Font Size: a A A

Systematic nonlinearities in the production and perception of temporal intervals

Posted on:1998-10-16Degree:Ph.DType:Thesis
University:Brown UniversityCandidate:Crystal, Jonathon DavidFull Text:PDF
GTID:2460390014974678Subject:Psychology
Abstract/Summary:
This dissertation examined the empirical relation between physical time and the psychological representation of time, a relation that has typically been observed to be linear (linear timing hypothesis). In time production experiments, intervals (10-160 seconds) changed by two seconds on successive trials (Experiments 1 and 2). Temporal representation was assessed by examining the times at which rats start and end bursts of lever presses. Linear timing predicts that start and end times are linearly related to interval durations. Start and end times were approximately proportional to the timed intervals. However, departures from the linear expectation were systematic. Departures from linearity from two overlapping ranges of durations superimposed when plotted as a function of interval duration, suggesting that the nonlinearities are associated with timing interval durations. The rats anticipate the current interval, rather than the previous interval. Departures from linearity for start and end times superimposed as a function of intervals, suggesting that the source of nonlinearity in timing is in the memory representation of time rather than in a decision process.; In time perception experiments, rats judged short and long noise durations in a two-alternative forced choice task (Experiments 3 and 4). The duration of the long stimulus was adjusted to maintain accuracy at 75 percent correct for short durations from 0.1 to 50 seconds. Linear timing predicts that approximately equal ratios of short and long durations are equally discriminable. Sensitivity to time, from signal detection theory, was nonlinear, with local maxima at 0.3, 1.2, 10, 24, and 37 seconds. Similar nonlinearities were observed in two groups of subjects that received different orders of short durations, suggesting that the nonlinearities are based on the absolute value of the durations.; The generality of nonlinearities in the representation of time is demonstrated by consistency across subjects, interval ranges, and experimental procedures. A connectionist timing model qualitatively predicts nonlinearities in the memory representation of time. A scalar timing model fails to predict nonlinearities. The empirical nonlinearities in timing imply that equally spaced physical durations are not represented as equally spaced psychological durations.
Keywords/Search Tags:Nonlinearities, Durations, Interval, Time, Timing, Representation
Related items