A heuristic approach to extending the cognitive classification of English discourse markers

Abstract

The article contains the results of developing a heuristic approach to extend the cognitive classification of discourse markers by means of identification of discourse markers with a priori indeterminate form. The approach is based on contextual localization of discourse markers by their position in the utterance, peculiar features of accentuation in speech and punctuation in written text, part-of-speech attribution, and lexical co-occurence in discourse. The author argues that these specific contextual features of discourse markers are determined by their cognitive-functional potential and therefore open the possibility of adequately establishing their discursive status with a significant degree of correctness. The proposed approach is verified on the basis of authentic English corpus material contained in Open American National Corpus. Automated natural language processing was used as the key method of analysing the corpus material. The use of this method allowed the author to extend the cognitive classification of discourse markers to 33 more speech units actively functioning in contemporary spoken American English discourse