New Atheist philosophers Sam Harris and Dan Dennett are well known for their articulate dismantling of religious principles, Sam in his classical direct attacks on religious assertions and Dan tending towards more neutral but esoteric explorations of religious themes in the context of memetic viruses of the mind. Yet it is a mistake to think this philosophical duo necessarily presents a common underlying message. Indeed, their different thought processes seem to have created a bit of a stir recently when it comes the concept of free will.
Harris has just released his book on this topic, aptly titled Free Will in keeping with his directness. In this 96-page inkling, the newly neuroscientific doctor of philosophy Harris exorcises the demon of free will in both flesh and spirit. He demonstrates in clear language that the evidence unmistakeably precludes the possibility of a free will and this is not surprising because the concept of free will itself is inherently incoherent, at least as classically defined.
Much of Harris’ argument will sound familiar to free will philosophers. A decent portion of this work revisits the classic Dilemma of Determinism, recognizing that if thought processes follow deterministic laws then they are directly caused by prior events and hence aren’t free, and if they don’t follow deterministic laws then they are random by definition and hence also aren’t free, where “free” here means a personal choice that elicits first cause responsibility. (There is room for statistical patterns in this dichotomy but that only results in a mix of the two, where the pattern is caused but the variation is random.)
What makes the discussion timely and interesting is new technology. Neuroscientists, including Harris via his recent Ph.D., are able to monitor people’s brains via fMRI and see just what processes are involved during thinking. To cut to the chase, in some cases it is possible for such scientists to know what decisions you will make before you are even aware you made them. This is because your conscious brain only really observes the decision. In other words, you aren’t choosing an option, you are observing your choice (for useful definitions of you).
Free will therefore not only can’t exist, it doesn’t exist. (This backward sounding phrasing is rather common in science. Theorists, who are essentially philosophers armed with math, often “know” what can or can’t exist long before an empiricist observes it to be true, unless it is observed to be false in which case the theorist was merely mistaken about the details. Read Lawrence Krauss’ A Universe from Nothing for a more comical exploration of this phenomenon in cosmology.)
While I highly recommend reading Sam Harris’ Free Will, you can get a decent overview from his recent Caltech presentation as part of The Great Debate event, particularly if you are cheap or lazy or both. (Come on, it’s only 96 pages, and only about 60 are text of the main essay.)
So that’s it then. No free will. Well, except that Dan Dennett seems to disagree. His books Freedom Evolves and, going much further back, Elbow Room, look at free will in terms of what kinds of free will are worth wanting. Dennett also has a video summarizing his thoughts on this topic, taken at Edinburgh University in 2007. (Given Dennett’s rigor and verbosity I would never accuse you of being lazy for watching the video instead of reading the books. But judging by the fading faces in the audience, his talk is also quite esoteric.)
While I find Dennett’s talk quite interesting, his opinion is best summarized in the question and answer session at the end. It is here that he clearly states that he fully believes that the “Laplacian demon” with perfect knowledge could exactly predict people’s thoughts. He states that free will is complex determinism. His talk is not aimed at convincing us that free will is real, but that the type of free will worth wanting is real. If this sounds odd, it is.
Dennett distinguishes between free will as it is classically defined and the reality of complexity. In complex systems, particularly those governed by non-linear and chaotic behaviour, a small variation in a state at one time can result in a very big difference at a later time even though every step along the way is deterministic. This is essentially the Butterfly Effect that describes the unpredictability of weather, noting that the flap of a butterfly’s wing can deterministically be linked to, say, a hurricane much later that would never have happened had it not flapped its wing.
In this version of free will, your choices may be deterministic but very small changes in circumstances can greatly affect those choices. This may seem like an odd definition for free will, but Dennett addresses that at end of this talk too. Quoting from the book Net of Magic by Lee Siegel, he notes that “real” magic is not real, but illusionary magic is real. Likewise, “real” free will is not real, but illusionary free will is. Unfortunately it’s not very clear in his talk how that links back to practical uses of free will such as punishment theory for crimes, but his point eventually comes through.
I think I can best summarize the difference of opinion as follows: Sam Harris claims that free will is an illusion; Dan Dennett claims free will is the illusion.
There really is no disagreement. They are simply defining the term differently. Harris means classical free will, and claims it is a illusion of complex but deterministic processing. Dennett agrees classical free will is an illusion, and re-defines the complex processing itself as the “real” free will. Harris’ version is a fairly common belief these days and one I’ve held since my first course in non-linear dynamics and chaos. Dennett’s version, while an interesting perspective on the topic, seems to muddy the waters by re-defining the term such that readers may be prone to homonymous confusion. If you pay close attention to his competing chess algorithm example, and his Q&A description applying them to moral questions, Dennett is talking about a complex mechanism that essentially evaluates game theory problems (as in moral dilemmas). I agree with this entirely. I just wouldn’t call it free will.