An assessment of the existence of muscle synergies during load perturbations and intentional movements of the human arm.

A cross-correlation analysis was performed on EMG activities in elbow and shoulder flexors evoked by force perturbations acting in different directions on the forearm and during intentionally generated movements with the purpose of characterizing the temporal relationships between patterns of activation of different muscles. Qualitatively it was found that the shape of the cross-correlation function differed from one experimental condition to the next. A principal component analysis permitted a quantitative assessment of this point. In general it was found that two principal components could account for the data. Furthermore when the cross-correlograms are represented in principal component space, there was no clustering of the data points. Several possible definitions of 'muscle synergies' are discussed from the perspective of this finding. It is concluded that the most restrictive definition is incompatible with the data. Less restrictive definitions, while compatible with the data, do not lead to a simplification of the control problem.