Generality

Another basis for evaluating a theory is its generality. This refers both to its ability to account for behavior generated by a large number of different procedures and its ability to account for many different measures of response.

Scalar timing theory has been applied to many different procedures that include both classical conditioning and instrumental training, both perceptual and behavioral, and both appetitive and aversive (Church, 2002). It has been extended from simple procedures to more complex ones involving several stimuli (involving simultaneous temporal processing) and several responses (involving choice). It has also been extended from the analysis of steady-state behavior to the study of dynamics of behavior, including acquisition, extinction, and transition effects (Gallistel and Gibbon, 2000).

Scalar timing theory has been applied to many different dependent variables, including latency, relative response rates as a function of relative time from stimulus onset, and the distribution of the times of transitions from low to high response rates (Church et al., 1994). But it does not account for many other dependent variables. For example, in the fixed-interval schedule of reinforcement, it does not account for the absolute response rates as a function of time since stimulus onset (Figure 1.3) or for the distribution of interresponse times (IRTs). Thus, as it is presently formulated, scalar timing theory could not pass a Turing test (Church, 2001).

0 0

Post a comment