The Joker Among Our Cards

During the exchange concerning my recent post in support of libertarian views, an idea occurred to me that I would like to follow up, namely the analogy of a joker among the cards that are dealt us.

John Fischer has written that 'our behaviour may well be "in the cards" in the sense that we simply have to play the cards that are dealt us.' (Fischer, J. 'The cards that are dealt you' (2005), Journal of Ethics 10, 107 at 129.)  This drew the apt comment from Kip Werking that it misleadingly suggests there is a player of the cards distinct from the hand that is dealt, whereas in truth human beings simply are the cards that are dealt them by genes and environment.

My view can be understood as accepting this, but as suggesting that each of us includes, in the hand of cards that is dealt us and that constitutes us, along with  particular cards like aces, tens, jacks and so on, one powerful and flexible general-purpose card, like a joker.  The particular cards engage with circumstances and laws of nature to limit our conduct to a spectrum of possibilities, while the joker, our capacity for conscious choice, can combine with our other cards to steer a course within this spectrum of possibilities.

I am not suggesting this joker is a self or soul that itself makes decisions, or that it corresponds with any particular region of the brain.  Rather, it is a capacity that operates only in conjunction with our other cards.  It is however powerful and flexible:  so long as our other cards are not seriously deficient, for example because of mental abnormality or senility, the joker enable us to make reasonable albeit fallible decisions about what to believe (including what to believe about right and wrong) and about what to do, for good or ill.  And these decisions can in turn affect what particular cards we come to hold for the future, for better or worse.  Since we all have this joker, we all have some ultimate responsibility for our conduct, again at least so long as our other cards are not seriously deficient.

Why then do I think our cards include this joker, this capacity to decide?  In brief, I say there are very strong reasons to accept that conscious experiences make a substantial positive contribution to our decision-making, and that this contribution is not one wholly constituted by rule-determined processes:  if it were, as Alan Turing's arguments show, consciousness would be a superfluity.  There is however a plausible account of how conscious experiences can make a contribution that is not rule-determined, namely by providing feature-rich gestalt experiences that cannot engage as wholes with laws or rules of any kind but to which, as wholes, we can respond reasonably.  On this view, the role of consciousness is to contribute this response to our decision-making, giving us a capacity to make decisions that are not wholly determined by the engagement of pre-existing circumstances (including our characters) with laws of nature.  That is, it gives us our joker.

I guess Tom Clark would say this is gratuitously introducing a non-natural and contra-causal black box into our decision-making, whereas other views hold the promise of transparency.  But I say there are good positive arguments in support of my position, that it takes a wider view of what is natural, that we should not assume in advance that mechanistic explanations can be sufficient, and that we can understand non-conclusive plausible explanations pretty well (for most things, they are the best we can do).

Comments

[Paul Torek]

As I pointed out before, if "conscious experiences make a substantial positive contribution to our decision-making" deterministically, it does not follow that "consciousness would be a superfluity." Even if it followed that non-conscious processes could accomplish similar results, it would not show that consciousness in humans is superfluous. The fact that a different car uses fuel cells and not internal combustion, does not show that we can remove the internal combustion engine from your car without killing it. Your internal combustion engine remains absolutely vital.

[Tom Clark]

Paul,

There's little question that neural processes associated with consciousness are essential for decision-making and behavior control. But whether there's a causal role for phenomenal consciousness per se, as distinct from its associated processes (if it is distinct, and that's the big question) is unclear. So it seems consciousness *might* be behaviorally superfluous in us, depending on how the mind-body problem gets solved.

[Mama Africa]

great post! thanks for sharing

[Paul Torek]

Tom,

My point is that lack of causal role can't be inferred from determinism. For that matter, distinctness (of phenomenal consciousness from neural process) also can't be inferred from determinism.

[David Hodgson]

Paul,

I agree it is possible that consciousness achieves results that could be, but are not, achieved by non-conscious processes; and that if so, consciousness would not be superfluous. However, I think this is implausible, at least unless some account is given of what those results are, of how they are achieved by consciousness, and of why they are not achieved without consciousness.

Tom,

I agree that neural processes associated with consciousness may do all the work, but again I think this is implausible, at least unless some account is given of why those processes are associated with consciousness, that is, of why they have these features of subjectivity that on this view do nothing.

[Neil]

David,

There is a view on which consciousness has no causal powers - all the causal work is done by subpersonal mechanisms - but on which it is not superfluous either. That is the global workspace view, according to which consciousness allows the subpersonal mechanisms to communicate with one another. It is widely believed by neuroscientists and there is some experimental as well as conceptual evidence in its favor. On this view, it is false that consciousness does nothng. The view is attractive, because (a) consciousness is biologically expensive, so we shouldn't expect it to be purposeless but (b) it bypasses the mystery of how a phenomenal property could have causal powers. So far as I can see, however, consciousness as global workspace has no jokers to play.

[Alan]

I have had experiences of akrasia, and even stronger instances where I consciously wish to do things that I do not eventually do. Why is my conscious experience causally otiose in such circumstances? The problem is that there are few if any canonical instances of conscious causality outside ordinary action that are demonstrably incompatible with unconscious causation, and plenty of canonical instances of causality of unconscious reflex, eventual akrasia, etc. The inductive evidence is not on the side of clear and unequivocal conscious causation. I fail to see why the empirical evidence should not be decisive here.

[Alan]

Whoops--I meant this instead--substituting "outside" with "inside"--

have had experiences of akrasia, and even stronger instances where I consciously wish to do things that I do not eventually do. Why is my conscious experience causally otiose in such circumstances? The problem is that there are few if any canonical instances of conscious causality inside ordinary action that are demonstrably incompatible with unconscious causation, and plenty of canonical instances of causality of unconscious reflex, eventual akrasia, etc. The inductive evidence is not on the side of clear and unequivocal conscious causation. I fail to see why the empirical evidence should not be decisive here.

[David Hodgson]

Neil,

I think the global workspace view may well provide part of the explanation of consciousness, and part of the explanation of its function. But it still faces the fundamental problem that if what happens in the global workspace is rule-determined, it could be done without consciousness -- and indeed I understand there are computer programs that seek to use the global workspace idea. And if what happens in the global workspace is not rule-determined, then how does it work? So I don't think the global workspace view gets to the heart of the problem.

Alan,

I agree that a lot of what we do is the product of unconscious processes, and that we often fail to carry out what we consciously decide to do. But I think the considerations that support the view that consciousness can contribute to what we decide and to what we do are very strong indeed.

[Alan]

Thanks for the reply David. In my FW class I often present contemporary libertarianism as an attempt to "fit inside" the small inductive space left by evidence that much of our mental-involved behavior is explainable by causal mechanism. (Question for those in the know: was C. A. Campbell the first to try this minimalist strategy? I've found no L earlier who takes this stance in outright opposition to the scientific evidence for biological determinism.) Certainly van Inwagen and Kane take this approach, and in your own way you seem to do so as well.

In any case I think your work is interesting and you're a damn clear writer to boot!

[Tom Clark]

David,

As you know, features of subjectivity such as phenomenal feels (qualia) and the sense of self are, according to some naturalistic accounts of consciousness, exactly what one would expect were one to consist of a neurally-instantiated representational system operating in service to survival. For instance Thomas Metzinger in Being No One presents a philosophically astute, empirically and clinically based theory of how phenomenal consciousness might be *necessitated* by a sufficiently ramified functional-representational architecture. This suggests that qualia and other subjective aspects of consciousness might well not have an independent causal role over and above the representational workings of their neural realizers (assuming that they are in some sense distinct from those realizers or their workings). Metzinger's theory and others along somewhat the same lines, e.g., Jaegwon Kim's (who argues for the epiphenomenal status of qualia) present fairly well substantiated reasons, although not definitive of course, for why features of subjectivity might do little or nothing that isn't being done by their (possibly deterministic) material substrates. The philo-scientific question of the causal role of phenomenal consciousness, not to mention its root nature, is very much open, so, as I’ll continue to urge, theories of agency can't be based on a premature, contentious answer to it. In particular, just because a theory shows that subjective aspects of consciousness might be a "superfluity" doesn't count against it, since it’s quite possible our standard intuitions about consciousness are faulty.

[Eddy Nahmias]

David, you say "it still faces the fundamental problem that if what happens in the global workspace is rule-determined, it could be done without consciousness." As Paul noted in the first comment above, just because a system could replicate the behavior or even functional structure of a conscious system without consciousness, that does not show that consciousness is causally irrelevant in the conscious system. Just because a wind-up clock can keep time as well as a digital clock does not show that an electric digital clock can work without electricity.

More importantly, why is it a problem if we turn out to be "rule-determined" systems? If those rules turn out to be based on, say, reasons and desires working in a conscious deliberation process, why does free will go out the window? (Forgive me if you have made this clear in previous posts or published work I have not read.)

[David Hodgson]

Eddy,

To your first comment, I'd respond as I did to Paul. Yes it's possible conscious processes are rule-determined but are not superfluous because what they do is not done non-consciously (although it could be). But what would these rule-determined conscious processes do, and why is it not done non-consciously, as it could be?

To your second comment, I accept compatibilism is a possible view. But I contend close consideration of how reasons and desires work in conscious plausible reasoning suggests this is not algorithmic. There are links to articles where I argue this in some detail in my previous post of May 23. There is one at http://users.tpg.com.au/raeda/website/why.htm

[John Fischer]

I have suggested the metaphor or "playing the cards that are dealt you" to seek to illuminate some issues pertaining to moral responsibility. Mainly, the idea is in contrast to the (purported) requirement of being a "causa sui". Kip helpfully suggests that this metaphor presupposes that the self is separate from the cards; otherwise it wouldn't be able to do the work it is supposed to do.

But I would wish to deny this. First, as Paul (and I believe others) point out, causal determination is compatible with--does not crowd out--agency or the self's having a role. When I use my three wood on the fairway, I hit the golf ball, even in a deterministic world (in which my hitting the golf ball supervenes on a complex set of events in the deterministic net). Similarly for my playing the cards that are dealt me in a distintive way--by exercising guidance control.

[Kip Werking]

I can't much disagree with John, because at least he grants that we can't choose otherwise than we do.

Of course, John insists that (most) people are morally responsible, at least most of the time, in virtue of guidance control. However, as Paul Russell and Derk Pereboom, etc., have pointed out, guidance control raises questions.

I think perhaps the best way to focus on the difference between John (and others who share his or similar views) and myself (and others who share my or similar views) is Mele's Zygote Argument. If you focus on the Zygote Argument, you see that hard compatibilists like John are willing to grant that people are still morally responsible for their actions, provided they exercise guidance control.

To myself, and those (very few?) people who share my sympathies, the fact that such people have guidance control counts for very little considering that their entire life story was written before they were born by a Designer God/Alien (or whoever). To myself, the idea that these people could be morally responsible---in the sense at stake in the free will debate throughout the centuries---doesn't even begin to move me. It's not even a close call.

That's because, at least for me, the whole point of free will, the VERY reason people started talking about free will, and the moral responsibility people associate with it, was to deny exactly the sort of possibility the Zygote Argument contemplates. (Consider, as just one example, Lucretius' reaction towards strict determinism by positing a "swerve".) So, to say that those people have fw, or can be mr, is like saying the Pope doesn't have to be Catholic. To ignore the Pope's Catholicism ignores the very essence of what it means to be a pope, and to ignore the aspect of free will that let's ourselves, as opposed to somebody else, write our own destiny regardless of what the rigid laws of physics may say, is to ignore the very essence of what free will is and was meant to be.

To back up that claim, I would probably have to do a thorough program of research into the history of the usage of "free will" (and perhaps I could get a Ph.d doing it), so right now it is just a hunch, and I can't claim certainty. But it's the hunch I've always had, and I think it would be strange if people, throughout the ages had referred to fw, and the mr associated with it, as something that everybody agrees already exists (as everyone agrees, and has always agreed, that most people exercise guidance control).

All of this makes me wonder if John and I use "free will" and "morally responsible" in the same way, if anybody uses these terms in the same way, if there is even a strong enough consensus about the definitions of either of these terms to answer the questions we ask of them. And answering these questions would probably involve another research program, which I also haven't conducted.

But I would be very interested to hear John's thoughts on what exactly he thinks free will is, whether it exists, and how confident he is that his definition of moral responsibility as involving guidance control is the one, true definition.

[Tom Clark]

Kip,

A functional/operational definition of having fw/mr could be framed as the sort of response people think is warranted on the assumption that agents are acting freely. Fischer, Dennett, most other compatibilists and revisionists, *and* libertarians such as Kane and Hodgson, all of whom claim we have fw and robust mr (although they have very different definitions of fw), generally think retributive, non-consequentialist responses are warranted so long as a (wrongful) action is "up to us" (the theory-neutral definition of free will). Skeptics about fw/robust mr such as yourself, myself, Tamler, Derk, Joshua Greene and a few others, think not. Actions might still be up to us and free in a perfectly definite sense (no coercion, etc), but we don't suppose retribution is warranted. At any rate, this seems to me one of the central demarcations between dueling ideas of agency.

I know this is far afield from David's original post, but since you adverted to your favorite theme, I couldn't resist dragging in mine!

[John Fischer]

Kip,

Thanks, as always, for your great comments. I'm traveling in Philadelphia, so this has got to be quick. (Lucky for me, eh?)

I try to explain my answer to the Zygot Argument in my MIND review of Mele's book. Check it out. You are correct that we disagree about this. I guess I think that if the Molinist picture of God's providence were correct, then an agent (God) would have distally created the world knowing and intending all the details--but that this in itself would not vitiate our moral responsibility.

I take it that "Free Will" is a term that is quite flexible. I think it is typically used to pick out that freedom or control required for moral responsibility, whatever that freedom or control turns out to be.
For me, it is guidance control.

[Paul Torek]

David,

To answer, or mostly dodge, your question on why cognition is not done non-consciously if it could be: that partly depends on the answers to a host of thorny philosophical questions. Does intentionality define consciousness? Is the right sort of causal or evolutionary relationship between objects/properties and a being's internal informational structures, sufficient for intentionality? If such a being is intimately familiar with certain qualities of its own perceptual system, which are distinct from yet play a role in its perception of the world, do those qualities ipso facto count as qualia? Or on the contrary, are qualia a natural kind that bears no obvious relationship to intentionality and intentional systems?

Depending how we answer these questions, it might not be possible for a cognizer to be capable of the sophisticated self-descriptions and cognitions that we are, and yet lack phenomenal consciousness. Sufficiently rich and complex causal webs within cognizers and between cognizers and world would then suffice for phenomenal consciousness, though it still would not follow that such qualia would be anything we could relate to.

Or, not -- see my last question from paragraph 1. (By the way, if you want my views on phenomenal consciousness, read Owen Flanagan's Consciousness Reconsidered, the only philosophy book outside of a logic text I've ever agreed with from cover to cover.) So, supposing that human-like behavioral smarts are possible in principle without phenomenal consciousness, why did mother nature design us to have it? Why not evolve robots instead, who could survive and reproduce just as well?

The scientific part of the answer goes something like this: robots require complex arrangements of silicon and copper. A very simple arrangement of silicon and copper fails to reproduce at all. By contrast, a very simple arrangement of organic chemicals does successfully reproduce and occasionally mutate. Like the old joke about the farmer's directions, you just can't get there (robots) from here (organic starting point).

The philosophical part of the answer is to imagine a community of robots, who lack phenomenal consciousness, but have instead "quantomenal compsciousness". These robots directly register certain qualities of their own perceptual systems, and call these qualities "quantomenal". A 'bot named David argues against a 'bot named Paul, that the latter's views seem to have no explanation for quantomenal compsciousness. Paul admits that yes, in some sense it IS just a matter of cosmic luck, and aren't they lucky they aren't bags of organic chemicals (yuk!) who lack quantomenal compsciousness and merely have what the meat-bags call phenomenal consciousness (pity on them!).

[Paul Torek]

By the way, I realize I haven't addressed the question of how consciousness achieves its results. Part of the answer flows from my denial of distinctness (of phenomenal feels from certain associated neural processes). Type B materialism for me, thankyouverymuch. But you are concerned over to what extent conscious thought is or is not rule-determined.

Without trying to answer that yet, let me say that I'm very keen on the distinction between rule-following, AKA normative guidance, versus causal-law-governed processes, which laws are merely descriptive. There is plenty of room for conscious thought to be law-governed in the spaces in which it is not rule-following.

[David Hodgson]

Paul

I don’t think “intentionality defines consciousness.” A central case of consciousness is pain, and although pain may perhaps refer to things other than itself, it need not do so. I think essential to consciousness are subjectivity and qualia (broadly understood to extend to all contents of consciousness) and unity or all-at-onceness. Examples are visual and auditory experiences, feelings (including pain), conscious thoughts, and conscious doings. Intentionality is a feature of some but not all of these conscious experiences.

I have and I’ve read Consciousness Reconsidered, and also Self Expressions, and I think they have limitations. Flanagan does consider epiphenomenalism, and in that context refers to pain, but does not address how pain and other conscious experiences could operate in human motivation if they are not epiphenomenal. He does not discuss plausible reasoning in either book, or address how human reasoning works if it is not algorithmic.

In your story, you need to say just how the reasoning of meat-bags differs from that of the robots, if it does differ; and whether the robots have experiences with subjectivity, qualia and unity.

In my account, when I talk about something being rule-determined, I not talking about rule-following in the sense of someone or something recognising the rule and following it. However, I make no assumption about whether laws of nature are prescriptive or as Hume would say merely descriptive.

[John Fischer]

David,
Thanks for (another) fascinating contribution.
I like the idea of our possessing a joker, a flexible card that works in combination with the other cards in certain ways. Your association of such a card with the "capacity to decide" is suggestive, but I do not see why one would need to go on and insist that the whole process must be indeterministic. That is, more carefully, nothing in the metaphor seems to require indeterminism. Can you say more about what you think that the joker cannot work in a deterministic game?

[Paul Torek]

I don't keep entirely up-to-date on the latest neuroscience. Still, I don't think anyone knows how human plausible reasoning works, and I don't understand why it is imperative for those who hold certain philosophical views to say more about it.

Suppose evidence A, B, and C is available, and that there are at least two different conclusions either of which would be a reasonable response. Say, "Probably D but possibly E" is one, and "Either D or E, I have no idea which" is another. Suppose one creature "plays its joker" to reach one of these conclusions, while another deterministically (but not algorithmically, i.e. not by rule-following) reaches the same conclusion. So what? Both are, by stipulation, reasonable. Where's the problem? Again, suppose a second joker-playing creature, and a second deterministic creature, both reach the alternative conclusion, in contrast to their brethren. Where's the problem?

"How pain and other conscious experiences could operate in human motivation if they are not epiphenomenal" is a bizarre question. The question is how they could operate if they were epiphenomenal - or rather, the answer is that they would be inoperative. Pain feels bad, we want bad experiences to end ... how is that motivating?? Nothing could be clearer.

Finally, what does "subjectivity" add to "qualia" plus "intentionality" plus "unity"? If nothing, then I have already specified that the robots lack subjectivity, since they lack qualia. They do have intentional states, and they do cognize a unity of their own internal state-at-a-time. If something, tell me more, and I will try to flesh out the robot scenario accordingly.

[John Fischer]

Paul,
I'm not sure if you were addressing me in your first par. If so: I never said it was "imperative" for anyone to say more. Basically, given the point you make in your own par. 2, I simply invited David to say more about why he thinks indeterminism is required. Actually, I don't see why algorithms would be problematic either.

But you don't have to say anything--I just was raising a question in my mind...

[Kip Werking]

John,

Thank you for the kind words about my comments!

Two questions:

1. Do you believe that designed agents would still be mr just because they still satisfy your preferred account of mr (in terms of guidance control, etc.), or do you have an argument in addition to that, to explain why your account suffices?

In other words, I think many or most people would regard the fact that somebody could design my entire life, and predetermine every decision I would make in my entire life, before I was ever born, and write it down on a piece of paper (for example), as prima facie evidence against my having fw and/or mr. It's a pretty big bullet to bite. Hence the "hard" in "hard compatibilism". If any compatibilist view suggests that such agents might still be free and morally responsible, then this should give us pause---it should make us wonder if those requirements are actually sufficient. In particular, how can we be so confident that we have the right requirements for mr?

2. Should I understand, from your comment, that while you regard "free will" as being quite flexible, you take "moral responsibility" as being less flexible?

[John Fischer]

Hi Kip,

I have an independent argument that causation by such an agent in the remote past shouldn't matter to responsibility attributions: please see my review of Mele in the (I believe) current issue of MIND. (I enjoy the blog, but I have an aversion simply to repeating what I've said in print; please bear with me on that.)

Yes, I grant you that it is prima facie evidence of the lack of free will and moral responsibility. But in the end, all things considered, I do not think it is decisive or that it trumps other considerations. Suppose God created the universe, and Molinism is true; then God knew in advance what each free creature would do in all possible circumstances, and he created this world (actualized this world) precisely because he believed that this was the best of all possible worlds. Ok, the picture is contentious in many ways. But I find no inclination to think that no one could be free or responsible, given this sort of causation by God in the distant past (and then hands off from there on). See also Susan Wolf's Leibnizean Story (about Rose) in her book, Freedom Within Reason. Now I'm not convinced by Wolf, but it is interesting that she thinks it is just obvious that Rose is free, despite a Molnist God.