Encoding time is universal to human cognition and experience, yet languages vary in tense-aspect conceptual categories and associated exponents. For instance, Chinese lacks grammaticized past tense whereas German does not differentiate between progressive and non-progressive aspect. Presumably, bilinguals and second language (L2) learners are required to undertake acquisition routes specific to their L1-L2 pairings. The influential Aspect Hypothesis (e.g., Andersen & Shirai, 1994) and ample research evidence in the past decades suggest otherwise, however. Learners tend to follow semantics-based prototypical associations between lexical aspect and tense-aspect categories in initial and subsequent stages of language development. The influence of L1 seems minimal. This raises interesting questions about the longstanding debate concerning language-specific and universal factors in accounting for learners’ interlanguage development over time. In this talk, we will examine recent acquisition data from the English present perfect and the English progressive as it relates to Imperfective Paradox — areas that are known to be problematic for many L2 learners. Findings suggest Cantonese-speaking proficient L2 English learners were more prone to L1 transfer than previously assumed, especially in certain tense-aspect domains involving nuanced L1-L2 contrasts. Implications to the Aspect Hypothesis, the role of L1 in second language instruction, and acquisition of the Japanese -teiru will be discussed.