Taking Fingerprints of Speech-And-Gesture Ensembles. Approaching Empirical Evidence of Intrapersonal Alignment in Multimodal Communication

摘要: Co-occurring speech and gestures of natural language dialogues composes into meaning units, that is, they jointly describe discourse referents. We start from the idea interlocutors tend to re-use this cross-modal information units if referent is referred again: co-occurring gesture are assumed “align into” bimodal ensembles (BMEs). further hypothesize due principles dialogical economy will exploit impact a BME’s shorten its linguistic part BME. If hypothesis right, we expect words in multimodal communication exhibit different frequency distribution written texts, whose known obey Zipf’s law. This tested for 24 direction-giving using two fits, rank complementary cumulative distribution. According first fit, can be confirmed, according second one, it has rejected. In addition, also propose way measure strength informational association.