3 PSEUDO-PARADOXES IN QUANTUM DECISION-THEORY - APPARENT EFFECTS OF OBSERVATION ON PROBABILITY AND UTILITY

In quantum domains, the measurement (or observation) of one of a pair of complementary variables introduces an unavoidable uncertainty in th e value of that variable's complement. Such uncertainties are negligib le in Newtonian worlds, where observations can be made without appreci ably disturbing the observed system. Hence, one would not expect that an observation of a non-quantum probabilistic outcome could affect a p robability distribution over subsequently possible states, in a way th at would conflict with classical probability calculations. This paper examines three problems in which observations appear to affect the pro babilities and expected utilities of subsequent outcomes, in ways whic h may appear paradoxical. Deeper analysis of these problems reveals th at the anomalies arise, not from paradox, but rather from faulty infer ences drawn from the observations themselves. Thus the notion of 'quan tum' decision theory is disparaged.