A corpus-based functional study of shi…de constructions

Existing research on shi…de constructions has often referred to its purported “emphasis” or “focus” functions. This paper reexamines shi…de by analyzing 787 examples of shi…de extracted from a spoken corpus. Positive evidence was found for several earlier proposals that rely on intuitively-generated data. Meanwhile, additional features of the construction can be observed. Results indicate that the preferred form of shi…de takes stative predicates, and some examples occur in such high frequencies that they form formulaic expressions. The construction always achieves stative predication, is often associated with subjectivity, and expresses the speaker’s certainty in stancetaking. To explain all the data, a unified “emphasis” function is proposed to integrate the traditional analyses of “constructive focus” and “affirmation.” Also, the overlap between the copula and the emphasis/focus function of shi in shi…de suggests that the construction is a form grammaticalized from shi + nominalization. This paper thus shows that corpus data can enable us to tackle an old issue with new evidence by making all subtypes of the construction available for quantitative and qualitative analysis, which in turn helps us redefine and reconceptualize otherwise ambiguous notions.