An intermodal approach to the complementizer that in constrained and unconstrained language

Abstract

Abstract This study investigates the factors conditioning the that/zero alternation in four varieties of political discourse: translated and interpreted English as constrained varieties, and original written and spoken English as unconstrained varieties. Distinctive collexeme analysis and multivariate analysis were employed. The results show that constrained varieties use the complementizer that significantly more often than unconstrained varieties, with a significantly higher rate in translated than interpreted English. Distinctive collexeme analysis reveals subtle yet systematic differences in the matrix verbs associated with that-complement clauses, suggesting that distinct semantic and constructional profiles contribute to variation in the use of that. Multivariate analysis shows that all four varieties are largely conditioned by the same conventionality- and complexity-related factors, although production mode interacts with specific linguistic constraints. Translated English favors that even in contexts where omission is favored, a pattern consistent with pragmatic risk avoidance. Its broader preference for that may also reflect a tendency toward analyticity associated with cognitively demanding bilingual activation. In interpreted English, however, explicit marking is more plausibly attributable to analyticity than to heightened risk avoidance. Taken together, these findings provide intermodal empirical evidence of increased explicitness in constrained language based on Chinese–English language pair, extending previous findings derived primarily from European language pairs.