Chapter 5. Examining associations between lexis and textual position in hard news stories, or <i>according to a study by&#8230;</i>

MyBook is a cheap paperback edition of the original book and will be sold at uniform, low price.
This Chapter is currently unavailable for purchase.

It is claimed within the theory of lexical priming that text-linguistic phenomena are as deeply involved in our lexical primings as are more obvious phenomena such as collocations. One of the textual claims is that people are primed to associate certain words or phrases with certain recognised discourse positions in particular genres and/or domains (an association referred to as textual colligation). One kind of textual colligation is hypothesised to be the psychological association that a language user makes in a genre between a word or phrase and text-initial position. This paper seeks to demonstrate that it is possible to use the Key Words procedure implemented in WordSmith Tools not only to identify over 1000 words with strong text-initial associations in hard news data but also to examine the way in which phrases or clusters of words also may have textual priming. In particular we focus on one three word lexical cluster which suggests how lexis and textual position interact. We conclude that our findings are fully compatible with Lexical Priming theory, even though they cannot directly be used as evidence for it. The paper also makes use of a number of novel or under-used methodological strategies, including the utilisation of complete texts as a unit of analysis, the creation of positional subcorpora, the use of key words to identify candidates for textual colligation, the comparison of an item&#8217;s behaviour in a subcorpus with its behavior in other subcorpora derived from the same texts (following Gledhill, 2000), the extension of key word procedure to collocate tables (key collocates) and the investigation of textual semantic association and textual collocation using concordance line grouping.


This is a required field
Please enter a valid email address