Running head: POSTURE CONTROL AND MOTOR IMAGERY

26 Imagined movements are thought to simulate physical ones, with similar behavioral 27 constraints and neurophysiological activation patterns, and an inhibition mechanism that 28 suppresses movement execution. When upper body movements such as reaching with the arm 29 are made from an upright stance, lower body and trunk muscles are also activated to maintain 30 body posture. It is not clear to what extent parameters of imagined manual movements are 31 sensitive to the postural adjustments their execution would necessitate, nor whether such 32 postural responses are as effectively inhibited as the imagined movements themselves. We 33 asked healthy young participants to imagine reaching movements of the arm while in upright 34 stance, and measured their self-reported movement times and postural sway during imagined 35 movements. We manipulated mediolateral stance stability, and the direction of arm 36 movement (mediolateral or anteroposterior). Imagined arm movements were reportedly 37 slower when standing in a mediolaterally less stable stance, and the body swayed more when 38 arm movements were imagined in the direction of postural vulnerability. The results suggest 39 that the postural state of the whole body, not just the involved limbs, informs trajectory 40 planning during motor imagery, and that measurable adjustments to body posture accompany 41 imagined manual actions. It has been suggested that movement is suppressed during motor 42 imagery by a premotor inhibitory mechanism operating at brain stem or spinal level. Any 43 such inhibition must be incomplete because, for example, it does not eliminate autonomic 44 arousal. Our results suggest that it also does not effectively suppress postural adjustments 45 planned in support of imagined movements. 46 47 48 49 50 Running head: POSTURE CONTROL AND MOTOR IMAGERY 3 Introduction 51 Activities of daily living are frequently accompanied by thoughts about past, present 52 and future actions. Such imagined movements retain so many characteristics of their physical 53 counterparts, that they have been described as simulations of physical actions (Jeannerod 54 2006). They exhibit temporal scaling of movement duration to distance (e.g., Decety et al. 55 1989; Papaxanthis et al. 2002; Sirigu et al. 1996), the same speed-accuracy trade-off as 56 expressed in Fitts’ law (e.g., Decety and Jeannerod 1996; Stevens 2004), the same adherence 57 to biomechanical constraints (e.g., Frak et al. 2001; Johnson 2000), and the same pattern of 58 simulated effort (e.g., Cerritelli et al. 2000). Neurophysiological evidence suggests a unitary 59 mechanism for movement representation and execution (Bonnet et al. 1997; Clark et al. 6