This dissertation project examines how children learning English as their first language acquire the grammar of English questions. The ultimate goal of this study is to contribute to an understanding of why some types of errors, but not others, occur in language acquisition, and how these errors relate to input frequency and the properties of individual words. The research will include both cases of under-generalization of subject-auxiliary inversion (failure to invert the subject and the auxiliary in direct questions, e.g., "Why you are laughing?") and also cases of over-generalization (over-application of subject-auxiliary inversion in indirect questions, e.g., "I don't know what time is it.") Since such forms are entirely absent from the speech of adult native speakers, these are linguistic patterns created by the language learning mechanism. The outcomes of the proposed studies will shed light on the acquisition of interrogative structures and on the nature of generalization mechanisms in language acquisition.
This dissertation examined why young children learning English as their first language make mistakes when asking direct and indirect questions. For example, children say "Why you are laughing?" instead of "Why are you laughing?" Children also say "I don’t know where is it" instead of "I don't know where it is." Adult native speakers never produce such questions. Thus, children never hear the incorrect form, but they produce it. Word order errors are extremely rare in children's speech, so the occurrence of errors in questions is particularly interesting. The ultimate goal of this study was to understand why some types of errors, but not others, occur in language acquisition. To achieve that goal, the research investigated how children's errors are related to a) how frequently specific combinations of words occur in speech directed to children by parents and b) the linguistic properties of individual words, i.e., words that complete the meaning of a verb (who, what) versus words that add optional information about events (when, why). We created two tests to elicit direct and indirect questions from children. An example of our experimental protocol can be found online at www.sas.upenn.edu/~lpozzan/Publications/Questions%20Protocol.%20wmv. The results of the two tests confirmed the impression gained from children’s spontaneous speech: children produce word order errors in both direct and indirect questions. Moreover, some interrogative words (why, when) are associated with higher rates of word order errors than others (what, which). Errors were more frequent with words that added optional information than with words that supplied necessary information about an event, regardless of how often such questions occurred in the speech of adults talking to children. Thanks to its systematic design and careful methodology, this investigation contributes reliable data on children’s production of direct and indirect questions. The study demonstrates that frequency input patterns are insufficient to explain children’s errors. Instead, children are sensitive to the linguistic properties of the words and sentences they hear. The project has also created a successful protocol for the difficult task of eliciting direct and indirect questions experimentally. Other experimenters working with children speaking different languages, and experimenters working with other populations, such as atypically developing children and adults learning a second language, can use our protocol.
