Tuesday, April 20, 2010

Balaguer on Free Will, Part Three

The last post discussed Mark Balaguer’s argument that that the question of libertarian freedom reduces to the question of indeterminism at the point of decision.  In this post, I’ll summarize his argument that the presence or absence of the indeterminism in question is an open scientific problem.

If there is indeterminism at the point of many decisions, we’re L-free (specifically he calls this “TDW-indeterminism” referring to torn-decisions being wholly undetermined at the point of decision).  So does the required indeterminism exist?  Balaguer says we don’t have good enough evidence on either side; it is an open empirical question.

In sections 4.2 and 4.3 of his book, he argues there are no good a priori arguments for determinism or indeterminism; I’ll skip these (there don't seem to be any that are viewed as very persuasive).

In 4.4, he begins to consider the empirical evidence.  He takes it as clear that we don’t yet have good evidence for TDW-indeterminism.   However, some people think we have good evidence for its falsity, so he will focus on arguing why there is no such evidence.

First, there is no good evidence for universal determinism in nature.  We know micro-events cannot be convincingly argued to be deterministic since the advent of QM (hidden variable theories are not ruled out absolutely, but they are a very long shot.)

Now, some would argue that despite quantum mechanics, the macro-world is still deterministic or virtually so.  A typical argument goes that, apart from a naïve interpretation of our own mental events, everything else we experience seems determined by causes.  But, actually, there are lots of things in nature without clear causes; this attitude is a hangover from when we used to think that even mysterious macro events had to be determined since we knew that the micro-events that composed them were determined.

OK, but at the level of neurons and such, things are determined, or virtually determined, right?  Well, it appears that as a matter of fact (at least from consulting a textbook and phoning up a few neuroscientists) that synaptic transmission and spike firing are typically described probabilistically.  The point isn’t that neuroscience treats these as definitely indeterministic either; but at this level of description the phenomena are best described as stochastic processes (could be theoretically consistent with either a determined or undetermined underlying situation).  In particular, the opening and closing of ion channels seems random/indeterministic.

What about the fact that large scale quantum coherence is ruled out by the hot and wet brain (cf. Tegmark’s paper in response to Penrose/Hameroff)?  Large scale quantum coherence is not necessary to the right kind of indeterminism:  as above, neural decisions could go one way or another due to micro-level indeterminism.
Balaguer quotes a few neuroscientists, including Christof Koch (p.154):  “At this point, we do not know to what extent the random, i.e., stochastic, neuronal processes we observe are due to quantum fluctuations (a la Heisenberg) that are magnified by chemical and biological mechanisms or to what extent they just depend on classical physics (i.e. thermodynamics) and statistical fluctuations in the underlying molecules.”

What about Libet’s work?  This is not a good argument against the presence of the relevant kind of indeterminism either.  In Libet’s experiments, we don’t know what the unconscious processes indicated by the readiness potential are for.  We don’t have reason to think it’s relevant to making a choice between options, as opposed to being a part of processes needed to set up the moment of decision and determine which options are live.

What about all of the recent psychological experiments which indicate our introspective conscious perceptions are very fallible guides to what’s going on?  Well, these are interesting and important but they don’t argue for determinism, per se, especially in the case of torn decisions.  Now, they do provide evidence that our decisions are influenced by non-conscious causal factors.  But these still don’t refute the idea that indeterminism is present at the point of decision in many cases.

So whether there is such indeterminism is present is an open question.  And if it is present, then we are free.

(Note that much of the discussion in ch.4 of his book is drawn from this earlier paper: "Why there are no good arguments for any interesting version of determinism")

My personal view is that the "odds" are much more in favor of indeterminism playing a key role given the ubiquity of such processes in nature and the fact that all physical processes are constituted by quantum micro-events.  It's surprising to me that the vision of nature as a deterministic machine is still so pervasive 80+ years into the quantum mechanical era of science.

1 comment:

Doru said...

Hi Steve
I’ve been reflecting a lot on the idea that at the thought level, we could have at least some nondeterministic choice. After reading through Penrose/Hameroff papers I came to the conclusion that generating a pure random conscious thought is statistically possible but the chance is so small is like one event for trillions of brains in one trillion years (or like shaking the little pieces of a watch in a box and catch that shake when all the pieces randomly assemble together into a working wrist watch).