More spoken or more translated? Exploring the known unknowns of simultaneous interpreting from a multidimensional analysis perspective

This article explores the features of simultaneous interpreting (SI) from a multidimensional (MD) analysis perspective (Biber 1988), drawing on a newly built comparable intermodal corpus, the LegCo+ corpus. The corpus incorporates Cantonese speeches that are both interpreted and translated into English, as SI and written translation (WT), respectively. Additionally, a third English corpus consisting of English native speeches (NS), without mediation, serves as a benchmark comparison. We aim to examine the extent of similarities and differences between SI, NS and WT in terms of the linguistic patterns they display. Our findings show that: (1) SI is a hybrid language mode, exhibiting features that lie between those of non-mediated spoken language and mediated written language; (2) in terms of its spoken nature, SI resembles NS in certain dimensions where typical features are associated with orality, suggesting a strong modality effect; and (3) in terms of its mediated status, SI demonstrates similarities with WT, despite their perceptibly distinct modalities, pointing to a potential mediation-specific effect. These empirical findings emphasize the necessity of understanding the multidimensionality inherent in interpreted language.