| This study provides a systematic analysis of how devices referring to people and time are developed by longitudinally investigating the process by which two Korean ESL speakers acquired personal and temporal reference in the context of both spoken and written English discourse. The data for the study consist of approximately 44 hours of NS-NNS spontaneous conversations and the written corpus elicited from the journal entries over twelve months.; The qualitative and quantitative analyses offer insights into an emerging pattern regarding the use of personal reference forms. The learners in general displayed a gradual development from deviant use of referential forms to target like use, but the qualitative analysis of the data reveals that they had difficulties in applying appropriate referential forms in discourse contexts. The ability to choose the appropriate linguistic forms by considering the discourse context seems to carry a heavy cognitive load, and thus it is not acquired until fairly late in the developmental continuum.; The results of the learners' expression of past time reference confirm that there is a stage prior to verb morphology---even though the appropriate use of the target language forms is not acquired in the earlier stages---where the learners depended on the discourse structure of the topic and on adverbials to express past time events. Further, in their use of regular past forms, learners' data represent a disproportionate distribution by oral and written modes, which seems to stem from the pronunciation difficulties the learners have due to L1 phonological structure.; The examination of the association between verb morphology and lexical aspect supports the Aspect Hypothesis. The analysis of both token and type counts reveals that learners marked past tense first on telic and punctual predicates, present tense first on state predicates, and progressive marking first on activity predicates. The predicted extension of past tense marking, i.e., from telic and punctual to activity and state, was also observed. However, the spread of present tense and the progressive marking was not observed. The native speakers' data confirms the Distributional Bias Hypothesis, but represents a weaker association between verb morphology and lexical aspect than that of the learners. |
