The Perception of Temporal Order: Fundamental Issues and a General Model

How do people judge the order of two nearly simultaneous stimuli, such asa light and a tone? We consider this question in the context of a generalindependent-channels model that incorporates most existing models of orderperception as special cases, and which has been implicitly assumed whentemporal-order judgments are used to study perceptual latency. In the model,a "decision function" converts a difference in central "arrival times" of twosensory signals into an order judgment. The psychometric function for order isregarded as a distribution function, and can be represented additively in termsof the central arrival latencies and the decision function. Various distinct decisionfunctions correspond to various previously proposed mechanisms involving a"perceptual moment," attention switching, a threshold for arrival-time differences,and so forth (Section II).One test of the model is to compare reaction-time measurements with orderjudgments (Section III). Discrepancies can be understood by an analysis of theconcept of perceptual latency that recognizes the internal response to a pulseas being spread out in time (Section IV).An alternative test is to determine whether experimental factors thatinfluence two signal channels selectively have additive effects on the mean,variance, and higher cumulants of the psychometric function for order, as thegeneral model implies. Some data confirm the model when examined in this way,but others do not. We consider extensions of the model from order judgmentsto other perceptual domains in which the relative arrival time of a pair of signalsis thought to determine the percept. An additivity test applied to binaurallateralization favors extending the model to that phenomenon, and suggests anew method for tracing information flow in sensory channels by analyzing timeintensitytrading relations at different "levels." Such an analysis revealsdifferent effects of stimulus intensity on latency for different visual tasks, andleads to speculations about the locus of stereoscopic depth perception inrelation to other processes (Section V).The influence of attentional bias on the point of subjective simultaneitymakes tests of the model difficult. However, the model suggests how to studythis "prior-entry" phenomenon and determine whether attention influences thesensory channels or the decision mechanism (Section VI).Implications of transitivity of perceived order are examined, particularlyin relation to the idea of a single multisensory "simultaneity center" in thebrain; some experimental tests of transitivity are reviewed (Section VIII).The problem of perceived order of three or more stimuli bears on severalimportant questions, including transitivity. But existing experiments withmultiple stimuli shed little light on these issues (Section IX).Several aspects of experimental method are considered (Section X).