An MEG/EEG study of tempo production

Yann Cojan, Line Garnero & Viviane Pouthas

LENA-CNRS, Paris, France

Although many studies have investigated rhythmic motor production, its neural basis involved in this task is largely unknown. In an MEG/EEG study, twelve right-handed participants (age = 25.3) underwent a rhythmic tapping task involving a synchronization phase (S) and a continuation phase (C). During the S phase participants had to synchronize their finger taps to 25 ms tones (15) delivered every 600 ms and after the cessation of tones they had to continue to tap at the same tempo (~20 taps). Both phases were performed with each hand (right hand (RH) and left hand (LH), the order was counterbalanced across participants).

Different behavioural indices were analysed: inter tap intervals (ITIs), synchronization error (SE), coefficient of variation (CV = std/ITI), motor and clock variance (applying Wing and Kristofferson model, 1973) and tap duration (TD). There was a clear effect of the condition (S vs C) on ITIs; they were significantly shorter in the C phase (582ms 8ms) than in the S phase (599ms 1ms), reflecting an acceleration of the tapping rate in the C phase. SE corresponded to an anticipation of the tap onset over the tone onset (-59 ms 10ms). Interestingly results showed a correlation between SE and acceleration in the C phase (r = .75, p<.05), reflecting the fact that the more participants anticipated the tap over the tone, the more they accelerated their tapping rate in the C phase. In addition, CV was lower in the S phase than in the C phase. Applying Wing and Kristofferson model, we showed a significant effect of laterality on clock variance (483ms 167 RH vs 632ms 232 LH), but no difference between hands on motor variance (181 55 RH vs 185 75 LH); this is in agreement with the multiple timer model (Ivry and Richardson, 2002) and suggest that timing processes are specific to each effector system. Furthermore, a strong effect of laterality on tap duration (right TD < left TD) was found. Propioceptive afference properties may explain this difference between dominant and non dominant hand.

We are currently analysing the relationships between performances and electrical brain activity in order to shed new light on mechanisms underlying motor timing.