The explanatory model of cross-cultural miscommunication, or crosstalk, is extended here through a multi-feature, multi-dimensional analysis of Soviet and American speakers' discourse in two three-hour audio/video “spacebridge” meetings. The study demonstrates that variation between speakers' uses of co-occurring sets of lexical and syntactic features can contribute to crosstalk. Crosstalk is shown to be functionally motivated by interlocutors' different constructions of the speech-event context and norms of interpretation.

A factor analysis of lexical and syntactic features in spacebridge participants' discourse identifies two main dimensions of stylistic variation; use of the sets of co-occurring features that constitute these dimensions are interpreted as performing different discourse functions, and thereby indexing different aspects of the speech-event context. Soviet floor turns that exhibit the greatest stylistic divergence from American stylistic behavior and expectations are shown to correlate with communicative breakdowns in the spacebridges, and thereby to contribute to crosstalk. The analysis suggests that, at some level, stylistic contextualization cues are quantitatively “analyzed” in real time by discourse participants, and it demonstrates some of the explanatory potential of quantitative modeling of complex indexical processes such as stylistic accommodation, divergence, and opposition. I wish to thank Doug Biber, Elinor Ochs, and Barbara Johnstone for invaluable feedback on this study.