probabilistic indeterminism [i.e., quantum mechanics] is true, then given L & P, the state of the universe a moment from now will be either N1, N2, N3, . . . or Nn (with an objective probability for each possible state). To simplify, let us say either N1 or N2 occurs (with certain probabilities), where N1 includes your doing A and N2 includes your doing B.
(2) You have no choice about (control over) whether L is true or P obtained.
(3) You have no choice about (control over) whether probabilistic indeterminism is true and so you have no choice about the objective probabilities that obtain between L & P and, respectively, the occurrence of N1 or N2 (which include, respectively, your doing A and your doing B).
(4) (Modified Transfer Principle) If you have no choice about (control over) L & P, and L & P implies that there is an objective probability you will do A and an objective probability you will do B, then you have no choice about (control over) how likely it is that you do A nor about how likely it is that you do B. (You cannot influence how likely it is that one or the other action in fact occurs.)
(5) So, if probabilistic indeterminism is true, then you have no choice about (control over) the likelihood that you will do A nor the likelihood that you will do B.
[I go on to explain why this challenges free will as much as the conclusion of Con arg, but here I'll just add:]
(6) If you have no choice about (control over) the likelihodd that you will do A nor the likelihood that you will do B (and those are the only possibilities), then you are not free (or morally responsible) in doing A or B.

(As with the Con arg, I think one can challenge premise 2 or 4 or better, the relevance of the type of choice (control) used in these premises to conclusions about FW and MR.)

[R. Clarke]

I'd like to see an argument for (6).

[Bob Doyle]

Barry,

Of course you are quite right. Bohr insisted that quantum measurements are made by macroscopic measuring systems that are themselves classical. John von Neumann and others said there must be some cut (der Schnitt) between the quantum and the classical. Then Eugene Wigner and others asked what would happen if the quantum measurement, the macroscopic apparatus, and the observer were in a box (cf. the Schrödinger's Cat paradox and Wigner's Friend).

The Information Philosophy solution to the problem of Schrödinger's Cat is here:

informationphilosopher.com/solutions/experiments/schrodingerscat/

[Barry Loewer]

Hi Eddy,

Your argument is quite similar to the one I sketch in my paper. Randy is right that the premise that does the work (what is new in the indeterministic case)is 6. An argument for 6 is that it is presupposed by accounts of indeterministic causation. Another deeper argument is that violating it leads to a violation of the usual connection between objective probabilities and rational degrees of belief; Lewis' PP. Suppose that the objective chance at t1 of A occuring at t2 is x while my choice at t0 guarantees the occurence of A at t2 then by the PP my degree of belief at t1 that A occurs at t2 should be x but rationally it should be 1. The situation is a bit more subtle than this but the above captures the main idea.

[Bob Doyle]

Eddy,

Thanks for your version of the Consequence Argument.

Let me try to distinguish what I call the Cogito model of free will from yours/van Inwagen's and Kane's accounts.

I believe my model will appeal to many determinist/compatibilists because the results are only different where the randomness allows us to come up with an unpredictable and creative new idea.

PvI pointed out that a key discriminator is to see how free will models handle "exactly the same circumstances," something Robert Kane calls "dual rational control."

Given that A was the rational choice, how can one defend doing B under exactly the same circumstances?

Please see http://www.informationphilosopher.com/freedom/same_circumstances.html

PvI's description, for example, assumes that chance is the direct cause of action.

Please see http://www.informationphilosopher.com/freedom/chance_direct_cause.html

van Inwagen says:

"Now let us suppose that God a thousand times caused the universe to revert to exactly the state it was in at t1 (and let us suppose that we are somehow suitably placed, metaphysically speaking, to observe the whole sequence of "replays"). What would have happened? What should we expect to observe? Well, again, we can't say what would have happened, but we can say what would probably have happened: sometimes Alice would have lied and sometimes she would have told the truth. As the number of "replays" increases, we observers shall — almost certainly — observe the ratio of the outcome "truth" to the outcome "lie" settling down to, converging on, some value. We may, for example, observe that, after a fairly large number of replays, Alice lies in thirty percent of the replays and tells the truth in seventy percent of them—and that the figures 'thirty percent' and 'seventy percent' become more and more accurate as the number of replays increases. But let us imagine the simplest case: we observe that Alice tells the truth in about half the replays and lies in about half the replays. If, after one hundred replays, Alice has told the truth fifty-three times and has lied forty-eight times, we'd begin strongly to suspect that the figures after a thousand replays would look something like this: Alice has told the truth four hundred and ninety-three times and has lied five hundred and eight times. Let us suppose that these are indeed the figures after a thousand [1001] replays. Is it not true that as we watch the number of replays increase we shall become convinced that what will happen in the next replay is a matter of chance."
("Free Will Remains a Mystery," in Philosophical Perspectives, vol. 14, 2000, p.14)

I think this agrees with your points 4), 5) and 6), no choice about - or control over - the decision, thus no FW or MR.

For more on van Inwagen, please see http://www.informationphilosopher.com/solutions/philosophers/vaninwagen/
________________
Robert Kane is well aware of the problem that chance reduces moral responsibility, especially in his sense of Ultimate Responsibility (UR).

In order to keep some randomness, Kane says perhaps only some small percentage of decisions will be random, thus breaking the deterministic causal chain, but keeping most decisions predictable. Laura Ekstrom and others follow Kane with some indeterminism in the decision.

Let’s say randomness enters Kane’s decisions only ten percent of the time. The other ninety percent of the time, determinism is at work. In those cases, presumably Alice tells the truth. Then Alice’s thirty percent of random lies in van Inwagen’s first example would become a mere three percent.

But this in no way explains moral responsibility for those few cases.

For more on Kane, please see
http://www.informationphilosopher.com/solutions/philosophers/kane/
___________________
Compare the Information Philosophy model, which agrees with compatibilism/determinism except in cases where something genuinely new and valuable emerges as a consequence of randomness.

In our two-stage model, we have first “free” – random possibilities, then “will” – adequately determined selection of the best option.

Alice’s random generation of alternative possibilities will include seventy percent of options that are truth-telling, and thirty percent lies.

Alice’s will evaluates these possibilities based on her character, values, and current desires.

In the Information Philosopher’s Cogito model, she will almost certainly tell the truth. So it predicts almost the same outcome as a compatibilist/determinist model.

It is not identical, however, merely adequate determinism.

It is possible that among the genuinely new alternative possibilities generated, there will be one that determinism could not have produced.

It may be that Alice will find that option consistent with her character, values, desires, and the current situation she is in. That might include a pragmatic lie, to stay with van Inwagen’s example.

In a more positive example, it may include a creative new idea that information-preserving determinism could not produce.

Alice’s thinking might bring new information into the universe. And she can legitimately accept praise (or blame) for that new action or thought that originates with her.

For more on my model, please see http://www.informationphilosopher.com/freedom/cogito/
______________
To summarize the results:

................ / Van Inwagen / Kane / Doyle / Compatibilism /

Alice tells truth /.... 70% /.... 97% /.... 100% /.... 100% /

Alice lies....... /.... 30% /.... 3% /.... 0%* /.... 0% /

* (unless a good reason emerges)

______________
Let’s consider the Moral Luck criticism of actions that have a random component in their source.

Alfred Mele would perhaps object that the alternative possibilities depend on luck, and that this compromises moral responsibility.

Mele may be right with respect to moral responsibility.

In any case, free will and creativity may very well depend on fortuitous circumstances, having the new idea "coming to mind" at the right time, as he says.

The universe we live in includes chance and therefore luck, including moral luck, is very real.

[Bob Doyle]

Barry,

I could not find your article in Philosophical Topics, 1997, Spring or Fall issue.

I am accessing it through Harvard library and Philosophy Online.

Can you help me find it?

[Eddy Nahmias]

Randy, the brief version of my argument for 6 was, unsurprisingly, (1) to distinguish between "open universe" (OU = given the exact state of the universe at the moment of choice and the laws, two or more actions are possible for the agent) and "dual action" (DA = open universe *plus* some agents sometimes have some control over *which* of the possible actions occurs), (2) to argue that OU adds nothing to free will in terms of moral responsibility (though it may add some other attractive things)--and also DA seems to be what most libertarians want, and (3) to argue that my conclusion 5 above leads to the conclusion that quantum indeterminism is incompatible with DA.

I'd be curious to hear about some of the many ways I've gone wrong ;-)

[Brian Parks]

Eddy,

"Brian, I see your point. But here I've just been emphasizing that the kind of determinism you describe (or that Mark describes or that Bob describes) is different in important ways from the determinism defined by, say, van Inwagen."

How so?

If, at the time that I make some choice, the choice is closed off to alternate possibilities, why does it matter *when* in time the choice gets closed off?

What difference does it make, for example, to PVI's consequence argument to say that my choice was closed off to alternate possibilities starting at the big bang, or 50 years prior to my making it, or .000001 femtoseconds prior to my making it?

"And it's not obvious how the "determinism" you describe can be used in an argument for incompatibilism (or whether it should count as an argument for *incompatibilism* rather than an argument for the incompatibility of FW and some other thesis X)."

The determinism I describe is just as hostile to free will as Laplacian determinism. In fairness to the compatibilist, that may not be very hostile. But the equality remains.

All my variety of determinism does is filter out the *irrelevant* part of Laplacian determinism, i.e., the part that says that *everything* in the universe was set from moment one. When it comes to free will and determinism, what we need to worry about is not whether *everything* was set from moment one, but whether this choice that I made right now was set from some infinitesimally small amount of time prior to my actually making it. Even on the assumption of indeterministic quantum mechanics, I can almost guarantee you that it was ;-)

[R. Clarke]

Well, now I'm curious about the argument that (5) leads to no dual action.

In my original post, the view that I described didn't focus specifically on microphysical laws. The claim was that (with incompatibilism assumed) coverage of all actions by laws of nature (of any level) precludes free will.

Barry says that Eddy's (6) is presupposed by accounts of indeterministic causation. I'm skeptical. The principle (P) that Barry appeals to in his paper concerns how one event causes another chancy event, viz., by altering the chance of that effect (or of some causal intermediary) prior to the effect's occurrence. It's supposed to follow from (P) that "the only way that [agent] A can influence what choices she makes at time t is to influence the chances of those choices at prior times" (p. 105 of "Freedom from Physics"). However, I don't see how the principle about how EVENTS cause chancy effects gets one to the claim about how an AGENT influences whether her action occcurs. Certainly, by making an undetermined choice, an agent thereby influences the chance of that choice being made!

In any case, what if the laws in question aren't causal laws? O'Connor's view, for example, has probabilitically structured agent causation. As I understand it, there would be laws about which prior events are followed (with which probability) by which agent-caused effects, but the prior events wouldn't be causes of these effects.

[Barry Loewer]

Bob,

sorry...the Article is in Phil Topics 1996

Barry

[Mark Smeltzer]

Eddy,

You've made me very curious with this comment:

Brian, I see your point. But here I've just been emphasizing that the kind of determinism you describe (or that Mark describes or that Bob describes) is different in important ways from the determinism defined by, say, van Inwagen.

It was my intent to remain consistent with PvI's use of determinism. Have I erred?

[Eddy Nahmias]

Just to address Mark for now (and I think this gets at Brian's concern too), notice that PvI's Consequence argument crucially involves premises that refer to the state of the universe before the agent existed (or even before any humans existed). That claim is supposed to help ensure the truth of the crucial premise, NPast (that the agent has no choice about the distant past). If I am understanding Mark and Brian, their view of determinism would not allow this move, because there would *not* be a state in the distant past that, combined with the laws, logically entails the state of the present (and the agent's action). One of my points was that, without defining determinism in this way, such that you get a Consequence-style argument with an NPast premise, it is less clear how exactly the incompatibilist argument is supposed to go or whether it is really an *incompatibilist* argument rather than an argument for the incompatibility of FW and some other thesis (such as reductionism).

[Mark Smeltzer]

Eddy,

Earlier you asked whether standard incompatibilist arguments can be formulated using "near-enough determinism".

After reading the comment you just posted, I now take you to mean be asking whether we could replace the word "determinism" with the phrase "near-enough determinism" in these arguments and still make them work. Is that more or less right?

If that is the question, I would like to change my answer to a "no". Given any argument that depends on NPast-type premises, the argument won't go through if we muck with the definition of determinism.

I took your question more liberally to be asking whether arguments for agential impotency (aside from the specifics of the formulation, that's all the consequence argument is) could be developed that take indeterminism into account. To that question, my answer is still a "yes".

[Ghenadie Mardari]

I have a question. Suppose there is a probabilistic law of nature that is responsible for the large scale distribution of choices with regard to any particular situation in social contexts. Should a libertarian be committed to rejecting such a law?

[Bob Doyle]

Hi Randy,

In your original post you said three important things:

1) All of the writers I have in mind take Libertarianism to be at least metaphysically possible.

2)) Thus, they think there is some possible solution to the problem of luck.

3) Perhaps the idea is that a solution to that problem requires that undetermined free actions be lawless. But why think that?

My most direct answers are:

1) With due respect to the many Libertarians, since Kant at least, who take free will to be metaphysical and outside the laws of nature, and perhaps a gift of God to humans exempt from those laws because they are above the other natural animals, I think our best answer will be found in a physical solution.

Perhaps, as my post on Martin Heisenberg suggests, it will be not "metaphysical free will," but a "biophysical free will," one that makes us continuous with the animals.

We can limit human exceptionalism to our extraordinary information communication abilities.

2) I see no solution to the "problem of luck." It seems to me to be only a problem for moral responsibility. I admit that at least since Peter Strawson's Freedom and Resentment FW and MR are all tangled up, but I believe they can be separated again.

The randomness needed for genuine alternative possibilities must also give us the real chance that entails luck as real.

3) I do not think we need to describe undetermined free actions as lawless. Perhaps this would be more reasonable if our actions were directly caused by chance. But even then, there are laws of chance - Laplace's "calculus of probabilities," - which become near certainties for large numbers of chance events.

So my carefully considered answer to your 3) is that undetermined free actions must be parsed into two parts - first "free" and then "will."

The first is the "free" undetermined random generation of alternative possibilities.

The second is the "will" - an "adequately determined" selection process that chooses possibilities consistent with our character, values, and desires.

It may not be the metaphysical free will that the Libertarians you are discussing are looking for, but it should appeal to many Compatibilists, because it is adequately determined, yet supports human creativity and complete freedom from the fixed past.

For more, please see the Problem of Free Will.

http://www.informationphilosopher.com/freedom/problem/

[R. Clarke]

Ghenadie, that's pretty much what I was asking. (I was wondering why one might think a libertarian must reject such a law.)

[Ghenadie Mardari]

Randolph,

I consider myself an agent-causal libertarian at the moment. From my point of view, physical laws must have “gaps” in order to create openings for agent causality. This means that some physical events may be insufficiently determined by physical laws (including the probabilistic ones). However, even I need non-physical laws in order to take advantage of those “gaps”. Therefore, I do not see the need to oppose all laws of nature, provided that non-physical laws qualify as laws of nature. I wonder if there is something more subtle that I might be overlooking.

Special laws (like the one described in my previous question) are a different matter, but even they seem to have room for agent-causality. Suppose the probabilistic distribution is enforced by some sort of “social pressure”, which increases as more and more people make up their mind. This law might limit the freedom of those who choose later, but it could still be observed if, say, the first 25 percent of the agents had no constraints. Of course, if it was a deterministic special law for single choices, I would have to resist it.

[Barry Loewer]

Suppose the answer to Ghenedie's (and Randy's ?) question is, as Randy seems to think, "No". Now imagine those probabilities of alternative choices at various times changing to all be 1,0 just prior to (or in the limit approaching) the time of the choosing. Is the answer still "No". If not is there a point at which libertarian free will is ruled out? a vague boundary? Of course probability one is not the same thing as determinism but I find it hard to believe that the libertarians I know would be happy to keep the answer "no" when the probabilities go to 1 and 0. And if the libertarian says that the choices are still free then she will also say that it is possible for an agent to make a choice whose probability is 0 over and over again. This is not a contradiction since it is possible for events with probability 0 to occur. The problem comes when the libertarian says that the choice has an explanation; that it is explained by the agent's so choosing. that is in conflict with there being laws that cover every event since if they do there cannot be another explanation or if there is it is a case of overdetermination. Not a happy result for the libertarian.

[R. Clarke]

Suppose the choice is made at time t. An incompatibilist might say that the choice is not free if at any time prior to t the probability of that choice being made at t was 1. (I'm bracketing the possibility that the choice might be derivatively free, because brought about by some prior free action by that same agent.)

Barry, can you point me to something that explains the possibility of events with probability 0 occurring?

[Barry Loewer]

Randy,

if there are infinitely many points at which a spinner can stop and they are equally probable then then each has probability 0. But the spinner will stop somewhere. The principle that only logical (or necessary) truths get probability 1 is called "regularity" and has some advocates when the probability is epistemic. I don't like it even there but it seems flat our wrong for physical probabilities.

[R. Clarke]

Barry, thanks.

[Ghenadie Mardari]

Barry:

Let us also consider the case from my previous comment, in which probabilities change for different people. The early birds have more free will, and the rest of the people have less. By implication, choosing later can put one beyond the boundary of free will. Similarly, we can think of a person whose will is less and less relevant as he gets closer to his object of temptation. Doing something to produce abstinence is possible early on, but not later on. Is there something wrong with having a boundary for free will? After all, responsibility holds for someone who knows in advance about the existence of the boundary. It was up to him or her to decide before getting to the point of no return.

Also, if the probabilities go to 0 and 1, then it seems to matter how they are formed. Suppose that two alternatives are equally possible from a physical point of view (no determinism), but the agent is temporarily insensitive to one of them. For the time being, one particular distribution holds, but it is up to the agent to do something that reverses the terms in the future. If the distribution is not written in stone, then it sounds compatible with libertarian principles to me (and with Randy’s comment about derivative freedom). By the way, is there an argument to suggest that probabilities must necessarily converge on 0 and 1 before every actual choice?

[R. Clarke]

I have a draft of a paper addressing the question raised by the initial post of this thread. Anyone interested in seeing it, let me know.