(A-)W{O,A}K(-EN)(-EN)(-ED) (UP)

Dennis Preston

Sat. 9:00-10:40 C

This paper throws down the gauntlet, challenging others to cite a more morphologically variable element in English than wake. The facts are these:

1) It may or may not be a- prefixed (awake).

2) Its infinitive form may be wake or waken (the -en here from an Old English augmentative).

3) Its preterit may be formed by stem-changed -o- for -a-, suffixed -ed, or both (woke, waked, awokened, etc...).

4) Its participial (passive or perfective) form may be stem-changed o, -en, -ed, or various combinations (e.g., woke, awoken, awakened, etc...).

5) It may be followed by the particle up.

Although some of the interdependencies of this multiplicity of forms may be open to intuition (e.g., a- prefixing and up seem incompatible), other preferences cannot so easily be intuited. Additionally, preferences for one form or another in various grammatical environments (e.g., transitive versus intransitive) do not seem to be open to intuition at all. Finally, although some status or other demographic constraints seem clear (e.g., wokened appears to be nonstandard) others (e.g., sex, region, style) are not clear at all.

A questionnaire given to over 100 respondents form southeastern Michigan provided a 'story' format which elicited various forms of wake in a variety of grammatical environments. The respondents were divided by ethnicity (European- or African-American), age, sex, and locality (rural versus urban).

Since a number of forms were used only infrequently, a VARBRUL format was used for data analysis. In general, the various forms of wake outlined above were tested for their likelihood according to demographic identity, to their interdependencies on one another, and to their relationship to other (linguistic) environments (e.g., tense-aspect, voice).

The results are compared to claims in earlier and more recent dialect surveys of American English (e.g., LAGS, DARE, LAUM) and are carefully examined for the emerging (dominant) patterns of current, local use. In general, southeastern Michigan tendencies show both an unusual likelihood for mixed ('strong' and 'weak') forms of inflection in the verbal paradigm and a mood-sensitive (i.e., passive) preference for '-en.' This exercise shows the value of careful quantitative analysis in changing morpho-syntactic patterning.