One of the most worrisome aspects of determinism, in many people’s minds, is that it means that our deliberation — all our agonizing about the choices we must make in our lives — is a sham. In Daniel Dennett’s book Elbow Room, which I think is one of the best expositions of the “compatibilist” view of free will that I am attempting to defend in this series of posts (and which I have drawn upon liberally all along), he offers three common opinions (pp. 102-103) about this problem. I will paraphrase and summarize them here.
One might point out that living in a deterministic world means that, for any state of the world at time t, only one state is possible at time t + 1. This, in turn, means that any deliberation on our part — the point of which is to select between available futures — is an illusion, because only one future was ever possible. Deliberation and determinism are simply incompatible.
But how can that be so? After all, we plainly do deliberate, all the time. So this is obviously reading too much into the pernicious effect of determinism.
Refining the objection somewhat: OK, fine, we “deliberate”, but it is not effective deliberation; it makes no difference to the outcome. Dennett gives an example of some truly ineffective deliberation:
The prisoner who spends his days and nights concocting vengeful schemes but dies in his chains has engaged in deliberations whose only effects are, let us say, to deepen the furrows in his brow and exacerbate his high blood pressure.
But then he offers a contrasting instance:
His cellmate engages in similar deliberations, and acting on them, escapes his cell and puts all his plans into tumultuous effect. If this is what “making a difference” comes to, then some of the deliberation in a deterministic world will make a big difference and some of it will not. There will be cases of premeditated murder, for instance, of which one can truly say: had the premeditation not occurred, the victim would still be alive.
So the effect of deliberation is not nullified by determinism; after all, we have no reason to doubt that if that if we decide to do X we will do X, and if we decide to do Y we will do Y.
So already we have moved some distance from the original, naive concept that in a deterministic world all deliberation is pointless; indeed, it is hard to imagine how we would react to such a revelation even if it were true. How could we not deliberate? (Dennett quotes Peter van Inwagen as suggesting that we would “either move about in random jerks or scuttles, or would withdraw into catatonia.”)
So, a further refinement: well, OK, our deliberation is indeed effective, in that it plays a necessary role in determining the future. But our deliberation is itself determined, no? That’s not real deliberation.
But is this so? As Dennett asks:
…isn’t a determined thunderstorm a real thunderstorm? Isn’t a determined traffic accident a real traffic accident? It begs the question to declare, without support, that determined deliberation is not real deliberation. But — comes the reply — in real deliberation there is a genuine opportunity for the agent, with both branches “open to the agent”. The agent’s deliberation closes off one of these as it selects the other. If the outcome of the deliberation were itself determined, then it would have been determined “all along” — so there wouldn’t have been a real opportunity in the first place, just an apparent opportunity.
This is a more substantial objection, and one to which we must return. For now, though, I hope we can agree that it is clear that our deliberation is no illusion, and that our acting as the deliberators we so obviously are plays an essential causal role in what actually happens. But before we can understand whether it actually matters in any meaningful or perceptible way if our deliberation is deterministic or not — in other words, whether we have any good reason to care one way or the other — there is more work to do.
- Who’s In Charge?
- What You Mean “We”, Kemosabe?
- Wagging The Dog
- The Weakest Link
- Causes and Reasons
- The Choice Is Yours
- Causes and Cans
- Stopping The Buck
- Do True Scotsmen Have Free Will?
- The Buck Stops Nowhere
- More From Sam Harris On Free Will
- Facts Of The Matter
- Whither The Buck?
- Daniel Dennett on Sam Harris on Free Will
- Determinism And Predictability
26 Comments
Oh, man, I wish that I had read this before replying to Gypsy. You say this all very well. I feel another book purchase coming on. Between JK and reading his military recommendations and your philosphical, and my friend the herpetoligist( don’t know the spelling! snakes) my book shelves are groaning.
Jeanie
I keep wondering what happens to the concept of randomness in your deterministic world view. Perhaps you could expound on that at some point.
Jeanie,
Let ’em groan!
Hi Addofio,
I haven’t said much about randomness in this series because I don’t think it has much bearing on the free will issue.
There does seem to be “genuine” randomness at the quantum level. Stephen Wolfram has argued, however, in his fascinating and provocative book A New Kind of Science, that real randomness can also emerge from the action of cellular automata.
Randomness may not have anything to do with free will directly–but it definitely has to do with the questions of whether or not the universe is deterministic. If at time t there is only one state possible for time t + 1 (which, incidentally, also implies non-continuity), then how is there any role for chance? Which implies more than one possible outcome–that is, only one outcome will occur, but which one it is is not pre-determined, so more than one is possible as of time t.
Quite so, Addofio; I think nobody would disagree with that. But as you note, neither a random nor a determined future at time t seems to be what most people are hoping for when they imagine themselves having free will.
Malcolm et al –
It’s just my opinion, but I think the “real” problem here lies in the “alienation” of self from causal externalities. ‘Freedom’ smacks of the “personal,” while causation is, famously, impersonal. So when we view deliberation as a segment in a potentially infinite chain of causes, it becomes very strained to think of it as yours, mine or ours.
Dennet, of course, does try to ease that strain — primarily by “talking around” it, again and again and again, thus managing to de-sensitize readers. But de-sensitization doesn’t provide a rational explanation leading to insight and/or understanding. It just makes the question seem so familiar that one can “live with it.”
Hi Bob,
But that’s exactly what we do: live with it. Unless we are going to hold out, against all the available evidence, for some sort of mysterious and apparently incoherent notion of agent-causation, it does seem that our deliberations are indeed a segment in a potentially infinite chain of causes – yet we quite rightly go on deliberating, and consider the result of those deliberations to be “our” choices – as, I believe, they are, in any sense that ought to matter to us.
I agree with you about alienation; I think there is a lot of confusion about where to draw the boundary around our “selves”. Looking at the obvious physicality of our sensory (“inbound”) and motor (“outbound”) apparatus, we tend to shrink our locus of agency to an infinitesimal point, which is actually nowhere to be found.
Malcolm –
But… As I noted, coming to be able to “live with” a problem/strain isn’t the same as arriving at a rational solution to the problem. If I’m right, Dennett’s version of compatibilism doesn’t show that personal agency and causation are compatible in the sense of being reconcilable in reason. And that’s the sort of compatibility that matters here. After all, being able to “live with” flat out inconsistency isn’t all that unusual. Hypocrites do it all the time.
Bob,
Well, something obviously has to give. Either we insist on mysterious “unmoved mover” agent-causation (in the face of the way the world, and our brains, actually seem to work, and for no reason other than that it makes us more comfortable), or we abandon the notion of agency and responsibility altogether, or we re-examine what we mean by a number of commonsense concepts that we tend to take for granted – in particular what we mean by “freedom”, “having a choice”, and indeed what we mean, also, by “self”.
I don’t interpret compatibilism as meaning that we can accept the apparent fact that we live as embedded parts of a causal world and still hang on to naive definitions of freedom; rather, I think what is necessary is to see that some of the things we thought it was necessary to insist on either don’t make much sense in the first place, or aren’t nearly as important as we thought. Admittedly, this may not be entirely satisfactory to everyone, but after all, the real world doesn’t owe us anything. If the way things are isn’t perfectly in line with our illusions, well, then, the problem is not with the world, but with our illusions.
Malcolm –
I agree that something has to give. And I strongly incline to the view that our notions of personal agency have to do a lot of the giving. But the ‘having to’ here is explicitly normative and, to that extent, in tension with strictly causal stories about deliberation. From within that strictly causal framework, what will be will be, including our efforts to make sense of our place in this world, as well as whatever products those efforts might yield, as well as our assessment of those products, etc, etc, etc. There’s just no obvious room in that world for a normatively loaded ‘have to.’
Bob,
This gets at the nub of the issue. The way I’d look at that is that the “have to” arises from the way our inquiries into the problem affect our subsequent deliberations on the matter.
Malcolm –
I guess the way I’d approach the normative ‘have to’ in a strictly causal framework is to view all norms as hypothetical rather than categorical. In other words, we have to do x IF we are attempting to realize end E. I think it likely that attempting to realize end E can be sorted out in causal terms. (Of course, this approach isn’t going to be acceptable to purveyors of categorical norms.)
Bob,
That’s pretty much what I was getting at. We are in a state where we are attempting to make sense of this issue; new information or analytical results come to light; this affects the subsequent course of our deliberations.
Sorry, but I think my just previous post probably leaves the impression that I think all norms are hypothetical. I don’t. I was just sketching how I think a “strict causalist” might (would “have to”?) try to handle normativity. But I don’t think normativity can be “reduced” to attempts to realize various ends.
Why not? What is normativity other than advising a course of action in light of a desired goal?
“Because we want X, we ought to do Y.”
Malcolm – Why not? Because I think the common inference you state presupposes a positive answer to the question “Ought we to want X?” I don’t think what ought to be the case can be reduced to what we (in fact) want to be the case, any more than what is true can be reduced to what we (in fact) believe to be the case.
On the other hand, I don’t want to encourage those who posit unexplained explainers willy nilly. What I do want to encourage is the sober assessment of the extent of our ignorance, and a curiosity that compels serious questions about unknowns. But this sort of “living with” questions should never be mistaken for having answered them.
Well, Bob, I’d say our “oughts” depend on our interests, and the appearance of interests in the world — there once having been a time when there were no creatures with interests at all, and therefore no “oughts” at all — was an evolutionary process.
It is hard (some will say impossible) for us to make a radical reassessment of all our valuations; there seems to be nowhere for us to stand.
But what question are you saying we are living with here, rather than answering?
Well, one question is whether, prior to deliberating and choosing, the course of events is actually determinate. Another question, given the seeming incompatibility of intuitions about causality and personal agency, is whether either or both sets of intuitions are mistaken, and if so, how. And, since I’ve raised the issue of normativity, all sorts of questions can be posed about that phenomenon, some of which are alluded to in our conversation. I certainly don’t have the answers.
Hi Bob,
The position I’m staking out in this series is that the course of events inside our heads is most likely as causally determinate as the rest of the world seems to be — and the intuition that I think needs the closest examination is our feeling that causality deprives us of meaningful agency, or, in particular, responsibility.
Malcolm – My raising the question of normativity is directly relevant to the issue of responsibility. And the notion of responsibility relevant to concerns about personal agency just isn’t going to yield to analysis in terms of familiar notions of causation. One can, of course, reject notions of personal responsibility — but that’s much too steep a price to pay.
Bob, while I think many would agree with what you say about responsibility and causation, I don’t: I think this is exactly the “something” that “has to give”. I think our notion of what it means to be a responsible agent (and therefore our expectations of what we think is “required” for us to be such agents) is not entirely coherent, and is the source of the difficulty and confusion that this subject always raises. But I intend to get at that in more depth in future posts; it’s really the nub of the matter.
As regards the relevance of normativity to responsibility, you’ve lost me, I’m afraid: how would you describe the relationship?
Malcolm – The notion of personal responsibility is inherently normative, in contrast to the notion of causal responsibility, which is “merely” descriptive. Someone who is personally responsibile for x is on the “moral hook” as regards the doing or not doing of x; i.e., either they ought to x or they ought not to x. That’s about as clear as normativity can be.
As for the need for something to give, why not spread the need around? Or do you think our intuitions about causation are so much more securely grounded than our intuitions about normative matters that we don’t need to worry about tweaking the former? There’s actually quite a lot at stake in how we deal with these questions.
Thanks, Bob, for clarifying that. I don’t see it quite that way, though: I’d say I’m responsible for what I do regardless of whether I “ought” to be doing it or not, which I think is a separate matter.
I have to say that I do think that our opinions about causation are more firmly grounded than our introspective intuitions about our choices being uncaused; after all, we don’t have much reason to question causation when it comes to any of the other machinery we work with all the time, or even other parts of our own living body.
On the other hand, our notion of ourselves as unmoved movers, as original loci of “agent-causation”, is not only hard to make sense of upon close logical examination, but also rests only upon what it “feels” like to deliberate — and that such an introspective approach could be depended on to give accurate information about the micro-level facts of physical causality in the brain hardly seems likely, I think. It’s also, apparently, at odds with recent results in neuroscience that indicate that decisions are made before we become conscious of them.
So, of these two options:
A) that we can be so confident in our subjective “feel” for where our decisions come from that we have adequate grounds to be confident that our brains are somehow uniquely exempt from having to work according to “ordinary” causal rules; or
B) that our brains work just like all the other physical systems we’ve ever observed, and that our intuitions about being “prime movers”, which are based on nothing more than what it “feels like” to make choices, are simply wrong;
I am inclined to choose B, especially given that, as I mentioned in this post, we never really observe the decision-making we are trying so jealously to defend from causality anyway.
I do realize that causality is a tricky business, as you have pointed out before, and consciousness certainly is also. There may be surprises in store. But option B above seems the most parsimonious for now, given what we know; it seems to me that the only reason we have for resisting it is that we don’t like some of its implications. But the truth doesn’t really depend on whether we like it or not, and anyway, I think we have nothing to fear from B anyway, as I will argue going forward.
Malcolm –
When I referred above to intuitions about normative matters I was talking about notions of personal/agential responsibility (e.g., that agents are subject to norms), but not intending that reference to include intuitions about choices being uncaused. Frankly, I’ve never had any intuitions of the latter sort, since it’s always seemed that my choices were obviously caused, as were my deliberations, as were my perceptions. So I’ve always taken it more or less for granted that some sort of compatibilist story has to be right, and that dichotomous accounts of causality and agency must be playing on false dichotomies. But for a compatibilist story to be right, it will have to get the substories of causality and agency right. My guess, from within a dense cloud of ignorance, is that if we get the causality story right, agency will fall into place very neatly (again, it’s always seemed obvious to me that agency is integral to the “causal nexus”). But trying to devise a compatibililst story that doesn’t involve deep revisions in our understanding of causality is, I think, futile.
(Harking back to conversations about causality, perhaps here, perhaps at Bill V’s, I’m not at all sure that causality and agency aren’t one and the same — or perhaps I should say, the double aspects of this Janusian world.)
Well, in these posts I’m musing on the implications of our current understanding of causality on the issue of free will. As you mention, each of us is a complex causal nexus, and I think that fact is sufficient for us to regard ourselves as responsible agents — a position I will be trying to clarify in the days ahead.