Three computational models that track patterns in semantic change over time are described. The models were used to investigate whether certain word properties (e.g., frequency of use, accessibility) and the relationships between them would be sufficient in themselves to produce some patterns of change. (15 references) (MSE)

Historic linguistic consequences were studied in five separate areas of Atlantic Canada's Acadian population. Focus was on certain verbs in the third person present plural and the hypothesis that "allent" and faisent" are the result of hypercorrection. Results suggest complex reasons for their use and an indication of some social…

Acoustic analysis of two vowels thought to be merged in Utah English suggest that there are small but consistent differences between them. A matched guise experiment provides evidence that when the vowels are merged hypercorrection is involved. (33 references) (LB)

A study explored the ability of children to comprehend expressions of uncertainty in varying degrees (e.g., "not likely, possible, probable"). Subjects were 42 French students aged approximately 9, 12, and 15. Results, including age and gender differences, and implications for classroom communication are discussed. (MSE)

The overall role played by hypercorrection in the literature on language change, language variation, and second language acquisition is reviewed. The paper argues that hypercorrection is not a completely unified phenomenon, citing an empirical study showing that quantitative methods applied to qualitative hypercorrection necessarily involve…