Monday, February 7, 2011

A Critique of Compatibalism

Part of an ongoing conversation I am involved with on a friends blog got into the philosophy of free-will (can I ever escape this topic??) and I was asked to offer my thoughts on compatibalism.  Given that these thoughts are a little off-topic for that discussion I decided to post them here rather than there to avoid too many rabbit trails.

So here very briefly ("very" is a relative word in my vocabulary, be ye warned) is my critique of compatibalism:

Compatibalism purports to make the following two statements "compatible" with one another:
  1. Determinism is true.
  2. We have free will.

Truth be known, they have already lost me at statement one.  But we'll leave my more general critiques of determinism aside for now and talk about how compatibalists attempt to relate these two statements.  The claim is that so long as my "will" or "desires" correspond with the determined actions in which I take part, then I can still say that I am "free," at least in the way that really matters, which is the way that makes me morally responsible for what I do.

So what the compatibalist is doing, then, is proposing a division between our "will/desires" and our "actions."  Then they are arguing that so long as these two planes correspond with one another- so long as the "will/desires" side of the equation equals (or gives approval to or lines up with) the "actions" side then I am free in the sense of being morally responsible for my actions (because I at least condone my own actions even if I couldn't have chosen any other course of action).

I am going to offer two critiques here.  The first has to do with the relationship of the "will/desires" side of the equation to the "actions" side.  The second has to do with the origin of the "will/desires" side.

First, how do these two sides relate?  It seems like we have three options.  The first and the most commonsensical option is that the will causes our actions.  The second would be the opposite of that- that our actions have the epiphenomenal effect of causing our will- in other words, our "will" is really just a perception that is caused by our actions, but it doesn't exist independently of those actions.  The third option is that the two don't actually have any sort of causal relationship, they exist parallel to one another without interacting with one another.

For the claim of compatibalism to be distinct from that of strict determinism, it seems like the relationship between the "will/desires" side and the "actions" side of the equation is going to have to be such that they are not causally interconnected.  Basically the claim that correspondence between these two sides of the equation makes free will and determinism compatible is, I want to argue, compromised by any causal interrelatedness of the two.

To see this, lets start with epiphenomenalism.  If our will is actually the result of our actions then while the two will always coincide, this is not because the will is freely condoning our actions.  In this case, the will is every bit as determined as our actions are because of the causal relationship between the two.  In other words, while the two will always coincide on this scenario, that is at best a trivial observation and has no bearing on our "freedom" because the will is itself dependent on our actions and thus must coincide with them (is itself determined to coincide with them).  Something similar happens in the case of the will causing our actions.  In that scenario, we again have a causal dependence of one plane on the other.  If we want to maintain that our actions are determined then this interdependence will lead us to say that the will is likewise subject to determinism.  In both of these scenarios we end up with a will that is subject to the same determinations as we are positing for our actions.  While this certainly implies that our two planes will always line up with one another, it seems very difficult to say what value this has for making determinism compatible with free will.  In fact, in both of these scenarios we end up with strict determinism.  Compatibalists want to say something more than this- strict determinism is not enough, the will needs to be independent of our actions so that it can freely condone them and make us responsible for these actions.  If there is any causal relationship between the two planes, this independence is not possible.

This leaves us with the option that our will and our actions have no interaction with one another.  Thus, the "action" side of the equation could be determined and the "will/desires" side free and the two coincide with one another in a way that makes free-will and determinism compatible.  The problem with this scenario is that for our will and our actions to line up we are going to have to believe in a tremendous chain of cosmic coincidences, which strikes me as wildly unlikely.  Perhaps the compatibalist wants to bite that bullet, but I think there are probably simpler explanations in either the strict determinist or free-will camps than to claim that every free will that has ever existed has somehow magically lined up with every determined chain of actions performed by the bodies those free-wills were attached to.

Second, where does the "will" originate?  Here it seems we have two basic options: that the will is either "internal" to us, meaning it is in some sense generated by us, or that the will originates from some source "external" to us.  This question is more or less independent from the previous one- we could imagine a will that originated in either of these ways relating to our actions in any of the above considered ways (with the possible exception of matching an externally generated will with the epiphenomenal interaction scheme above), but when we ask this question in the context of  compatibalism we will discover some considerations that result in a bit of a dilemma.

Lets consider the "external" origin option first.  If we choose this option, I contend that we have to give up the notion of freedom altogether.  If our actions are determined and our "will" is imposed upon us from an outside source, then while our will may "condone" our actions, this is meaningless in so far as establishing our responsibility for these actions goes.  So if this is the angle we take, then the compatibalist goal of harmonizing free will and determinism has failed because we have entirely eliminated free will.

The "internal" option- where we ourselves generate our own will- poses its own challenges for compatibalism.  First, there is going to be a serious question of how this could happen and what it would look like without the first statement- that determinism is true- being undermined.  Second, this is going to connect back into the first critique: how can a will that we have ourselves generated (assuming this can be done and determinism still be true) interact with our (determined) actions?  My thought, for what its worth, is that the only way for the compatibalist to consistently do this is going to be very similar to the conclusion of the above argument- the will must come into existence on a metaphysical plane that has no interaction with the tangible (and determined) world, it must develop entirely on that plane, and experience the tangible world as a passive observer on that plain that coincidentally has condoned everything that it observes.  Any other solution, I think, is going to collapse back into a strict determinism.

Let me try and summarize all this and perhaps conclude with one succinct critique.  The often heard slogan of compatibalism, that we are free so long as we do what we desire/will, is not very helpful in my opinion.  This is because on any of the systems we have touched on here- either strict determinism, libertarian freedom, or compatibalism we could be said to do what we desire/will.  People act in accordance with their will (unless we want to postulate some sort of voo-doo principle of action).  What I have tried to demonstrate here is that without the compatibalist also arguing that the will must be independent of our determined actions then there is nothing separating their theory from that of a strict determinist.  The problem is that to make that kind of independence work we end up creating a metaphysical gap within people- we create what is known as an "interaction problem" where it is impossible for  one part of the person to interact with the other.  In this case, the result is what I think is the rather absurd conclusion that for compatibalism to work we have to believe in a cosmic chain of coincidences between what we will and what we actually do.  Here would be a good place to note this: I don't think Theism helps the compatibalist on this issue.  This is part of the point of the second objection- if God makes the coincidences work then this is either because God has imposed our will on us (thus eliminating freedom and collapsing back into strict determinism) or because God has planned history around our free will (the middle-knowledge hypothesis, which I also have some suspicions about but also think is the only version of compatibalism that works), a conclusion most compatibalists would want to adamantly deny.


  1. i thought you were kidding when you said "very" is relative for you. you were not.


Blog has moved, searching new blog...