2

Many times in discussions elsewhere and in answers here, certain distinctions and claims hinge on Ability to Do Otherwise. However, whether those distinctions are meaningful or claims likely or able to be true, depends on what exactly the term means.

Perhaps one of the more important lack of clarity is whether the aforementioned ability is meant in a causality-adhering-to or causality-independent manner.

In a context 'Entity X produced result A, but could have produced result B instead', is this intended to mean:

  • 'The entity could have produced B had its internal traits been different, such as to produce result B instead of A' (a counterfactual outcome requiring counterfactual preconditions), or
  • 'The entity could produce either A or B based on a non-causal true randomisation, with an unchanged set of traits both internal and external being able to randomly produce either, and it just so happened to have produced A in the event under discussion'?

Unfortunately, it seems like the searches I conducted gave me texts which discuss something hinging on ability to do otherwise, but treating the term as something with which the reader is very familiar, and not explaining the basic definition.

vicky_molokh
  • 121
  • 4
  • 1
    I think it depends; I've seen the first meaning explicitly stated, and I'm almost certain that the second is sometimes meant. Are there any specific instances you're thinking of? – wizzwizz4 Jun 23 '22 at 17:22
  • @wizzwizz4 Mostly ones in which discussions of choice and the like dig deeper below the surface, which inevitably seems to lead towards discussions bringing up the topic of alternative possibilities or lack thereof. (Note that I'm trying to keep the question focused on the term, and avoid anything that risks getting on a more superficial tangent that I'm deliberately not mentioning.) – vicky_molokh Jun 23 '22 at 17:35
  • As far as I understand, which one is meant depends on whether somebody's taking a "deterministic computational" or "extra-universal soul" perspective on free will. I've seen both taken in discussions of _decision theory_, so I for one can't answer this without more context. – wizzwizz4 Jun 23 '22 at 17:49
  • @wizzwizz4 I was so much hoping to avoid bringing up *that* ubiquitously debated thing, precisely because some of the debate's point hinge depending on whether Ability to Do Otherwise is causal or 'purely' stochastic. But here we are, getting on a a FW tangent, leading to circularity of definitions. Could we please take a step away from back and return to the more focused scope of the question? – vicky_molokh Jun 23 '22 at 17:52
  • Are you asking "which meaning is meant in every instance that this phrase is used?" – wizzwizz4 Jun 23 '22 at 17:55
  • @wizzwizz4 No, I'm asking what is the definition of the term as used in philosophy, because it seems like a very specific term, on which reasoning is build much like theorems are built upon axioms. – vicky_molokh Jun 23 '22 at 18:09
  • 1
    There is no "the definition of the term as used in philosophy". It is indeed ambiguous and disambiguation typically depends on philosopher's beliefs about free will. For example, libertarians and compatibilists define "freedom/ability to do otherwise" differently (with variations even within each camp), and not quite as you do in both bullets. On various definitional variants and their implications see [SEP, Free Will](https://seop.illc.uva.nl/entries/freewill/#FreeDoOthe). It is more straightforward for libertarians since they reject determinism, but is still distinguished from randomness. – Conifold Jun 24 '22 at 01:40
  • It seems to me definition 1 leads to infinite recess: what exactly would have made the entity's internal traits different from what they were? As it assumes determinism by implying that outcomes can differ only if initial conditions differed too, those conditions too must have been caused by different conditions, and so on. Definition 2 includes the word random, I think if you are referring to discussions about free will "random" does not cut it, as "will" implies self determination, which is not randomness (although I'll admit it might look purely random to an external observer). – armand Jun 24 '22 at 02:24
  • @armand - Are you saying that determinism in general leads to infinite regress because the explanation for any past state must make reference to a state prior to that? (not necessarily true since an initial Big Bang singularity is compatible with determinism, but even if time did go back forever I don't see that this sort of infinite regress presents any basic philosophical problem with the idea) Or are you saying that defining "ability to do otherwise" in terms of behavior under different prior conditions creates a special problem not found in determinism in general? – Hypnosifl Jun 25 '22 at 16:59
  • The phrase the ability to do otherwise simply means there is more than one possibility or more than on solution to the problem that works. Science is full of possibilities and not absolutes. Sciences do not offer absolute answers. The topic of discussion is likely a science so that is why you get the phrase ability to do otherwise. Absolute topics would not allow such wording because there is only one answer. No possibility talk with absolutes. – Logikal Jun 25 '22 at 20:45
  • @Hypnosifl mostly the latter. "The entity could have produced B had its internal traits been different" seems to imply determinism (it takes different initial conditions to produce a different result -> same conditions give same result -> determinism). But then the initial conditions couldn't have been different, and then "Ability to Do Otherwise" as defined in 1 does not make sense. – armand Jun 26 '22 at 01:35
  • @armand Why couldn't the initial conditions have been different? Deterministic laws produce later states as a mathematical function of earlier states, so you can just as well use these laws to analyze a counterfactual of how things would have evolved from a different initial state. Classical statistical mechanics, which assumes deterministic dynamical laws, depends critically on this sort of counterfactual analysis, based on the idea that a system observed to be in a given "macrostate" at some initial time is equally likely to have been in any of the possible "microstates" consistent with it. – Hypnosifl Jun 26 '22 at 03:17
  • @Hypnosifl the big bang could have been different, why not. But then we are not speaking about the ability of one person to have done otherwise, but more the possibility for the whole world to be different since its inception. If we have to go back in the causal chain to the big bang each time we imagine someone doing otherwise, it can hardly be called their own "ability to do otherwise". – armand Jun 26 '22 at 03:27
  • @armand - Of course a determinist doesn't believe you have the ability to examine different possible worlds and pick which one you prefer, but if one believes in some notion of [transworld identity](https://plato.stanford.edu/entries/identity-transworld/), that individuals in different possible worlds can all be versions of "you" in a meaningful sense (perhaps they are identical to you in all details some sufficiently 'macro' level up until the moment of choice), then "ability to do otherwise" can be understood as "there are versions of you that did in fact choose otherwise". – Hypnosifl Jun 26 '22 at 03:31
  • @Hypnosifl sure, if people believe in nonsense anything goes. – armand Jun 26 '22 at 03:52
  • @armand Do you have an argument as to what's nonsensical about it, or is it just that you find it counterintuitive, a version of what philosophers sometimes call the "incredulous stare" reaction to non-commonsensical views? Again I would point to the analogy with statistical mechanics, where the same macrostate can be compatible with different possible microstates--transworld identity could be understood as defining "you" as a macrostate in this sense, one could in principle make this as precise as desired in terms of what macro-characteristics are the defining ones. – Hypnosifl Jun 26 '22 at 04:11
  • @Hypnosifl any reasonning that has to bring up the baseless speculation of the multiverse to try and make sense is not making any. I have no time to entertain such musing. – armand Jun 26 '22 at 07:11
  • @armand The definition I referred to doesn’t require a multiverse in the sense of actually existing alternate worlds, it just requires that there is a meaningful truth about what would happen in different possible worlds with the same laws but different initial conditions. Again, statistical mechanics in physics is based on this kind of analysis, do you think that’s nonsense as well? – Hypnosifl Jun 26 '22 at 12:16

1 Answers1

1

I think that the most "intuitive" illustration of the concept is in the form of gamebooks. Consider:

You enter the tavern, eyeing the other customers warily. You pick a secluded spot, away from as much of the garish music and gibbering of the inebriated guests as possible. Eventually, a waitress approaches you. She seems amiable enough, and asks you if you would like a drink.

  1. You ask for a shot of vodka. Go to page x.
  2. You ask for a glass of mead. Go to page y.
  3. You say nothing, but you smile and shrug to indicate you don't intend to drink anything right now. Go to page z.

Your "experience" of this narrative structure would be same initial conditions, different possible outcomes. That the kind of disjunctive imperative!! at issue would ever only be resolved "at random" might also be your "experience," and so you might reject the whole sense of the "situation" as a (transcendental) illusion. However, if free will is supposed to be neither predeterminate nor random, and if there is a difference between chance and randomness too, then it is not clear (to me) that strong free will (as in the strong ability to do otherwise) is impossible: maybe it's as different from chance and randomness as those are different from determinism, and as similar to those as it is to determinism: it's its own thing, it's a unique concept in the same family, irreducible in the limit.

I wanted to quote something from John L. Austin (I believe was the guy), from the SEP article on him in which they go over his (positive) treatment of this strong "ability to do otherwise" concept. But the SEP is not working on my laptop for some reason right now, so... This question of mine on this SE contains the desired quotes, though.

Now, consider that mathematics by now either has no absolute foundations, or the best theory of its foundations is a category-theoretic amalgamation with multiversal set theory. If entire mathematical universes, upon which the mathematics of probability would in turn be grounded, are themselves objects of unlimited intellectual will, formed and mediated and traversed by whatever abstract interiority Hamkins and his colleagues are adverting to, then it is not normal game theory, neither probability theory, that covers the mathematics of free will. (An abnormal game theory, or games on the multiverse of sets, could be in play, but here we verge on the debate between semiotic formalists and ante rem realists, among other things.)

Instead, it is the mathematics of the set-theoretic multiverse that is itself the closest thing possible to the mathematics of (strong) free will. One might turn this into a transcendental argument for such will: if strong free will did not exist, we would not know what we know about the set-theoretic multiverse. Since we do know such things, we apparently must have such will available to us. QED

Lastly, though, I don't know that I've ever seen it asked whether some of our actions might be determined, some random/chancy, and others the result of free will. Kant said that free will proper doesn't even directly manifest in the empirical world; it is intelligibly attributed to the meaning of the world from a transcendental source. At any rate, I am not an aggressive enough person to indelibly condemn anyone I see doing something I believe is wrong, neither to overly praise those who I see doing what I believe is right. The question of moral luck weighs on my mind a lot, seeing as it seems like it's due to moral luck that I didn't join the conspiracy cult that is trying to destroy my homeland. They tempted me with drugs, sex, money, weapons, and political power, and I'm pretty surprised I didn't fail the test. Schizophrenia-spectrum issues, of all things, shielded me, because according to my pre-existent delusions in this case, according to my own conspiracy theories, that conspiracy cult is the incarnation of a (metaphorically) demonic evil, and so I resisted them (and continue to resist them) with all my might. Was this an example of my free will? Did my schizotypal character predetermine my resistance? Did my mental health problems lead to a random disavowal of the cult? I don't quite know.

So I won't pretend to have "solved the problem of free will." I will add in this, though:

Consider the following imperatives:

  1. Do A or B.
  2. Do ~A or do B.
  3. Do ~A or do ~B.

If you never have a pure free choice, then none of these imperatives should be intelligible. So if they are intelligible to you, then you have some "sense" of the strong "ability to do otherwise." But note also that (3) is not, in the limit, entirely meaningful: a choice between two negations is an empty choice.

!!In another post on this site, I asked about "disjunctive imperatives," and one respondent claimed I was misusing the word disjunctive, since this supposedly is only "meant" to apply to assertoric functions. However, in his deontic writing, Immanuel Kant spoke of hypothetical and categorical imperatives, translating those concepts from his theoretical to his practical realm. "Originally" (in the first Critique), categorical and hypothetical logical determination is also grouped with (you guessed it!) disjunctive logical functionality. If someone doesn't understand the notion of a disjunctive imperative, that's on them.

Kristian Berry
  • 9,561
  • 1
  • 12
  • 33
  • Why is set theory closest to free will? Why not constructive math or any other foundation to some useful math? – J Kusin Jun 24 '22 at 02:08
  • It has to do with how the logical background is integrated into the substantial mathematics. Hamkins has explained in extraordinary detail how to situate multiversal set theory modulo modal logic, and if we bring deontic/justification logic into the picture (to justify the axioms, which can't be justified by deduction from other mathematical principles), we end up with a sort of "supererogatory" multiverse all over again. Even if God doesn't exist, the concept of God does, and involves absolute free will, so to explain what free will is supposed to be means explaining how God would have it. – Kristian Berry Jun 24 '22 at 05:06
  • As far as constructivism goes, it depends on a faulty concept of mathematical intuition, first of all. Second of all, its denial of things like double-negation elimination seems absurd (*intuitively*, canceling out a cancellation restores whatever was eliminated in the first place, so right off the bat our intuition defies the "intuitionism" at issue). – Kristian Berry Jun 24 '22 at 05:08
  • Thirdly, the refusal to accept relative and absolute infinity is *superstitious*. People who don't accept transfinite set theory and who huff and puff about potential-vs.-actual infinity are the mathematical equivalent of Flat Earthers. – Kristian Berry Jun 24 '22 at 05:09
  • does any other human activity require as strong of free will as set theory, say philosophy, creative writing, or music? – J Kusin Jun 24 '22 at 05:42
  • Something depends on or is caught up in strong free will if and only if it is a matter of alternative cosmological possibilities. Or, at least, if first with Kant we say that this will is situated in a quasi-noumenal context, and then second that physical worlds are concrete empirical models of mathematical worlds, and so third that the range of possible mathematical worlds corresponds to the alternative possibilities of the principle at issue re: the OP, then it is modal mathematics that specifically requires strong free will. Anything else, IDK but I suppose not so much? – Kristian Berry Jun 24 '22 at 18:02