Learning to time (LeT) or scalar expectancy theory (SET)? A critical test of two models of timing

Two theories of timing, scalar expectancy theory (SET) and learning to time (LeT), make substantially different assumptions about what animals learn i n temporal tasks. In a test of these assumptions, pigeons learned two discr iminations: On Type 1 trials, they learned to choose a red key after a 1-s signal and a green key after a 4-s signal; on Type 2 trials, they learned t o choose a blue key after a 4-s signal and a yellow key after a 16-s signal . Then, two psychometric functions were obtained by presenting them with in termediate durations (1 to 4 s and 4 to 16 s). The two functions did not su perpose, and most bisection points were not at the geometric mean of the tr aining stimuli (contra SET); for most birds, the function for Type 2 trials was to the left of the function for Type 1 trials (contra LeT). Finally, t he birds were exposed to signals ranging from 1 to 16 s and given a choice between novel key combinations (e.g., red vs. blue). The results with the n ovel key combinations were always closer to LeT's than to SET's predictions . Observations of the birds' behavior also suggest that, more than being a mere expression of an internal clock, behavior constitutes the clock.