Je pense, donc je pense.
~ Not Descartes
Ten years ago, Daniel Wegner (2002) published a provocative book in which he claimed that conscious will is an illusion. Specifically, he aimed to refute the idea that the conscious experience of willing something can cause that something to happen. This castration of conscious will dealt a blow to the concept of human agency, according to which humans do many of the things they do because they want to do them and know that they want to do them. Wegner made his argument in the tradition of the Scottish philosopher David Hume. He noted that the experience of conscious will and correspondent behavior are often correlated and that we mistake correlation for causation. In his elegant experimental work, Wegner showed that he could increase or decrease this correlation at (his) will and change people’s sense of agency accordingly.
Though intrigued by the experiments, I was skeptical of Wegners’s conclusions. I could not see how the results supported the much stronger claim that the entire sense of agency is an illusion. The misperception of agency does not entail the impossibility of agency. If agency were impossible, then the experience of conscious will would be an epiphenomenon. According to this view, the brain prepares and executes both action and the conscious intention of acting at the same time. Conscious will is thereby correlated action without having any causal power. The correlation is due to the common cause of brain activity. If conscious will does not (ever) cause action, it is a dead end in the stream of causation. Conscious will is, as it were, an organism without offspring in the great genealogical tree of nature. Just as future generations are begotten by other organisms that mate, future actions are caused by other brain activities that are not causally dependent on conscious will.
This argument seemed illogical to me. If brain activity B causes both conscious will C and action A, then by modus tollens, we can infer that by denying C, we also deny B (Krueger, 2004). Now I have to concede that if B is denied, A is still possible. The action A might be caused by some other type of brain activity B*. Only if the argument is “If and only if B, then C and A,” then modus tollens also denies A. Either C is not an epiphenomenon, but a necessary part of the causal structure, or nature was wasteful when evolving C in spite of the existence of direct causal paths from B to A. In the latter case, we return to the question of why we should have conscious will if it doesn’t do anything?
Instead of asking “Why do we have consciousness?” we may want to ask “Can we imagine [consciously -- heehee] a world in which we are unconscious while everything else is the same?" If we can, we have proof that consciousness is unnecessary. This is where Wegner is taking us. His claim suggests (implies?) that a world without consciousness (not just without conscious will) is possible. In such a world nothing happens that requires consciousness. To some this may seem a bit outlandish, but it takes a specific case to disprove it. All of your behaviors, Wegner would say, could occur if you were an unconscious automaton. You could sing, swim, make love, take the garbage out, and pay your taxes. Direct paths from B(rain) to A(ction) would do. There is one exception, though: You could not discuss consciousness without experiencing it. An unconscious automaton discussing consciousness is a matter of science fiction, not nature. It is a logical contradiction. If a conscious automaton discussing consciousness is impossible, then consciousness must exist. In syllogistic terms, we say that “If there is no consciousness, there will be no discussion of consciousness,” and conclude with modus tollens that since there is discussion of consciousness, there is consciousness.” Je pense, donc je pense.
You might object that automatons can be designed to talk about consciousness. This is true, but the designing can only be done by someone who knows the experience of consciousness. If this is another automaton, the question is begged who designed this automaton, and so on. No one likes an infinite regress. It cannot be automatons all the way down. There must be the experience of consciousness at the bottom.
Wegner does not deny consciousness itself, but its causal power. I argue that our talking about consciousness requires consciousness as a cause. Therefore, consciousness causes at least something, and therefore it is not an epiphenomenon.
Perhaps you think there’s something fishy about my argument; or you smell a rat. You could say that if modus tollens can prove the existence of consciousness, it can prove anything. How about this syllogism: “If there is no god, there is no theology. But there is theology, therefore there is a god.” I have argued against proofs of god elsewhere, and I should be particularly troubled by this analogy.
But I’m not troubled because this is not a good analogy. The modus tollens proof of god would be empirically valid only if the proposition were empirically valid. It could just be the case that god does not exist and yet we are talking about her (i.e., have theology) [see note below]. The question is now whether this disarmament can be applied to the consciousness syllogism. We could charge that the proposition “If there is no consciousness, we will not talk about consciousness” cannot be true. We could, in other words, discuss consciousness without having experienced it (at least once). I tried to disable this line of thinking with the automaton example. Why does this not apply to the god case? We can talk about god just like we can talk about anything imaginary, but we cannot talk about imagination (consciousness) without having imagination (consciousness). God, if she existed, would presumably exist outside of our consciousness. But consciousness cannot exist outside of our consciousness. That would be a self-contradictory claim.
If this is the stuff of headaches, I sympathize. This is why the good lord created ibuprofen and the ability to read text a second time. It is also possible that I am wrong, in which case I can edit this post if the errors are small, or delete it if the errors are great. At least, do yourself a favor and think about it. When you do, you have proof not only that you exist, but that you do so consciously.
Note. This is the inverse of the so-called belief bias. Belief bias means that we believe a logically invalid conclusion because it sounds (and perhaps is empirically) true. Here we have a case in which we might believe an empirically false conclusion because it is logically valid.
Krueger, J. I. (2004). Experimental psychology cannot solve the problem of conscious will (yet we must try). Review of ‘The illusion of conscious will’ by Daniel M. Wegner. Behavioral and Brain Sciences, 27, 668-669.
Wegner, D. M. (2002). The illusion of conscious will. Cambridge, MA: MIT Press.