Аннотация:This study explores the organization of multimodal systems, mediated by two types of hierarchical regulations: systemic regulations governing each mode (speech and gesture) and multimodal regulations operationalizing mode alignment. To identify these regulations, we examine the variability of the multimodal speech-gesture system, modulated by the cognitive factor of vague reference in expository dialogue. The data were collected in an experiment involving native Russian speakers explaining the differences between close synonyms. The paper contrasts the distribution of vague reference speech cues used to shape referents in names (placeholders) and point the way to the referent in predications (approximators), both aligned with functional gestures (deictic, representational, and pragmatic).The findings reveal several regulations that constrain the distribution and alignment of speech cues and gestures. First, the prevalence of approximators is observed, indicating that a higher input of predication expresses vague reference. Second, the prevalence of placeholdersin the Request communicative move was found, accounting for the Request’s role in both initiating a new act and completing the previous one by renaming the vague referent. Third, a more frequent use of pragmatic gestures with approximators was identified, suggesting thatwhile nominations primarily evoke iconic and indexical representations in gestures, predications are strongly linked with pragmatic manifestations. Finally, the study reveals that vague reference serves as a cognitive principle regulating the speech and gesture system ininteractional discourse.