Logical stress and gesture synonymy in the cephalic channel

In this article, based on the material of the Russian-language reference subcorpus RUPEX and the Spanish-language monologue subcorpus CAFE (comunicacin de los artistas flamencos espaoles) annotated in ELAN, we analyze head gestures that mark logical stress and consider all the variations that occur in this position.When determining the position of coincidence of a gesture and stress, we took into account the principle described by Grishina about the anticipation or delay of gestures depending on the strategy chosen by the speaker.In 20% of cases, the gesture anticipates the logical stress, starting a little earlier, and in 13%, on the contrary, it occurs immediately after the stressed syllable, highlighted by the accent of the word.According to our research, some of these head gestures are synonymous with each other in other positions as well.Some gestures were chosen by the subjects according to the characteristic features of their cephalic portrait (for example, moving the head forward with a sideways tilt), while others turned out to be typical that highlight significant words in a monologue and are characteristic of both Russian and Spanish (for example, the pragmatic gesture Down).When compared with the data of the MUMIN corpus research group and the authors of the Spanish-language corpus collected from "spontaneous" speeches on talent shows, principles were developed for describing gestures in the cephalic channel in marked positions from the point of view of different functional approaches, taking into account the following factors: the influence of other kinetic channels, the presence/absence of a listener, and the superposition of gesture functions on top of each other.In the Russian language, in 73% of cases, the opposite movement is adjacent to or behind the gesture, which indicates the visualization of "emphatic tone curvature" or different types of "skid".In the remaining cases, an intensification of the gesture is observed by anticipating it with the same one, but of smaller amplitude.Testing the annotation principles on two corpora showed their effectiveness as a basis for developing automatic head gesture annotation.