The Acquisition of Tense–Aspect: Converging Evidence From Corpora and Telicity Ratings STEFANIE WULFF University of North Texas Department of Linguistics and Technical Communication P.O. Box 311 307 Denton, Texas 76203-2171 Email: Stefanie.Wulff@unt.edu KATHLEEN BARDOVI–HARLIG Indiana University Department of Second Language Studies 1021 East Third Street Bloomington, IN 47405 Email: bardovi@indiana.edu NICK C. ELLIS University of Michigan English Language Institute 500 East Washington Street Ann Arbor, MI 48104 Email: ncellis@umich.edu CHELSEA J. LEBLANC University of Michigan Department of Computer Science 1200 Fairground Plymouth, MI 48170 Email: leblancc@umich.edu UTE R ¨ OMER University of Michigan English Language Institute 500 East Washington Street Ann Arbor, MI 48104 Email: uroemer@umich.edu The aspect hypothesis (Andersen & Shirai, 1994) proposes that language learners are initially influenced by the inherent semantic aspect in the acquisition of tense and aspect (TA) morphol- ogy. Perfective past emerges earlier with accomplishments and achievements and progressive with activities. Although this hypothesis has been extensively studied, there have been no ana- lyses of the frequency, form, and function of relevant types and tokens in the input. This article reports the results of 2 corpus-based studies investigating how various features of the input— frequency distributions, reliabilities of form–function mapping, and prototypicality of lexical aspect—affect TA morphology. Study I determined the relative frequency profiles of exemplars of English TA and employed various statistics to determine the associations between particular verb–aspect combinations. Study II expanded the aspect hypothesis, examining whether native speakers judge the most frequent forms in isolation to be more prototypical in lexical aspect. Analyses were then matched against acquisition data for different TA patterns by adult learners of English (Bardovi-Harlig, 2000) to determine whether the verbs’ acquisition order is deter- mined by their frequency, form, and function in the input. Rather than testifying to the effect of 1 factor alone, the results suggest that frequency, distinctiveness, and prototypicality jointly drive acquisition. THIS ARTICLE EXPLORES THE ACQUISITION of tense–aspect (henceforth TA) morphology from a constructionist perspective (e.g., Bates & MacWhinney, 1987; Croft, 2001; Goldberg, 2003, 2006; Ninio, 2006; Robinson & Ellis, 2008; The Modern Language Journal, 93, iii, (2009) 0026-7902/09/354–369 $1.50/0 C 2009 The Modern Language Journal Tomasello, 2003). The basic tenets are as follows: Language is intrinsically symbolic. It is consti- tuted by a structured inventory of constructions as conventionalized form–meaning pairings used for communicative purposes. Constructions are of different levels of complexity and abstraction; they can comprise concrete and particular items (as in words and idioms), more abstract classes of items (as in word classes and abstract gram- matical constructions), or complex combinations