Resumen de On the interpretation of negation in mandarin chinese
Abstract This dissertation aims to experimentally investigate the interpretation of negation in Mandarin Chinese (MC), namely, when multiple negative expressions combine in a sentence, when negative expressions are used as fragment answers to negative questions, and when native speakers express rejection to a negative assertion or a negative polar question.
It first examines whether a single negation (SN) reading may be possible under certain conditions, despite the fact that MC has been characterized as a language in which two negative expressions within the boundaries of a single sentential domain cancel each other to yield a positive reading. To test this hypothesis, an online perception experiment was conducted with native MC speakers. The results showed that SN readings were indeed obtained, particularly when the first of the two negative expressions was an adjunct (i.e., cóngláibù/cóngláiméi(yǒu) ‘never’) or there was stress on the second negative expression (i.e., the negative markers méi(yǒu) ‘not’ and bù ‘not’).
Next, this dissertation explores the mismatches in the interpretation of MC argumental negative expressions (namely, méi(yǒu)rén ‘no one’ and méi(yǒu)shénme ‘nothing’) when they are used as fragment answers to negative wh-questions. The results of our production experiment showed that the acoustic correlates that characterize these fragment answers are identified when they convey not only double negation (DN) but also SN meanings. The results of the production experiment support the conclusion that in MC DN is prosodically marked and that DN and SN readings display significantly different acoustic correlates in speech production. More specifically, DN readings show shorter duration, more pitch variation, higher maximum pitch, and larger rising pitch excursion. The results of our audio perception experiment further showed that native speakers of MC perceive these prosodic correlates and reliably use them to distinguish between DN and SN readings of argumental negative expressions used as fragment answers.
Finally, this dissertation addresses the central question of whether MC is a canonical truth-based language, a language in which the speaker is expected to express agreement to a negative proposition by means of a positive particle followed by a negative sentence and disagreement using a negative particle followed by a positive sentence. A production experiment was conducted, in which native speakers participated in an oral Discourse Completion Task that elicited confirming/rejecting responses to negative assertions/questions and broad focus statements (control condition). The results showed that MC speakers convey confirmation/rejection by relying on a combination of lexico-syntactic strategies (e.g., negative particles such as bù/méi(yǒu) + positive sentences; positive particles such as shì(de) + negative sentences) together with prosodic (e.g., mean pitch) and gestural strategies (mainly, the use of head nods). Importantly, the use of a positive or a negative particle, which was the expected outcome in truth-based languages, only appeared in 82% of the confirming answers and in 52% of the rejecting answers, respectively. Our results bring into question the macroparametric division between truth-based and polarity-based languages and calls for a more general view of the instantiation of a CONFIRM/REJECT speech act that integrates lexical and syntactic strategies with prosodic and gestural strategies.
Consequently, this dissertation provides a new understanding of the interpretation of negation in MC as a so-called DN language and as a so-called truth-based language.
