PDF《Title children》

, '

why'

and '

when'

. It is because wh-pronominals are relatively simple syntactically since they ask for major sentence constituents that they replace (pp. 1086);

wh-sententials do not substitute major sentence constituents but ask for information that pertains to the semantic relations among the constituents in a sentence (pp. 1086). Although Bloom et ai. (1982) was a production study showing the linguistic demand in asking wh-pronominal questions is easier than that in asking wh-sentential questions, it can be inferred that similiar processes operate in children'

s ability to comprehend wh-pronominals and wh-sententials as

1 well as formulating answers to them. The effect of syntax on wh-question comprehension was also documented by Ervin-Tripp (1970) and Tyack &

Ingram (1977). They found that the transitivity of the verbs used in the sentence affects children'

s answers. From the discussion above, if we want to investigate children'

s ability to respond to different question-words and to explain their position in the developmental sequence with discussion of their underlying conceptual prerequisite, the sentence complexity of the questions asked should be carefully controlled. From all the comprehension studies stated above, only three studies fulfilled these cntena (Tyack &

Ingram 1977;

Gullo 1981;

Gullo 1982). In Tyack &

Ingram'

s study, although syntactic complexity of the questions was controlled, it was not without criticism. It was stated in their study that the questions were constructed with rigid control that some of the questions (e.g. What is riding the boy ?) would not occur in real life. This lack of realism reduced the validity for comparing performances of different wh-words (Gullo 1982). Another factor, suggested by Gullo (1982), which might have affected results in Tyack and Ingram'

s study was that the pictorial stimuli used depicted many wh-questions. Some of them were in the foreground and some in the background. Gullo (1982) recalled Elkind, Van Doornick &

Schwarz (1967) that children recognize the pictorial arrays more on the central figure rather than utilize the entire array. Therefore children'

s response accuracy may have been affected by the position of the answer located in the picture. Gullo (1981) and Gullo (1982) were more satisfactory studies since they controlled syntax as well as depicted realistic situation in the pictorial stimuli. Only two wh-questions were depicted in each picture and so both could be placed in the foreground. This prevoited the failure or success in answering wh-questions to be a function of where the answer was located in the picture. (Gullo

1982 pp. 126) These studies seemed to be good candidates to explain the developmental sequence of children'

s response to

2 wh-questions. However, the major focus of their studies was to compare the social class difference. No detailed investigation was on different wh-question answers as well as the errors patterns;

and not to say thorough discussion on the underlying conceptual prerequisites for different question-word responses. All studies above focused mainly on whether children'

s response fulfilled category agreement (Ervin-Tripp 1970), i.e. the category of information required by the particular question form. In addition to category agreement, Parnell et al. (1984) and Pamell et al. (1986) also looked at children'

s ability to provide factual, logical and acceptable information, i.e accurate response. The present study attempted to investigate responses to