Here's a straightforward question: how many positions are currently held on the nature of free will and moral responsibility? Most would not (I'd think) reply with an actual number, but some variation on the metaphor of a cart of animal waste. And what if I press further how many logically possible positions there are on this issue? I think the clear answer to that is that they are not easily denumerable, especially given that in the future new positions will no doubt be carved out of the then-existing block of literature. But aren't there conceptual limits on the types or kinds of positions available, so that while we can't answer the question about logical possibility with numbers of specific positions, we can with respect to the conceivable number of types or kinds of positions? Well, anyway, I'm going to suggest that we take a very familiar framing of the free will problem--the dilemma of determinism--to achieve just that answer, by using it as a sort of conceptual management tool involving bivalent logic.
Beforehand, I wish to apologize up front if this will all seem obvious or simplistic or naive to many. It may well be that no one besides me sometimes feels a little lost in the inflationary nomenclature of our subdiscipline. I often find myself needing a "big-picture" grasp of where a particular view might fit in relationship to others. My hand is forced in this in part because I have to present the free will problem to my 101 students in a comprehensive and complete way even if I lack the time to tell them how e.g. Wolf's view differs from Fischer's. But the framework has to be robust enough so that if they do go on to study Wolf and Fischer, then they will be able to dovetail intelligibly what they learn later with what they've gotten out of my class. I don't wish to emulate the (nonetheless really excellent) professor I had in grad school who, after teaching a particularly grinding seminar on phenomenalism , only half-sarcastically quipped: "Of course in my undergraduate courses I teach a completely fallacious Berkeley!"
Readers know the dilemma, I gather, but I will reproduce it for reference, using my previous general model of freedom (here FWx as some idea of free will, where x is singular or plural ability/opportunity) as explanatory of the relevant entailments about moral responsibility (MRx, where x is fixed by entailment from FWx or by stipulation in some way that at least is compatible with some FWx):
(1) If determinism of human nature is true, then because we have no appropriately metaphysically plural abilities/opportunities of choice, then we have no incompatibilist FWa (thus specifically dubbed hereon as that form of abilities/opportunities-based FWx), and so no moral responsibility of any entailed MRa-type.
(2) If indeterminism of human nature is true, then although we at least sometimes have metaphysical opportunities of choice, by the most direct interpretation of indeterminism--randomness/chance--we have no appropriate ability(-ies) to control them, so no incompatibilist FWa, and again no MRa.
(3) Either the "if" proposition of (1) or (2) is true (law of excluded middle--LEM).
(4) Therefore one way or another we lack incompatibilist FWa and MRa.
Using the old lingo of these being "hard" (negative) on freedom and responsibility as connected in a certain way, (1) and (2) with the "ifs" dropped may be called hard determinism and hard indeterminism. They are both negative versions of incompatibilism, accepting FWa as the correct concept determining MRa, but which in the case of (2) acknowledges an insuperable luck problem given the truth of indeterminism.
The dilemma is valid and pertinent, based on LEM and the inter-definition by negation and metaphysical relevance of determinism and indeterminism as descriptions of the world. Logic alone is futile here as criticism, and one may not easily dismiss determinism and indeterminism as LEM incoherent or senseless (like a vacuous term that I contrastively use in LEM fashion: "either this table is geschmerfberg or non-geschmerfberg").
So, questioning the nature of the entailments of (1) and (2) is the only option. These in turn will yield "soft" (positive) positions on freedom and/or responsibility.
Thus:
(1') ) If determinism of human nature is true, then though we have no incompatibilist FWa/MRa, either we have some form of compatibilist FWb (in open-ended ways and only commonly defined as not incompatibilist FWa), and so we also have MRb from FWb→MRb, or we have MRb of some type consistent with FWx→MRb.
Alternately:
(2') If indeterminism of human nature is true, then either we at least sometimes have opportunities of choice, and we at least on significant occasions also have sufficient (morally-relevant) ability(-ies) to control them, so we have both incompatibilist FWa and MRa from FWa→MRa, or we have MRa of some type consistent with FWa→MRa .
Dropping the conditionals here and taking the disjunctions to be exclusive yield "soft" determinism and "soft" indeterminism of various types, some directly FWx→MRx related, some mainly MRx related.
(1') contains versions of FWb compatibilism if the first disjunct of the consequent is affirmed, or some form of semi-compatibilism , revisionism, pragmatism, etc. (of some normative MRb-type) if the second is affirmed.
(2') contains versions of FWa incompatibilism (usually some version of libertarianism) if the first disjunct of the consequent is affirmed, or some form of illusionism, revisionism, pragmatism, etc. (of some normative MRa-type) if the second is affirmed.
The key assumptions here are: FWx is a proper account of freedom as a kind of relationship between ability- and opportunity-relata in any given case (that's a big one for me but only arguably necessary here); FWa and FWb are LEM interdefined via incompatibilism and compatibilism; MRa and MRb are likewise interdefined (typically by ultimacy-desert versus non-ultimacy-desert criteria); determinism and indeterminism are empirically stable concepts LEM interdefined; FWx entails MRx or not, and in the latter case MRx might have reasonable normative grounds for specification apart from FWx.
As you see, bivalence and interdefinition by negation play major roles throughout.
Only one logical move left. Deny that FWx/MRx are semantically or epistemically or normatively justified --and apply LEM again. So treat FWx/MRx as (e.g.) like geschmerfberg or not--vacuous/unstable/underdetermined, etc. or not. Free will/moral responsibility skepticism or agnosticism or the like is the first result, where the latter disjunctive alternative leads back to the ins and outs of the dilemma.
Double's skepticism, for instance, is that first kind of geschmerfberg denial about the objectivity/meaningfulness of FWx. If I could make one suggestion for current FWx parlance, it would be to reserve the term "skepticism" for only positions about FWx/MRx based on their semantic indeterminacy, epistemic uncertainty, or moral underdetermination, and refer to hard-incompatibilists as FWa-deniers (or maybe "afreeists" to coin a phonophilic term). The loose use of "skeptic" to refer both to denial of the existence of and denial of meaning or relevance of FWx/MRx is confusing (at least to me).
So, parsing things on the determinism/indeterminism distinction and with my restriction on "skepticism", I see five general areas of possible FWx/MRx positions: hard and soft determinism, hard and soft indeterminism, and FWx/MRx skeptics. Interests in the significance or insignificance of the meaningfulness and/or logical relatedness of the two relevant FWx/MRx concepts determine everything, and I'd say all possible positions are specifiable as some subset of those five. (There are some positions that seem to slide from one area to another depending on what's emphasized about them. For example Beebee and Mele's Humean compatibilism appears to be a form of both hard indeterminism (the luck part across all worlds) and soft determinism (the strong Lewisian ability part in deterministic worlds) depending on what one stresses.)
If I'm just wrong about all this that's ok; in that case I figure that the big-picture perspective I'm taking might motivate others to try and marshal better resources to see what's possible about free will and responsibility, or to say why this kind of project is futile. Of course I conclude that as well by applying LEM to this higher-order issue.
Last observations. Since FWa→MRa-views are specifically incompatibilist metaphysically, and FWb→MRb views are interdefined by negation against that, then the latter constitutes an open class of contrary views susceptible to future innovation , as of course are any number of free-standing (sorry) MRx views not entailed by FWx. Skepticism (as I define it) about FWx/MRx is likewise open, and besides Double's view includes Mele's agnosticism, perhaps PvI's mysterianism, Waller's responsibility skepticism, and Levy's "hard luck" skepticism that swallows (Strawson's sense) more narrow metaphysical and axiological considerations. I admit to great sympathy for these latter two points of view, though I wonder if pragmatism of some sort is the best (workable?) response to them. But that's for my final post. So after this, on to the zombies!
I count more than 5 positions even given your schema. You can combine either of hard/soft on determinism, with either on indeterminism, while remaining uncertain on determinism itself. If one rejects meaning-skepticism, there are 3 questions with 8 possible combinations (not counting agnosticism, leanings, etc.). Adding meaning-of-FW-skepticism back in, that's 9 basic positions.
Posted by: Paul Torek | 02/16/2015 at 12:55 PM
Hi Paul and thanks. I agree that there's much mixing and matching to be done with essentially two issues in contest--free will and responsibility--and subtleties abound. Mix in issues of logical (in)compatibility, epistemic problems and normative realism, anti-realism, quasi-realism and we could probably get a lot more than 9. But that was my motivation here: try to provide *areas* of relatedness where it would make sense to group positions together logically but parsimoniously. That's why, for example, (1') and (2') have disjunctive consequents that allow FW and MR of similar types to be more or less closely associated depending on the motivations to do so. As well I throw together all skeptical positions I can think of--semantical, epistemic, normative--so they can be mixed or matched as adherents see fit. My recommendation about restriction of the use of the term "skeptic" to that grouping is based on how I view even just a fairly strong atheism (my own view): such a view takes a strongly negative position--amounting to denial--of certain kinds of concepts much as hard incompatibilists do. Yeah, sure Pereboom is a skeptic about incompatibilist free will in some loose sense, but so is Mele, and in quite another sense. Better I think to call Derk a denier and Al a skeptic (pardon my assumed familiarity).
But again, thanks for engaging me on this.
Posted by: V. Alan White | 02/16/2015 at 06:45 PM
If 3 Quarks starts a "Went Over Like A Lead Balloon" award, then this post has the clear inside track, and Flickers will be able to boast zenith and nadir in that regard!
Before the weekend I will post my Zombies farewell for this short month of posts.
Posted by: V. Alan White | 02/18/2015 at 09:37 PM