Timing, Sequencing, and Executive Control in Repetitive Movement Production Ralf Th. Krampe University of Potsdam Ulrich Mayr University of Oregon Reinhold Kliegl University of Potsdam The authors demonstrate that the timing and sequencing of target durations require low-level timing and executive control. Sixteen young (M age = 19 years) and 16 older (M age = 70 years) adults participated in 2 experiments. In Experiment 1, individual mean-variance functions for low-level timing (isochronous tapping) and the sequencing of multiple targets (rhythm production) revealed (a) a dissociation of low-level timing and sequencing in both age groups, (b) negligible age differences for low-level timing, and (c) large age differences for sequencing. Experiment 2 supported the distinction between low-level timing and executive functions: Selection against a dominant rhythm and switching between rhythms impaired performances in both age groups and induced pronounced perseveration of the dominant pattern in older adults. Complex rhythm production requires the precise timing of in- tervals, such that their absolute and relative durations as well as their adequate sequencing are respected. Theories of action control (e.g., Jordan & Rosenbaum, 1989; MacKay, 1987) refer to these performance constraints as the problems of timing and serial order control. Many theorists would agree that this distinction corre- sponds to dissociable mechanisms (e.g., MacKay, 1987; Vorberg & Wing, 1996). Indeed, there exists a large body of work on the neural bases of simple timing processes (e.g., Hazeltine, Helmuth, & Ivry, 1997; Ivry, 1997; Wing, 2002) as well as on the psycho- logical (Collard & Povel, 1982; Rosenbaum, Hindroff, & Munro, 1987; Rosenbaum, Inhoff, & Gordon, 1984) and neurologic bases of sequential processing (e.g., Grafton, Hazeltine, & Ivry, 1998; Hazeltine, Grafton, & Ivry, 1997). However, empirical evidence from within one experimental paradigm that involves both a sys- tematic variation of timing and sequencing complexity is missing so far. Our basic assumption is that a critical component of the serial order control problem consists of the adequate selection of action plans that specify parameters for the hierarchical programming of the low-level timing mechanism while countering the interference from competing plans. So conceived, the serial order control problem during the production of temporal sequences is just one type of a more general class of executive control problems. We also assume that the low-level timing mechanism can function in relative independence of executive control mechanisms in tasks with low representational demands on target durations and their sequencing. We call this set of assumptions the executive control hypothesis of timing. We test this hypothesis with two methods. First, we determine for individual participants the most salient characteristic of timing processes, the systematic increase in variability with produced interval duration (mean-variance functions) for a number of ex- perimental conditions. These conditions comprised the repeated production of identical target intervals (isochronous tapping) as well as the sequencing of multiple target intervals into rhythmic patterns. The most general prediction of our model is that slopes of mean-variance functions in complex rhythm tasks should be steeper than slopes for low-level timing obtained from isochronous tapping. At a more detailed level, the executive control hypothesis predicts specific mean-variance patterns for the different target intervals within a rhythmic sequence, which identify hierarchical control over and above low-level timing. Our second approach to dissociate low-level timing and se- quencing mechanisms capitalizes on the observation that low-level timing remains functional until older adulthood (e.g., Duchek, Balota, & Ferraro, 1994; Greene & Williams, 1993), whereas specific aspects of sequential processing exhibit marked age- related decline (Kramer, Hahn, & Gopher, 1999; Mayr, 2001). In sum, we aimed to demonstrate task-specific and age-specific dis- sociations between low-level timing and executive control. Ralf Th. Krampe and Reinhold Kliegl, Department of Psychology, University of Potsdam, Potsdam, Germany; Ulrich Mayr, Department of Psychology, University of Oregon. The project was funded by the Grant INK 12 (Project B1) of the German Research Foundation. Parts of this research were also supported by Na- tional Institute on Aging Grant R01 AG19296-01A1 awarded to Ulrich Mayr. This research was part of the interdisciplinary project “Formal Models of Cognitive Complexity” conducted at the Center for Cognitive Studies, University of Potsdam, Potsdam, Germany. Comments by Asher Cohen, Richard Ivry, Steven Keele, and Dirk Vorberg at early stages of the project are gratefully acknowledged. We thank Petra Gruettner, Andrea Kliewe, and Antje Fuchs for their invaluable help with data collection. Correspondence concerning this article should be addressed to Ralf Th. Krampe, who is now at the Department of Psychology, Katholieke Uni- versiteit Leuven, Tiensestraat 102, B-3000, Leuven, Belgium. E-mail: ralf.krampe@psy.kuleuven.ac.be Journal of Experimental Psychology: Copyright 2005 by the American Psychological Association Human Perception and Performance 2005, Vol. 31, No. 3, 379 –397 0096-1523/05/$12.00 DOI: 10.1037/0096-1523.31.3.379 379