Which Forces The Addition Of A New Matching Process Type To The Grammar

By Author: Alex Lee
Total Articles: 366
Comment this article

Christian Matthiessen presents ongoing work around probability profiles of key lexico-grammatical systems. Through a diagram of 20 major clause systems (e.g. declarative, marked theme, and mental process), he relates profiles for two individual texts (a spoken interview and a written news report), to corpora of spoken interviews and news reports, to spoken and written language, and to his larger, still-growing archive. Thus, profiles of corpora (created by systematic sampling) describe registers (e.g. 'the language of news reports'), which shed light on the nature of individual texts and on the nature of language itself.Thomas Sabo

Pointing to current limits of corpus analysis, Matthiessen argues that 'the more significant a system is in the organisation of the lexicogrammar..., the harder it will be to automate the analysis' (p. 141). He then illustrates work in progress, based on manual analysis of 6,500 clauses. Simple frequencies of clause complex types can be probed when recursion and nesting are included to reveal differences by level between hypotaxis and parataxis ...
... not reflected in logico-semantic type. Examples show further intersections between pairs of systems. Thus, material processes and effective agency attract each other; different process types are attractive to certain circumstances and can be ranked. Subsequent examples explore intersections between pairs of mood, process type, and polarity. This suggests that systems from the same metafunction interact differently (variable ranking) than those from different metafunctions. Matthiessen then explores implications for syntagmatic and paradigmatic dimensions. For example, material processes are not only most frequent, but also account for the highest number of verbs and verb classes, and have the highest share of circumstances. Thomas Sabo Charms

This work has enormous potential for descriptions of English, SFL theory, and methodology, although the quality of some of the 22 figures suggests that greater attention is needed to presenting results in publications. In contrast to Matthiessen, Amy Neale's analysis starts with the lexis, specifically individual verb senses, which forces the addition of a new ‘matching' process type to the grammar? Her clearly exemplified paper shows how secondary corpora can reduce the need for manual analysis of large amounts of text, to inform functional description and theory development. Building on evidence from West's manual frequency analysis of verb senses, COBUILD dictionary entries and Verb Pattern frequencies, she explains the development of her living Process Type Database (PTDB). This most useful resource includes 5,400 Processes with associated Participant Roles.