Journal of Experimental Psychology: Human Perception and Performance 1999, Vol. 25, No. 6,1793-1812 Copyright 1999 by the American Psychological Association, Inc. 0096-I523/99/S3.00 Strategy Differences in Oscillatory Tracking: Stimulus-Hand Versus Stimulus-Manipulandum Coupling John F. Stins and Claire F. Michaels Vrije Universiteit, Amsterdam In 3 experiments, participants matched the rotations of a unimanually grasped wheel to a visual oscillation. Two coordination modes were studied: in-phase coordination (no phase difference between stimulus and movement) and anti-phase coordination (180 degrees phase difference). The hand grasped the wheel at either the 12:00 or the 6:00 position. Stimulus frequency, hand placement, phasing, and visibility (whether the hand and wheel were visible) all affected movement amplitude and stability. There were large individual differences especially at the 6:00 position; some participants appeared to couple movements of the wheel to stimulus oscillations, some coupled movements of the hand, and some did both. The results parallel stimulus-response compatibility effects in a similar choice reaction time task and reiterate J. A. S. Kelso's (1995) emphasis on studying intrinsic coupling dynamics at the level of the individual, where apparent differences in strategy can be observed. Students of human interlimb coordination have repeatedly found that certain rhythmic coordination patterns between two limbs can be maintained more stably than other patterns. A variable that captures the order of the coordination in many examples involving rhythmic movements is the rela- tive phase (4>) between the interacting components. In general, patterns produced at 0-rad phase difference between two oscillators (defined as in-phase coordination) have greater stability than movements involving u-rad phase difference (defined as anti-phase coordination). For ex- ample, Kelso (1984) asked participants to oscillate their left and right index fingers; he observed spontaneous transitions from the anti-phase movement pattern (simultaneous activa- tion of nonhomologous muscle groups) to the in-phase movement pattern (simultaneous activation of homologous muscle groups) but not the other way around as the frequency was gradually increased. This transition is as- sumed to result from the loss of stability of the anti-phase movement pattern at a critical frequency, after which the in-phase pattern is the only stable solution of the system (see also the model by Haken, Kelso, & Bunz, 1985; the HKB model). The patterns of differential stability obtained with in- phase and anti-phase coordination are not limited to inter- limb coordination but also apply to cases involving between- person coordination (Schmidt, Carello, & Turvey, 1990), coordination of finger movements with an auditory metro- John F. Stins and Claire F. Michaels, Faculty of Human Movement Sciences, Vrije Universiteit, Amsterdam, the Netherlands. The Faculty of Human Movement Sciences at the Vrije Universi- teit, Amsterdam, participates in the Institute for Fundamental and Clinical Human Movement Sciences. Correspondence concerning this article should be addressed to Claire F. Michaels, Faculty of Human Movement Sciences, Vrije Universiteit, van der Boechorststraat 9, 1081 BT Amsterdam, the Netherlands. Electronic mail may be sent to c_f_michaels@ fbw.vu.nl. nome (Carson, 1996; Kelso, DelColle, & Schoner, 1990), and coordination involving manual tracking of a rhythmi- cally moving visual stimulus (Wimmers, Beek, & van Wieringen, 1992). Note that what constitutes an in-phase or anti-phase movement pattern can sometimes be determined only a posteriori by measuring the system's stability under different phasing relationships. This became apparent in an experiment performed by Kelso et al. (1990), where it was found that when participants established a certain phasing relation between a rhythmic auditory stimulus and an effector, the terms in-phase and anti-phase meant different things on different occasions. In that experiment, partici- pants had to coordinate a finger flexion with the pulse of an auditory metronome in one of two modes. They either had to synchronize (flexion "on the beat") or syncopate (flexion "off the beat") their finger movements with the stimulus train. These modes were labeled in-phase and anti-phase coordination, respectively. When the driving frequency was gradually increased (from 1.0 Hz to 3.5 Hz), a transition from syncopation to synchronization was often observed, which paralleled the Kelso (1984) findings. However, 1 participant was able to maintain the anti-phase coordination mode at even the highest frequencies. A postexperimental interview revealed that this participant had apparently discovered a special strategy for performing the movements. Instead of producing finger flexion between two beats (anti-phase coordination), this participant performed finger extensions on the metronome beat. In other words, this participant's strategy apparently transformed a movement pattern involving anti-phase finger flexion into a pattern involving in-phase finger extension, which presumably was easier to maintain. This observation led Kelso et al. (1990) to address the issue of how the relative phase between two oscillators should be defined. They argued, in short, that the coordination strategy determined the meaning of the relative phase and hence the dynamics of the pattern (see also Kelso, 1994). In this article we further investigate differences in coordi- 1793