Friday 3 December, 2010, 11:30

The Location of Subjective Information

Dr Michael Beaton, University of Sussex

Using Tononi’s Information Integration Theory of consciousness as a
jumping off point, I will critically examine the claim that internal neural
states carry information for the subject. I will review Edwin Jaynes’
presentation of Cox’s Theorem, which demonstrates that the laws of
probability are the only consistent way to mathematically capture reasoning
in the face of uncertainty. Also following Jaynes, I will argue that
Shannon information is intimately bound up with probability, and that
information for a subject exists whenever subjective probabilities exist.
Therefore, I will suggest, information for a subject inheres in the
structure of a subject’s rational behaviour. I will argue that internal
information in neural states is always derivative (information for a
theorist) – even though dynamic neural complexity can be a crucial part of
the explanation of richly structured rational behaviour. Some important
clarifications will be noted. Firstly, the rationality in question here is
practical rationality (coherent, meaningful behaviour in the world) and not
abstract theoretical rationality. Secondly, whilst probability theory
provides rules which a rational subject should not breach, it does not
provide a full set of rules for ‘how to be rational’. Indeed, it remains
possible that rationality as such can never be fully captured in rules.
Finally, even though rationality inheres in a subject’s pattern of
behaviour, there are good reasons to think that the structures of the world
responsible for a subject’s rational behaviour extend beyond the subject’s
body; we should not expect to find rules of rationality encoded anywhere
inside a rational subject.