Can a libertarian account deal with worries about manipulation?

I have just become a contributor to GFP at the invitation of Neal Tognazzini.  My homepage (not the most elegant I’m afraid) is here.

I want to take up the libertarian cause by responding to two points made by Joe Campbell in his comments on Haji’s “The Manipulation Argument”.  First, he says “no libertarian account has successfully illustrated the key to free human action,” and suggests that a libertarian response to worries about manipulation can be no better than a compatibilist response; and he criticises agent causation as being a “primitive unanalyzable concept”.

I’ve been trying to develop a libertarian account that does focus on how and why human action is free and that can deal with manipulation worries, as well as with Galen Strawson’s responsibility argument.  One article, an edited version of which was published in Times Literary Supplement last year, was the subject of some discussion on this website, but this discussion tended to skate over the main arguments.  (The article was written in part as an answer to Galen Strawson’s argument, and since he as philosophy editor for TLS okayed it for publication, I was encouraged to hope he saw it as having some merit.)   I also published an article (“Making our own luck”) to similar effect in Ratio in September last (which I can’t put on the internet until next September), and another on the role of consciousness in Philosophy Now in Jan/Feb this year.

The (bare) gist of my position is that, given our circumstances and laws of nature, the way we are provides available alternatives, inconclusive reasons (and how they appeal), and unconscious tendencies, and also the capacity to decide between the alternatives on the basis of the reasons; and what we do is what we decide in exercise of that capacity.  The constraining effect of the way we are is limited to determining alternatives, reasons and unconscious tendencies, and our decisions are not otherwise constrained by any distinguishing features of the way we are (we are all alike in respect of our capacity to decide):  to this extent, we are truly responsible for our decisions.  That applies whether the way we are is due to manipulation and/or natural causes and/or prior decisions.

I support this by arguing (1) that conscious experiences can make a positive contribution to decision-making; (2) that this contribution is not one wholly determined by rule-based processes; (3) that we can respond to whole feature-rich gestalt experiences that cannot engage as wholes with rules or laws of any kind; and (4) that the role of consciousness (and its advantage) is to enable us to contribute this response to decision-making.

This may be seen as a kind of agent causation, but I would suggest it is far from being primitive and unanalysable.

The focus in the free will debate at present is so much on compatibilism and incompatibilism that there seems to be little concern about what I see as of central importance, namely the role of consciousness in decision-making, and especially in plausible reasoning.  So I would really appreciate comments that squarely address the arguments of these articles, particularly the four points listed above.  And if anyone would like to investigate my arguments further, they could look also at an article in Philosophy in 2001 which first introduced my gestalt argument, and another in Journal of Consciousness Studies in 2002 which has the fullest exposition of it.

Comments

[Joe Campbell]

Here I am posting again without reading all the relevant literature but I just wanted to point out that the following seems false:

"... given our circumstances and laws of nature, the way we are provides available alternatives, inconclusive reasons (and how they appeal), and unconscious tendencies, and also the capacity to decide between the alternatives on the basis of the reasons; and what we do is what we decide in exercise of that capacity."

It seems to me that the strongest claim that one can make is that there is no reason to doubt that the way we are provides "available alternatives" etc. That is, in part, what worries about libertarian luck are all about -- getting from the lack of determinism to the reality of things being up to us is problematic.

What, for instance, would it take to show that we have "the capacity to decide between ... alternatives on the basis of ... reasons"? Terms like "capacity" and "alternatives" are up for grabs. What could possibly show this short of a near-consensus view about the meanings of these terms -- which is doubtful.

Also, I want to go on record as saying that as soon as folks stop assuming that free will is incompatible with determinism, I'll stop caring about the compatibility problem!

[David Hodgson]

Joe,

You ask 'What, for instance, would it take to show that we have "the capacity to decide between ... alternatives on the basis of ... reasons"? Terms like "capacity" and "alternatives" are up for grabs.'

I agree, if this is your point, that determinism is not necessarily inconsistent with the existence of alternatives (in some relevant sense) or with the existence of capacity to decide (in some relevant sense). However, neither is indeterminism (as to which alternative occurs) inconsistent with capacity to decide between these alternatives, such that which alternative occurs is not a matter of chance but something selected by the decision-maker. Most arguments to the contrary assume what they set out to prove.

What I am trying to do is to give an account of that kind of capacity that is consistent, plausible, and supported by plausible arguments; and I really would like to have my arguments attacked.

Of course the question of compatibilism is important, but it is not the only important question.

[Joe Campbell]

David,

I agree with you that arguments which try to show that free will is incompatible with indeterminism are problematic. My own view is not this strong. Rather it is that indeterminism can't help us in understanding the nature of free will. And if it can't help, any problem for compatibilism can be turned into a problem for free will in general.

For instance, one could interpret Galen Strawson as starting from the assumption that the compatibilist can only provide us with "the existence of alternatives (in some relevant sense) or with the existence of capacity to decide (in some relevant sense)" and then moving on to show that we can't have either in THE relevant sense (whatever that is), even if indeterminism is true. Pereboom's argument is not too different (in general respects) other than that he draws a slightly weaker conclusion.

I agree, too, that there are other interesting issues besides the compatibility problem -- and I'm glad you and others are working on them! I'll try to take a look at your work and make some more constructive remarks.

[Tamler Sommers]

David,

Can you explain what the 'we' refers to in premise (3) as well as the 'us' in premise (4)? It seems to be something different from consciousness and conscious experiences, since you say that "we" can respond to gestalt-rich conscious experiences and that conciousness enables "us" to contribute to decision-making.

Without a clearer account of this 'we'--and crucially how it could be a responsible 'we'--I still can't see how this argument is supposed to undermine Strawson's basic argument. Strawson does not assume that human beings lack consciousness or that the consciousness is rule governed.

[David Hodgson]

Tamler

The 'we' is the conscious subject, which I generally equate with the whole person, while reserving judgment on what is necessary and sufficient about a person to constitute a conscious subject. How can this be a responsible 'we'? I say because each 'we', as well as having properties that distinguish one 'we' from another and that (along with circumstances and laws of nature) determine alternatives, reasons and tendencies, has a property as to which all 'we's' are the same, which can be called capacity to decide; and what we do is what we (that is, the whole 'we' including this capacity as well as distinguishing properties) decide in exercise of that capacity (in respect of which no one is at any advantage or disadavantage). This is the bare bones, elaborated in my articles.

[Tamler Sommers]

So if this property is shared by all, why does one person exercise this capacity to decide to behave morally while another person decides to behave immorally? That's the key question, right?

The plausible answer, it seems to me, is that the way the person is right before the decision, in conjunction with a host of situational factors, will determine how the person exercises that capacity. If not, what else could possibly play a role? (The word 'determine' might be misleading there, since I'm open to the possibility that there may be indeterminism in either the situational factors or in the way the person responds to them. But as Joe says, this indeterminism doesn't really help when it comes to moral responsibility, since it's unclear how this kind of indeterminism that could make us more blameworthy or praiseworthy than if the indeterminism were absent. If you disagree, then it seems like the burden is on you to show how you can spin that indeterminism into responsible choices. But in your remarks to Joe, you claim not to think that the indeterminism is essential anyway.)

Now, we're not responsible for the various situations we find ourselves in unless an earlier responsible decision has led us to be in that situation, right? So then in order to be responsible for our decisions, we either have to be responsible for the way we are when exercise our capacity to make the decision, or we have to be responsible for the situation we're in. (Like an alcoholic who decides to go to his friend's bachelor party in Las Vegas and then can't help himself from getting drunk and blowing his child's college fund.)

But here's where Strawson's regress comes in. At some point--I think you'd agree, no?--we were not responsible for either our situtation or the way we were as people. (Say, when we were two years old.) So on your account, how is it possible that we made our first responsible decision? Just acquiring the capacity to reason consciously wouldn't be sufficient. We'd still have a character we weren't responsible for when we acquired this capacity. And we couldn't be responsible for our situation then either. (Because the only way to be responsible for our situation was to be responsible for an earlier decision that led us to be in that situation.) So again, on your account, at what point could we possibly have made our first responsible decision?

And of course, if we can't make our first responsible decision, we can't make any...

[David Hodgson]

Tamler,

You say I 'claim not to think that the indeterminism is essential anyway'. What I intended to say is that I recognise compatibilism as a reasonable and defensible position. In my own account, however, indeterminism is important.

You also say 'The plausible answer, it seems to me, is that the way the person is right before the decision, in conjunction with a host of situational factors, will determine how the person exercises that capacity.' And although you also say that there may be indeterministic factors in the way a person responds, I don't think you recognise how I say such indeterministic factors may work.

For one thing, I don't equate indeterminism here with randomness.

I give what I think are very strong reasons for holding that a person's conscious experiences have a role in decision-making, that this is a role unlikely to be wholly rule-determined, and that this role involves responses to gestalt experiences grasped as wholes that cannot engage with laws or rules of any kind.

So I say that, given laws of nature and circumstances, the way a person is does not wholly determine the decision, except in the sense that it limits alternatives, reasons and tendencies and provides the capacity to decide between the alternatives. You may say, well, it is still the way the person is that exercises that capacity; but I say that the person is not constrained in the exercise of that capacity by the way the person is, beyond the constraints of alternatives, reasons and tendencies. And I also say that since all normal adults are alike in their capacity to decide, and to this extent are not at any relative advantage or disadvantage, the person has some ultimate responsibility for the decision made by the person, a totality comprising both the distinguishing features of the way the person is and the indistinguishable capacity to decide.

Again, I'm trying to address particular comments promptly and briefly, and refer you to more detailed arguments in the papers.

[czrpb]

Hi!

I am a simpleton and do not understand these:

(4) that the role of consciousness
(and its advantage) is to enable us
to contribute this response to
decision-making.

And:

The focus in the free will debate
at present is so much on
compatibilism and incompatibilism
that there seems to be little
concern about what I see as of
central importance, namely the role
of consciousness in decision-making,
and especially in plausible reasoning.

I do not understand how you/anyone can use consciousness to explain anything when consciousness is basically incomprehensible to us right now?

[Tom Clark]

David,

If we imagine two identical agents in identical situations, on your account there’s no reason to suppose they would make the same decision (although they might), since the decision is ultimately determined not by any feature of the agent or her situation, only by a decision-making capacity centrally involving consciousness. This means there can’t be any explanation of the choice in terms of the agent’s constitution or available alternatives, nor is there any explanation stemming from the conscious decision-making capacity that’s equally shared by all agents, since according to you this isn’t rule governed. So it seems to me that you purchase ultimate responsibility at the cost of explanatory transparency and self-knowledge.

On your account, we can’t really understand *why* we make the free choices we do, since it seems to me such understanding necessarily involves citing antecedent determining conditions and/or seeing how behavior conforms to various regularities (e.g., belief/desire/action schemas). So I have to agree with Joe’s criticism of libertarian agent causation in your case: it seems to be a primitive unanalyzable concept, and it must remain so since otherwise it would necessarily exhibit the causal provenance of an agent’s actions and/or show how those actions fall under law-like regularities. So I’m left wondering if we can really explain the actions of an agent who has buck-stopping, ultimate responsibility in the sense you think we must have in order to hold each other responsible. If we can’t, then free choices are a mystery and can’t be subsumed into a naturalistic account of the world.

[David Hodgson]

czrpb,

I entirely agree that consciousness is a mystery that is far from being solved. But that's not a reason either for not investigating it, or for assuming it has nothing to do with free will.

I think in fact there are powerful reasons for holding that consciousness and free will are components of the same mystery.

For one thing, I think that although continuing increases in computing power will produce computers that are increasingly reasons-responsive, these computers will not without a radical change in their mode of operation either have consciousness or anything that could reasonably be called free will.

Tom,

I think you too readily assume that the only worthwhile explanations are conclusive mechanistic ones, in terms of rule-governed processes, and that any suggested process that does not have that kind of explanation is 'non-natural' and therefore unbelievable. That is not the view of all scientists: see for example Stuart Kauffman's recent essay that appears on the Edge website and elsewhere.

As I've said, I've given what I believe are strong reasons for thinking that our grasping of whole feature-rich gestalts has a role in our reasoning and decision-making - and if that is so, there must be something about our decision-making that is not rule-determined. It is this argument, which I've developed in a series of articles since 2000, that I would particularly like to have examined for weaknesses.

[czrpb]

Hi again!

But that's not a reason either for not investigating it, ...

Whoa! Sorry about that; I did not mean to suggest that we should not investigate consciousness. But I do wonder:

... or for assuming it has nothing to do with free will.

And you have spoken about the "role of consciousness" in free will. I just worry about the correctness of attempting to explain one (mostly) unknown thing with another (mostly) unknown thing. But I certainly enjoy the efforts! Thanks!

[Tamler Sommers]

David,

I have read the TLS piece (in the earlier discussion and skimmed it again now). I still don't feel like I have a clear understanding of your account. Maybe I should just reemphasize Tom's point, framed slightly differently.

Let's say I grant you that consciousness plays a role in decision-making. And I grant you that the the "contribution is not one wholly determined by rule-based processes." And finally, I grant you that people as a whole can respond to "whole feature-rich gestalt experiences that cannot engage as wholes with rules or laws of any kind." In other words, I grant you your first 3 premises.

Two questions:

Bill and Dan have identical characters in identical situations. (Or Bill and Bill* in different possible worlds, as Mele puts this kind of objection). Bill decides to do the moral action. Dan decides to do the immoral action. What explains the difference, on your account? Clearly they've reasoned differently, but why did they reason differently, and how could the explanation be something they're MR for? It couldn't be anything about their characters or situations. And it can't be just a random indeterministic blip in one of their deliberation processes. So what is it?

Now can respond that indeterminism doesn't necessarily equal randomness. That's true, I'm not claiming it does. (Kane too has nicely defenned that claim.) But from this claim it doesn't follow that that indeterministic decision-making equals responsible decision-making. Random and responsible are not exhaustive alternatives. It seems like the burden is on your account is to explain the connection between non-rule-governed conscious decision-making and responsible decision-making. And that's what I don't see yet.

My second question is just a repeat of an earlier one you didn't address. Can you explain how we make our first responsible decision, on your account?

[David Hodgson]

Tamler

I acknowledge (indeed I claim) that Bill can decide to do the moral action and Dan can decide to do the immoral action, and that this is not because of any difference in their characters or situations, but is because they have reasoned differently. Why did they reason differently? I say they just did, not from any prior cause, but because this is how they did reason. To require any further prior cause at this point is already to assume that free will of the kind I’m supporting cannot exist, and (granting my first 3 premises) would be looking for a causal link that cannot be explained in terms of laws of nature. Any causal link that cannot be explained in terms of laws or rules must leave an area of indeterminacy in the result, and my premises support the view that what happens within this area is both not random and not explained in terms of causal inks from prior circumstances.

This does not make Bill or Dan a causa sui, I contend, because there are causes in their characters, circumstances and laws of nature that provide the constraints I’ve previously identified; but within those constraints I say the person (the whole person, including character and capacity to decide) decides and to this extent is an originator and therefore responsible as an originator. If anything is responsible for anything, such an originator must be responsible. Both Bill and Dan equally grasp moral imperatives and also grasp reasons inclining them to depart from them. Bill reasons to the conclusion that he is to do the moral action and Dan reasons to the conclusion that he is to do the immoral action. I say that makes them responsible.

Our early conscious decisions as infants would, on my account, be or at least include non-random selections between available alternatives; but I would not attribute moral responsibility until the person becomes capable of grasping and acting on moral considerations, a capacity that develops gradually over childhood. This is reflected in the law’s denial of criminal responsibility in young children. In Australia a child under 10 cannot be guilty of a crime, and for children between the ages of 10 and 14 the prosecution must prove beyond reasonable doubt that the child knew that what he or she was doing was seriously wrong.

[Cihan Baran]

Hello,

This does not make Bill or Dan a causa sui, I contend, because there are causes in their characters, circumstances and laws of nature that provide the constraints I’ve previously identified; but within those constraints I say the person (the whole person, including character and capacity to decide) decides and to this extent is an originator and therefore responsible as an originator.

One worry I have with what you say is that I don't see how having constraints can help mitigate worries about being causa sui. For instance, imagine an unusual rock that jumps, every now and then, at most 5 meters without a cause - it's causa sui. Surely, there are constraints on the rock's jump, it can only jump five meters, but still, it's causa sui.

You say that "the causes in their characters, circumstances and laws of nature" provide such constraints - so it's not causeless as I portrayed in my first paragraph. But then isn't it just random if these causes merely constrain the output? Suppose that the number of foxes in a forest were a genuinely random number. In that case, laws of nature would cause and constrain the number of foxes in a forest but then it would be just random.

Finally, you specifically point to causes in one's "character". But isn't this move likely to be prey to arguments like Galen Strawson's Basic Argument? If what one does is a result of one's character and if one's character is a result of what one does, then how could one be free?

(That's why I think that any free will realism ought to start with a denial of the character concept. I don't think it's too hard - at least to come up with common-day character talk, i.e. people have no characters - our way of classifying some people as "miserly", "grumpy" is just a way of describing some of their actions in the past.)

[David Hodgson]

Cihan,

I would argue it is a mistake to separate out the indeterministic element from the rest of the decision and ask what made the difference in just that element considered in isolation; that is, to say that reasons and character take you to the point of there being (say) two alternatives, so that a distinct something else must have given rise to the occurrence of one rather than the other alternative. I would argue that the decision comes from the whole person deciding for the reasons, but that the person’s differentiating features (together with circumstances and laws of nature) can only get to the alternatives (perhaps weighted with quantum mechanical probabilities). That does not mean that it is a distinct something else that gets you the rest of the way: rather it is the totality that gets you the whole of the way, the decision made by the person for the reasons. And I rely on my gestalt argument to show there is an indeterministic but non-random element in the whole process, albeit not one that can be severed and considered in isolation.

I don’t think we can deny the character concept: we do have more or less settled characteristics that affect what we do, that can be called our characters or ‘the way we are’.

[Tamler Sommers]

David, thanks. You write:

"..but within those constraints I say the person (the whole person, including character and capacity to decide) decides and to this extent is an originator and therefore responsible as an originator. If anything is responsible for anything, such an originator must be responsible. Both Bill and Dan equally grasp moral imperatives and also grasp reasons inclining them to depart from them. Bill reasons to the conclusion that he is to do the moral action and Dan reasons to the conclusion that he is to do the immoral action. I say that makes them responsible."

We may have reached one of those dialectical stalemates. To my mind, unless I know more about why they reasoned differently, it seems to me like a matter of pure luck that Bill took the moral path and Dan the immoral path.

If the explanation for the different decisions has nothing at all to do with the way they are right up to the moment of the decision, then I can't see how it's Dan's FAULT that he took the immoral path. Dan just drew a bad card on the river, it's a bad beat.

Maybe you're right when you say that "if anything is responsible for anything, than such an originator must be responsible." But it doesn't follow that such an originator is responsible. Strawson denies the antecedent of your claim. Who says anyhing has to be responsible for anything? The disagreement between you and Strawson is over that very question.

[Tom Clark]

David wrote:

“Why did they reason differently? I say they just did, not from any prior cause, but because this is how they did reason. To require any further prior cause at this point is already to assume that free will of the kind I’m supporting cannot exist, and (granting my first 3 premises) would be looking for a causal link that cannot be explained in terms of laws of nature. Any causal link that cannot be explained in terms of laws or rules must leave an area of indeterminacy in the result, and my premises support the view that what happens within this area is both not random and not explained in terms of causal inks from prior circumstances.”

Tamler responded:

“To my mind, unless I know more about why they reasoned differently, it seems to me like a matter of pure luck that Bill took the moral path and Dan the immoral path.”

David,

You’d say no, it isn’t luck since there’s a definite causal link between each agent (consisting of his characteristics and his decision-making capacity) and his respective choice, but one that can’t be explained in terms of laws and rules or in terms of causal links from prior circumstances. I take it that the existence of such a causal link is an empirical claim on your part – a hypothesis that could conceivably be falsified, one we could conceivably find evidence for or against. But it seems that by its very nature your hypothesis can’t make a prediction about behavior that would distinguish it from other hypotheses about choice-making, in which case it’s difficult to adduce evidence for or against it, at least in this fashion. As you put it, “they just did” reason the way they did and on your account there’s no further concrete explanation forthcoming for the choice. This means, as far as I can tell, that there’s no basis on your hypothesis to predict future choices that would do any better than either tossing a coin or looking at the same factors involved in ordinary deterministic, rule-governed belief/desire/action schemas. So I guess I don’t see how we’d distinguish between your account and other more transparent and concrete explanations of decision-making involving mechanisms, rules, laws and links to prior circumstances. In which case the naturalist would quite reasonably stick with the latter sort of explanation.

[Paul Torek]

David,

I prefer to deny what Tamler calls your third premise. I will base my comments on the Three Tricks of Consciousness paper.

In section 2 you argue that in a universe implementing Conway's Game of Life, a composite form called a "glider" has no causal oomph. The argument for this surprising conclusion seems to turn on the idea that if a composite's behavior can be explained by its components and the laws governing those components, the whole and its parts are in a competition which the whole must lose. I believe an adequate reply to such arguments can be found in Stephen Yablo, “Mental Causation”, repr. in David J. Chalmers (ed), Philosophy of Mind: Classical and Contemporary Readings (OUP, 2002). On the other hand, I'm not sure whether this point is critical to your overall argument.

It's section 3, where you claim that chunked qualia cannot engage general laws (i.e. - i.e.? - thereby having causal powers that can be explained with general laws), where I really want to drag my heels. You consider an objection along the lines I'd start with:
"For example, the chunking of two large dots side-by-side, a central vertical line below them, a horizontal line below that, and a surrounding circle, may disclose a general pattern suggestive of a human face. But general patterns of this nature can be detected without chunking, by a system set up to do so: our computers can do it, and it seems clear that most of our own basic recognition of types like human faces is carried out pre-consciously, by evolution-selected algorithms, without the assistance of either the qualia trick or the chunking trick."

Have you ever seen a face without experiencing any qualia? If not, how can you say that you receive no assistance from the qualia trick? I'd wager that if the neural circuits responsible for your visual qualia - and only those circuits - were destroyed, your facial recognition abilities would be significantly impaired.

The fact that similar performance can be attained by different means, in a computer lacking qualia, in no way proves that your abilities do not depend on qualia. There are cars which drive around powered only by fuel cells, with no internal combustion. That does not show that we can remove the internal combustion engine from your car without affecting its performance.

Similar points, I believe, could be made vis-a-vis chunking.

[David Hodgson]

Tamler,

Maybe it is a stalemate, but I’ll keep trying! Most of us probably start off with the idea that people are responsible for what they do. Then we think that can’t be the case because we’ve become what we are because of things outside our control, and in that respect we’ve all drawn different straws, been dealt different hands, or whatever. One thing I’m saying is that in one very important respect we haven’t been dealt different hands, perhaps you could say that along with our other cards we all have a joker that we can play again and again and that can have a large impact on how things turn out. We can’t help having that joker, our capacity to decide on the basis of inconclusive reasons, we’re not responsible for that, but we all have it equally, so we can fairly and reasonably be considered responsible in our use of that joker.

Tom,

Two points:
(1) I say that my gestalt argument gives powerful reasons for thinking that we do make non-random selections within the constraints of spectra of possibilities determined by circumstances, character and laws of nature; and if you accept that, then explanations limited to mechanisms, rules, laws and links to prior circumstances will be incomplete. Of course if you don’t accept it, then so be it. But as I’ve said, I want to be directed to weaknesses in my gestalt argument.
(2) If I’m right, and if the capacity to make these non-random selections is an important element in our ability to engage in plausible reasoning, then one would expect that the outcomes of these selections would on the whole be more reasonable than the outcomes of purely mechanistic processes, and would generally be capable of being understood by others exercising the same ability. It may be that eventually this could be experimentally tested.

Paul,

I agree that if the neural circuits responsible for my visual qualia were destroyed, my facial recognition abilities would be significantly impaired. But I don’t think this affects my argument. For one thing, I doubt if there are entirely distinct circuits for pre-conscious processes and for qualia, so that I think pre-conscious processes would inevitably be affected.

I am making the concession that recognition can be achieved mechanistically, whether or not that is the full explanation of our recognition capabilities; while insisting that grasping of whole gestalts (achieved by the chunking of qualia) requires consciousness, and that this grasping has an impact on our decision-making that can’t be achieved mechanistically.

[David Hodgson]

Tamler,

One other thought I’ve had concerning our stalemate. Returning to Bill and Dan, I think it will be rare that there are diametrically opposed possibilities between which a choice is made in an instant. I think Bob Kane argues that free will kicks in only occasionally, in considered decisions. I don’t agree with that, but I do think that for a lot of our lives we are pretty much on autopilot with conscious control being limited to fine-tuning and readiness to do more if the occasion arises. When we make considered decisions, plausible reasoning operates incrementally over time; so that if Bill and Dan are faced with a choice requiring a considered decision, by the time they are getting close to the final decision they are unlikely to be in identical states, if the final decisions are to be different. Of course it sometimes happens that we are required to make an important decision in an instant, and while I would want to say that the ability to respond to whole conscious experiences is important here (that’s one reason why our full conscious attention is engaged when an emergency happens), responsibility is less and sometimes we are in the position of mentally tossing a coin. Premeditated murder is considered worse than murder committed in the heat of the moment, because responsibility is greater.

[Tom Clark]

David,

You hypothesize that we’re free and morally responsible because our choices are in some sense insulated from rules, laws and prior circumstances, including our *very own dispositions*. That makes us ultimate originators in some sense, therefore robustly MR. But if our decisions don’t stem ultimately from our beliefs and desires, that makes them ultimately arbitrary from the standpoint of those beliefs and desires, in which case they aren’t rational. Rationality, including what you call “plausible reasoning,” is always in service to desired ends (Hume’s familiar point) and requires a reliable, dare I say deterministic, link between such things as goals, logical inference, and evidence and the resultant behavior, such that the behavior reliably reflects these constraints. As rational actors, we want to be *fully constrained* by such things, since to rise above them would simply and necessarily introduce something extraneous to desire, logic and evidence in shaping our behavior.

If there is some further basic consideration that rational action should and does reflect, what is it? You say it's our ability to grasp and respond to whole gestalt experiences, such that “if the capacity to make these non-random selections [via the gestalt-grasping ability] is an important element in our ability to engage in plausible reasoning, then one would expect that the outcomes of these selections would on the whole be *more reasonable* than the outcomes of purely mechanistic processes” (my emphasis). If this ability reliably contributes to rational behavior (plausible reasoning) then it *governs* behavior, in which case it constitutes a general rule, a reliable constraint that somehow contributes to rationality. But on your account the crucial point is that our response to gestalts *isn’t* rule-governed. Whatever else it does, this ability doesn’t operate on inputs and produce reliable outputs in service to our beliefs and desires (and if it did it would ultimately reflect our character and dispositions, which you say it doesn’t). So I don’t see how it could contribute to rationality, and therefore to being morally responsible in the sense of being a rational, reasons-responsive locus of control that can be guided by responsibility practices. The price we pay for being an ultimate originator on your account is that there’s a non-rule-governed, disposition-independent, situation-independent element of arbitrariness about choices that to my way of thinking undermines ascriptions of responsibility. But I’m glad we at least agree that whether that element exists or not is an empirical question in need of testing.

Btw, I have engaged with your argument about conscious gestalt-grasping abilities here and here. Your responses were 1) Since we’re conscious, it’s very likely that consciousness plays a causal role over and above that of its neural realizers. But of course this is one of the major open questions about mental causation. The phenomenal aspect of consciousness might very well not play a separate, essential role that its realizers can’t accomplish. 2) Plausible reasoning can’t be formalized. Again, this is at least an open question; it’s certainly possible that what we *experience* as a non-algorithmic process might turn out to be realized by neural algorithms. 3) It’s clear that algorithmic computers can’t (and will never) do things we conscious creatures do, e.g., grasp gestalts, and such grasping can’t be rule-determined since “rules cannot engage with feature-rich wholes but only their constituent features.” Proving these negatives decisively, which your account depends on, is a very tough row to hoe, since they are empirical claims about the nature of consciousness and our neurally-realized cognitive capacities. Should we stake a theory of free will and moral responsibility, and all its practical ramifications about blame and punishment, on definitive answers to questions that are so very much open at this stage of the game?

[David Hodgson]

Tom,

A quick response.

I think your argument depends on attempting to isolate the indeterministic element from the rest of decision-making. I say the indeterministic element is not independent of reasons, but makes use of them in a way that mechanistic causes cannot. And I strongly disagree with Hume's sharp distinction between reason and emotion: I say emotion can be irrational but is also an important part of our rationality.

And yes I agree that there are rules or laws that give us the capacity to choose: in my 2001 Philosophy article I called them laws of empowerment, but I say these rules do not determine the outcomes of the exercise of this capacity.

[Tom Clark]

David,

A quick response (after a run and some frisbee on a gorgeous spring afternoon in Boston) to your quick response:

On your account neither rules nor agent characteristics finally determine how the capacity for choice is exercised, although they may influence it. The deciding factor is something indeterministic, something not determined by antecedent conditions (the agent and her situation) plus laws. If so, then this factor can’t reliably reflect the agent or her situation – it can’t further her interests in the reliable way a deterministic process would (although it might on some occasions). Nor does it link the agent tightly to action, which is what we ordinarily look for in ascriptions of responsibility. So again, what’s lacking for me in your account is 1) a transparent explanation of how the indeterministic contribution of the capacity for gestalt recognition to choice-making makes us more rational than mechanistic processes, 2) an account of how this capacity can plausibly ground ascriptions of responsibility (overcoming Tamler’s luck objection), and 3) a convincing empirical demonstration that such a capacity exists and how it works that doesn’t presume answers to very deep and open questions about consciousness and cognition.

[Tamler Sommers]

David,

You've moved on to a new aspect of the discussion but just one remark about the proposed stalemate resolution. You write:

"When we make considered decisions, plausible reasoning operates incrementally over time; so that if Bill and Dan are faced with a choice requiring a considered decision, by the time they are getting close to the final decision they are unlikely to be in identical states, if the final decisions are to be different.'

Fair enought, but all that means is that the crucial moment occurs when Dan and Bill's reasoning processes diverged, leading one later in the process to the moral choice and other to the immoral choice. Why did that happen, given their identical characters (at that moment) and situations?

You might respond by saying Ok, maybe it was pure luck that Dan and Bill diverged at that particular moment but that there was still plenty of time for Dan to get on the right path, the one leading to the moral decision. But then I'd introduce a new identical person, Ivan, who went along on Dan's path at the moment they diverged with Bill's reasoning process. At some point before the decision, however, Ivan got back on the path leading to the moral decision (the way Dan should have done). Given that Ivan and Dan had identical characters and situations about to the point of THEIR divergence, what accounts for change, and how could it be something for which Ivan and Dan are responsible? And so forth...

Hope that makes a modicum of sense.

[Paul Torek]

David,

As Tom nicely puts it, "The phenomenal aspect of consciousness might very well not play a separate, essential role that its realizers can’t accomplish." That is exactly my argument, to wit, that phenomenal consciousness does play a role precisely as its realizers play the very same role. These are two descriptions of the same events. The whole and its parts are not in competition. The fact that a different set of parts comprising a different whole could play a functionally similar role (in facial recognition) is neither here nor there.

You say "this grasping [of chunked qualia gestalts] has an impact on our decision-making that can’t be achieved mechanistically." I certainly agree that it impacts our decision-making, and that we could not achieve similar decisions without consciousness (if that is what you mean by mechanistically). Whether future computers might achieve similar feats of decision-making without consciousness, I remain agnostic, but I'll grant for the sake of argument that they cannot. Still, the conclusion that chunked qualia cannot engage general laws simply does not follow. They engage via their component qualia and (probably also) some of the non-conscious characteristics of the person, which together are causally sufficient for the gestalt and the person's appreciation of it. At least, for all we know, that remains a distinct possibility.

As to the purpose of our abilities to grasp chunked qualia and make gestalt judgments, it is not hard to guess. As a person who works with machine vision systems for a living - sorting metal parts for recycling - I have an enormous appreciation for the awesome superiority of the human mind. Engineers can spend years programming a machine to sort parts by visual appearance, only to be put to shame by a human being who spends a half hour learning the same task. Ask that human being for criteria, and you get some useful tips that can be programmed (in some crude approximation, after much thought and reformulation), like, "if a piece has a rough edge like this, it's probably cast aluminum." But beyond those tips, there remains the comment "you just get a feel for it". And indeed, you just do. You just get a conscious gestalt qualia-laden perception.

Our flexible, rapidly-learned pattern-recognition abilities have obvious evolutionary value. In contrast, the evolutionary value of the abilities you postulate is debatable. Indeed, if our reason-underdetermined gestalt judgments reliably led to greater evolutionary fitness, wouldn't that be the very sort of engagement with constraining laws which is prohibited? I suppose it may depend on the degree of reliability. But then, to the extent that the postulated abilities leave avenues for low-fitness behavior, wouldn't you expect mother nature to cut those avenues off?