In this section, procedures of a pilot study, the formal study and the way of scoring will be reported.
3.3.1 Pilot Study
The pilot study investigated the acquisition of Hakka passives of eight children aged from four to seven. Five properties were examined: subject animacy, reversibility, no truncation, verbal transitivity, and the bi-status pronoun in a special construction. All of them were included in the two tasks: the puppet selection task (PS Task) and the picture-cued production task (PP Task). The PS Task was conducted with two puppets.
The children had to select one puppet, whose utterance was correct. In the PP Task, the subjects were asked to produce in the bun construction after seeing pictures with the researcher’s short descriptions.
The results of both tasks showed that the children performed better as their ages increased. The critical point of fully comprehending Hakka passives fell on the age of five; their correct production of the construction was late, possibly between ages five and six.
The results also indicated that factors of subject animacy and reversibility were influential only on the youngest subjects (aged four) but not on the older children. The
results on reversibility generally agreed with the literature (cf. Bever 1970, Herriot 1970, Baldie 1976, Sudhalter and Braine 1985) that there was a tight relationship between reversibility and subject animacy. The 4-year-olds were inclined with reversible passives with animate subjects.
Our subjects of all ages performed worst on experiential verbs, showing that this type of verbs was not typical in Hakka passives. About the ungrammaticality of truncation in Hakka passives, the subjects generally did not perform as well as they did on full passives. With regard to the bi-status pronoun, it was found that the expletive in passives should be the default setting. Neutralization of bun and tung in Hakka was acquired at latter stages. The expletive interpretation can thus be assumed as an unmarked form of the bi-status pronoun in the special construction.
Nevertheless, the pilot study was inevitably deficient in certain aspects. First, there were less than ten children in the study. Only two subjects in every age group might not be representative enough. We thus proceeded with a qualitative study, but with limited time the study was not thoroughly detected.
In addition, not all properties of the bun construction were examined in the pilot study. The bi-status pronoun in a special construction, for instance, was confusing for adult native speakers (cf. Chiang 2006), let alone for young kids. In this special construction, the real referent of the pronoun was usually out of concern by adult speakers, and thus, both interpretations for the bi-status pronoun were only mentioned by the linguistic hypothesis. Moreover, “reversibility,” in fact may be constrained by animacy of the subject or object. The overlapping designs of the pilot study failed to clarify the real issue. In view of these deficiencies, a thorough investigation is needed.
3.3.2 Formal Study
Before the formal experiment was administered to the subjects, a training section
was given to them to insure their understanding of the task. The subjects heard eight active practice sentences and were asked to act upon pictures to familiarize them with the procedures. The protocol is as follows:
Table 3- 5: Sample Dialogues Used in the Training Section
Subjects saw: Subjects heard:
Experimenter:
Lia tieu gieu don ok, gi voi dui chale.
this CL dog very mad it MOD chase car
Liha gieu zo makgai?
now dog do what
‘This dog is so angry that it is chasing the car. What is the dog doing?’
Subject: Gieu Dui chale.
dog chase car
‘The dog is chasing a car.’
Experimenter:
Don kian. Liha ngi tang masa
very good now you listen which-one
gon gai he cok gai.
speak CSC be correct CSC
Gon chok gai ngin ngi si mia
speak correct CSC person you SI touch
gia tieuna i ha.
his head one time
‘Very good. Now you listen, and decide which one is correct. You can touch the head of the puppet who says the correct
Subject saw:
sentence.’
Piggy: Chale dui gieu.
car chase dog ‘The car chases the dog.’
Monkey: Gieu dui chale.
dog chase car
‘The dog chases the car.’
Experimenter:
Masa gon gai he chok gai?
which-one speak CSC be correct CSC
‘Which one is correct?’
Subject: [Touch Monkey’s head] Experimenter: Don kian.
very good ‘Very good.’
After ensuring their abilities, the formal experiment began, with the two tasks applied in sequence. It proceeded in one of the classrooms of the subjects’ school. In the classroom, the experimenter interacted with the subject individually. In completing the task, the subject may ask questions if s/he was unsure of the lexical items. The picture-cued production task (the PP Task) was conducted first where the subject needed to describe the behavior of the patient in the picture. The subject was asked to answer questions with Hakka passives. The subjects’ answers were recorded and then were analyzed later. The PP Task took each subject about fifteen minutes to complete.
In the PS Task, the two puppets were introduced again, and the subjects were invited to play a game named “Which puppet is smarter?” with the experimenter. In the game, each subject needed to select a puppet out of the two. Sentences were randomly ordered
and uttered by the two puppets in the experiment. This task took each subject about fifteen minutes to finish.
3.3.3 Scoring and Statistical Analysis
In this section, we will explain how the subjects’ responses will be scored and how the results will be analyzed.
In the PS Task, sentences of all types were examined. If the subjects made a correct choice from the two puppets, they would obtain one point for every property of the bun construction. No point would be given if they did not make response1 or made a wrong choice.
In the PP Task, sentences of all types of test items were elicited. The recorded utterances of the subjects would be examined. For every target item the subjects produce, one point would be given. If not, no point would be given. For errors made by the subjects, a qualitative analysis would be conducted later.
The data obtained from each task was entered into SPSS (Statistical Package for Social Science) and processed by the computer. To sum up, the procedures of the present study is shown in the following figure.
1 If the subjects show hesitations and do not respond, then they will not be given any point.
Designing the experiment
Setting up scoring criteria
Selecting participants of a pilot study
Conducting the pilot study
Revising the experimental design
Selecting participants of the formal study
Conducting the formal study:
1. the picture-cued production task 2. the puppet selection task
Analyzing results and errors
Figure 3-1: The Procedures of the Overall Research