| Language production involves three stages of processing,including the stage of conceptualizer,seen as a generator of preverbal message which includes all the conceptual information of what speakers want to include,the stage of formulator that consists of grammatical encoding and phonological encoding and the stage of articulating the final sentence.Among them,grammatical encoding consists of lexical retrieval which involves retrieving lemmas corresponding to the concepts in the prepared message and building syntactic structure which involves the elaboration of early sentence structure.It is generally accepted that production proceeds incrementally.Therefore,to produce fluent speech,speakers often separate the longer utterance into several units,and prepare the subsequent unit as speaking unfolds in time,that is,the subsequent unit X+1 is prepared when articulating the unit X.This comes to an important research question,how large is actually the planning scope X in speech production? Planning scope refers to the processes happening before overt articulation of a specific speech unit.This is still pretty controversial in L1 sentence production,because some studies claimed that planning scope may be extensive enough to cover a clause,and some other studies claimed that planning scope can also be limited to phrases or a single word.Recently,with more and more focus shifted to second language production,many linguists attempt to identify bilingual language production model and spot some differences between first and second language production.However,this concerning issue seemingly has seldom been touched in empirical studies in L2 sentence production.The present study,therefore,prepared to touch upon this controversial question in L2 sentence production,and came up with the following research questions: 1)What is the size of the planning scope at the grammatical encoding in L2 sentence production? 2)Do the factors,namely lexical frequency,structure availability,semantic relatedness and proficiency,affect the planning scope of L2 sentence production? If yes,how they affect?Two experiments were included in the current study by adopting Agnieszka E.Konopka’s picture-naming paradigm,in which participants were asked to describe the position of three pictures by using required sentence structure,such as “The A and the B are above the C” in the target display.“The A” and “the B” are either semantically related or unrelated.The magnitude of semantic relatedness served as a predicator to show whether there is temporal overlap when retrieving the complex noun phrase.The semantic interference from the second word indicates the retrieval of two nouns in parallel.To manipulate the ease of word retrieval,both the prime and target sentences either began with a high frequency or a low frequency word.To manipulate the ease of building a sentence frame,the target sentence was either structural primed or unprimed.In Experiment One,48 high-level Chinese ESL speakers and 48low-level Chinese ESL speakers were recruited.In Experiment two,to improve the accuracy rate of low-level speakers,another 48 low-level speakers were recruited,and they were arranged to learn the target words for three times before starting Experiment Two.In both experiments,speech onsets were then compared and analyzed.The two experiments together found that: 1)a reliable effect of word frequency was only observed in high-level participants.When the frequency of the first word is high-frequency,the planning scope is phrasal,while the planning scope shifted to subphrasal when the first word is of low-frequency.However,no effect of word frequency was found in low-level participants,so their planning scope is subphrasal.2)In both experiments,there was no effect of structure,and also no interaction effect on structure and semantic relatedness in both high-level and low-level participants,indicating that structure did not impose influences on the planning scope of L2 speakers who adopted a subphrasal pattern of planning.. |
PDF Full Text Request