Preview

PDF, English Download (8MB) | Terms of use

Abstract

Utterances are minimal ostensive stimuli produced by a speaker for an interlocutor, who interprets them by decoding linguistic input and by carrying out inferential processes (Sperber & Wilson 1995[1986]). Inferencing implies performing computations to obtain and connect mental representations with each other and with the context and obtain cognitive effects from the processed utterances (idem). These inferential processes are often guided by linguistic expressions with procedural meaning, among which are discourse markers. Discourse markers act as instructions during discourse processing by constraining contextual access (Blakemore 1987, 2002; Portolés 2001[1998]; Loureda & Acín 2010). As procedural-meaning expressions, their semantics is rigid and asymmetric as to concepts: instructions encoded in the meaning of a discourse marker must necessarily be executed (Leonetti & Escandell Vidal 2004; Escandell Vidal et al. 2011). In this dissertation, we investigate experimentally how concepts and instructions interact in discourse processing, and how such interaction is managed by speakers with different degrees of competence in an L2, as compared to native speakers. Processing data were gathered in an eye-tracking reading experiment for four discourse-related phenomena and three groups of readers. The two participant groups consisted of speakers an intermediate level (B1 CEFR, n = 58) and with a proficiency level (C1 CEFR, n = 49) in Spanish as an L2; the control group consisted of native speakers of Spanish (n = 102). At the discourse level, we compared processing of different argumentative discourse relations (causality versus counter-argumentation signaled by the Spanish connectives por tanto ‘therefore’ and sin embargo ‘however’, study 1); how the presence of a procedural interpretive guide influences processing of causal relations (implicit versus explicit causality marked by por tanto, study 2); and how congruency between procedural meaning and mind-stored assumptions impacts discourse processing (plausible versus implausible causality marked by por tanto, study 3, and plausible versus implausible causality marked by sin embargo, study 4). In general, results show that discourse relations are approached differently in cognitive terms depending on an individual’s degree of linguistic and pragmatic competence. Most frequently, the patterns obtained point to a direct correlation between proficiency and degree of nativelikeness in L2 performance, both in the strategies deployed, and in the effort allocated in processing of causality and counter-argumentation and in the resolution of pragmatic mismatches. Specifically, feasibility and relevance in discourse overrides discursive differences (the type of discourse relation at issue) from a certain degree of communicative competence on. By contrast, when pragmatic and linguistic competence are not sufficiently developed, relevance and discourse feasibility do not seem to offset the higher cognitive complexity of a certain discourse relation (study 1) and the absence of processing instructions (study 2). Communicative competence is also determinant of whether and how the accommodation strategies needed to process utterances in which procedural meaning leads toward recovering of a communicated assumption that clashes with mind-stored assumptions are performed. Accommodation is cognitively demanding and, therefore, effortful, but only given a certain degree of communicative competence to perform a certain task (studies 3 and 4). In complex and highly complex tasks processing by less proficient language users is shallow (cf. Clahsen & Felser 2006a, 2006b, 2006c) compared to more proficient and native readers. As a result, when the cognitive constraints imposed by the task are very high (study 4), readers fail to carry out the accommodation processes needed to recover the assumption communicated in the utterance. From a theoretical perspective, this study may contribute to the refinement of theories on L2 discourse processing, particularly in respect to how non-native language users cognitively manage discourse marking; from an applied perspective, the experimental evidence provided may serve as a basis for future studies to determine if empirically observed processing strategies in an L2 correlate with the thresholds and the content-sequencing established in frameworks of reference for the teaching and learning of second languages in relation to discourse marking and, in general, to contents at the discourse level for any language skill.