What does it mean for our traditional conceptions of free will and moral responsibility that all our thoughts and all our behavior are caused by physical processes in the brain?
There is a long-standing philosophical debate (see Free will.) Determinism is a fairly clearly defined term, whereas Free will is not. In Freedom Evolves, Daniel Dennett finds considerable philosophical recreation in considering different sorts of free will and whether they are “worth having.” Our position is that because Prediction is Intractable, the threat to any sort of free will is overblown.
We can define determinism in two similar ways:
Right away we see that determinism also has a rather philosophical character, since neither of these things could ever actually be done. The first requires time travel, and the second requires a privileged observer outside the universe (as well as some really impressive measurement technology.)
For the concept of determinism to have any practical relevance we must limit the discussion to small parts of the universe that can be isolated in a controlled environment so that they can be subjected repeatedly to the same stimulus. If we look at Atomic physics (likely the smallest scale relevant in brain function), then we find that atoms are in many ways highly deterministic. Each atom seems to be utterly interchangeable with other atoms of the same isotope, and properties such as excitation energy are so invariant that they are now used to define fundamental measurement units such as the second. However, if we look at phenomena such as the timing of Spontaneous emission, we find only randomness, and Quantum mechanics tells us that this non-determinism is fundamental.
So then we're done, right? No determinism, so having a physical brain does not constrain free will. Hold on… First of all, as Dennett points out, randomness is not a “form of free will worth having.” Second, as discussed in Mind, the brain is designed to overcome the fuzzy imprecise behavior of the goop it is made out of, keeping our hearts beating at an appropriate rate in spite of atomic scale randomness such as Brownian motion. Even so, since the brain is basically an Analog computer there is surely some level of true random nondeterminism that leaks into our decisions.
But even without this, human decisions would still be highly unpredictable due to the sheer complexity of the brain and the unpredictable inputs that each person constantly receives from the larger environment (which includes other unpredictable people reacting to our own behavior, adding to the Physical Chaos.)
In spite of underlying randomness and irreducible complexity, in our common experience we do see that people are somewhat predictable, showing a Personality or temperament that is reasonably consistent over time, as well as habits or other characteristic behaviors.
This differs from the sort of precise “I know what you will do next” determinism of philosophy thought experiments because it is only true in a statistical sense: “Joe tends to go and work out at the gym when he's had a bad day at work.” This sort of human predictability is real, and can at times seem like a sort of suffocating mechanical control that defeats our free will, especially when we are consciously trying to change how we behave. Though this sort of human determinism may be frustrating, it is familiar and unthreatening.
What is new is that fMRI and other improved measurement techniques are revealing more of the physical workings of the brain, undermining our naive dualism and suggesting the possibility that human behavior could be predicted much more accurately using these new means. Indeed, in the lab it has been found possible to predict simple decisions some seconds ahead of the apparent time of conscious decision.
Due to the fundamental intractability of complex systems, predictions of individual behavior will always be short-term, statistical, or both (as are weather predictions, and for the same reason.) However, improved instrumentation and analysis may allow more accurate and less subjective predictions of behavior. The most dramatic area of practical application of behavior prediction is in the control of socially unacceptable behavior, especially in the legal system. See The Law and Neuroscience Project and Neuroethics. There is also considerable interest in exploiting regularities in human behavior for financial and political gain (marketing, spin control.)
Since free will is traditionally understood as being a conscious process, see also Consciousness.