論分佈分析對早期語言習得的解釋能力：個案研究

本論文探討輸入語言（input）的分佈分析（distributional analysis）對兒童語言習得過程的解釋能力。對於語言能力為先天賦予或後天習得的爭論而言，此類討論有其重要性，因多數反對語言天賦的理論（non-nativist theory）認為跨領域的模式發現能力（domain-general pattern finding skill）是促使語言發展的一項動力。因此，若輸入語言能對一位特定兒童的語言發展作出良好的預測，則反對語言天賦的理論將獲得支持，但此種理論也必須有能力解釋成人語言與兒童語言之間的差異。本論文的研究材料為一位習得國語的男童與其母親的口語語料（自男童1;11.9歲起至2;5.28歲止，每週收集一次）。本研究以三種方式評估這對母子的語言相似度：（i）以結構（construction）出現的頻率比較個別動詞在兒童語言中的語法發展以及成人語言中的分佈形態；（ii）搜尋兒童語言中經常出現在動詞周圍的字詞，並檢查這些字詞在輸入語言中是否為提示動詞出現的可靠線索（cue）；（iii）搜尋兒童模仿成人的語句時經常刪略的字詞。研究結果顯示，分佈分析僅能解釋部分語言發展過程，且兒童語言與成人語言有兩項主要差異：（i）在兒童語言中，虛詞（functional word）大部分發展較晚，即使在輸入語言中虛詞的出現十分頻繁；以及（ii）輸入語言無法預測男童大致上「偏好」說出受詞而非主詞的傾向。本研究的發現顯示，反對語言天賦的理論不應只依賴頻率計算為語言發展做解釋。除了計算頻率，此類理論亦應考慮輸入語言中分佈上與功能上的多元程度（distributional variety and functional variety），並且提出計量「多元程度」的尺度以做出可供檢驗的預測。另外，此類理論應有能力解釋早期語言發展中抽象語法知識（例如論旨結構）與各別字詞特有現象並存的情形。

關鍵字

兒童語言習得；分佈分析

並列摘要

This thesis explores to which extent distributional analysis of the input can account for a child’s language acquisition process. Such an examination is important for the nature-nurture debate in the linguistic field (whether there is innate knowledge specific to language) because most non-nativist theories recognize domain-general pattern finding skill as a driving force of language development. Therefore, a non-nativist theory will be supported when the input makes good predictions of a specific child’s language development, but it should also be ready to explain the mismatches between adult’s speech and child’s speech. The spoken data of a Mandarin-speaking boy (collected weekly from age 1;11.9 to 2;5.28) and his mother are chosen as the research material. Three kinds of methods are used to evaluate the similarity of their speech: (i) comparing the child’s syntactic development of individual verbs with the adult’s speech by counting construction frequencies; (ii) searching for frequent items neighboring verbs in the child’s speech, and seeing if the items are reliable cues for verbs’ appearance in the input; (iii) searching for frequently dropped items when the child imitated the adult’s speech. The results show that distributional analysis can only account for part of the child’s language development process, and two major discrepancies between child’s speech and adult’s speech are found: (i) most functional items are lagged in the child’s part despite their high frequencies in the input; and (ii) the boy’s general “preference” of realizing objects but not subjects is not predicted by the input. The findings of this study suggest that non-nativist theories should not solely rely on frequency counts to account for language development. Besides frequency counts, these theories should take distributional variety and functional variety in the input into consideration, and make a measure of “variety” to make testable predictions. Also, these theories should be able to account for the coexistence of abstract syntactic knowledge (e.g. argument structure) and item-specific phenomena in early stages of language development.

並列關鍵字

child language acquisition ； distributional analysis

參考文獻

Lieven, Elena V. M., Heike Behrens, Jennifer Speares, and Michael Tomasello. 2003. Early syntactic creativity: A usage-based approach. Journal of Child Language 30: 333-370.

Van Valin, Robert D. and Randy J. Lapolla. 1997. Syntax: Structure, meaning, and function. Cambridge: Cambridge University Press.

Tomasello, Michael. 2001. The item-based nature of children’s early syntactic development. In Language development: The essential readings, eds. Michael Tomasello and Elizabeth Bates, 295-306. Malden: Blackwell.

Abbot-Smith, Kirsten, Elena Lieven, and Michael Tomasello. 2001. What preschool children do and do not do with ungrammatical word orders. Cognitive Development 16: 679-692.

Abbot-Smith, Kirsten, Elena Lieven, and Michael Tomasello. 2004. Training 2;6-year-olds to produce the transitive construction: The role of frequency, semantic similarity and shared syntactic distribution. Developmental Science 7: 48-55.

國際替代計量

論分佈分析對早期語言習得的解釋能力：個案研究

全文下載

主題瀏覽