Abstract

Writers, editors, and everyday language users look to dictionaries, style guides, usage guides, and other published works to help inform their language decisions. They want to know what is Standard English and what is not. Commentators have been prescribing and proscribing certain usages for centuries; however, their advice has traditionally been based on the subjective opinions of the authors. Recent works have analyzed usage by relying wholly or partly on statistical and descriptive data rather than traditional opinion alone; however, no work has presented statistical usage data in a user-friendly and consistent format. This study presents a statistically based methodology for analyzing the standardness of disputed English usage points that can be presented in a dictionary-like format useful to writers and editors. Using data from the Corpus of Contemporary American English, this study determined the percent of use of several disputed usage items. Percents of use were then applied to a statistically based "standardness" scale with several levels. The scale presented in this study is adapted from scales that have been used previously to study language change. In addition, returns from the Corpus of Historical American English were used to present historical trends, if any, for each usage item. It was found that traditional sentiments about certain prescribed and proscribed usage items differ markedly from actual observed usage. Corpus data make it clear that even usage guides that purport to rely at least partly on descriptive data are often wrong about the prevalence and acceptability of usage items. To produce truly objective and accurate analysis, usage advice must depend on corpus data and use a standard usage-trend scale that accounts for how language changes.

Degree

College and Department

Humanities; Linguistics and English Language

Rights

http://lib.byu.edu/about/copyright/