by Eric Hart
The last issue of Noesis, compiled by Chris Cole, contains some interesting reading. I hope no one will mind if I construe an invitation to enlarge on some of its themes. Consider it preparatory reading for this issue, the conceptual continuity of which may be enhanced by the following preliminary observations.
The thesis that we are somehow analogous to "brains in a vat" is
independent of all axioms
I therefore give a brief account of the ongoing search for the
While I'd never heard of the Allais paradox before Chris wrote about it, I
happen to be familiar with some of the
Cerebral functionability has structural, dynamical, and programmatic phases. If a certain interpretation of the Allais paradox - that its subjects are acting irrationally - is valid, then most human beings have rationality-preclusive cerebral parameters and/or irrational neural programming. We will see, largely on the basis of an advanced perspective on economics, that this interpretation is improperly derived. The data on which it centers are consistent with another interpretation: that the subjects are in fact acting rationally (relative to the distributional tendencies roughly described in issue no. 40), and that the contrary analytic perspective is itself incomplete and so irrational.
The distinction between rationality and irrationality implies a distinction between those who cannot reason correctly in the Allais content, and those capable of either resolving the paradox or following and internalising the reasoning of one who can. This suits the thesis that not all human brains are equal in potential. It does not. however, confirm the inherent irrationality of neural networks or of the analog reasoning by which they function. Neural computation, in theoretic equivalence to its digital counterpart, is not inimical to the arbitrary implementation, interpretation, or recursive extension of logical grammars.
The second article addresses only those aspects of the Allais paradox
explicitly mentioned in Noesis no. 40. Since I wrote it, it has come to my
attention that Allais himself also denies the premise, as derived from certain
controversial axiomatisations of rational preference, that the data imply
subjective irrationality. Specifical1y, he disputes the risk-insensitive "
Thus, both Allais and his neo-Bernoul1ian adversaries appear to overlook factors critical to the issue. Again, the spectre of irrationality looms: no sooner have we vindicated the participants in the test surveys than we indict the analysts themselves. This is all the more disquieting as said analysts might ordinarily be considered "more intelligent" than the subjects of their samples. Either the demands of rationality escalate through rising levels of thought, and/or the analysts of these data are unable to defeat irrational assumptions inherited from their tutors (an hypothesis seemingly embraced by Allais, who condemns the "dogmatic and intolerant, powerful and tyrannical domination over the academic world" exercised by his opponents). This serves only to amplify the premise that some brains are not as rational as others. More importantly, it marks a potential for improvement in the structure and programming of even above-average humah brains.
The final article enunciates and extends this theme in a socio-politically motivated discussion of eugenics, an agendum many will not support for fear of being somehow associated with certain unpopular historical figures. I should point out that I consider unacceptable the abuse or belittlernent of anyone by reason of low, average, or insufficientiy above-average intelligence. The subject nonetheless requires honesty, and I have tried to limit my assertions to those I consider obvious as well as demonstrable. While the material is intended to place no one in a defensive posture. I can only be so apologetic for its somewhat homiletic tone. My regards to the members, who I hope will forgive the bad print.
Copyright © 1989 by the Mega Society. All rights reserved. Copyright for each individual contribution is retained by the author unless otherwise indicated.