Conditions for the disappearance of long-range apparent movement were investigated. In an experiment on beta motion, critical interstimulus intervals (ISIs) of downward simultaneity thresholds for stimuli presented in continuous alternation were determined for exposure durations (EDs) varying from 3 to 160 msec. Each subject performed each test twice. Data were collected in three sessions, each for one of three angular separations (30, 6 0, and 12 0) and the full set of EDs. The distribution of critical ISIs collapsed across subjects, EDs, and angular separations shows sharp maxima at regular distances within a range of 0-110 msec lSI. Significant or near-significant peaks were found at ISIs of5, 9, 22, 27, 43, 55, and 107 msec. Although mean critical ISIs shifted with spatial separation, no essential shift of the main maxima occurred. Evidence of a periodic modulation with a period duration of 4.5 msec was obtained from the distributions of differences between critical ISIs of the first tests and their replications, which exhibit extremely low standard deviations (<10 msec). These results agree well with previous analyses (Geissler, 1987(Geissler, , 1992) that led to a taxonomic model of quanta! timing, briefly summarized in this paper. Further consequences are discussed and related to earlier developments (Geissler, 1991(Geissler, , 1992(Geissler, , 1997.In this paper, we consider simultaneity thresholds of apparent movement as a tool for exploring phenomena of temporal quantization in perception on the millisecond level of real time. Before introducing the experimental setup, we present a brief outline of the theoretical rationale from which our expectations about fine-grained temporal structures in perception derive.
MotivationWhether or not mental processes exhibit a discrete quantal temporal fine-structure is an old question in psychology. But to date, none of the proposed answers has proved to be satisfactory. Early theorizing on the issue was based on the notion of a psychological moment as a universal building block of mental time (Stroud, 1956;Baer, 1864). Evidence from the last four decades, however, indicates that approaches adopting universal principles for mental timing cannot safely assume one unique basic unit of time, if a theory of this type can be maintained at all (Block, 1990;McReynolds, 1953; for an over-
