Strawson on Free Will: What are the most persuasive challenges to his position?

Question

There are arguments against free will and moral responsibility which rely on strict causal determinism and/or determinism modified by quantum randomness. Criticisms of these views raise doubt as to our ability to adequately measure the incredibly complex deterministic processes in question, and point out that our limited understanding of mental processes - more specifically our relative ignorance regarding the mental phenomena - prevents us from a clear understanding of the extent to which what we call volition is actually truly voluntary.

Strawson pessimistic view takes a different tack, one which seems logically robust. Strawson's argument is rather long to quote here, but Benjamin Studebaker paraphrases him as follows (as far as I can tell, reasonably accurately) in this article:

When we make conscious decisions to act a certain way, we make these decisions on the basis of reasons, and we find these reasons persuasive because of the way we are (our values, beliefs, desires, principles, physical capacities, etc.).

So if one is to make decisions in a self-determined way, one must also self-determine the way one is.

But to self-determine the way one is, one must have made conscious decisions to be a certain way, and made those decisions on the basis of reasons, and found those reasons persuasive because of the way one is.

So for self-determination to be possible, one would need to be able to self-determine the process by which one self-determines, which would require that one self-determine the process by which one self-determines the process by which one self-determines. This produces an infinite regress.

Therefore self-determination is logically impossible.

Therefore, insofar as free will requires self-determination, free will is impossible.

As far as free will goes, it doesn’t matter whether we are the way we are because of natural laws, heredity, environment, randomness, or some combination of these things, what matters is that we cannot be the way we are because we self-determined it, and because we make our decisions based on the way we are, we cannot make our decisions on the basis of self-determination.

I admit to a a bias here; a bias that I wish to interrogate by opening this issue up for discussion. It is Strawson's pessimism (almost commonsense?) I find most persuasive; so persuasive that I wonder at why it is not more widely understood and taught (outside of the strictly philosophical realm, in more general areas of academia study) given its relevance to one of the most fundamental aspects of our experience and the ramifications it presents for the ways in which we view ourselves and treat others. Challenges to the pessimistic view are outlined here, but I find myself agreeing that these challenges are inadequate.

Experience to date suggests that this stack often furnishes views which are not apparently addressed by the usual first-glance sources I resort to; views which argue from resources I haven't considered or which rely and valid/(sound?) logical presentations. It is any such views I am interested in, for I remain surprised that Strawson's view is not met with a more widespread acceptance and I wonder at what I might be missing.

Note: I acknowledge that logic and reason often aren't the key drivers for our arrival at beliefs, and that the free will debate in particular presents significant challenges for minds conditioned to believing the intuition we typically possess of being in command of our own decisions; of having what we colloquially understand as 'free will'. I am not asking here for reasons as to why we don't more readily accept we don't have free will, but for any challenges to Strawson's view not already contained within the resources I've provided.

Answer 1

Strawson's argument is about ultimate moral responsibility not stritly about free will (although related).

The tricky phrase here is "ultimate". One can argue that free will is not about ultimate responsibility, as this is equivalent to omnipotence and wishful thinking not free will as we know it.

Strawson's argument (extended to free will) implies that free will can only exist as omnipotence in a maximal absolute sense or not exist at all. This is not what we experience as free will. We experience making choices towards goals but that does not mean that our goals will certainly be reached or be reached as we expect them. We do experience making choices but do not experience omnipotence in this sense. The fact that a certain goal may not be reached as expected, is irrelevant to us making a choice (if we indeed made a choice).

There is an additional caveat: when we talk of ultimate responsibility, should we mean all possible reasons up to that time, no matter the amount of influence they might have or only the sufficient reasons up to that time? It seems that talking about sufficient reasons only, makes the argument lose much of its apparent force, since then infinite regress is not bound to happen. In other words, it doesn't matter if we can't choose and control everything as long as we can choose just enough.

Furthermore Strawson's argument has an assumption that reasons at time t1 uniquely determine only one possible course of action or choice at time t2. If this is not the case, if reasons at time t1 can be equally compatible with more than one course of action or choice, the argument fails. It is one thing to say that what an agent has at time t as dispositions or reasons will be used in the justifications/explanations for a choice, any choice, and another thing that these dispositions or reasons cause and fully determine a choice. There is no a priori reason that dispositions/reasons are necessarily compatible with only one course of action or choice.

It has also been argued that even accepting Strawson's premises fully does not necessarily lead to infinite regress thus the basic argument fails. The point made is that a choice amounts to more than one thing: choosing one action over another (ie action-commitment vs action-forfeiture). Not both of these actions/choices fall prey to the infinite regress suggested by Strawson while both contribute to self-determination as distinct choices/actions.

At this point, it can be shown that Ronny’s sandwich-eating action-forfeiture (and Chinese-food-eating commitment, for that matter) evades the problem of infinite regress inherent in the Basic Argument. One of the primary reasons that make up Ronny’s rational explanation for not getting a sandwich is Ronny’s action-commitment to eat Chinese food. In order for the regress to start, then, the rational explanation of Ronny’s action-commitment to eat Chinese food would have to be traceable, in its entirety, to reasons Ronny had prior to the action-commitment itself. This would not be enough for an infinite regress, of course, but it is how such a regress would begin. As it turns out, many of the reasons Ronny has for eating Chinese food do precede his action-commitment, such as Ronny’s belief that the Chinese food will taste good. But not all of Ronny’s reasons for committing to eat Chinese food appear prior to the commitment itself. Ronny’s action-commitment is also explained by Ronny’s sandwich-eating action-forfeiture. Ronny was not committed to eating Chinese food until he gave up the idea of eating sandwiches, but he did not give up the idea of eating sandwiches until he committed to eating Chinese food. These two distinct choices appear simultaneously, and they both factor into the rational explanation of the other. Thus, a full rational explanation of Ronny’s sandwich-eating action- forfeiture (and Chinese-food-eating action-commitment) includes reasons that came into being only at the moment the decision was made. There is no regress.

One can attack Strawson's arguments in other ways as well (SEP has various pointers to critiques of Strawson's argument).

There have been numerous replies to Strawson’s argument. Mele (1995, 221ff.) argues that Strawson misconstrues the locus of freedom and responsibility. Freedom is principally a feature of our actions, and only derivatively of our characters from which such actions spring.

..Clarke (2003, 170–76) argues that an effective reply may be made by indeterminists, and, in particular, by nondeterministic agent-causal theorists. Such theorists contend that (i) aspects of ‘how one is, mentally speaking’, fully explain an agent’s choice without causally determining it and (ii) the agent himself causes the choice that is made (so that the agent’s antecedent state, while grounding an explanation of the action, is not the complete causal source of it).

Arthur Schopenhauer and others have made similar a priori arguments against free will being able to will what it wills (leading to some infinite regress).

However these arguments are misleading and beg the question in the sense that they initially restrict what can be willed and then require an additional level to will to will. One can attack all these arguments together through either of two routes:

1. There is absolutely no a priori reason for a will to be initially restricted in what it can will so that it requires additional levels in order to will something else.
2. Even if will is restricted in levels, one in order to will something in practice will go at most some finite levels deep to reach the appropriate level, and there is no infinite regress in practice.

Further references:

1. The Impossibility of Moral Responsibility, Galen Strawson
2. Concerning the resilience of Galen Strawson’s Basic Argument, Michael Istvan
3. On an argument for the impossibility of moral responsibility, Randolph Clarke
4. Agent causation and the alleged impossibility of rational free action, Chris Tucker
5. Two Arguments for Impossiblism and Why It isn’t Impossible to Refute them, Joseph Corabi
6. Why Strawson’s Basic Argument Is Not Impressive: an Answer from Frankfurt, Christman and Ekstrom, Fei Song
7. Free Will and Ultimate Explanation, Boris Kment (alternative 1, alternative 2)
8. The Possibility of Moral Responsibility, Nelson Zhang
9. Freedom and Forfeiture: Responding to Galen Strawson's Basic Argument, Eli Benjamin Kelsey

Answer 2

Strawson’s full arguments are more careful and I’m not deeply familiar with them, but in the Studebaker paraphrase you quote, the central paragraph claiming to “logically refute” self-determination/free will is a basic logical fallacy (or at best, a large gap) — it’s an old paradox of causation applied to “self-determination”. As you quote, Studebaker writes:

So for self-determination to be possible, one would need to be able to self-determine the process by which one self-determines, which would require that one self-determine the process by which one self-determines the process by which one self-determines. This produces an infinite regress.

Therefore self-determination is logically impossible. Therefore […] free will is impossible.

One could just as well claim to refute determinism by the same argument:

So for determinism to be possible, some cause would need to be able to determine the process by which actions are determined, which would require that some cause determine the process by which the first cause determines the process by which actions are determined. This produces an infinite regress.

Therefore determinism is logically impossible.

The fallacy here is assuming that infinite regresses are impossible. Valid logical arguments should be well-founded, without infinite regress — but the regress Strawson is constructing here is a totally different thing. We have known since Zeno that many kinds of infinite processes and regresses can and do occur in nature.

There can of course be reasons why a particular infinite regress in a particular context is contradictory, or refutes some other expectation — but that is a further argument that needs to be made. You can’t just conclude “X would lead to an infinite regress; so X is impossible!”

Answer 3

Strawson's argument is a highly persuasive one, but suffers from the problems I pointed out previously. Proof for the absence of free will? I will revise that answer, to be more explicit to your current rephrase of your question.

We humans have strong evolutionary tendencies. We have inclinations to believe in a caused/random dichotomy. We also have inclinations to believe in an intuitive form of logic, which while flawed, if it is applied to itself in self-correction will produce classical logic. We also have a strong evolutionary inclination to believe in willing, and the agency of our willing, and that we have choices, which our willing can usefully choose between. We also have strong evolutionary inclinations to universalize -- to extend local observations and make global conclusions about our universe.

These inclinations clash, and you and Strawson argue that causation, plus classical logic, plus universalizing of each of the above, trump and prevent freedom of willing.

I note in your argument above, you have not identified the universality of the caused/random dichotomy, or the universality of classical logic, as explicit assumptions that you need to justify.

We KNOW that over-universalizing is one of the most common and harmful of human reasoning errors. Over universalizing is super useful in lots of circumstances, as a plausible working hypothesis, and gives us a good starting point to guesstimate how things work in a new environment. But is NOT actually true in LOTS of cases, and the dogmatic insistence on unjustified universalizing is behind basically all blind dogmatism. The essence of the philosophic enterprise is to learn to identify and question our unrecognized walls to our thinking, and questioning these universalized assumptions is something you should try to do.

We know that the caused/random dichotomy is not an actual dichotomy. There are two cases we have found that explicitly violate it. In the micro-sphere, quantum mechanics is not EITHER caused OR random but instead an interesting fusion of the two. And in the macro sphere, we cannot reduce any of our physics well enough to eliminate all unknowns, so basically all physics operates with identified effects, superimposed over a random draw of background noise. So -- our physics operates with two different approaches that violate the dichotomy, and conflict with the classical law prohibition of the excluded middle.

We ALSO know that classical logic is not the only possible logic, and that instead there are infinite logics, and that all sorts of other logics apply better to certain aspects of our world. SO -- we know that each of the assumptions I identified above:

• Your caused/random dichotomy
• classical logic
• universalizing

are very often not valid assumptions.

You have not shown that they apply, and therefore trump free will.

I take the evolutionary evidence for freedom of will to be convincing evidence of its reality. We have a characteristic of Agency, which does not fit into your caused/random categories.

HOW Agency works as a logic/causation principle, we still don't know yet. Given infinite logical pluralism, there will be SOME form of logic that can explain how Agency based libertarian free will works, but I have not found it yet. But someday, we will.

Agency causation is, of course, different from caused/random causation. These will be incompatible logics, and how they fit together, if they ever do, I can't say. Empiricists accept that there are many aspects of our world that are not fully understood, and may not be reconciled with each other, yet appear to be true. We cannot have logical coherence in worldview today, and very plausibly, never will.

Answer 4

[Note: I am revising my entire response but undeleting it just the same, at the site's own behest (I clicked "Add Another Answer" and it was like, "Why not just edit the one you already posted?"). So none of the comments on this answer pertain to the current version, but to the previous one.]

If Galen Strawson's argument does not seem "persuasive," much less conclusive, it is because for one, there are "compelling" arguments for self-determination (of the relevant disputed form) as well. So the whole local dialectic has the flavor of a Kantian antinomy, with Strawson's horn of the dilemma resembling the antithesis, negatively mirrored by e.g. J. L. Austin's counterpart argument. With respect to the first of Kant's antinomies, for example, the reasoning for the thesis is intuitively very strong: "If time goes backwards forever, then for the world to have evolved from then to now would require an infinite amount of time to have already elapsed, which is impossible." That's a plausible bastard of an inference. But just the same, so too is the antithetical deduction: "If the world in time didn't go backwards forever, then there would have been a point where a total void—a state of reality with no matter and no causation—spontaneously filled itself in and began chugging along like a well-oiled machine."

Now, Strawson's argument seems like a clever repurposing of the thesis from Kant's third antinomy. There, Kant reported on the doctrine in favor of free will due to almost the very same no-infinite-regress argument. Later, in the second Critique and then Religion within the Limits of Reason Alone, Kant would even say that the regress of practical reasoning tended towards an ideal and an infinity:

Now, when we abstract from a law all matter, i.e., every object of the will (as a determining principle), nothing is left but the mere form of a universal legislation... We can become conscious of pure practical laws just as we are conscious of pure theoretical principles, by attending to the necessity with which reason prescribes them and to the elimination of all empirical conditions, which it directs. [Critique of Practical Reason, theorem III, remark on problem II]

Now this moral evil (transgression of the moral law, called SIN when the law is regarded as a divine command) brings with it endless violations of the law and so infinite guilt. [Religion within the Limits of Reason Alone*, "Concerning the Legal Claim of the Good Principle to Sovereignty Over Man," sec. C (emphasis added)]

(The relevance of the "infinite guilt" thesis is: according to Kant, the process of fixing our maxims requires tracing backwards along a line of infinitely many wrong maxims all the way to an ideal limit, an ultimate wrong maxim over the others; but of course, per his overall theory, we can never "remake ourselves" altogether in empirical time.) So is one person's reductio ad absurdum another's iron-clad proof of the same claim? Perhaps; at least, that the same line of reasoning can be used to defend two relatively incompatible notions suggests a reason for Strawson's thesis, here, to find less purchase in the intellectual world than you might expect.

Answer 5

Strawson's argument begs the question. To paraphrase it, he assumes that our mental state at one moment mechanistically determines our mental state the next, so if you follow that chain backwards our mental state today arises as a series of mechanistic incremental changes from the way our brains were wired at birth.

Until we have a proper explanation for consciousness, we cannot say whether we have genuine free will or whether our thoughts are the consequence of something akin to the operation of software in a computer, so that they play out according to pre-determined factors.

As far as we can tell, our thoughts seem to arise from electrical activity in the brain, or at least to be associated with such activity. To have genuine free will, it must be possible for our thoughts not to be mapped one-to-one with configuration states of the particles that comprise our brain. If they were always so mapped, and if the configuration of the particles in our brain is governed by the laws of physics (albeit in a way that might include quantum randomness etc) then our thoughts would be governed by the laws of physics.

Answer 6

"So if one is to make decisions in a self-determined way, one must also self-determine the way one is."

This is one big problem, a non-sequitur.

Being able to self-determine what you do does not require the ability to self-determine what you are.

Schopenhauer said: "A man can do what he wants, but not want what he wants?"

This means that we can act according to our preferences, but we cannot choose our preferences. That would be illogical, since we need preferences in order to choose anything. You cannot prefer to prefer something else than what you currently prefer.

Answer 7

Strawson's argument seems to be stronger version of Arthur Schopenhauer's (A man can do as he wills, but not will as he wills). If, contra Shcopenhauer, a man can will as he wills then he hasta also will the will that wills, but then to self-determine he has to will the will that wills the will that wills and so on ad infinitum/ad nauseum (infinite regress).

However, as a tentative solution to Strawson's amazing argument, I'd say the infinite regress seems kinda illusory. The set of likes & dislikes I possess (the will, that which we want to have a say in) cannot be radically different from the set of likes & dislikes I will then choose. For example if I like chocolate & I don't like smelly feet, what I select as my next set of preferences will be either identical or very similar. That is to say the infinite regress, instead of looking like ... selects W selects X selects Y selects Z, is actually ... Z selects Z selects Z selects Z ... In short, the infinite regress collapses, is Z, has been all along Z. All we need now for free will is the ability to choose any set of likes & dislikes (preferences), including but not limited to Z.

Note:

1. A to Z represent groups/sets of preferences (likes & dislikes) that determine the choices we make. Control them and one is master of one's destiny i.e. one has free will.

2. Z represents my current preference set.