On the possible and stable psychophysical laws (Q908879)

Experimental data collected by S. S. Stevens show that when subjective magnitudes are related by a power law \(y=kx^ b\) then b turns out to lie between 0 and 3.5. R. D. Luce showed that power laws are the necessary form for such relations but his result does not give the range for b. This paper applies the notion of structural stability from the theory of dynamical systems to argue that b should be 1,2 or 3. The apparent counter-example of Stefan's law \((u=\sigma T^ 4)\) is claimed to be a consequence of the fact that this is a mathematically derived law rather than one based upon physical observation.