Contextual dimensions of conversational implicature

This research adapts a methodology for the analysis of context for pragmatics research and explores the dimensions of contextual variation of texts associated through previous research with conversational implicature production. The study adapts the activity types analytical framework for the categorization and interpretation of linguistic contextual markers as part of the adaptation of the corpusbased multidimensional approach to register analysis to the study of aspects of pragmatic context. The methodology was used in this study to determine the pragmatic contextual characteristics of registers classified according to their expected adherence to Grice's cooperative principle (CP) and maxims of conversation. These registers were represented by sample texts of a grammatically annotated corpus of approximately 21 million words, 1094 texts, prepared for the study. Frequency counts for 37 linguistic markers of context were obtained through the use of a parser written in the Perl computer programming language, developed as part of the study. These frequency counts were then converted to normalized frequencies based on 1000 words of text. Factor analysis was done on the frequency counts following multidimensional analysis procedures. The factor analysis resulted in an eight-factor solution. These factors were interpreted as contextual dimensions which were defined through the analysis of the salient linguistic features (markers) loading on each factor. Functional interpretation was based on the activity types functions. All of the 37 features were found to be salient. Results describe the contextual characteristics of sample groups of text representing seven registers. Relations between the registers were used to verify the dimensional interpretations, as well as to examine their contextual relationships on the dimensional scales. Internal coherence of the registers and the power of the dimensions to determine register distinctions were measured. Results largely supported the accuracy of the method in determining pragmatic characteristics of the sample texts. Non-parametric statistical comparison was made on two groups of text based on their association with conversational implicature production—one corpus representing registers in which Gricean maxims may be observed, and one in which they may be violated. Results of the statistical comparisons suggest that the two groups show significant differences on several contextual dimensions determined through the study. Thus, the study identified functionality associated with constraints on utterance production associated with non-observance of Gricean maxims through empirical quantitative analysis.

Preview PDF FBMK 2007 16R.pdf Download (949kB) | Preview

Actions (login required)

View Item