La réinterprétation des prédicats statifs dans les supplétives en DO SO

Abstract

The VP-substitute DO SO is usually said to corefer with a dynamic predicate. Corpus studies have shown that this is apparently not always the case, that DO SO can also have ‘stative antecedents’.This paper examines DO SO with purely stative antecedents, here defined as the sum of four negative features: [− process, − culmination, − control, − perform]. It is here shown that DO SO may have a stative antecedent (unlike do it/this/that/the same), but only if the do so clause or VP induces a different (and more dynamic) representation of the ‘event’ represented in the antecedent clause or VP. For DO SO to be an acceptable option, this ‘reinterpretation’ must imply the transformation of at least one of the negative features into its positive equivalent. The two features most often implied are [+/− process] and [+/− perform].Four syntactic configurations are taken into account. In the first two (coordination of two clauses + do so in a subordinate clause), DO SO has its own grammatical NP-subject that can have its own distinct semantic (or thematic) role. The third is a coordination of two VPs and so with only one grammatical NP-subject (factorized for the two VPs); and in the fourth, do so’s antecedent VP is embedded within do so’s own NP-subject. Seeing that the stative antecedent VP has to be reinterpreted for DO SO to be an acceptable option, this brings me to hypothesize that in these structures, DO SO must have a covert semantic subject with its own distinct semantic role.

Keywords

• do so
• lexical/auxiliary verb
• stative/dynamic predicates
• VP-substitute
• thematic roles
• reinterpretation