Monthly Archives: May 2013

Notes on Haji’s “Deontic Morality and Control” Ch. 6

In Chapter Six of Deontic Morality and Control, Ishtiyaque Haji moves his discussion of moral responsibility and deontic anchors from the premise of determinism and examines how these concepts fair under the premise of indeterminism. One thing this lets Haji do is test the durability of his asymmetry thesis – the proposition that ascriptions of moral responsibility require only singular control while deontic ascriptions of rightness, wrongness and/or obligatoriness require dual control – in a world where fixed laws and a fixed past do not entail a fixed future.

Equipped with this thesis, Haji proceeds to referee a debate between Modest Meleian Libertarianism and Fischer’s Compatibilism. Not too surprisingly, Haji tends to favour Mele’s view as it, by his lights, seems most congenial to Haji’s own asymmetry thesis.

Mele says that in order to secure moral responsibility, it is enough if an agent’s actions issue from states that are within his proximal control even if what causes these states to occur when they do is the result of an indeterminate process. We can call this Meleian view, following Haji, ‘doxastic indeterminacy’. Under doxastic indeterminacy, action-causing beliefs occur in the mind of an agent in a process much like that of quantum decay, though not necessarily with the same fixed range of probability. Fischer responds critically by insisting that such an account erodes and diminishes the very agency we would need to assume in order to make any meaningful judgments about moral responsibility at all. For Fischer, what constitutes a free action is not merely that the action is caused by an agent’s beliefs but that these beliefs themselves stem from the agent’s own causality, perhaps from his nature or ‘essence’. (I will examine later what this latter term could mean in the given context).

However, Haji’s rejoinder to Fischer’s criticism of Mele is insightful: since Fischer’s compatibilism concedes, under the thesis of determinism, that our beliefs and any action-causing states that make up our proximal control develop from a process that is ultimately external to that very control, it should make no difference, not to our moral responsibility, whether this process is deterministic or indeterministic given that it is, in either case, outside of the agent’s own power. In this respect, Haji’s response to Fischer is very reminiscent of Pereboom’s central argument in Living without Free Will that what is damning to an agent’s bearing of moral responsibility is not whether his actions are deterministically or indeterministically caused but that upon either worldview, his actions are the result of an “alien” causal process that is beyond his agential, or executive control.

And yet, Haji recognizes that Fischer’s intuitions reveal a discrepancy between the cases. More specifically, Fischer has the intuition that Mele’s account “smacks of alchemy” (Haji. p. 95, citing Fischer) as it secures an agent’s moral responsibility at the cost of mutilating what we normally mean by ‘agency.’ Indeed, Mele’s account does leave a somewhat fishy smell in the air: S is only responsible for action A if A issues from a belief B that is both grounding-line and conductor of moral responsibility for S even though B’s occurring to S is entirely outside of his proximal control. For Fischer, this presents us with an impoverished view of agency that is at best a Pyrrhic victory for any honest proponent of free will. Fischer’s concern seems to be that Mele’s solution succeeds in grounding moral responsibility at the cost of destroying the very matter through which that responsibility would be conducted across time, namely, the unity of the agent himself.

Nevertheless, Haji’s point against Fischer is well-taken:

If compatibilists are willing to grant that we can have proximal control over beliefs once they come to mind despite the fact that the coming to mind of typical beliefs is something not in our proximal control, then, it seems, doxastic indeterminacy need not erode proximal control [of the kind needed to underwrite moral responsibility]. (Haji. pp. 100-101)

Haji chalks up Fischer’s misgivings about doxastic indeterminacy to the fact that it doesn’t live up to “the Libertarian’s guiding principle that free will is the power of agents to be the “ultimate buck-stopping originator[s]” of their actions” (Ibid. p.100, citing Strawson): a criterion of freedom that Haji terms “ultimacy”. But to ask for ultimacy, in Haji’s mind, is tantamount to asking for dual-control, and it is unlikely for Haji that a compatibilist like Fischer or a modest Libertarian like Mele can hope for this much, assuming that Haji’s asymmetry thesis obtains. Recall that the asymmetry thesis states that while singular control – S’s ability to do A – is sufficient to ground moral responsibility, deontic anchors require that S be able to do A and not do A, that is, to exert dual-control over his actions.

This means that, for Haji, S ought to do A at T iff at T S can do A and not do A. But on the Meleian doxastic indeterminacy view, the occurrence of belief B that grants S the power to do A does not grant S the power to not do A. In other words, even though S can do A and not-A at time T0, as soon as belief B occurs to S at T1, S no longer has dual-control over A. Any obligation held by S at T0 then is a “remote obligation” (Ibid. p.102, citing Zimmerman). It might be thought that a remote obligation is better than no obligation at all, but in line with Zimmerman, Haji reasons (and rightly, I think) that the notion of a remote obligation depends on a coexistent immediate obligation. This follows naturally from Haji’s modified version of K which states that:

S ought, at T0, do A at T*, iff A, at T0, can do A and not do A at T*. I am only obliged at T0 to discharge a duty at T*if it is within my power at T0 to discharge or not discharge my duty at T* (Where T* may or may not be later than T0).

This Meleian view has the effect of leaving S in a strange bind: up until the time that belief B arrives on the scene, it is, supposedly, well within the range of possibilities that S can do A and not do A since it is an open question whether the belief that occurs to S will cause him to do A or not-A. But as soon as B shows up, S can’t help but do one or the other but never both.

However, I see a deeper problem here: it is a strange kind of power that S has at T0 if he can do A and not do A and yet have no capacity for action at all until belief B shows up. This seems to me to fundamentally obscure the concept of an ability or power. For a power is a disposition to respond to a stimulus according to an internal principle or law and to respond differently to different stimuli (see, eg. Sebastian Rodl, Infinite Explanation, 2010). But the way this doxastic indeterminacy construes the situation, S plays no active role in belief B issuing in the performance of A; A simply follows from B and nothing in S determines this connection. Accepting doxastic indeterminacy, so construed, entails the following: there is no unity between S, B and A.

In short, the strange bind is in fact a double-bind. The double bind is this: at T0, S has the opportunity to do A and not-A but not the ability, while S at T* has the ability but not the opportunity.

The double-bind can be relieved if we instead, view beliefs and actions under a concept of causality that is not mechanical but teleological.

But this will have to wait…

Notes on Haji’s Deontic Morality and Control” Ch. 5

In Chapter 5 of Deontic Morality and Control, Ishtiyaque Haji issues the following challenge:

“[1] Develop a prima facie plausible account of ‘S can do X.’ Then [2] show that in this very sense of ‘can’, the agent can both do X and refrain from doing X in a Frankfurt-type case, where X is the sort of action that the counterfactual intervener wants the agent to perform. I suspect that [3] any such account of ‘can do X’ will yield the result that in a Frankfurt-type case, the agent won’t be able to refrain from doing other than what she in fact does. [sic] This is because (i) [4] such an account will, in all likelihood, include a proviso to the effect that there are no external or internal impediments that prevent that agent from doing X or from refraining from doing X; (ii) [5] in Frankfurt-type cases, the counterfactual intervener ensures that the agent will lack opportunity to do otherwise even is she has the ability to do otherwise; so there will be an external impediment that prevents the agent from refraining from doing X; and (iii) [6] refraining her doing X requires both the ability and the opportunity to refrain from doing X. In turn, if this is all correct, then [7] we will have strong reason to believe that, in the proposed sense of ‘can’, it won’t be true that the agent will be able to refrain from doing what she in fact does if determinism is true.” (Haji, p. 69)

Here, I would like to accept the challenge and outline a response step by step.

1) Jones can ask Smith to marry her.

2) Suppose Black wants Jones to ask for Smith’s hand in marriage and has the power to generate in Jones’s mind the thought that she will propose to Smith because Jones loves Smith and has learned that she is pregnant with his child. But it just so happens that before Black can work his magic Jones asks for Smith’s hand in marriage because she loves him and is pregnant with his child.

3) Contrary to Haji’s suspicions, it is not that case that Jones cannot refrain from asking for Smith’s hand in marriage. This is so because FAP – ‘it is within S’s power to do X iff it is within S’s power not to do X’ – obtains. (Ibid. p. 45, cited in Zimmerman 1996:86-7)

4) In accord with Haji’s anticipation, it is true that we must include a proviso that there are no “external” impediments (like a broken telephone connection) or “internal” impediments (like either Jones or Smith falling unconscious) that prevent Jones from exercising her ability to propose to Smith or to refrain from doing so. But such a proviso only serves to hold the relevant variables fixed. If either type of impediment (both of Haji’s examples seem ‘external,’ to my lights, by the way) obtained, then neither the decision to propose to Smith nor the decision not to propose to Smith would be live options for Jones. But such a proviso does not rig the game; it merely sets the terms of play.

5) Though it is true that in FTCs the counterfactual intervener, here Black, robs the agent of the opportunity to do otherwise than she does, it does not follow from this, I argue, that Black’s presence in any way hinders the exercising of her power to do otherwise. For, according to FAP, the very ability to do other than X is exercised in the doing of X. For Jones, the very possibility of not asking for Smith’s hand in marriage (and the consequences that would follow from that course of action) plays a determinate role in her decision to ask Smith to marry her. If Jones could not entertain this possibility — if she could not understand what it would mean for her to refrain from marrying Smith — then she equally could not be said to have the power to make marriage proposals, to Smith or to anyone.

6) Despite Haji’s adamant insistence on this point, I do not think that having the power to refrain from doing X requires both the ability and the further opportunity to do so. I feel this way because on my view, those alternative possibilities that are sufficient to underwrite free-will, deontic anchors and moral responsibility are built-in to any power that satisfies the concept of a rational ‘ability’ in the first place. (I suppose this puts me on even ground with Fischer and Dennet, though I haven’t worked out the details of any specific alliances yet).

7) Since determinism — taken as the view that states that all the “non-relational” or “hard” facts of the past together with the laws of nature produce one definite future (Ibid. p. 60) — is only devastating to those who take opportunity to be the gold-standard of alternative possibilities against which we measure one’s moral responsibility and/or the existence of deontic anchors it matters not to my view of freedom, obligatoriness and responsibility whether or not determinism so construed obtains.  This is because on my view (as laid out so far), an agent is only morally responsible for what she has the ability to do and not for ensuring for herself that she have, in every possible situation, the opportunity to express this ability.  Likewise, according to what I’ve said thus far, the ‘can’ in ‘ought implies can’ refers only to ability and not to opportunity. And so, although determinism seems to render it preordained whether or not any given agent will have the opportunity to exercise her ability to do X and not do X, I offer a perspective on moral responsibility and obligatoriness that corresponds to only that range of an agent’s ability that is ‘called upon’ in any given situation.   

I have hereby met Haji’s challenge and I have done so by holding fast to FAP, the principle that states that an agent can only do X if an agent can do not-X. However, Haji asked not just for any account, but for a plausible one, which would preserve deontic anchors in either a FTC or in a deterministic universe. And though I think that the view I have offered here is not only plausible but also true, I have yet to argue for its plausibility or to demonstrate its validity.

If I can show that a FAP-like principle holds not only for actions but for any and all determinations that range over a gradient of possibilities, including the “non-relational” or “hard” facts of determinism then I will shift the onus onto advocates of Haji’s style of moral nihilism to show why FAP should be rejected in the sphere of ethics while its equivalent principle — for now what I will call determinate negation — is not contested in the same way in other fields of metaphysics.

Stay tuned…

Brandom vs. Rödl

Notes on Pereboom’s “The Contours of Hard Incompatibilism”


In Living Without Free Will Derk Pereboom advances the view that the kind of free will that we would need in order to underwrite the existence of moral responsibility is impossible in both a deterministic universe and in an indeterministic universe.  Pereboom’s thesis is stronger than the traditional doctrine of Incompatibilism for the latter is thought to obtain only if determinism is true.   Hence Pereboom introduces:

The two theses of Hard Incompatibilism:

i)                    All of our actions lie on a “responsibility-undermining” continuum from alien-deterministic to truly random events


ii)                  An action is free in the sense required for moral responsibility iff it is not an alien-deterministic event, nor a truly random event, nor a partially random event. (pp.127-8 Living Without Free Will, Derek Pereboom)

Pereboom’s reasoning for (i) can be grasped if we take the following to be true:  all phenomena are natural phenomena thus our actions are natural phenomena and natural phenomena are either (a) determined by physical laws and micro-physical states (which combine to create causal forces that are “alien” to us as agents who want to be self-determining) or (b) determined by anomic processes that are equally beyond our agential control.

Similarly, Pereboom’s reasoning for (ii) can be understood if we assume that the following propositions obtain:   (c) there is no way to make sense of the concept of a capacity for moral-responsibility-conferring free will that is subject to either alien-deterministic or truly random events and (d) there is no way to make sense of a concept of moral responsibility outside the context of such a capacity for free will.

Propositions (a) and (b) can be subsumed under the broader heading of ‘Complete Physicalism’ (“complete” because it includes deterministic and indeterministic presuppositions) while propositions (c) and (d) can be recognized as standard, garden variety Incompatibilist assumptions.   Consequently, Hard Incompatibilism can be undermined by refuting Complete Physicalism, as is done by some Libertarians, for instance, by insisting on the existence of a non-physical (e.g. noumenal) “agent-causal power” – a Kantian-inspired view which Pereboom considers and rejects as implausible though not impossible – or by refuting Incompatibilism by demonstrating the plausibility of a view of moral responsibility and/or freedom that is compatible with either deterministic versions of Physicalism or indeterministic versions of Physicalism.


For his own part, Pereboom is concerned with defending a notion of morality that need not refute either thesis of Hard Incompatibilism for he is also concerned with sketching the contours of a moral theory that need not be tethered to a workable notion of moral responsibility in order to be intelligible.   In this respect, Pereboom offers a deflationary theory of morality as a competitor to Haji’s more robust conception of morality that ties the latter to the existence of deontic anchors such as obligatoriness, rightness, and wrongness.   For Pereboom, moral ascriptions such as right and wrong survive even if Hard Incompatibilism obtains and moral responsibility, with its attendant notions of praise- and blameworthiness, is dissolved.   In other words, morality for Pereboom is not anchored to a conception of free will that itself stands or falls with the principle of alternative possibilities as defined by Haji in his K and OW principles:

K:   S has a moral obligation to do A [not do A] iff it is within S’s power to (i.e. S can) do A [not do A]


OW:   S has a moral obligation to do A [not do A] iff it is morally wrong for S not to do A [do A] (ibid. pp. 143)

I will here paste over the details of Pereboom’s reductio ad absurdum argument against Haji – basically, Pereboom thinks we should defer to the least counterintuitive explanation, and that his is less counterintuitive than Haji’s – and say but briefly that the upshot of Pereboom’s refutation renders the following axioms false:

Ought implies can.

Right implies can.

Wrong implies can.

If by ‘can’ we mean something like Haji’s dual-control thesis which stipulates that ‘can’ includes both the ability to do otherwise and the opportunity to do otherwise. (Deontic Morality and Control, Ishtiyaque Haji, p. 27)

For Pereboom, these deontic anchors still might track what we might call “real properties” or real facts even if Hard Incompatibilism is true – (we won’t examine here why Pereboom is more sure about the latter two axioms than the first.  My hunch is that “ought” is deeply connected to an intuition that ‘ought’ implies opportunity whereas ‘right’, for reasons we will see shortly, is coherent even only in the context of ability).  In addition, Pereboom wants to leave open the precise definition of these deontic anchors, reserving judgment on whether or not his theory entails a Consequentialist (a la J.J.C. Smart), Nicomachean (a la Michael Slote) or a Kantian ethics.   To his mind, Hard Incompatibilism – and the consequent non-existence of free will – is compatible with a view of the Right as the maximization of utility, the nurturing of Eudaimonia , or the Categorical Imperative in either of Kant’s formulations.  

We are left with a view of morality that is supposedly coherent without relying on a belief in free will (understood as the freedom to do otherwise).  About his view, Pereboom makes the following interesting observations:

Morality is more like aesthetics than is typically thought and,

The Hard Incompatibilist position implies that human immoral behavior is much more similar to earthquakes and epidemics than it would be if we were morally responsible.  (Pereboom, pp. 154)

The thought is this: The transformation of ethics into a branch of aesthetics would be justified because aesthetics is the only other domain of thought that employs a concept of rightness or goodness in its evaluation and appraisal of objects that does not tie these concepts to freedom but rather to happiness and beauty.  Aesthetics is a fascinating island for morality to find itself shipped wrecked on and Pereboom may be correct to insist that his theory is preferable to Haji’s insofar as it is his view which salvages some ground for our moral intuitions whereas Haji’s leaves them floating at sea.   But before we explore this island in further detail let’s take stock of what we’ve established so far:

Pereboom offers a sharp objection to traditional views that tether morality to freedom and moral responsibility because:  if Hard Incompatibilism is true, no agent can live up to any kind of freedom (understood as a power to be autonomously determined) because he has no originary control over his actions or over his power to act for reasons (even if these actions flow from his faculty of responsiveness to reasons) and thus it is out of his power whether or not he will be able to for example, act according to his rational capacity to distinguish the Good/Right from the Evil/Wrong in any particular scenario that calls for moral judgment.   Just as we do not blame those who are born handicapped for their rational and physical disabilities, we cannot coherently blame agents who are born “morally handicapped”, so to speak.  Thus even if we attenuate the demands of freedom to include a mandate for only those duties which an agent can be said to ‘have the power or capacity’ to dispatch successfully, the formation of such a ‘power’ is subject to the same alien-deterministic causes that everything else in the universe is (given Complete Physicalism).  Thus it is a factor of chance or luck that determines not only what happens to us but also any powers through which we might mitigate or negate those determinations, i.e. what we are.   This holds if we understand ‘chance’ as either stemming from a deterministic causal chain that we did not initiate, or if we understand it as stemming from an indeterministic process of partial or true randomness.   On either definition, ‘chance’ simply means that what we are and what we do are the products of an external power, an external necessity that we do not have a say in creating or negating.

I will say for now that I think Pereboom’s conclusions are sound but only if we do not peek too closely at his first thesis of Hard Incompatibilism, which was, if we recall that:

(i)                  All of our actions lie on a “responsibility-undermining” continuum from alien-deterministic to truly random events

The part of the proposition that seems dubious to me, and worthy of inspection is the concept of ‘alien-determination’.  Alienation, otherness, externality; all of these only make sense in relation to opponent concepts of, for example, selfhood, identity, and internality.  But it seems to me an open question – if we grant Hard Incompatibilism – whether or not we can understand these notions without connecting them to a concept of freedom.   In other words, where , if not over one’s free will, do we draw the line that is to non-dogmatically individuate agents from the rest of the causal (either nomic or anomic) whole of which they are a part?  If we are not truly individuals but are rather pieces of a process stretching back through time to the beginning s of the universe, then it makes no sense to say that we are determined by an alien or external power, since we are part of whatever power exists, unless of course we can show somehow that we are in some non-trivial sense free from it.

But if it is no longer true that we are determined by an alien-causal power, since we are in no causally efficacious and objective way individuated from the entire cosmos then Spinozism seems to win the day, and all we get is nature naturing itself.  But what if, in fact, Hegel is the real victor on this account since he showed that the concepts Spinoza relies on to build up his system only make sense from a transcendental perspective that includes both practical reason (as Kant maintained) but also theoretical reason?  What if, in other words, Hegel shows that the two-standpoints view that allowed Kant to make sense of first personal practical reason applies equally to theoretical reason about ‘objective’ facts?

More on this to follow

Notes on Haji’s “Deontic Morality and Control: Ch. 3”

Dual-Control Requirements lead to ontic and not just deontic nihilism.

In chapter 3 of Deontic Morality and Control,  Ishtiyaque Haji uses Frankfurt examples to argue that while moral responsibility may not require the principle of alternative possibilities, the concept of moral obligatoriness, rightness and wrongness do in fact stand or fall with the coherence, or lack thereof, regarding the notion of the ability to do otherwise. It makes no sense, Haji maintains, to say that an agent should or must do something if it is the case that he cannot do that thing.  For Haji, “cannot” means both that the agent does not have the ability to do otherwise and/or that he does not have the opportunity to do otherwise.  The former refers to logical possibility while the latter refers to the material or actual possibility in the weightier sense.

There are three principles that ground Haji’s argument for the view that deontic anchors such as rightness, wrongness and obligatoriness require ‘dual-control’:

i) it is obligatory for one to do A only if it is wrong to refrain from doing A

ii) if it is obligatory for an agent to do A then that agent can do A and,

iii) if it is obligatory for an agent to do A then that agent can refrain from doing A

Now, in Frankfurt-type cases, it is stipulated that though the agent may enjoy the (logical) ability to refrain from doing A, the presence of the counterfactual intervener ensures that the agent lacks the opportunity to refrain from doing A, and thus, ensures that his doing A is externally determined.  Haji argues that though the agent in question may still be open to moral appraisals of responsibility and even praiseworthiness and blameworthiness, and even though the act in question (not the agent) can plausibly retain an ascription of goodness and badness, nevertheless it is incoherent in circumstances such as these to say that the agent is obliged to refrain from doing A or that he must or should refrain from doing A or that it is right to refrain from doing A and wrong to do A.  All of this follows from the three principles and ‘dual-control’ requirements stated above.

It goes without saying that the fallout of this argument is striking, for it entails that whenever an agent lacks the opportunity to do otherwise, he is no longer judicable as a moral object and deontic ascriptions of rightness, wrongness, oughts and obligations go out the window. Furthermore, if it is the case, as Haji believes, that metaphysical determinism obtains, then it follows that we are always in Frankfurt-type scenarios and (inconvenience and absurdity aside) that the history of the universe is one big Frankfurt example. Therefore, Haji insists, the concepts of deontic anchors can never be coherently used in judgment.  Moral concepts never apply to the real world and nothing anyone has done, is doing, or will do is right, wrong or obligatory if determinism is true.

I want to argue here that Haji’s argument is devastating not only for deontic anchors but also for ontic anchors as well.  If we accept his view, we should be nihilists not only about morality but also about ontology too, for the modal concept of alternative possibilities is built into ontological categories no less than in moral concepts. (Incidentally this is perhaps why Slavoj Zizek insists, for example, that to be a materialist is to believe that at a fundamental level the world does not exist).

Let’s jump right in to some examples of how modal actuality is determined by modal possibility:

J.J.C. Smart introduces the case of the falling plate.  You are washing the dishes, an ordinary porcelain plate slips from your hands and hits the floor but does not break.  You say to yourself “Phew! I am lucky it didn’t break; after all, it could have broken!”

Smart says that we are perfectly correct to say that the plate ‘could have’ broken even though it did not break because, and insofar as, it is part of the concept of plates that they are breakable because they are generally made of glass or porcelain and porcelain is brittle, or fragile, and that ‘brittle’ is just another way to say ‘breakable’ or ‘able to break’.

But under Haji’s dual-control criterion this would be false since the plate that fell did not break and though it had the logical (conceptual or analytic) possibility of shattering when it hit the floor, in fact, it did not have the opportunity to break because circumstances intervened and ensured that the plate remained intact when it hit the floor.  Thus, if it is true that plates are brittle, and brittle things are breakable, it is not true that whatever hit the floor is either breakable, brittle or a plate, insofar as the counterfactual intervener guarantees that there is no “genuinely” possible world in which the plate breaks when it drops to the ground.  In a merely logically possible world the object that fell to the floor may have been a brittle plate but in the real world it does not bear this predicate with integrity.

But if the object that hit the floor is not a brittle plate then what is it that hit the floor and did not break?  If we remove the predicate of brittleness or breakability from the object then its unity as a particular object disintegrates since for example, brittleness does not apply to the plate’s component molecules taken on their own and it is only true of porcelain that it is brittle in certain amounts and certain shapes.  But if we are no longer talking about the breakable thing, then we are no longer able to talk about the thing that was dropped and hit the floor for it is not the case that the entirety of the object hit the floor (P.S. what floor!? you mean this spot of tiling?). In this respect, it is unclear how the predicates that unify the object of judgment can be accurately applied since to predicate platehood of something is, it seems, bound up with predicating breakability of that very same thing such that if the latter predication is false then so is the former.

Or consider the case of Mary the shut-in colour scientist from Frank Jackson’s famous thought experiment. Mary is kept in a room her whole and is not allowed to see any objects or images that reflect red light. Though it is true that she never enjoys the opportunity to see a red object, is it also the case that she does not have the ability to see red? If we say that she has the ability to see red – insofar as we would be willing to say that had she died without having ever seen red it is still accurate to say that she “could have” seen red – is this the same ‘could have’ that Smart uses to say that the fallen plate “could have broken”?  In other words, is the unbroken plate’s breakability equivalent to colour-deprived Mary’s ability to see red?  It would seem that in both cases, the ability requirement is satisfied, but the opportunity requirement is not.

What happens if we use Haji’s argument for dual-control to critique the ontological predication of an object like a porcelain plate?

i) an object is only porcelain if it is breakable

ii) if any object is breakable it must be able to break

iii) if an object is able to break it must be possible for it to refrain from (or resist) breaking

But in Frankfurt-type case, the counterfactual intervener ensures that the object does not break. Therefore, if the universe is a big Frankfurt example then the object is not breakable, is not porcelain and it is not true that it could have broken when dropped because had the circumstances been different (e.g. colder temperature, faster velocity) the counterfactual intervener would come to the rescue and make it the case that the plate still does not break, perhaps by lowering the force of gravity or by cushioning the fall with denser air. Thus, in this scenario it can be said with absolute certainty that the plate will not break. The plate is therefore unbreakable.


However, the sense in which this porcelain plate is unbreakable seems peculiarly different from the sense in which an aluminum plate is said to be unbreakable.  What is the cause of this peculiar difference?  The difference is this: the porcelain plate is disposed to break when dropped and the aluminum plate is not.  Mary is disposed to see red whereas a shark is not. This is getting at the dual control distinction introduced by Haji when he discusses what the meaning of “can” is in “ought implies can”.


For Haji then, it appears that there will be certain predicates that can be ascribed to certain objects only if the latter satisfy one half of the dual control requirement, namely the logical half, which corresponds to ability (in an agential sense) or disposition in a physical sense. But there will be other predicates, specifically deontic anchors,  that require that the object of which they are predicated satisfies both requirements of dual control, namely, ability and opportunity to manifest that predicate.  This asymmetry makes sense. For it is only meaningful to speak of the opportunity to do otherwise in the context of an ability to do otherwise.  But in a deterministic, Frankfurt-type universe, material, or genuine alternative possibilities are ruled out of hand.  But this means that ontic concepts go the wayside with moral concepts since there will be very few things that have the opportunity to instantiate or manifest all of the abilities that we infer of them from their predicates. Many brittle things never break, after all.

But does ability make sense once wrested from the context of opportunity?  It does, but only if we are willing to grant a degree of reality to the potential or, in other words, to the category of essence.

More to follow…

Notes on Frankfurt’s “Alternate Possibilities and Moral Responsibility”

In his paper, Alternate Possibilities and Moral Responsibility, Harry Frankfurt argues that the principle of alternate possibilities (PAP) – the principle stating that an agent is only morally responsible for an action if there were alternative possibilities open to him at the time the action is executed – is false. Frankfurt reasons to this conclusion on the basis of a shrewd observation: there are cases in which an agent is morally responsible for his action even though it is in some sense true that he “could not have done otherwise”. Though I agree with Frankfurt’s observation, I believe that his reasoning to this conclusion is muddled by a misunderstanding of the relevant notions of action, causality and intervention.
Frankfurt wants to leave room for the following case in which an agent is morally responsible for an action even though it is not true that he could have done otherwise:

Suppose someone-Black, let us say-wants Jones to perform a certain action. Black is prepared to go to considerable lengths to get his way, but he prefers to avoid showing his hand unnecessarily. So he waits until Jones is about to make up his mind what to do, and he does nothing unless it is clear to him (Black is an excellent judge of such things) that Jones is going to decide to do something other than what he wants him to do. If it does become clear that Jones is going to decide to do something else, Black takes effective steps to ensure that Jones decides to do, and that he does do, what he wants him to do. Whatever Jones’s initial preferences and inclinations, then, Black will have his way. (Harry G. Frankfurt, Alternate Possibilities and Moral Responsibility, The Journal of Philosophy, Vol. 66, No. 23, (Dec. 4, 1969), pp. 829-839)

The metaphysics of Frankfurt’s intervener case will depend on what it is that the intervener is intervening on. If it is an event like ‘the killing of Smith’ or a state like ‘Smith is dead’ then these are the kind of things that can brought about through external intervention but these interventions will have no effect on the freedom or responsibility of the agent through whose action the intervener brings about the desired state or event. This is because it is actions and not events or states that confer upon an agent moral responsibility and that unlike the latter two metaphysical categories, actions are partly constituted by the self-conscious reasons in light of which they are performed. I want to argue that though external events and states can catalyze actions, they cannot cause actions.
Consider the following: If Black wants Jones to perform the action of ‘killing Smith’ then he wants something more than the event of Smith’s murder and the state of Smith’s being dead, namely he wants Jones to kill Smith and this action is not separable from the state of Jones’s being morally responsible for the killing of Smith. This is something Black can want or hope for but not something that he can himself bring about for it is part and parcel of the concept of action that it is performed self-consciously, that is, for reasons internally represented and motivating for the agent in action. But if the event of Smith’s murder and the state of Smith’s being dead are caused by Black’s puppeteering of Jones then the action ‘to kill Smith’ was performed not by Jones but by Black and the event of Smith’s murder and the state of his being dead are both things that Black, not Jones, will be morally responsible for since they will be traceable to Black’s action rather than to Jones’s. This is all the case because actions are not the kind of things that can be externally caused whereas events and states are.
Another way to put this is to say the following: an action is that which can only be caused in the first-person and which can be thwarted or provoked but never stopped or started by solely external , third-personal conditions. This is because it is internal to the concept of action that it be precipitated by the rational will of the acting agent. Thus, no matter what the scenario we can say generally that if an action is the expression of a rational will, then the action is necessarily free and the agent is necessarily morally responsible for that action.

The metaphysics of this case do not change whether the intervener is thought to be another person, an evil demon, or a material malignancy in Jones’s brain. For interveners can only be either events or states and neither has the power to cause an action though they may catalyze (or provoke) an action. Jones can only be responsible for events and states caused by his actions and his actions can themselves only be called actions if they were performed for reasons internal to Jones’s rational thinking, or, in other words, because he willed or intended to act in just this way in light of the facts he is able to judge rationally.

If Jones is able to examine the facts and decide for himself that he must murder Smith then Jones is morally responsible for the action of killing Smith, the event of Smith’s murder and the state of Smith’s being dead. For if Jones is the kind of thing of which actions and moral responsibility can be predicated, i.e., a rational agent, then it is necessarily and a priori the case that Jones’s actions will be caused by reasons represented self-consciously as necessary grounds for action. If Black wants Jones to kill Smith then Black must be able to make it the case that Jones has reasons which are self-consciously motivating for him to believe that he should, and is going to, kill Smith. But if Black actually has the power to do this – that is, by making it a fact that Jones has good reason to kill Smith – then he succeeds only in facilitating and ‘luring out’ Jones’s rational autonomy but then nothing Black does here prevents Jones from acting freely or from incurring moral responsibility for his actions.

Consider what happens if Black is an evil demon and has the power to make Jones decide to perform the following action: ‘to kill Smith because Smith is in intense pain and suffering due to a critical, incurable illness’. If Black has the power to make Jones believe that he should kill Smith for this reason then Jones is still exercising his rational ability to act for good reasons and thus his autonomy and therefore moral responsibility are intact.

Now, results differ if Black, the demon, has the power to make it the case that Jones decides to perform the following action: ‘to kill Smith because the Vancouver Canucks lost the Stanley Cup to the Boston Bruins’. Here, Black has the ability to destroy one agent, call him Jones, and create ex nihilo another agent, call him Evil Jones. But even in this scenario Black is still powerless to cause Jones to perform an action since he must erase Jones’s original (shall we say healthy) rational faculty of responsiveness to reasons in order to get a new agent, Evil Jones, whose rational faculty is corrupted, to freely decide to kill Smith. That we can be certain we are dealing with a new agent is ensured by considering the fact that Jones, previous to his molestation at the hands of the demon, would not have taken the loss of a hockey game to be good reason to murder someone, (even if Jones is a Canuck and Smith is a Bruin). On the other hand, if Jones is and always has been the kind of person to go on a killing spree whenever his team loses an important game then nothing the demon did jeopardized Jones’s acting autonomously and from his own nature. Here, Jones is simply a wretched individual with a murderous, violent character. Black merely put him in a state whereby he could freely express his own nature.
The point to realize is this: if Jones is the kind of person to kill people for no good reason then Black’s intervention has no effect on Jones’s freedom and responsibility since nothing Black does has prevented Jones from exercising his natural autonomy, wretched though it is. Conversely, if Jones is not the kind of person to kill people for no good reason then the fact that the demon is able to make him do so means that Black has robbed Jones not only of his agency but also of his identity and has in fact created a new agent, Evil Jones, who is responsible for killing Smith. Either way, Black has not succeeded in “getting his way” if we understand the latter to be intervening to bring about the fact that Jones (and not Evil Jones) kills Smith. For neither Black nor Evil Jones can bring this about, only Jones can.

Our reflections tease out a hard fact to swallow about human beings: our identities as rational agents are fragile, perishable and to a large extent externally conditioned. We can lose our identities as persons though maintain our identities as animals, bodies, and collections of molecules largely through the contingencies of circumstance and outrageous fortune. We also begin to see why the mind and body have been strongly differentiated in the tradition of metaphysics and why it is not an easy matter to put them back together. But this is a topic for another time.
Our reflections also illuminate Frankfurt’s substitute for the principle of alternative possibilities, namely, the principle that an agent is not morally responsible for an action if he performed that action only because he could not have done otherwise. For now we are in a position to understand what is meant by “only because”: it is a vague attempt to articulate the essential relevance that self-conscious representation of reasons has in relation to the concepts of freedom and moral responsibility. I can only be morally responsible for killing someone if my reasons for doing so are other than the fact that I am coerced to do so for good reason. If Smith dies at the hands of Jones only because (i.e. only in light of the knowledge that) Black has made it the case that Jones is powerless to prevent Smith’s death then Jones is not morally responsible for killing Smith because in fact Jones did not kill Smith; Black did. And if Smith dies at the hands of Jones only because or in light of the knowledge that the Canucks fell to the Bruins in game 7 of the Stanley Cup Playoffs then again, Jones did not kill Smith; Evil Jones did, for Jones is not the kind of person to kill ‘for sport’, so to speak.
Here, we get a picture of the causality of action that is distinct from the causality of mechanism for though it is true that Jones is in some obvious way the efficient cause of Smith’s death, it is still not the case that Jones killed Smith. Contemplating action theory should give us clues leading to the discovery of a unique concept of causality that goes beyond the limits of naturalistic explanation that appeal only to mechanical or efficient notions of cause and effect. But this is a topic for another time.

Notes on Slote’s “Ethics Without Free Will”

The consequences of our actions are largely determined by forces external to our control as agents and thus whether or not one’s actions can lead to good or bad results is thought by many philosophers, and justifiably so, to be outside of one’s power. This entails that whether or not one’s actions are to be deemed morally right or wrong hinges on elements beyond the executive purview of the acting agent. Following Thomas Nagel, Michael Slote calls this phenomenon ‘moral luck’.
  If moral luck is an accurate picture of moral ascriptions then freedom to do otherwise is neither necessary nor sufficient as a condition for one action to be labeled right or good and another wrong. Slote points out that though this seems to vitiate deontic theories of morality it leaves utilitarian theories as well as virtue ethical theories in tact, for neither of these positions equates the right and the good with what an agent has as a duty to freely choose.
  For his own part, Slote takes it upon himself in his paper “Ethics Without Free Will ” to explicate briefly the virtue ethical side of this forked alternative to deontology.  However, it is unclear whether or not this alternative will do the work Slote would have it do.
  Slote implies that whereas deontic acts presuppose agential control virtue ethical acts do not. Slote believes we can coherently speak about morality (under a Nicomachean rubric) without tethering our concept of morality to our concept of control. Whether or not his argument is successful will depend then on our understanding of the notion of ‘control’.
Leaving for later an examination of the concept of control, we can say presumably,  that Slote thinks that virtue ethical acts presuppose conceptual and rational differentiation.  That is, in order to speak of morality within a virtue ethicist framework one must be able to distinguish what is prudential from what is rash and what it is to be a good father from what it is to be a bad one, for example. It was indeed Spinoza, whose ethical theory Slote looks upon favorably in this essay, who said that “omnis determinatio est negatio”. This means that for something,  anything, to be determinable even descriptively, it also must be distinguishable from any and all of those determinations that it is not. We must be able to say what something is not, in order to be able to say what something is. To be able to say what a red thing is requires that we are able to say that it is not green, blue, colorless etc. Applied to virtue concepts,  we likewise must be able to say that an act is not rash and not impractical if we are to say that it is prudential.
  If we assume Spinoza’s metaphysical principle, call it the principle of determinate negation, then it is also the case that a thing is only what it is insofar as it is distinct from what it is related to negatively, i.e., what it is not. Red is only a determinate quality in negative relation to other colours on the colour wheel; removing these rival colours entails that redness loses its qualitative determination as a colour. (It may retain its quality as a shade, however, but this would be to judge it under a different qualitative scheme which has its own continuum of possible determinations and negative relations, judged in this case under the magnitude of opacity rather than colour). The ability to see red things only makes sense in the context of the ability to see blue things, green things and yellow things. Red is only red iff it can be distinguished from the colours that it is not.
  Similarly,  when the qualitative scheme is virtue instead of colour, what it will mean to instantiate any one virtue, say, prudence, will include by necessity the ability to instantiate behaviour or demeanor that is not rash, not impractical,  not frivolous and so on, and to distinguish these instantiations. This principle of determinate negation only makes sense within the context of a power or ability to differentiate what something is from that which it is not.  That is, we can introduce the idea of a ‘virtue wheel’ that, like the colour wheel, allows us to measure determinations of qualitative virtue in relation to its ‘rival’ or alternative determinations flanking it and opposing it on the virtue wheel. A being unable to relate determinations on the wheel in this negative way would be the moral equivalent of colourblind.  To lack distinctness is to lack any quality whatsoever. Thus we can say that a theory of morality that exchanged prescriptive concepts of duty and moral obligation for ‘merely’ descriptive concepts of virtue still requires the capacity to judge qualities on the virtue wheel by relating them (comparing and contrasting them) to their relevant alternatives.  Such a capacity presupposes then the ability to discern what is actual (prudence) in any circumstance being judged by comparing it to what are its possible alternatives in that circumstance (rashness, impracticality etc.). Having this capacity to “tell the difference” is essential to the coherence of purportedly merely descriptive moral schemes such as Slote’s virtue ethics since this comparitive capacity is essential to any and all determinative (i.e. information-processing) mechanisms generally. If we do not have the ability to say that an apple could have been green or yellow then we can’t say that it is red. And biologically speaking, an animal that cannot see blue and green and yellow cannot see red either. It may see bright things and dark things and maybe even coloured things and uncoloured things but it won’t see them as red.
Likewise, if we can’t say that an agent’s actions could have been rash or impractical then we simply do not have the ability to say that the agent instantiated the virtue of prudence. When we say that a moral choice implies a free choice and that a free choice implies the ability to do otherwise we simply imply that a moral ascription is the kind of judgment that we can apply to agents that have the capacity to “tell the difference” between the moral quality of actions in the same way we know that colour ascriptions are the kind of judgments we can apply to instantiations of a power to tell the difference between a range of wavelengths and frequencies of light visually. This means that ‘control’ is nothing other than one’s having the ability to judge determinations that actually obtain by comparing them to determinations that do not obtain but could. 
  If the universe is determined in such a way as to allow for this kind of determinate negation, then this is the only concept of control that we need to rescue both moral prescription and free will from the clutches of certain morally nihilistic accounts of determinativist metaphysics.