This is the third part of the talk ‘Free Will / Free Won’t’. Notes in the text in the form [ABC] can be found by following the above link.
13. The Unavoidable Daniel Dennett
In his 2003 book ‘Freedom Evolves’, Dennett revisits the issue of free will that he previously considered in ‘Elbow Room’ as part of his life-long project of providing a philosophy of mind grounded in science. Dennett is a compatibilist – and much of his argument is about how determinism can be made compatible with free will. His solution can be summarise as:
Determinism operates at a physical level whereas freedom operates at a design level.
By ‘design level’, in the case of robotics he would be referring to the design of the robot; in the case of us, he is referring to evolution. I’m not completely satisfied with his solution, but there is a key idea within this, as follows. People have a problem with determinism because it implies that things are inevitable. Inevitable is another word for unavoidable. For Dennett: freedom is the ability of an agent to avoid bad things – such as being eaten by lions. We have evolved increasingly complex mechanisms to be able to avoid bad things, including bigger brains, and social factors (group cooperation and language). Patrick Haggard summarizes his own position as [HAG]:
Freedom is just a product of complexity.
And this is a fair summary of Dennett’s position.
Again, I think many would feel disappointed about this conclusion: but to be fair to Dennett, his book is entitled ‘Freedom Evolves’, not ‘Free Will Evolves’. It does what it says on the cover [CON].
I think it is not about ‘free will’, but ‘freedom’.
14. Quantifying Freedom
One criticism of Dennett is that he discusses the concept of freedom in purely qualitative terms which I think is unsatisfactory. To paraphrase Lord Kelvin [KEL], we don’t really understand something if we can’t measure it.
So: how might we quantify ‘freedom’? The very idea might seem strange to you. But similarly, quantifying the concept of ‘information’ was strange before Claude Shannon’s theory of information which defined it in terms of entropy. Our ability to talk to one another via ridiculously cheap mobile phones and make secure payments over the internet are just two consequences of his work.
Regarding a ‘theory of freedom’, I haven’t been nearly so successful as Shannon (perhaps not surprisingly!) [QUA] but I would like to make a case that freedom arises from not just functional complexity (Dennett’s and Haggard’s case) but a combination of complexity and time [ENV] in a comparable way to how the cryptographic security of internet payments relies upon complexity and time.
15. Freedom and Cryptography
The clever maths of the cryptography works by having a key (a very large number) that is very easy to create but very hard to decode. It is a bit like it is easier to multiply than divide, but far, far more asymmetric [SOP]. Given a supercomputer and enough time, anyone would be able to decode the key – but by that time, it’s too late for it to bother us.
It is the same with freedom: freedom is the inability of other agents to predict what you are going to do, because of a combination of:
- The complexity of the agent itself (a robot, an animal, you!), and
Consider an example: a fly improves its freedom through its random-like movements. As a superior being, I could model the fly. I could do this right down to the atomic level, like Laplace’s demon, except that this would be a model just of a very small part of the universe, around the fly. I could then predict the fly’s behaviour. But by the time I would have done that, it will have moved.
A fundamental idea within cryptography is that the security does not rely upon the method of encryption being kept secret, it just relies on the particular private key that is used in the method being kept private. In a similar way, there is no need for the method of freedom to involve any ‘free will’ magic; we can still be free from the control of others in a common, deterministic universe.
16. Freedom and Time
Staying with the example of the fly, that fly can react very quickly. This is the evolutionary consequence of ‘Free Won’t’ adaptations, just like with the antelopes on the savannah. But, even so, it can only react so quickly. The fly’s freedom is a function of time: it can react fast enough to me trying to swat it with my hand, but has very little freedom if I direct a spray at it. Similarly with us: I have some freedom if someone runs towards me wielding an axe but much less if someone shoots me with a gun. I can’t dodge bullets.
[This is the end of Part III. Follow the link at the top of this blog entry for the next part.]