Methods of Morpheme Quantification: Their Effect on the Interpretation of Second Language Data

Abstract

Production of English grammatical morphemes has provided a fruitful data base for language acquisition research. However, questions have been raised as to how these data can best be analyzed to represent subjects' actual use of morphemes (Andersen 1977, Hakuta 1976, Hatch and Wagner-Gough 1976, Lightbown, Spada, and Wallace 1980, Stauble 1981). In the following study, application of different methods of morpheme quantification to the spontaneous speech data of 18 adults acquiring English as a second language led to conflicting interpretations regarding their morpheme production.Application of two alternative methods of morpheme quantification—one by suppliance in obligatory contexts, the other by target-like use—resulted in substantially different percentages of accuracy for subjects' production of the morphemes progressive -ing, progressive auxiliary, and past irregular. Another set of methods, used to analyze plural -s—quantification by noun tokens and by noun types—yielded contradictory findings regarding subjects' production of this morpheme.Results of this study demonstrate that data can be described in various ways depending on how morphemes are quantified. As a consequence of applying different analyses to the same data base, conflicting interpretations about subjects' interlanguage can occur.