Alternative Possibilities as 'Sufficient' for Free Will?

I'm puzzled by a claim in Robert Kane's new book, A Contemporary Introduction to Free Will.  In chapter 11, where Kane is talking about the importance of ultimate responsibility (UR) for free will, he attempts to show the connection between UR and the alternative possibilities condition for free will (AP).  Here is the passage I have in mind:

"To understand the connection between AP and UR (alternative possibilities and ultiamte responsibility) we have to return to a claim made earlier in this chapter--that having alternative possibilities is not sufficient for free will, even if the alternative possibilities are undetermined.  Some incompatibilists have thought that all one needs for free will is alternative possibilities plus indeterminism: in other words, it is sufficient for free will that we be able to do otherwise in a way that is not determined by our past" (124).

Now, I'll agree that often the almost exclusive focus on AP over UR may have given the impression that AP was more important than UR.  And surely Kane (and others) are to be thanked for showing that something along the lines of UR needs our attention.  But I'm puzzled that anyone would claim that satisfying AP is sufficient for free will.  Does anyone know who Kane might be talking about here?

Comments

[Mark Smeltzer]

Ginet?

[V. Alan White]

I would say that such remarks echo the sentiments of William James in his "Dilemma of Determinism". It's pretty clear there that any AP account--including any version of simple indeterminism--would suffice for a kind of freedom that stood in conceptual opposition to determinism. It's minimal and doesn't address something like van Inwagen's Mind argument, but there are still some that would posit SI as a kind of freedom nonetheless. APs can't be relinquished by incompatibilists, and that could be construed by some as a sign of weak sufficiency.

[Mark Smeltzer]

Hmmm... "APs can't be relinquished by incompatibilists, and that could be construed by some as a sign of weak sufficiency."

I'm not sure this is good reasoning. Suppose we say that in order to be rich, someone has to have a lot of money, equity, valuable stocks, etc. Then we take your reasoning and say that because Smith has a million bucks, this could be construed as a sign of richness. However, the truth of the matter is that though Smith does have a million bucks, he is actually ten million in debt! That's *far* from rich, isn't it?

Smith isn't even close to being rich! He does satisfy the necessary condition, yet this condition is hardly sufficient for determining richness -- let alone the likelihood of richness!

Considering AP as a necessary condition for freedom, why think that is any different in nature than the condition discussed here for richness?

[Mark Smeltzer]

Hmmm... "APs can't be relinquished by incompatibilists, and that could be construed by some as a sign of weak sufficiency."

I'm not sure this is good reasoning. Suppose we say that in order to be rich, someone has to have a lot of money, equity, valuable stocks, etc. Then we take your reasoning and say that because Smith has a million bucks, this could be construed as a sign of richness. However, the truth of the matter is that though Smith does have a million bucks, he is actually ten million in debt! That's *far* from rich, isn't it?

Smith isn't even close to being rich! He does satisfy the necessary condition, yet this condition is hardly sufficient for determining richness -- let alone the likelihood of richness!

Considering AP as a necessary condition for freedom, why think that is any different in nature than the condition discussed here for richness?

[V. Alan White]

But Mark--if I may--that example takes the available bucks to be effectively cancelled out by debt, which is a conceptual correlate to something like a Frankfurt intervener in ruling out APs. I grant that many take themselves to be to solvent enough if they have good credit--forgetting that they only are making payments to cover their basic interest on the debt that they have run up in maxing out their cards. But such self-deceivers are their own Frankfurt interveners, so to speak, in ignoring the cancelling effect of debt-to-credit. If they think they have monetary APs after maxing out their cards, a waiter's "card rejected" message should correct any misjudgements of supposed economic freedom.

[Eddy Nahmias]

To borrow an example similar to one in Mele (1995). Suppose an agent satisfies whatever sort of simple indeterminism a libertarian wants (e.g., Kane's). Now, imagine a manipulator who is able to control that agent such that she develops two "self-networks" that compete when making a particular choice, one of which most desires to kill her father and one of which most desires to kill her mother, only one of which can be acted on (I see nothing incoherent in this possibility). Before the manipulation, she would never have dreamed of killing either of her parents. She can only act on one of her "self-network" desires and there is about a 50% chance (objective, indeterministic, based on quantum events in just the right place in her brain, etc.) that she will choose to kill her father and about 50% that she will kill her mother.

If she is free and responsible just because of the indeterministic process, then libertarians have a very counter-intuitive position. The fact that manipulation cases can be developed for act-libertarian theories (and I think agent-causal theory too, but it may take more work) leads me to believe that libertarians cannot use such arguments against compatibilists. I'll work on developing the argument fully.

[V. Alan White]

Hello Eddy (again if I may)--and thanks after the fact for inspiring my latest song! (Or if that's a point of shame--sorry!)

What you suggest ala Mele is promising as another variation of Mind-style argument against indeterminism, and I have no quibble with that. The fact that the SI process is imposed seems to blend in more intervener problems, but one could just fancy that up with a more "natural" source of it within human nature (a genetic variation that produced the needed quantum-sensitive decisions when triggered by certain thoughts of mom and dad).

Still, to get back to the original question, there have been those who think that bare-bones AP SI of mind yields a form of freedom, even if questions of responsibility are problematic. I guess the question could be put this way: if an AP SI system of mind isn't really "free" in some sense, then what is it?
Why NOT call such a system "free"?

[Mark Smeltzer]

"I guess the question could be put this way: if an AP SI system of mind isn't really 'free' in some sense, then what is it?"

Why not just call it a hodgepodge of random event outcomes?

"Why NOT call such a system 'free'?"

Unless one believes that random composition is, in some sense, equivalent to freedom, there would be clear motive to reject that ascription.

[V. Alan White]

I guess the SI's intuition is that if determinism entraps, then the opposite of determinism unleashes. I have some sympathy for that, though it's a pretty minimal sense of being free. In my classes I use a more neutral description of "open" vs. "closed" futures to contrast such systems of events as they evolve through time.

[Eddy Nahmias]

Here's a simplified summary of my view on this issue: simple indeterminism is sufficient (and necessary) for a certain type of freedom--namely, and tautologically, freedom from determinism, or from there being one future GIVEN a particular past and laws of nature. This freedom has a sort of appeal if one thinks of determinism as limiting possibilities. However, I think this freedom (a) is manifestly NOT sufficient for moral responsibility (as my example above suggests)--nor do I think it is necessary (for standard compatibilist or semi-compatibilist reasons); and (b) it can be obtained as long as indeterminism affects human action, so that it does not matter if it happens inside the agent at the point of choice (a la Kane) or at some point during deliberation (a la Nozick, Dennett, Mele, Ekstrom) OR outside the agent (e.g., a quantum event whose objectively indeterministic outcome influences which reasons come to an agent's mind during deliberation which thereby, perhaps deterministically, causes which choice she makes). IF one doesn't think this freedom gets you moral responsibility but IF one thinks it is interesting for other reasons, I can't see any reason why it matters whether the indeterministic process or event occurs inside or outside the agent's skin (I'm working on this argument for a paper).

This kind of freedom should be called, I think, "openness" and should be clearly distinguished from "free will" (the abilities relevant to moral responsibility). That is, it MAY be that openness is necessary for free will (of the sort relevant to MR) as incompatibilists argue, but it is certainly not sufficient (as most incompatibilists recognize--I have no idea who Kane may be referring to, Alan). Or it may be, as I and most other compatibilists believe, irrelevant to free will in this sense. Fischer's semi-compatibilism, as discussed in one of the first posts here at GFP, is the view that determinism precludes freedom in this openness sense but not the freedom (guidance control) required for moral responsibility. The terminological confusion leads people to think semi-compatibilism means determinism precludes *free will* but not moral responsiblility, but that's misleading unless one is willing to give up the term "free will" to incompatibilists, which would be a mistake.

Finally, I think it is important to note that just as libertarians (and skeptics) can say that "compatibilist freedom" is valuable in many respects (e.g., it's good to be reasons-responsive and susceptible to incentives), compatibilists can say that "openness" (i.e., the freedom simple indeterminism gets you) is valuable in certain respects (e.g., it may satisfy some people's hope that the universe is a "garden of forking paths" even if it doesn't give us control over which path the universe goes down). Personally, I have a certain penchant for this sort of "openness".

[Eddy Nahmias]

Alan, I hadn't read your composition notes for your song (though I'd read the lyrics which are great--I'll have to try my hand at a free will song; I've just been writing children's songs for my sons). For the record, I think van Inwagen label's himself a "mysterian" so I did not come up with that (the term--actually "new mysterian", however, was coined by my advisor, Owen Flanagan, to refer to people like McGinn regarding their view that consciousness is an insolvable mystery--I think Owen got the term from a band name).

[V. Alan White]

No doubt from "? and the Mysterians"--famous for "98 Tears". Hadn't thought of that! Thanks.

[V. Alan White]

Oh and by the way Eddy, I agree with most all you said in the post previous to the song reference. My only concern with introducing interveners in examples is that it can be a distraction from and deferment of the question of responsibility that can be more directly addressed in the way you indicate. Anxious to see your paper!

[Josh Narins]

I find the term UR problematic.

I don't need UR to tumble to the right or the left, having been pushed out of a plane.

I will hit the ground, and I don't have UR for falling, but I can stretch out my arms and legs and catch some air.

Or even, if I'm really lucky, open an parchute.

[Kip Werking]

Kevin raises a really good question here!

For example, Eddy Nahmias writes in his recent paper:

“However, some incompatibilists argue that the problem with determinism is not that it threatens the AP condition but that it conflicts with an agent’s ability to be the ‘ultimate source’ of her actions. So, while the confident agent thought experiment may suggest that the AP condition is not essential for free will, it does not challenge this ‘source incompatibilism’ since there is no reason to think she is the ultimate source of her actions (at least if she exists in a deterministic universe). If my thought experiment helps motivate more incompatibilists to take this position, as have Frankfurt cases, then that will be a significant result, since the arguments for source incompatibilism are, I think, even more controversial than those relying on the AP condition. I also think the confident agent can help us see why this is. Even though she is not described as initiating some choice such that there are no complete causal explanations for her choice, she clearly endorses the reasons that are essential proximate causes of her actions. As long as those reasons themselves arise through her own deliberative process and are not manipulated in some autonomy-undermining way, it is not clear why we should demand that, if she is to be considered free, those reasons themselves cannot have sufficient prior causes.”

Nahmias considers this a "significant result" because it forces incompatibilists to move towards a sourcehood based position and he considers these positions (for reasons he doesn't mention) weaker than those based upon AP.

But as far as I can tell, most of the prominent incompatibilists upon free will have already granted that sourcehood is essential. Pereboom and Kane, in particular, explicitly emphasize the importance of sourcehood. The notable exception here is Van Inwagen, who may also acknowledge the importance of sourcehood (I am not sure what his position on the AP/UR distinction is)?

Perhaps Nahmias is referring to Van Inwagen's Consequence Argument, and considers it stronger than Pereboom's four-step argument for incompatibilist (does Kane have an argument for incompatibilism)?

[Eddy Nahmias]

A quick response for now (this Sunday morning). In this article my target is the intuitive appeal of alternative possibilities. That's why I focus on AP rather than ultimate sourcehood. Though I think the Consequence argument is problematic, I think Pereboom's and Strawson's arguments are even more problematic (as is the notion that ultimate sourcehood is necessary for moral responsibility--except perhaps for "ultimate moral responsibility", the sort that would make it appropriate to be subjected to *eternal* damnation). If I'm not mistaken, Kane relies primarily on Consequence-style arguments to motivate his incompatibilism, even though he also emphasizes ultimate responsibility (UR). I do think that compatibilism is bolstered by any arguments or thought experiments that force incompatibilists to rely on (a) ultimate sourcehood as a necessary condition for free will/moral responsibility or (b) Pereboom- or Strawson-style arguments rather than the Consequence argument (briefly, the former is undermined by Mele's response that shows manipulation not determinism to be the relevant threat in the scenarios, and the latter relies, I think, even more problematically on questionable transfer of non-responsibility principles than does the Consequence argument).

[Joe]

In his English version of the Consequence Argument, van Inwagen writes:

“If determinism is true, then our acts are the consequences of the laws of nature and events in the remote past. But it is not up to us what went on before we were born, and neither is it up to us what the laws of nature are. Therefore, the consequences of these things are not up to us” (van Inwagen 1983, 16).

In his formal version—at least the Third Argument (1983, 93-104, and 1989)—he utilizes a modal operator to (apparently) convey the concept of things-being-not-up-to-us: ‘Np’ stands for “p and no one has, or ever had, any choice about whether p” (1989, 404). Later, this is taken to mean ‘no one is able to act so as to ensure that not-p’ (van Inwagen 1983, 67-8, and fn. 31, 233-4).

My take on all of this is that van Inwagen adopts a kind of source view of free will (S has free will only if something is up to S) but holds that in order for something to be up to S it must be the case that S could have done otherwise.

I’d be interested to know what anyone thinks about this interpretation.

The interpretation is important, for it suggests that the Consequence Argument is a kind of basic argument for incompatibilism. One may interpret the phrase ‘something is up to S’ as ‘S is the source of something’ or as ‘S could have rendered something false’ or a combined view, like the one I am suggesting that van Inwagen has. (My thoughts about Kane and Ginet are similar on this.) Even in the case of Pereboom’s four case argument, there is an appeal to a lack-of-source transfer principle. For otherwise why would manipulation cases be threatening to our freedom? The principle here is something like: the manipulated event was not up to S, nor was it up to S that the action resulted from the manipulated event, so the action was not up to S. (Though this is expressed in terms of events rather than propositions.)

[Kip Werking]

I agree, Joe. And this suggests to me that most or all incompatibilists are motivated by concerns about sourcehood, and the quest for AP is an (in my opinion, misguided) effort to achieve that sourcehood. So compatibilist arguments which only address AP-type incompatibilism, without addressing more fundamental concerns about sourcehood, will not have much persuasive force with most incompatibilists.

[Mark Smeltzer]

"So compatibilist arguments which only address AP-type incompatibilism, without addressing more fundamental concerns about sourcehood, will not have much persuasive force with most incompatibilists."

I agree with the incompatiblists on that regard: I think sourcehood is crucial to freedom (and responsibility); and I'm sure other compatibilists would agree. The area where we disagree is how on to interpret the "up to us"-type phrases. Compatibilists (well, at *least* semi-compatibilists, if not all compatibilists) interpret these phrases according to an ideal of self-expression and incompatiblists interpret those phrases according to an ideal of self-definition. So, I agree with Kapitan’s analysis of the consequence argument: in the end, it does not provide the incompatibilist with the necessary means to circumvent the need to deal with the various compatibilist accounts of “up to us-ness”.

(In fact, I'm so enthusiastic about sourcehood that I've started calling myself an agent-causal compatibilist, which has really annoyed a few of my agent-causal libertarian friends! But, I’m not totally sold on the label just yet.)

[Joe]

Mark: Check out Ned Markosian’s “A Compatibilist Version of the Theory of Agent Causation.” Pacific Philosophical Quarterly 80 (1999): 257-77. I know that Ned has a follow up paper to this but I’m not sure where or if it is published.

I want to go on record as agreeing with most of what Eddy says above, save for three points noted below. He notes what I take to be the central problem for incompatibilism in the switch from the traditional view to the source view: they must hold that “ultimate sourcehood is necessary for moral responsibility.” But the compatibilist is free to respond that sourcehood need not be ultimate. Actually, the situation is even worse since both Markosian and Kant seem to provide a compatibilist account of ultimate sourcehood.

Three points for Eddy, though.

1/ You say that “Mele's response … shows manipulation not determinism to be the relevant threat in the scenarios.” But the source incompatibilist may contend that the real issue is whether some source-transfer principle holds. Given that it does, cases of determinism are, for such an incompatibilist, kinds of manipulation cases. As I understand him, that is the conclusion of Pereboom’s argument. (I don’t think that this argument ultimately works but not for the reason given here.)

2/ You suggest the use of the term ‘openness’ to classify the kind of alternative possibilities that the traditional incompatibilist favors. But I don’t like that use of the term in this sense. As Ginet uses the term, for instance, it is clear that something more than this is meant. Whether or not, for Ginet, a given action is open to an agent is not merely a matter of whether or not the action is possible given that we hold fixed all of the facts about the broad past. For an action to be genuinely open to an agent it must be possible for the agent to be its source (as I read Ginet), and that need not be the case just because the action is possible given the broad past. Again, Ginet like van Inwagen seems to offer a combined view, which recognizes the importance of both sourcehood and alternative possibilities of action in a theory of free will.

3/ This is really for everyone, since everyone is likely to disagree. It still strikes me as odd to say that S is the source of something even though S could not have done otherwise. I think that it is important to note that the intuition behind the kind of combined view sketched out above is that alternative possibilities of action are necessary for sourcehood. We all know that Frankfurt-style examples cast doubt on this link but both Ginet and van Inwagen are critical of those examples. Though I’m a compatibilist, I tend to side with them on this point.

By the way, a longer version of my “Compatibilist Alternatives” is coming out in the Canadian Journal of Philosophy in September. Here is the closing paragraph, which indicates a new section of the paper not included in the version posted previously on this blog (one that discusses Wallace (1994) in more detail).

"Traditional compatibilism claims that alternatives are essential to moral responsibility and that moral responsibility is compatible with determinism. In light of the traditional argument for incompatibilism and the argument against PAP, these claims are contentious. According to my two-‘cans’ view both arguments suffer from the fallacy of equivocation: the general reading of PAP is true but the all-in reading is false. This claim is supported by appeal to mitigating cases. These cases are of two kinds: exemptions and excuses. All of these cases involve situations in which one of the powers of reflective self-control is absent, lost, or hindered. ... Certainly the Frankfurt-style examples prove something but what they prove is that the all-in reading of PAP is false."

Perhaps I should have posted this in the shameless self-promotion thread!

[Mark Smeltzer]

Joe,

A friend told me about Markosian's paper after I suggested that I had coined a new compatibilist category (Smack down!! It really is hard to be innovative these days...), but I have not yet had a chance to read it. I'm curious to see how his concept of agent-causal compatibilism compares.

Regarding your third point, I support the modal analysis of 'can' in which S would have done differently at T if the world at T had been different. However, if the world at T and S's nature at T are held fixed, then S could not have done otherwise at T and yet I find no cause to think that S is not the source of S's action at T.

The main idea is that S's features at T in relation to the state of the world at T joinly determine S's action at T (given the presumption of determinism, S's particular action at T happen necessarily). This seems like genuine (compatibilist) sourcehood to me, even though when it is put this way it is not dependent on the inclusion of a modal "could have done otherwise at T" clause. Since I believe this to be true, I fall into the semi-compatibilist camp by definition. So, even though I ultimately support the traditional modal reading of "could have done otherwise", I don't think it is actually relevant to the question of sourcehood.

Based what you've said so far, I can't tell whether you'd find this analysis problematic... maybe I should read your paper :)

[Joe]

I don't think that your analysis is wrong, Mark. I just think that the word 'can' is equivocal. Borrowing an example from Unger (Philosophical Relativity), suppose that there is a pianist on a train. Can the pianist play a piano? Since he is a pianist, it is clear that there is some sense in which he can play a piano. Yet the world would have to be quite different indeed in order for it to be the case that he would have played the piano at the time in question. Any theory of free will that rests on this sense of 'can' is problematic.

[I should note that Unger uses the example to distinguish between contextualist and invariantist analyses of 'can' and not to suggest that the word 'can' is ambiguous.]

I distinguish between the all-in sense of 'can' and the general sense of 'can.' (I am not claiming to have originated this distinction, of course.) Your analysis of 'can' is one possible analysis of the all-in sense of 'can.'

Most philosophers believe that the all-in sense of 'can' is where the action is, that philosophers who hold the traditional view of free will (where S has free will only if S could have done otherwise) ALWAYS intend the all-in sense, and that to suppose that the general sense of 'can' plays a role in a theory of free will is to be deeply confused. I think that Frankfurt-style examples show that the all-in sense of 'can' is irrelevant to attributions of moral responsibility.

In an effort to show the importance of the general sense of 'can,' I appeal to Strawson's distinction between excuses and exemptions. Strawson also makes a subdivision between what I call I call ‘transitory exemptions’ and ‘enduring exemptions.’ Examples of enduring exemptions are: “‘He’s only a child’, ‘He’s a hopeless schizophrenic’, ‘His mind has been systematically perverted’, ‘That’s purely compulsive bahaviour on his part" ("Freedom and Resentment"). It seems clear to me that such examples illustrate the importance of the general sense of 'can' in a theory of free will and moral responsibility.

The rest of the story gets pretty complicated but I'd be glad to send you a copy of the paper. But I think that the above is enough to show that a general reading of PAP survives Frankfurt-style examples.

[Mark Smeltzer]

First off, thanks for sending me a copy of the paper. I look forward to reading it.

Second, I wanted to comment about some of the points made above. I recognize a distinction between an agent who *is* responsible and an agent whom other agents have epistemic warrant to treat *as* responsible. The two do not always line up.

It seems like the general sense of 'can' being referred to in your post relates to: 1) agents whom we know to be minimally competent within given contexts and 2) agents whom we have reason to believe were in a position to demonstrate their competence.

It is easy to see how the pianist on the train satisfies (1) but not (2) and that explains why we lack warrant to hold him responsible for the fact that he doesn't play the piano.

Moreover, a failure to meet one of these also explains the excusing conditions you mentioned: young children are not generally competent at making weighty decisions; a schizophrenic is often not in a position to demonstrate competence (assuming competence exists); etc. It is important to note that in all of these kinds of cases it is possible that the agent is in fact responsible nonetheless! But because we are not in a position to know, we lack warrant to treat them as if they were responsible.

However, the case of the person who has been systematically perverted is different in a fundamental way. It falls into the category of violating metaphysical responsibility: it is highly probable (or perhaps it is meant to be by definition) that this person is *not* responsible, even if we are warranted to treat them as if they were!

Thus, I don't think these alternative interpretations of "can" apply to the question of metaphysical responsibility. Interpreting "can" according to agent competence and ability to demonstrate competence grounds our (epistemic) right to treat others *as if* they are responsible -- I readily accept this.

However, I think it is clear that agent competence is irrelevant to the question of metaphysical responsibility (and metaphysical desert). With respect to this question, I believe that sourcehood is the primary concern and that competence (and demonstration) plays little to no role here.

[Joe]

Thanks, Mark! I understand the distinction that you are noting above. But my claims are metaphysical not epistemological.

It seems clear to me that a 1-year-old child, for instance, lacks the cognitive capacities that are necessary for moral responsibility. This is undeniable, as far as I can see. One might raise some skeptical worries about whether teenage children have or lack the requisite capacities or whether schizophrenics, psychopaths, etc. have or lack the requisite capacities. These are important worries but, as you note, they are independent of any metaphysical claims about the nature of morally responsible agents. As I see it, if an individual lacks the requisite capacities, then he or she is not an appropriate candidate for moral appraisal. Whether or not we know this -- or can know this -- is beside the point. So, as I see it, agent competence is NOT “irrelevant to the question of metaphysical responsibility,” so long as agent competence involves having the requisite capacities.

Clearly, more work needs to be done. In "Compatibilist Alternatives" I identify the requisite capacities with the “powers of reflective self-control” (Wallace 1994, 2) but it would be nice to have more detail than this.

I should add that on my view sourcehood is not a separate issue. Sourcehood is a power or a collection of powers and is, thus, included within the powers of reflective self-control. Once we've said all we need to say about these latter powers, we will have said all we need to say about sourcehood.

My view is very similar to the combined view that I claim Ginet and van Inwagen adopt, except for the fact that (1) I think that general abilities are necessary for moral responsibility but all-in abilities are not, and (2) I think that sourcehood is essential to moral responsibility but ultimate sourcehood is not.

[Mark Smeltzer]

Joe,

As I see it, the capacities necessary to make desert work at the metaphysical level are present from birth: agents have a moral quality and that moral quality *is* expressed through their actions -- and this is essentially how metaphysical desert works.

The problem is that until agents mature and develop competency to make informed moral decisions, agents will not be appropriately connected to their actions; that is their actions will not be adequately expressive. This is a metaphysical issue, but it also gives rise to the epistemic issues I mentioned previously.

The big point I'm making here is that metaphysical desert must come first in the chain of responsibility. If agents do not ultimately deserve anything, then the more localized sense of responsibility cannot get started. Not everyone is going to be inclined to agree with me here -- especially those who hold to the ideal of self-definition and those who deny moral realism and/or moral agency altogether.

The difference between our views seems to be that I believe the metaphysical conditions of responsibility can be explained primarily in terms of sourcehood, and that sourcehood itself is what gives rise to the secondary considerations you've mentioned, which are best expressed in terms of an agent's competency within a general context, an agent's ability to demonstrate competency within a specific context, and the agent actually being present in that specific kind of context.

The reason I believe these secondary considerations are necessary is because they are entailed by the prior commitment to sourcehood. So, I suppose my view could be described as a "combined" view. However, since the lack of these secondary conditions serves as an indicator that the proper chain of sourcehood has been obstructed, the agent in question is non-responsible primarily for that reason -- which is another way of saying that these secondary considerations are simply part of the whole "package" entailed by the concept sourcehood. I don't think these secondary considerations survive as independent concepts without being grounded in the concept of sourcehood itself.

So if my view is a combined view, it is so in a week sense because I identify these secondary considerations as essential "parts" of the concept of sourcehood, and to the degree that they are not, I identity them with the epistemic side of the equation.

In the end, our views may be quite similar. I suppose it turns on how you establish the basis for metaphysical desert. Do you think desert comes first and grounds responsibility as I do? Or do you think responsibility comes first and grounds desert? These are not questions common to the historical discussion, but I believe much rests on which stance we take.