Daniel Dennett. Edinburgh University. 3-Dec-2010

'My brain made me do it.'
Lecture about the 'environmental effects of ill-considered philosophy'.
1. There is a popular model that free-will depends upon indeterminism. That is, freedom of choice depends upon 'a miracle'. An event that cannot possibly be predicted.
Suppose we get our free will from a machine for generating quantum randomness. We're going on a trip and the machine isn't allowed on the plane. So, prior to the trip you use the machine to generate a list of the randomly generated numbers and you take these on the trip. Whenever you make a decision you check the list. Why is that different to the machine?

The feeling is that if if there is determinism it takes away opportunities. But:
If determinism is true then you can't change the future
If determinism is false then you can't change the future
Both are true because the future is what will happen. This can't be changed.

2. Agents:
eg chess playing program.
Two programs A and B.
A is a better player = better at producing possible futures and choosing the most advantageous.
Let's pick a game where A wins.
B COULD have won if it had castled. But it didn't because its evaluation routine isn't as good as A's.
If B's random-decision routine (using a pseudo-random number generator) was replaced by a real random number generator (quantum) would that improve B's chess competency. No. Randomness is irrelevant. Over-spec.
David Wiggins - 'cosmic unfairness of determinism'. But indeterminism is also unfair. We can never control everything.

3. Does neuroscience show there is no free will?
Core experiment is Benjamin Libet. Study in voluntary actions.
a. Move your fingertip at a random point. (spontaneous whim)
b. Notice on the clock when you made the decision.
Result is the 'Libet gap' of 350ms (about a quarter of a second) between the electrode firing for the message that makes your finger move and the clock position. The claim therefore is that the decision is prior to the conscious awareness. (Slightly complicated because there is also a smaller 'veto window' where you can override the electrical message.)
But surely there are also relevant events that could be measured prior to the 350ms gap. (and the veto, for tha matter.)
DD says all this is only relevant if we assume the Cartesian Theatre (the homunculus). (1. Not physically true 2. Gets into an infinite regression)
a counter example is
Model of Libet is:
1. Neural Event 2. Conscious intent 3. Action
Counterexample:
1. Political Event  2. Senate Debate  3. Law
Number 1 might happen first, but it is still the Senate that creates the law.

4. Responsibility of Agents
1. If not other agent is responsible for your actions then you are
2. Presumption of competence. (Needed for trade, contracts, promises, agreements)
If 1 and 2 hold, then the agent can be punished

Computers etc have shown that you can have 'competence without comprehension'

What was new for me, in my view of Dennett, was the ethical dimension of his work that he was hinting at during the lecture. Of course, in a way it is obvious there is an ethical dimension - as LW pointed out, you won't solve any problem of philosophy until all problems of philosophy have been solved.