Tian, Y; (2014) Negation Processing: a Dynamic Pragmatic Account. Doctoral thesis , UCL (University College London).

Preview

PDF thesis YTian 2014 UCL final.pdf Available under License : See the attached licence file. Download (2MB)

Abstract

This thesis investigates the processing of negative assertions. Psycholinguistic research shows that out-of-context negative sentences are more difficult to process than positive sentences. In the early stages of negation processing, the positive counterpart is often represented. Pragmatic research shows that negative sentences have richer pragmatic functions than positive sentences. These findings require a theory of negative sentence processing that can account for both the processing effects and pragmatic functions. Among current theories, a popular approach – rejection approach – attributes the processing effects to the processing of the linguistically coded meaning of negative sentences. They propose that negative sentences are represented as the rejection of their positive counterparts. They state that the representation of the positive counterpart is a mandatory first step of negation processing, and explain the processing cost in terms of the extra step of embedding. Arguing against current theories (especially rejection accounts), I propose the dynamic pragmatic account. In general, sentence processing – with or without explicit context- should not only involve processing the linguistically coded content, but also involve inferring pragmatically retrieved content such as how the sentence relates to the broader discourse. Specifically, when we interpret an assertion, we not only process the asserted meaning, but also the Question Under Discussion (QUD) addressed by this assertion, which can be retrieved and accommodated using linguistic and non-linguistic cues. Negation is a cue for retrieving the prominent QUD. Without contextual support or further cues, the most prominent QUD for a negative sentence ¬p is the positive question whether p. The projection of this positive QUD is due to the most frequent uses of negation, and is sensitive to other factors (e.g. frequency of the predicate and context) and other QUD cues (e.g. prosodic focus and cleft construction). I propose that the accommodation of a positive QUD contributes to the processing cost of negation, explains why the positive counterparts are often represented, and accounts for the pragmatic effects of negative sentences. The dynamic pragmatic account and competing theories are tested in three series of experiments in Chapters 3-5. In Chapter 3, I show that the representation of the positive counterpart is not a mandatory first step for negation processing. Rather it is likely due to QUD accommodation. When a negative sentence projects a negative prominent QUD (such as a cleft negative sentence “It is John who hasn’t ironed his shirt”), the positive counterpart is no longer represented. In Chapter 4, I investigate the verification of negative sentences against pictures. Previous studies have reported inconsistent results where verifying true negative sentences can take less, equal amount or more time than verifying false negatives. I argue that two strategies can be used in the task: the default strategy and the truth-functional strategy. The default strategy is to infer and represent the situation that makes the sentence true and compare it with the evidence. In addition, the accommodation of the positive QUD may encourage the development of a truth-functional strategy, in which participants answer the positive QUD and then switch the truth index. I show that when negative sentences project positive QUDs, there is a training effect: the reaction time pattern of true and false negatives change over time, indicating a development of a task-specific strategy; on the other hand, when negative sentences project negative QUDs, participants no longer develop the task-specific strategy. In Chapter 5, I investigate the time course of negative sentence processing in a visual world eye-tracking study. The results show that processing simple negative sentences is delayed compared to processing simple positives, but processing cleft negatives is no more delayed than processing cleft positives. Importantly, both QUD accommodation and the integration of the meaning of negation can happen incrementally. Overall, the findings speak against current models of negation processing (especially rejection accounts), and support the dynamic pragmatic account.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item