The legacy of Paul Fraisse: II. Two further studies of sensorimotor synchronization.

Abstract

This brief introduction precedes translations from the French of classic articles on sensorimotor synchronization: Fraisse, Oleron, and Paillard (1958) and Fraisse (1966). It points out how these pioneering but often overlooked studies anticipated findings of more recent research reported in Anglophone journals.Keywords: sensorimotor synchronization, tapping, multisensory integration, anticipation, feedbackThis projected series of translations of Paul Fraisse's seminal publications on rhythm and timing began in 2009, when my translations of short articles (Fraisse & Ehrlich, 1955; Fraisse & Voillaume, 1971) appeared in a special double issue of this journal devoted mainly to A History of Music Psychology in Autobiography. The article by Fraisse and Ehrlich was the first demonstration of the increasing difficulty of tapping in antiphase with a metronome as the tempo of the metronome increases, whereas Fraisse and Voillaume were the first to study how movement timing is affected by an overt or covert switch in control over the timing of an auditory rhythm, from control by a machine (synchronization) to control by the human participant's taps (auditory feedback).Fraisse's work is pioneering and important, but today's Anglophone researchers rarely cite it, and one reason for this neglect is surely the language barrier. The research was published in French, mainly in thejournal L'Annee Psychologique, which is now openly accessible (http://www.persee.fr/web/revues/home/prescript/ revue/psy); but this is little help for those who cannot read French. The aim of the present series of translations is to help reestablish Fraisse's preeminent place in the history of rhythm research. For this issue of the journal, substantial articles have been selected: Fraisse, Oleron, and Paillard (1958) and Fraisse (1966).Fraisse et al. (1958) conducted the first systematic study of sensorimotor synchronization in which the nature and timing of sensory feedback from the rhythmic movement were manipulated. In particular, they contrasted presence versus absence of auditory and tactile feedback, and they also introduced delays in the auditory feedback. Their main conclusion was that participants try to achieve a kind of equilibrium, such that different forms of feedback arriving at different times are grouped around the auditory pacing stimuli, resulting in minimization of the overall asynchrony. The equilibrium can be achieved in different ways, depending on the relative weights that participants assign to the different kinds of feedback, spontaneously or according to instructions.This research anticipated by several decades the systematic investigations conducted by Gisa Aschersleben and her colleagues at the Max Planck Institute for Psychological Research in Munich, in which tactile and auditory feedback were manipulated in various ways during synchronization tasks (Aschersleben, 2002; Aschersleben, Gehrke, & Prinz, 2001, 2004; Aschersleben & Prinz, 1995, 1997; Mates & Aschersleben, 2000). These authors mention Fraisse et al. (1958) only in passing or at all. Aschersleben and Prinz (1997) curtly dismiss the results of Fraisse et al. as being not very conclusive because of shortcomings in the control of the experimental situation (p. 36), but they do discuss those shortcomings. Instead, they give much attention to the so-called PaillardFraisse hypothesis (Fraisse, 1980; Paillard, 1949), according to which different neural transmission times of pacing stimuli and sensory feedback stimuli account for their relative asynchrony in synchronization. Although Aschersleben and colleagues were primarily trying to explain the commonly observed anticipation tendency in synchronization (taps preceding pacing sounds), their conclusions regarding the combination of different sources of feedback are quite similar to those of Fraisse et al. (1958). For example, Aschersleben and Prinz (1995) say two sources of feedback enter into a joint-event code whose effective onset point is computed as a weighted average from the onset points of its components (p. …