AbstractThis thesis explores neologisms in two distinct but related contexts: dictionaries and newspapers. Both present neologisms to the world, the former through information and elucidation of meaning, the latter through exemplification of real-world use and behaviour.
The thesis first explores the representation of new words in a range of different dictionary types and formats, comparing entries from collaborative dictionary Wiktionary with those in expert-produced dictionaries, both those categorised here as ‘corpus-based’ and those termed ‘corpus-informed’. The former represent the most current of the expert-produced dictionary models, drawing on corpora for almost all of the data they include in an entry, while the latter draw on a mixture of old-style citations and Reading Programmes for much of their data, although this is supplemented with corpus information in some areas.
The purpose of this part of the study was to compare degrees of comprehensiveness between the expert and collaborative dictionaries as demonstrated by the level and quality of detail included in new-word entries and in the dictionaries’ responsiveness to new words. This is done by comparing the number and quality of components that appear in a dictionary entry, both the standardised elements found in all of the dictionary types, such as the ‘headword’ at the top of the entry, to the non-standardised elements such as Discussion Forums found almost exclusively in Wiktionary.
Wiktionary is found to provide more detailed entries on new words than the expert dictionaries, and to be generally more flexible, responding more quickly and effectively to neologisms. This is due in no small part to the way in which every time an entry or discussion is saved, the entire site updates, something which occurs for expert-produced online dictionaries once a quarter at best.
The thesis further explores the way in which the same neologisms are used in four UK national newspapers across the course of their neologic life-cycle. In order to do this, a new methodology is devised for the collection of web-based data for context-rich, genre-specific corpus studies. This produced highly detailed, contextualised data that not only showed how certain newspapers are more likely to use less-well established neologisms (the Independent), while others have an overall stronger record of neologism usage across the 14 years of the study (The Guardian).
As well as generating findings on the use and behaviour of neologisms in these newspapers, the manual methodology devised here is compared with a similar automated system, to assess which approach is more appropriate for use in this kind of context-rich database/corpus. The ability to accurately date each article in the study, using information which only the manual methods could accurately access, coupled with the more targeted approach it can offer by excluding unwanted texts from the outset made it the more appropriate approach.
|Date of Award||2017|
|Supervisor||Hilary Nesi (Supervisor)|