In this article, we argue against the Representational Deficit Hypothesis, according to which second language (L2) speakers can never acquire functional categories or features that are absent in the first language (L1), suggesting that fossilization is inevitable. Instead, we support the Prosodic Transfer Hypothesis, which argues that the ultimate attainment of L2 speakers is constrained by L1 prosodic representations; these representations can, however, be minimally adapted to accommodate the needs of the L2 under certain conditions. We investigate the L2 acquisition of English by 10 Mandarin speakers, by means of an experiment involving judgement and production of tense and participial morphology. Mandarin lacks overt tense inflection, while English inflection is represented by adjunction to the Prosodic Word, an option not available in Mandarin. We show that Mandarin speakers have few problems interpreting English tense appropriately, contrary to the predictions of the Representational Deficit Hypothesis. A detailed analysis is offered of their production of the morphology, which motivates the claim that the prosodic representation required for regular inflection in English can be built by combining licensing relations available in Mandarin. We conclude that target-like prosodic representations are ultimately attainable for at least some functional material which is absent from the L1.