H. argues that two problems must be given high priority in the empirical analysis of face-to-face interaction: the problem of multimediality ('vox' and 'motus') and the problem of the differing semiotic statuses of the sign-systems involved.