Article by David Hodgson

You all might be interested to read this recent article by David Hodgson called "Partly Free", which appeared in the July 5th issue of the Times Literary Supplement.  David welcomes any comments you my have.

(You can find the original, unedited version of the paper on David's website here.)

Comments

[Tom Clark]

I posted this over at Memeing Naturalism:

Although he concedes the explanatory force of current physicalist accounts of human behavior, Hodgson opts for a kind of non-naturalistic mentalism:

"I believe there are stronger reasons for holding that, while our conscious experiences do correspond with physical processes, these experiences can themselves have effects beyond those explicable in terms of physical processes and laws of nature, and that this enables us to have free will and to be responsible for our actions."

He takes this line because he believes that determinism destroys moral and criminal responsibility, a belief that UPenn law professor Stephen Morse calls the “fundamental psycholegal error.” Hodgson also thinks we need retributive justice as a way of limiting punishment to only what people deserve, otherwise we risk over-punishing. Further, retributive justice can only rest on a sort of agent causation in which a deterministic, physicalist story cannot be traced from antecedent conditions, thence into and including the agent, and thence to the act. Fortunately such agent causation exists, he argues, so all is well. We are causal exceptions to natural laws and thus can take a full measure of what he sees as metaphysically real responsibility, that which justifies retribution.

But we needn't resort to human causal exceptionalism to remain responsible agents, nor do we need retribution. First, we don't need the concept of retributive desert to limit punishment. A central value in the West is personal liberty and autonomy, and it is this that limits punishment so that it doesn't become draconian. The criminal justice goals of deterrence and public safety are counterbalanced by our commitment to individual freedom such that legal sanctions remain proportionate to the crime.

Second, if we are naturalists there's no basis for retribution in Hodgson's agent causation since the requisite sort of undetermined, self-caused agents don't exist, which is to say we don't have contra-causal free will. Nor are there any convincing compatibilist grounds (that is, grounds compatible with not having contra-causal free will) for retribution; see for instance my critiques of Stephen Morse, Michael Moore, Ronald Bailey, Hoffman and Goldsmith, and David Hill at www.naturalism.org/criminal.htm . But since we don't need retribution to limit punishment, this isn't a problem.

Third, as natural agents, those fully subject to cause and effect, we remain moral agents in that we are responsive to the prospect of moral evaluation, rewards and sanctions. We don't need to be causal exceptions to the natural order to be held responsible. Indeed, if we were such exceptions, our responsibility practices, such as the threat of sanctions, wouldn't work. Our freely willing core wouldn't be responsive to moral evaluation - it would just do what it darn well pleased. So morality, minus its retributive component, survives without contra-causal agency. This has considerable implications for criminal justice.

Most of Hodgson's ideas in "Partly Free" appear in an earlier article, "A plain person's free will," which he wrote for the Journal of Consciousness Studies, critiqued at www.naturalism.org/hodgson.htm . He can only maintain his non-naturalistic notion of free will by dint of some very tenuous and contentious claims having to do with quantum mechanics, consciousness, rationality, evolution and human agency. Such implausibilities (by my lights) wouldn't be necessary but for his antecedent supposition that we need to somehow evade cause and effect to be moral agents and to keep punishment humanely proportionate. But there are far simpler conceptions of moral agency and humane criminal justice to be had within science-based naturalism, a worldview that accepts that human beings are fully included in the natural, physical order of things.

Btw, the Center for Naturalism July-August newsletter is just out at http://www.naturalism.org/CFN%20Newsletter.htm .

[David Hodgson]

In response to Tom Clark’s comment, while I do think that moral and criminal responsibility are important, I take my ‘line’ not because of this belief, but because of what I see as the strength of arguments that Clark does not address.

I really would appreciate a response to the actual arguments of my article, in particular that conscious experiences contribute positively to decision-making, that this positive contribution can’t be explained in terms of rule-governed procedures, and that there is a plausible explanation in terms of responding to gestalts that are too feature–rich to engage as wholes with general rules.

[Blake]

Tom: I have to admit that David's paper didn't rely on retribution but on the necessity of free will to any sort of reasoning process that is not dictated by a-rational causes. I don't see anything in your response that engages his actual argument. It seems like a fairly compelling argument to me.

[Tom Clark]

David wrote:

"I really would appreciate a response to the actual arguments of my article, in particular that [1] conscious experiences contribute positively to decision-making, [2] that this positive contribution can’t be explained in terms of rule-governed procedures, and [3]that there is a plausible explanation in terms of responding to gestalts that are too feature–rich to engage as wholes with general rules."

1) Conscious experiences contribute positively to decision-making.

I agree that, blind-sight notwithstanding, capacities associated with consciousness seem necessary for a wide range of higher-level behavior. What this says about the contribution of consciousness per se, however, is unclear since it seems the neurons are already doing the causal work.

2) This positive contribution [of conscious experiences] can’t be explained in terms of rule-governed procedures.

It’s difficult to prove the negative that the contribution of conscious experiences will never be explained in terms of rule-governed procedures. As you say: “Now I can’t altogether rule out the possibility that plausible reasoning might be explained as an expression of unconscious computational processes that do not have any validity on the basis of discovered or invented rules for good reasoning, but which have been selected in evolution for their effectiveness in promoting survival and reproduction.”

3) There is a plausible explanation [of the contribution of conscious experiences] in terms of responding to gestalts that are too feature–rich to engage as wholes with general rules.

You give the example of Gershwin: “I suggest that, in appraising the possibilities and the melody, Gershwin was responding to gestalts of the possibilities and of the melody and/or chunks of it, which because of their uniqueness and feature-richness could not engage as wholes with pre-existing rules of any kind.” I’d say that in responding to the consciously experienced melodic gestalts produced by his unconscious processes, it’s likely that Gershwin was applying definite selection criteria, for instance those which decided what melody or phrase sounded good to him when played after (or with, if it’s counterpoint) the ones already selected. As a composer with a recognizable style, these criteria were distinctively Gershwin’s, but for that very reason were rule-governed, namely by his musical personality (as were his unconscious compositional processes too, of course). So I don’t see why feature-rich conscious gestalts, especially those families of gestalts whose intra-family differences are along various dimensions (e.g., melodies, faces, letter forms, words, phrases, concepts) can’t be responded to in rule-governed ways that account for complex behavior.

[Mark]

Tom,

Regarding your response to (3), do you think Hodgson is making a probabilistic argument where the conclusion is seen as the most reasonable one? His argument seemed more inductive in nature to me, especially the premise you highlighted in your response to (3).

David,

In reading the unedited version, I didn't find anything specific that would make your account incompatible with the truth of physical determinism. For instance, the emergentist picture of causal powers can make sense in a deterministic universe: even if bare causal powers behave deterministically, it may be possible for relations between physical elements with bare causal powers to yield an irreducibly complex structure that has a new causal power. If this emergentist picture can work in a fundamentally deterministic universe, it seems like emergent powers could satisfy your worries about agent-powers being reducible to "general rules".

[Tom Clark]

Mark,

In this article and in his earlier JCS piece “A plain person’s free will,” David adduces a host of considerations, some philosophical, some scientific, to establish the plausibility that we have a capacity for choice that isn’t rule-governed and that grounds true responsibility. The plausibility of each consideration could be critiqued in turn, and I’ve done that in great detail for his earlier paper at www.naturalism.org/hodgson.htm (also published in JCS along with his rebuttal). Here I’ll simply advert to his conclusion in the present paper:

“The constraining effect of the way we are is limited to determining alternatives, reasons and unconscious tendencies. Subject to that, our decisions are not constrained by any distinguishing features of the way we are, and to this extent we are truly responsible for them.”

This of course reverses the standard intuition about responsibility: that we are truly responsible to the extent our decisions *are* constrained by distinguishing features of the way we are, including centrally our reasons for action. If it turns out that we are the sorts of agents that David describes, which contribute something independent of their characteristics and reasons (which he says are always inconclusive) then so be it. But I don’t see how being non rule-governed grounds responsibility since that divorces action from who we are as identifiable individuals. As I put it in my JCS response to his earlier paper: "What has free will and can be held ultimately responsible, on Hodgson’s theory, has nothing to do with the real, distinguishable people that show up in court."

On a related point, I think David conflates the difficulty of formalizing rationality with the idea that what he calls "plausible" and "informal" reasoning doesn’t follow rules of entailment, which of course it does most of the time even if the rules aren’t usually explicit. In fact, we *want* to be constrained by rules of logic and for our actions to be fully constrained by our reasons for acting and by the evidence of our senses. If the decision-making process inserts something non-rule-governed and unrelated to who I am and my situation, I’d worry about reaching valid conclusions and acting effectively. As it stands, the evidence suggests that evolution designed us to be good deterministic reasoners, and that conscious experiences as such are either epiphenomenal or identical to some behavior-controlling functional or representational process instantiated by the brain. It may turn out that consciousness plays a necessary, naturally selected behavior-controlling role that deterministic, rule-governed neural processes can’t, but I haven’t seen much evidence for it thus far.

[Kip Werking]

Like Tom, I find Hodgson's argument unconvincing. The argument seems premised on the idea that people process wholes or chunks in a way that, according to Hodgson, is incompatible with algorithmic processes. But, as Clark notes, just because it is difficult to articulate or identify such rules does not mean they exist. And, as Hodgson notes, neuroscience supposes that such rules do exist. So I do not see the reason why such rules cannot exist. Nothing about the Gershwin example suggests to me otherwise.

Hodgson's article does seem to serve as another data point showing that, according to common usage (if any), people use the term "free will" to refer to something more magical and extravagant than what compatibilists allege it means.

[Blake]

Kip and Tom: It seems to me that you're not addressing David's argument. As I understand his argument, it is somewhat of a variant on the problems raised by Fodor's position and critiques that observe that we have no rules for reducing semantics to syntax -- nor can we imagine what form such rules would take. More importantly, it seems that inductive reasoning and reasoning to the best explanation involve evaluative reasoning that could not be reduced to an algorithm or set of algorithms. It is the freedom to evaluate the evidence in a way not dictated by a-rational causal factors that is irreducibly essential to such reasoning -- and that is precisely what we cannot do if the conjunction of physicalism and determinism obtains. I just don't see anything you say addressing that point.

Further, if Tom is correct that reasoning in merely deterministic and consciousness is epiphenomenal, then our reasons and conscious thoughts don't enter into the process of reasoning at all. Thus, it seems that there is a great deal about human experience that is either denied (the eliminativist position) or left unexplained in the sense that we cannot even imagine how our reasoning is possible (compatibilism).

Tom -- it seems to me that basing moral judgments on our practices of holding others accountable or morally judging them is not the same as asserting that persons are actually morally responsible. For one thing, such a practice is circular. We justify our moral practices by what we do and yet our moral practices just are what we do. If we base morality merely on our practices, we miss the point that our practices may not be justified and given determinism cannot be justified on normative grounds. A mere practice cannot establish the existence of moral responsibility (we cannot derive ought from is).

[Mark]

Blake,

There are other ways to yield a deterministic system than through a deterministic algorithm. Two prime examples are neural networks and swarming.

Neural networks produce a deterministic output from a given input, yet they defy algorithmization given their very nature of applying a distributed system of interconnected, seemingly meaningless and arbitrary weights and relations. And yet, it is possible to derive a neural network that will correctly identify the letters in the alphabet from a section of an image that you've scanned into your computer in order to turn it into a document.

Swarming involves taking a large number of "robots" (either in software or hardware) driven by a seemingly simple algorithm for processing, storing, and sharing data. Individually these robots can't accomplish much of anything, but when they work collectively, unprecedented behaviors emerge from the group that makes it seem like they are behaving intelligently such that they are able to effectively and reliably handle complex problems. This emergent behavior also defies algorithmization. (The routing systems in modern telephone networks use swarming technology.)

Now, in both of these pictures we have a good reason to say that there is no master deterministic algorithm that governs the final behavior, and yet there are LOTS of little deterministic algorithms that work in conjunction to produce the final behavior.

I think Kip and Tom's point is simply that if there isn't a system isn't governed by a master deterministic algorithm, it doesn't mean a system doesn't behave deterministically according to the conjunction of a set of compartmentalized, deterministic algorithms.

[Blake]

Mark: It seems that what you mean by "neural network" is a binary system that reads algorithms. I have some background in neurodynamics, and I seriously doubt that neurodynamic theories can be reduced to this kind of deterministic reading without assuming determinism (and selecting pixels for an alphabet letter is worlds apart from reasoning to the best explanation or explaining the relation between conscious thoughts and the activity of neural populations). Further, it seems to me that the outcome of robotic algorithm populations are themselves explainable by a very complex algorithm that sums the algorithms (which admittedly we cannot complete because of the amount of information at issue).

Further, on what basis do we assert that there isn't something that is more than the sum of the parts, something truly novel or emergent from such complexity, that isn't deterministic? It seems that there are epistemic limitations to concluding that such complexity is deterministic.

[Blake]

Mark: Let me add that I believe what David is asserting is that reasoning involves a great number of indeterministic algorithms and that inductive reasoning in particular relies on a great number of open alternatives. Thus, to assume that we are dealing with merely deterministic algorithms misunderstands the argument and begs the questions. The question is whether deterministic algorithms with indeterministic algorithms could yield the kinds of inductive and macro-system reasoning essential to science.

[Tamler Sommers]

David wrote: "I really would appreciate a response to the actual arguments of my article, in particular that conscious experiences contribute positively to decision-making, that this positive contribution can’t be explained in terms of rule-governed procedures, and that there is a plausible explanation in terms of responding to gestalts that are too feature–rich to engage as wholes with general rules."

After which, Tom (after making such a response) wrote:

"If it turns out that we are the sorts of agents that David describes, which contribute something independent of their characteristics and reasons (which he says are always inconclusive) then so be it. But I don’t see how being non rule-governed grounds responsibility..."

To me, this is a key point that David fails to satisfactorily address in his paper. To put the question more tendentiously: how is the matter of our being rule governed or not rule governed IN ANY WAY relevant to the question of whether we can be morally responsible?

When discussing responsibility David addresses Strawson's basic argument as presented in "Luck Swallows Everything" (also from the TLS). But his interpretation of the argument is inaccurate. He writes:

"[Strawson argues that] we can’t be responsible for the way we are when we first make decisions in life (that must be all down to genes and environment), so we can never become responsible (through earlier decisions) for the way we are later in life."

As David interpretes it, Strawson's conclusion requires that genes and environment entirely determine our character. But that's just not true. David failes to mention the last two premises of the 'genes and environment' version of the basic argument, which are as follows (See "Luck Swallows Everything" on Tom's site):

"(F) This may not be the whole story, for it may be that some changes in the way one is are traceable to the influence of indeterministic or random factors. But (G) it is absurd to suppose that indeterministic or random factors, for which one is ex hypothesi in no way responsible, can in themselves contribute to one’s being truly or ultimately responsible for how one is."

Shaping this up a little to specifically address David's argument: let me put it like this:

(F): This may not be the whole story for it may be that some of the changes to our character are not rule governed, and are a result of feature rich gestalt processes.

But (G): But even if this is true, these "feature rich gestalt processes" cannot contribute to our being morally responsible for our actions or characters.

(H) So either way we cannot be morally responsible for our actions.

In response (perhaps) to something like (G), David writes that "(15) However, because our decisions are made in part in response to gestalts that cannot engage with rules, we have the capacity to make decisions that are not wholly determined by the engagement of laws of nature with our formed characters and our circumstances" and therefore, that we can become partly responsible for our actions and characters. But I see no argument for that claim, no connection between this capacity and moral responsibility. How would this capacity that "we" possess ground moral responsibility? How is it a responsible 'we'? Why doesn't luck swallow this capacity, and how 'we' exercize it, along with all the others?

This isn't to say that it wouldn't be interesting if we had the kind of freedom he says we have. It would. But it would be interesting in a sense unrelated to whether we possess (desert entailing) responsibility of the kind that is thought to ground retributive punishment.

[Tamler Sommers]

Here's another relevant section from Luck Swallows Everything to fortify the question of my previous comment:

"One challenge to [the basic argument] runs as follows. "Look, the reason why one can be ultimately responsible for what one does is that one’s self is, in some crucial sense, independent of one’s general mental nature (character or motivational structure). Suppose one faces a difficult choice between A, doing one’s moral duty, and B, following one’s desires. You deniers of moral responsibility] describe this situation as follows: Given one’s mental nature, you say, one responds in a certain way. One is swayed by reasons for and against both A and B. One tends towards A or B, and in the end one does one or the other, given one’s mental nature, which is something for which one cannot be ultimately responsible. But this description of yours forgets the self - it forgets what one might call 'the agent-self'. As an agent-self, one is in some way independent of one’s mental nature. One’s mental nature inclines one to do one thing rather than another, but it does not thereby necessitate one to do one thing rather than the other (to use Leibniz’s terms). As an agent-self, one incorporates a power of free decision that is independent of all the particularities of one’s mental nature in such a way that one can after all count as ultimately morally responsible in one’s decisions and actions even though one is not ultimately responsible for any aspect of one’s mental nature."

The [deniers of moral responsibility] are unimpressed: "Even if one grants the validity of this conception of the agent-self for the sake of argument", they say, "it cannot help. For if the agent-self decides in the light of the agent’s mental nature but is not determined by the agent’s mental nature, the following question immediately arises: Why does the dear old agent-self decide as it does? The general answer is clear. Whatever it decides, it decides as it does because of the overall way it is, and this necessary truth returns us to where we started: somehow, the agent-self is going to have to get to be responsible for being the way it is, in order for its decisions to be a source of ultimate responsibility. But this is impossible: nothing can be causa sui in the required way. Whatever the nature of the agent-self, it is ultimately a matter of luck. Maybe the agent-self decides as it does partly or wholly because of the presence of indeterministic occurrences in the decision process. Maybe, maybe not. It makes no difference, for indeterministic occurrences can never contribute to ultimate moral responsibility."

Substitute 'our non-rule governed capacity' for 'agent-self' and you have one key question I think David needs to address.

[Mark]

Blake,

A neural network just is a neural network, and I think it's something other than what you have in mind. See this article for more info. For more info about swarming see this article and also see the article on flocking.

I don't think there is any question about whether the complex behavior exhibited by swarms/flocks is emergent. If I design a system of robots that operate on a VERY simple set of instructions that and yet the collective swarm exhibits highly complex, seemingly intelligent behavior, it is clear that something that is more than it's parts has been created somewhere along the way. In fact, this is usually the kind of thing that one points to as an ostensive definition of emergence (whereas neural networks do not demonstrate emergent behavior).

Lastly, I never said either of these helps to account for consciousness. I merely said they may be ways that David can get at grounds for saying that there are "no general rules" that govern human behavior, without asserting the falsity of determinism.

[David Hodgson]

Thank you all for your comments. I’ll respond briefly.

Tom Clark:

>>I agree that, blind-sight notwithstanding, capacities associated with consciousness seem necessary for a wide range of higher-level behavior. What this says about the contribution of consciousness per se, however, is unclear since it seems the neurons are already doing the causal work.>>

Law-governed behaviour of neurons does a lot of ‘the work’, but part of this work is to give rise to conscious experiences, which I contend can do things in ways not entirely law-governed. If it could all be done without conscious experiences, why have them?

>>It’s difficult to prove the negative that the contribution of conscious experiences will never be explained in terms of rule-governed procedures.>>

There are powerful arguments that plausible reasoning can’t be formalised (Putnam etc), and there are also powerful arguments that anything that can be done by way of rule-governed processes can be done by a general-purpose computing machine (Turing), and without consciousness.

>>You give the example of Gershwin: “I suggest that, in appraising the possibilities and the melody, Gershwin was responding to gestalts of the possibilities and of the melody and/or chunks of it, which because of their uniqueness and feature-richness could not engage as wholes with pre-existing rules of any kind.” ... As a composer with a recognizable style, these criteria were distinctively Gershwin’s, but for that very reason were rule-governed, namely by his musical personality (as were his unconscious compositional processes too, of course).>>

Again, why the conscious appraisal, if it adds nothing? I think it is clear that we do (and computers don’t) grasp melodies (and other things) as unique and particular wholes, and my point is that this grasping either does nothing, or does something that cannot be rule-determined, because rules cannot engage with feature-rich wholes but only with their constituent features.

>>This of course reverses the standard intuition about responsibility: that we are truly responsible to the extent our decisions *are* constrained by distinguishing features of the way we are, including centrally our reasons for action. If it turns out that we are the sorts of agents that David describes, which contribute something independent of their characteristics and reasons (which he says are always inconclusive) then so be it. But I don’t see how being non rule-governed grounds responsibility since that divorces action from who we are as identifiable individuals.>>

I say we are responsible as identifiable individuals, but our distinguishing features don’t fully account for our decisions. To the extent that our decisions depend on something we all have in common, our capacity to decide, there are no advantages or disadvantages; but this does not remove responsibility from us as individuals, because each of us is an indivisible amalgam of distinguishing features and this common capacity.

>>In fact, we *want* to be constrained by rules of logic and for our actions to be fully constrained by our reasons for acting and by the evidence of our senses. If the decision-making process inserts something non-rule-governed and unrelated to who I am and my situation, I’d worry about reaching valid conclusions and acting effectively. As it stands, the evidence suggests that evolution designed us to be good deterministic reasoners, and that conscious experiences as such are either epiphenomenal or identical to some behavior-controlling functional or representational process instantiated by the brain. It may turn out that consciousness plays a necessary, naturally selected behavior-controlling role that deterministic, rule-governed neural processes can’t, but I haven’t seen much evidence for it thus far.>>

Yes we want to be constrained by rules of logic, but rules of logic often don’t give answers to real problems. We need in addition plausible reasoning, and I argue this is helped by our ability to grasp and respond to whole gestalt experiences.

Mark:

>>In reading the unedited version, I didn't find anything specific that would make your account incompatible with the truth of physical determinism. For instance, the emergentist picture of causal powers can make sense in a deterministic universe: even if bare causal powers behave deterministically, it may be possible for relations between physical elements with bare causal powers to yield an irreducibly complex structure that has a new causal power. If this emergentist picture can work in a fundamentally deterministic universe, it seems like emergent powers could satisfy your worries about agent-powers being reducible to "general rules".>>

There are two versions of determinism that make sense to me: the version that says everything that happens is determined by past circumstances and laws of nature, and the block universe. My take on free will is incompatible with the former at least, because it says that the link between one past circumstance, a person’s grasping of a gestalt experience, and something that happens, the person’s response to it, is not determined by rules.

>>Neural networks produce a deterministic output from a given input, yet they defy algorithmization given their very nature of applying a distributed system of interconnected, seemingly meaningless and arbitrary weights and relations. And yet, it is possible to derive a neural network that will correctly identify the letters in the alphabet from a section of an image that you've scanned into your computer in order to turn it into a document.
Swarming involves taking a large number of "robots" (either in software or hardware) driven by a seemingly simple algorithm for processing, storing, and sharing data. Individually these robots can't accomplish much of anything, but when they work collectively, unprecedented behaviors emerge from the group that makes it seem like they are behaving intelligently such that they are able to effectively and reliably handle complex problems. This emergent behavior also defies algorithmization. (The routing systems in modern telephone networks use swarming technology.)
Now, in both of these pictures we have a good reason to say that there is no master deterministic algorithm that governs the final behavior, and yet there are LOTS of little deterministic algorithms that work in conjunction to produce the final behavior.>>

In neural networks and swarming, an apparently reasonable global outcome emerges from the algorithmic behaviour, but I would argue it is still rule-governed and not the result of a response to a whole feature-rich gestalt experience.

Kip Werking:

>>Like Tom, I find Hodgson's argument unconvincing. The argument seems premised on the idea that people process wholes or chunks in a way that, according to Hodgson, is incompatible with algorithmic processes. But, as Clark notes, just because it is difficult to articulate or identify such rules does not mean they exist. And, as Hodgson notes, neuroscience supposes that such rules do exist. So I do not see the reason why such rules cannot exist. Nothing about the Gershwin example suggests to me otherwise.>>

See responses to Tom Clark

Blake:

>>More importantly, it seems that inductive reasoning and reasoning to the best explanation involve evaluative reasoning that could not be reduced to an algorithm or set of algorithms. It is the freedom to evaluate the evidence in a way not dictated by a-rational causal factors that is irreducibly essential to such reasoning -- and that is precisely what we cannot do if the conjunction of physicalism and determinism obtains. I just don't see anything you say addressing that point.
Further, if Tom is correct that reasoning in merely deterministic and consciousness is epiphenomenal, then our reasons and conscious thoughts don't enter into the process of reasoning at all. Thus, it seems that there is a great deal about human experience that is either denied (the eliminativist position) or left unexplained in the sense that we cannot even imagine how our reasoning is possible (compatibilism).
it seems to me that the outcome of robotic algorithm populations are themselves explainable by a very complex algorithm that sums the algorithms (which admittedly we cannot complete because of the amount of information at issue).>>

I agree.

Tammler Sommers:

>>how is the matter of our being rule governed or not rule governed IN ANY WAY relevant to the question of whether we can be morally responsible?
(Strawson) "Why does the dear old agent-self decide as it does? The general answer is clear. Whatever it decides, it decides as it does because of the overall way it is, and this necessary truth returns us to where we started: somehow, the agent-self is going to have to get to be responsible for being the way it is, in order for its decisions to be a source of ultimate responsibility. But this is impossible: nothing can be causa sui in the required way."
Substitute 'our non-rule governed capacity' for 'agent-self' and you have one key question I think David needs to address.>>

As I said in response to Tom, one feature of the way the agent-self is is that it has the capacity, as do other agent-selves, to make decisions that are to some degree unaffected by distinguishing features. We can’t help having this capacity, so we cannot be responsible for that, but we can be responsible for how we exercise it. Tom says this would mean that the real decisions are made by something featureless, whereas for responsibility we want to say that the decision is due to something particular to the decider; but my answer is that it is the whole combination (of what is distinctive and what is indistinguishable) that makes the decision.

[Tamler Sommers]

David,

With all due respect, I don't think you answered my question. The question was: How would this capacity that "we" possess ground moral responsibility? How is it a responsible 'we'? Why doesn't luck swallow this capacity, and how 'we' exercise it, along with all the others?"

We agree that we aren't responsible for possessing the capacity (assuming we do). So the disagreement focuses on our responsibility for how we exercise the capacity. You say we can be responsible for this, but I don't see the argument for that claim. What does the fact that "it is the whole combination (of what is distinctive and what is indistinguishable) that makes the decision" have to do with making the agent morally responsible for the decision? What's the connection? Why doesn't luck swallow the way in which the combination produces the decision?

[Eddy Nahmias]

Tamler (with one 'm') et al.:
If (as a compatibilist might argue) doing A using one's capacity to X (e.g., a capacity to do A-like actions properly responsive to the reasons relevant to doing A-like actions) just *is* what it means to do A freely and to be morally responsible for doing A, then isn't it begging the question (or asking too much or something) to ask whether, in doing A, one also is free and responsible for the way one exercised one's capacity to X? In light of the above analysis, this would be to ask whether one used capacity X in exercising one's capacity to X? If you want to set off an infinite regress in this way, you should not be surprised if the regress cannot be completed. (But Go Red Sox!)

[Tamler Sommers]

Eddy, I agree with you, I think--certainly about the Sox who are strugglilng a little with consistency right now.

My question then, stated as precisely as I can (and precision is not my strong suit), is this: what does the non-rule-governed way in which the capacity is exercised--as opposed to a rule-governed way of exercising the capacity--have to do with moral responsibility for decisions? If luck swallows rule-governed ways of exercising the capacity--as David concedes, you might not--why doesn't it also swallow this non-rule governing means of doing so?

That doesn't seem question-begging or too much to ask, because it's the non-rule-governing nature of decision-making that's suppposed to be grounding our moral responsibility.

And thanks for the name correction. I think everyone can agree that I'm not morally responsible for my parents giving me a bizarre name that seems to beg to be mispelled in one particular way (not that "Tammler" is any less bizarre than 'Tamler,' so I'm not sure why this happens. I wonder if I had been named Tammler, if everyone would write 'Tamler.' Is there any way to know that? OK, this is far beyond procrastination right now so I'll end my reflections on this matter.)

[David Hodgson]

Sorry for that extra m, Tamler.

I understand the problem of luck to be that if what we do is all down to things outside our control, then we can't be responsible.

What I say is that what we do is not ALL down to things outside our control, except in the sense that things outside our control have given us the capacity to make choices (albeit between alternatives that are limited by things outside our control and are influenced by things outside our control). So because what we do is not entirely down to things outside our control, to that extent we have ultimate responsibility.

[Mark]

David,

Granting that you reject the compatibility of determinism and responsibility, how do you suppose that indeterminism helps? I understand the significance and role of conscious experience in your argument, but I don't see how granting the truth of indeterminism helps your case. Your assertion requires at the least a minimal account of consciousness that demonstrates its incompatibility with determinism using a plausible set of cases and such. (Perhaps you've done this legwork else where?)

[Tamler Sommers]

David,

I think Mark's question is relevant to my own here. In your last response uyou write that:

"What I say is that what we do is not ALL down to things outside our control, except in the sense that things outside our control have given us the capacity to make choices (albeit between alternatives that are limited by things outside our control and are influenced by things outside our control). So because what we do is not entirely down to things outside our control, to that extent we have ultimate responsibility."

But what I still don't understand--and please tell me if you think I'm being thick-headed here--is how the fact that our deliberative processes are non-rule governed rather than rule governed would add in any way to our control and therefore to our ultimate responsibility. (In this, I might be aligned with compatibilists which doesn't happen very often.) You agree you have the burden to explain this, right? Since rule-governed decision-making would not, by your own lights, grant us the proper type of control for moral responsibility. Maybe, as Mark suggests, you've addressed this question elsewhere?

[David Hodgson]

Mark,

How does indeterminism help? As I see it, if what we decide is entirely determined by prior circumstances and laws of nature, that leaves no room for an efficacious input from the person (other than an input that is itself similarly down to things outside the person's control) - this is really van Inwagen's consequence argument. Of course, indeterminism in the sense of randomness would not help here, but if, as I argue, rational choice is not just a matter of rule-determined developments, then there is room for a positive contribution from the person that is not wholly down to things outside the person's control (that is, neither random nor wholly down to prior circumstances and laws of nature). I say a bit more about this in an article I published in 1999 called 'Hume’s mistake' (part of this was republished in Kane's Blackwell anthology), and in the article Tom Clark mentioned 'A plain person's free will'. My fullest development of the argument about gestalts is is an article 'Three tricks of consciousness'. These articles can be read on my 'Reasonable free will' website.

Tamler asks 'how the fact that our deliberative processes are non-rule governed rather than rule governed would add in any way to our control and therefore to our ultimate responsibility'. I am arguing that being completely rule-governed excludes the kind of control that is 'up to us'. Randomness would equally do so, but a contribution to rational choices that is neither rule-governed nor random is something for which the person could have ultimate responsibility.

[Mark]

David,

I'll check out those articles. Thanks for the references. Based on what you said above, would it be fair to classify your account of agency as an agent-causal account?

Sidebar:
I still speculate that your defense against Strawson's Basic Argument boils down to saying that the "up to us"-ness required for ultimate responsibility comes from having an agent-self and we are warranted to believe that we have agent-selves due to our conscious experiences.

I agree with this line of thinking insofar as "agent-selves" are equatable to "persons". However, I have found no reason to compel me to believe that persons and determinism are irreconcilable. Your account seems to demand an agent-causal view of action and though I am sympathetic with this point as well, I see no reason why an agent-causal view of action cannot be compatible with determinism.