The processing of pragmatic implicature is a critical topic in the study of second language cognition, and in daily communication, conversational implicature is conveyed indirectly via reading between the lines using pragmatic inferences in social interaction. Scalar implicature (SI) denotes that the weak terms in Horn scales (Horn, 1972) have literal and pragmatic interpretations, when a speaker uses a weak scalar word, it means the negation of a stronger term in the same scale. Based on Grice’s theory, the researchers developed two different approaches at the level of cognitive processing of scalar implicature. The Neo-Gricean (NG) accounts claimed that pragmatic interpretation is derived automatically (by default) when the trigger belongs to weak scalar words, and a literal meaning can be only obtained after the cancellation of pragmatic interpretation (Levinson, 2000). By contrast, the Post-Gricean accounts believed that context is the basis for the generation of pragmatic meaning, arguing that the appropriate comprehension of an utterance depends on the effort required for the relevance of context to derive pragmatic invited inference (Sperber & Wilson, 1986, 1995；Sperber & Wilson, 2002).
Focusing on the processing and the development of pragmatic comprehension ability of L2 Chinese scalar implicature, this dissertation aimed to explore the factors influencing the processing of L2 scalar implicature by varying the contextual conditions and cognitive resources in four semantic true-value judgment tasks and two self-paced reading tasks. The participants involved two L2 Chinese proficiency groups (i.e., intermediate-level group and advanced-level group) with English as L1 background, and a Chinese native speaker group served as control.

Adopting semantic true-value judgment tasks, the first part investigated the processing and pragmatic comprehension ability of L2 Chinese scalar implicature under the non-contextual conditions, including Experiment 1 and Experiment 2. In Experiment 1, the cognitive resources of the participants were not limited, whereas a response time of 1500ms was set to limit the cognitive resources of the participants in Experiment 2. By using the same true-value judgment task, the second part examined the processing and pragmatic comprehension ability of L2 Chinese scalar implicature under contextual conditions, involving Experiment 3 and Experiment 4. In Experiment 3, the cognitive resources of the participants were unrestricted. However, a response time of 3000ms was set to limit the cognitive resources of the participants in Experiment 4. The third part used self-paced reading tasks to explore the processing and pragmatic comprehension ability of Chinese scalar implicature in both neutral and non-neutral contexts, including Experiment 5 and Experiment 6. Among them, Experiment 5 was in a neutral context, while Experiment 6 was in a non-neutral context.

The results of the present study indicated that: First, the processing of L2 scalar implicature was influenced by an interaction of many factors, including the context, cognitive resources, and L2 proficiency, which all made contributions to the processing of pragmatic meaning, and there were individual differences in L2 processing of scalar implicature. Most importantly, the results showed that the context, cognitive resources,and L2 proficiency mutually affected L2 processing of scalar implicature. Specifically under the non-contextual condition, the processing patterns for both L2 groups were in the overall conformity to the relevance theory model, regardless of whether the cognitive resources were limited. However, under the contextual condition, with the
development of L2 proficiency, the processing of scalar implicature for advanced L2 learners was more consistent with the default model hypothesis in both self-paced reading tasks and the semantic truth-value judgment task without limitation of cognitive resources, while in the semantic truth-value judgment task with limited cognitive resources, the processing patterns of both L2 groups were in accordance with the relevance theory model. Furthermore, the results documented individual differences in L2 processing of scalar implicature, in that some L2 learners showed a processing pattern that was inconsistent with the overall processing pattern both in contextual conditions and in non-contextual conditions.

Second, the influence of L2 proficiency on the comprehension development of L2 scalar implicature was modulated by the context and cognitive resources. Although no significant difference was observed between intermediate and advanced L2 learners in the comprehension of L2 scalar implicature under the non-contextual condition when cognitive resources were restricted, higher comprehension ability of scalar implicature was found among advanced L2 learners both in semantic true-value judgment tasks and in self-paced reading tasks compared to that for intermediated L2 learners under the contextual conditions regardless of whether the cognitive resources were limited.

Third, the advantage in comprehension of scalar implicature for L2 learners was not observed when compared to that for Chinese native speakers in the present study. Although advanced L2 learners reached a native-like accuracy in semantic true-value judgment tasks, they responded slower than Chinese native speakers.

The findings of the present study provide insightful pedagogical implications for L2 scalar implicature instruction. To obtain the comprehension of pragmatic meaning,L2 learners need to acquire linguistic knowledge, such as lexical semantics, more cognitive resources from the context, as well as the induction of pragmatic inferences through contextual relevance. Therefore, in L2 scalar implicature instruction, activities such as the semantic truth-value judgment task, multi-media information input task, and action task can be utilized to enhance L2 learners’ attention to the pragmatic meaning information conveyed via underinformative words in the context, and promote L2 comprehension ability to generate pragmatic meaning.