I am mainly looking for information based on Dennett's work, I Could Not Have Done Otherwise- So What? because that is the only thing I am familiar with other than D'Holbach, but other works will do fine as well as long as your provide adequate context and resources.

Towards the end of his work, Dennett seems to talk more about how the CDO principle is irrelevant because when humans make a mistake, they are more concerned with how to improve themselves than whether they could have done otherwise, and I just don't understand how this is relevant to the larger conversation of determinism and free agency.

In this context, why do compatibilists believe that whether we act freely is independent of whether or not determinism is true?

  • 1
    They believe it because they interpret "free" differently, as in, roughly, in accordance with one's inclinations, see SEP, Compatibilism. The inclinations themselves may well be determined, hence compatibility of freedom with determinism. That humans are more concerned with improving themselves, i.e. bringing actions into better alignment with inclinations, presumably supports compatibilist view of freedom. Not that it is convincing, after all trying to improve oneself seems to manifest implicit belief in being able to act otherwise.
    – Conifold
    Nov 8, 2020 at 21:27
  • Welcome to SE Philosophy! Thanks for your contribution. Please take a quick moment to take the tour or find help. You can perform searches here or seek additional clarification at the meta site. Don't forget, when someone has answered your question, you can click on the arrow to reward the contributor and the checkmark to select what you feel is the best answer.
    – J D
    Nov 9, 2020 at 2:45

4 Answers 4


As I understand the principle — and I put it that way because the worldview is not one I share, for logical and philosophical reasons — the compatibalist position relocates 'choice' to be a property of mechanisms. In other words, a switch, a flipped coin, or a computer circuit are all capable of 'choosing' because they can be in this state or that state, and the outward (high-level) appearance is that of a choice. It doesn't matter, thus, what the underlying reality is: e.g., whether the switch was flipped by some deterministic physical force or whether the switch flipped itself. A choice was made, so the capacity to choose exists.

Going further, the human brain is conceived as just such a mechanism, albeit of a more complex design, and the human capacity for choice is just such an unattributed change of state. Thus the question of whether one could have 'chosen otherwise' becomes academic and meaningless. One didn't choose otherwise, so one only has to consider the state one is currently in (the de facto status quo) and the state one might next choose to be in. Think forward about how we can choose in the future and never think backward about what could have been done differently.

Of course, this completely sidesteps the question of will: of whether such state-change is a deterministic function of environmental forces or a semi-independent (free) decision by an agent (i.e., what we conventionally think of as 'choice'). Which is (obviously) what Dennett was going for. Dennett's worldview is explicitly anti-religious and anti-metaphysical. While he doesn't want to deny the capacity for 'choice' (because people find that strictly deterministic position difficult to swallow), he does want to preclude any form of material-world-independent agency, because such agency opens the door to discussions of transcendental, metaphysical principles. He's structured a carefully ambiguous philosophy: one that tries not to say anything on its own, but precludes others from saying things Dennett doesn't agree with.

If I wanted to sum up the idea — sourly — it's as though he asks the question:

  • "Do we have free will?"

and answers it with:

  • "Meh, we make choices, but we don't have souls."

... which merely deflects the problem onto the question of what we mean by the phrase 'make choices'. It doesn't really satisfy as an answer, but I'm not certain it's meant to be anything more than a rationalization.

  • Ignoring the negative comments, I think this answer is spot on regarding Dennett's goal: He wants to side-step the need for dualistic metaphysics and (I think correctly) concludes that the only way to do this is changing the (ontological!) perspective on what choice really means. This does in no way touch the ways in which there is a meaningful and rich psychological and pragmatic use of the term which may differ from that one, e.g. in the form of Davidson's primary reasons as causes. All he wants to avoid is taking Kant's path into metaphysical dualism.
    – Philip Klöcking
    Feb 4, 2021 at 12:21
  • But regarding the question, I guess this does not really answer the question as asked, since it misses the consequence: If we alter the understanding of free will in a way that does not lead to metaphysical dualism, ie. into a psychological category of accordance with inclination or a question of authenticity, it completely loses its impact on metaphysical questions. It is only because of that fact that the question of whether we act freely does lose the impact on metaphysical questions and vice versa.
    – Philip Klöcking
    Feb 4, 2021 at 12:28
  • @PhilipKlöcking: We could alter the understanding of 'food' to exclude 'processed sugar', too, but that wouldn't solve the problem of diabetes. This isn't a mere word-game; it is tied deeply to human praxis. If we eliminate metaphysics in the way D suggests (by consigning 'will' to the category of 'psychology'), we effectively eradicate moral theory, which depends on the notion of a moral agent. And if we tear up moral theory at its roots, what kind of society are we left with? Feb 4, 2021 at 16:51
  • @PhilipKlöcking: The question is "Why compatibalists believe...", and the answer (ai best I can see) is that they believe this because it's important to their ultimate goals, and they do not care to examine the philosophical details or ramifications too closely. I'm sorry that comes off so sourly, but I have no other way of explaining their position. Well... I suppose I could say that they have no agency: that the universe led us all to a point where they necessarily are in this state, and necessarily imagine that they got there by (psychological) acts of agency. But please, don't make me... Feb 4, 2021 at 16:58
  • Well I think it would be dishonest to suggest Dennett did not examine the philosophical ramifications too closely. As an external realist and materialist, there really are no consistent alternatives. This does in no way belittle the importance of moral discourse. It rather points out that there are discourses which imply their own ontologies without necessarily having an impact on metaphysics in the sense of what is part of the world proper. I myself think that this is an arbitrary exaltation of the ontology of one discourse above others without epistemic justification, though.
    – Philip Klöcking
    Feb 4, 2021 at 18:27

Compatibilists assume the truth in some sense of determinism and the truth in some sense of freedom. Their view is not 'independent of whether or not determinism is true'. Compatibilism assumes the truth of determinism but aims to persuade us that the truth of determinism is compatible with the reality of freedom - which here means free will. This looks to be some task, but that's the aim.

Take determinism to be the view that for any event or state of affairs, E, (including actions) there are preceding conditions, C, of which the occurrence is sufficient for E. In other words, given C, E will occur.

This formulation won't satisfy everybody but the main issue involved in the debate about compatibilism does not turn, I think, on the precise definition of determinism. What compatibilists hold is that our actions can be determined and yet still be free. They are able to reconcile freedom and determinism, not by offering a reconceptualisation of determinism but by offering a reconceptualisation of freedom.

For the compatibilist, I am free if and only if certain constraints on my actions are absent. Typically these constraints are coercion and compulsion imposed by another agent.

Assume that my actions are determined - say, causally determined by my genetic inheritance. I have a particular character and a pattern of motivation and intentions, all of which are determined and produce my actions. My genetic inheritance, C1, is sufficient for E1, my particular character and pattern of motivation and intentions. (This is a crude picture but it will serve to make the main point.) When I act I could not have done otherwise - so what? So long as my actions are the products of E1, without interference by others in the way of coercion or constraint, I act freely. Once coercion or constraint is introduced, I am unfree.

So free will is opposed not to determinism but to coercion or constraint. This is a perfectly intelligible position. The only trouble is that it reconceptualises freedom and renders the traditional problem of free will and determinism unrecognisable.

As the problem of free will has been traditionally understood, free will and determinism cannot both be true. They are incompatible because (on one standard approach) the human agent has a freedom of choice regardless of all and any preceding conditions, C.

While this formulation of free will has the merit of being traditional, it has the drawback that it seems to make human actions merely random (so not ascribably 'our' actions at all but merely things that happen) or to involve a non-empirical self, outside the realm of determinism (since it is not subject to determinism) but able to act within it. Is such a self a coherent possibility?

If compatibilism redefines freedom and sidesteps the traditional debate between free will and determinism, incompatibilism of the kind just described leaves the explanation of free action at least not transparently coherent. But this is to raise issues beyond the immediate range of your question.

For the record I have never been able to reach a stable view about free will and determinism. But this is autobiography, not philosophy.

  • Thank you very much for the balanced summary of the problem. I was particularly impressed by where you say of either the impersonal randomness or the unempirical self, a dilemma. In a traditional view (heavily grounded in natural science, too), in order an action (choice) to be own, it must belong a Being, only Being (something) is active and can possess it.
    – ttnphns
    Feb 4, 2021 at 21:37
  • (cont.) But what if Nothing is not just mere and abstract absence of Being, but is the real active annihilator of it? Next, freedom is traditionally felt as a merit, almost a "virtue" to break out the conditions. But what if freedom (and as a basis of free will, too) is a deficiency, the inability to be caused? Then it is what would "like to" get caused (founded), if only it could, but it can't engage Being for the causation. It could only constitute a loosen out-cast of Being instead - a possibility, to get caused by it.
    – ttnphns
    Feb 4, 2021 at 21:38
  • (cont.) Thinking along this or similar lines, it is perhaps feasible to find randomness truly own. Who knows?
    – ttnphns
    Feb 4, 2021 at 21:38
  • ttnphns: thought-provoking comments, I need to run them through my head. Must say I appreciate your giving time to considering my answer. Best - Geoffrey
    – Geoffrey Thomas
    Feb 4, 2021 at 21:50

As I was arguing in a recent discussion Testing Free Will position and momentum are a poor source of predictions about the future of a fellow human, even though a physicalist-materialist sees our behaviour as fundamentally, in principle, reducible to mechanisms that can be captured by position and momentum.

Just as we create a heuristic explanatory layer 'biology', which provides narrative groupings that are pictured as causal on the physics layer, in a way that is supervenient - but still reducible to physics. So with the heuristic explanatory layer with narrative groupings like 'identity', 'motivations', and 'intentions' or acts of will. In this way free will is subjectively real, as a heuristic for predicting others, but fundamentally emergent from biology, which is from chemistry, which is from physics m

Such a layer, say biology or psychology, reduces the combinatorial explosiveness of mistakes or errors about the initial conditions in a more fundamental layer, that more or less rapidly compromise predictions framed from there. Physics (& chemistry) predicts the behaviour of a dead body perfectly. A living body needs ideas like nerve impulses, and harvesting of Gibbs free energy to maintain local order. Predicting social interactions, uses ideas like 'character', in the same way biology uses a grouping like 'cell', to predict regularities, that enable useful predictions. And intentions or 'will'.

Now I link to uncertainty/randomness. In a classical universe, there is only sensitivity to initial conditions, preventing 'Laplace's demon' gaining unlimited predictive power, in principle. So quantum uncertainty is just like this general uncertainty about initial conditions, but contradicting even the 'in principle'. Uncertainty will have non-linear & unpredictable consequences for the accuracy of a model, based in the terms of the fundamental or irreducible layer of mechanism (physics).

But I think this framing that Dennett uses, where classical & quantum pictures are equated in relation to subjectivity, diminishes the power of minds, of biology. Exactly what they do, is find patterns, and the classical picture assumes there fundamentally is a single pattern that can predict everything (position & momentum of every particle at any 1 instant), whereas a universe with the uncertainty principle has to be treated not as the fuzzy edges around a definite reality, but as a set of realities, within a state space bounded by measurements. That has important consequences I explored here Why is the universe governed by very few laws of high generality instead of lots of particular ones?

Those who deny free will, are really after those who believe in libertarian free will (typically causal agency from a soul). Randomness cannot supply causal agency from within a being, any more than determinism. But quantum-woo means this is a red herring that must be dealt with by compatibilists.


Compatibilists redefine both freedom and determinism beyond recognition.

The freedom to determine one's own actions does not exist in determinism, where every action is causally determined by the previous event. There are no degrees of freedom at all.

Reconciling the conflict between freedom and determinism is simply not possible without redefining both. It is not a true freedom, if determinism rules. It is not true determinism, if it allows any kind of freedom.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .