ページ

2014年11月30日日曜日

Pancomputationalism in the Light of Peirce's Cosmogony

Warning: the following is highly speculative. Take with several grains of salt.

There is an idea gaining support among physicists recently called pancomputationalism (e.g. Gerard 't Hooft, Seth Lloyd, David Deutsch), according to which the universe is literally a giant quantum computer computing its own evolution. In this view, the laws of physics can be understood as algorithms which produce patterns, and these patterns in turn act as algorithms and produce further patterns, and so on ad infinitum. The initial algorithms are products of randomness, of quantum fluctuations in the early universe, which spread themselves via entanglement and reinforced themselves via the positive feedback feature of gravity. (Of course a theory of quantum gravity is here assumed to precede the beginning of the universe.) So physics is nothing but the study of information, of patterns of computation.

Now computation is only another word for the semiotic process. So we are led to Peirce's doctrine of pansemiosis, according to which the universe is an "argument, " inferring its own evolution through deduction, induction, and abduction. It need hardly be mentioned that the above description of pancomputationalism bears striking resemblances with Peirce's cosmogony. Quantum fluctuations are Firstness, the chaos form which the universe sprang, and the infectious nature of quantum entanglement corresponds to Thirdness. (Where Lloyd says that "information tends to spread, " Peirce writes that "ideas tend to spread". These are doubtless two expressions of the same fact). Secondness, the category of individuation, may be understood in terms of decoherence. In the near future, perhaps physics and semiotics will converge towards a single discipline, into a Physics conceived as Semiotics.

What is information?
Information is the anticipation of a certain outcome based on a correlation. A possesses information about B if an observer is able to make a prediction about B, with a probability higher than 1/2, on the basis of A. For example, a footprint in the snow (a sign) possesses information about the animal that made it (its object), if an observer is able to make a hypothetical judgment about the latter on the basis of the correlation that is assumed to exist between the sign and its object. The correlation is "assumed" only in the case that there is a real tendency for the hypothesis of the sort to be borne out in experience, and only in such a case will there be a flow of information, in which the observer is "infected" by the correlation. So information is essentially a Triadic concept, involving two correlated states and an interpreting agent.