Chapter 4 Results
4.1 Children’s motion event expressions in conversation
4.1.1 General results
About three hours of data was collected in each age group (ages 3, 4 and 5) for a total of nine hours of data. Within this data, 401 motion clauses were identified, constituting 5.7% of the total of all utterances. The percentage of motion expressions
Figure 1. Motion frequency in conversation
10%
20%
30%
40%
50%
3 4 5
M P D
Table 4. Token and percentage of motion verb in conversation
Verb M P D Total verbs
Age token % token % token % token %
3 76 35.0% 60 27.6% 81 37.3% 217 100%
4 68 34.3% 71 35.9% 59 29.8% 198 100%
5 85 29.8% 105 36.8% 95 33.3% 285 100%
Total
verbs 229 32.7% 236 33.7% 235 33.6% 700 100%
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
from each age group fluctuated between 5% to 6%. The results suggest that children have already begun to talk about motion as young as three years old. Generally speaking, the percentage of motion expressions in children’s utterances was not high.
The result is conceivable. This is possibly due to the nature of natural conversations.
Because natural conversations could vary widely in topic and context, children are not necessarily compelled to talk about motion events. Table 2 demonstrates the number of utterances, the number of motion clauses, and the percentage of motion clauses in total utterances.
3
There were several situations in which children talked about motion events.
Children might talk about the motions that they wanted somebody to do. In example (1), the mother (with the code MOT) of a three-year-old child wanted her daughter to draw a cat. The child CHI asked the mother to do the motion of “throwing” the paper toward her, so that she could draw with it.
Table 2. The number and percentage of motion clauses of children in conversation
Age Number of utterances3 Number of motion
clauses Percentage of motion
The utterances included clausal utterances, which were defined as a group of words containing
3
a noun and a verb in a subject-predicate relationship (Crystal, 1979). Ellipsis of the noun or/and verb in such relationship were also included as an utterance. However, utterances which
contained only interjection or a discourse marker (i.e ok “好” ok; “對” right, etc.) were excluded.
Example (1) [3-year-old]
%act: MOT takes out several pieces of paper
*CHI: 丟過來.
duiguolai
throw-pass-come
“Throw it over.”
Sometimes children described motions which they observed in the environment. In example (2), the mother asked the four-year-old child XUN about what people did in the zoo. The child replied with a general motion he observed from an elephant: it could roll things up with its nose.
Example (2) [4-year-old]
“The elephant would roll you up.”
Still another situation was when the child indicated what others should do or asked others to do something. In example (3), the mother and five-year-old child
LEE were playing card game. The mother said she had put out six cards on the table.
The child tried to tell her mother about how the game worked by telling her to take back the cards she didn’t like on the table.
Example (3) [5-year-old]
*MOT: 我 出 完 六張 了 耶.
wu chu wan liozhan le ye I exit finish six-piece ASP PRT
“I have finished putting out six cards.”
*LEE: 然後 再 把 你 不 想要 的 趕快 收回來.
ranhou zai ba ni bu xiangyao de gankai shouhuilai then again BA you not want hurry take-return-back
“And then take back those that you don’t want.”
4.1.2 Motion verbs
In this section, number of types and number of tokens of motion verbs are both investigated. The total number of verb types used by children of each age group is shown in Table 3.
Table 3. Total number of types of motion verb in conversation
Age M P D
3 25 13 2
4 33 13 2
5 33 14 2
Total Mean 30.3 13.3 2.0
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Overall speaking, Manner types outnumbered Path types irrespective of age.
Three-year-old children used more types of Manner verbs than Path verbs. Four-year-old children used even more types of Manner verbs than the three-Four-year-olds, while they used the same amount of Path types. Five-year-old children used the almost the same amount of Manner and Path types as the four-year-olds. Comparing with adults’ usage of verb types, children’s usage in all age groups was similar to adults’, meaning that the types of Manner were more than the those of Path. Some of the Manner verbs that were often found in children’s speech are zou ⾛走 “walk”, pao 跑 “run”, tiao 跳 “jump”, shen 伸 “stretch”, luo 落 “fall”, diao 掉 “fall”, etc.
Frequently used Path verbs include chu 出 “exit”, hui 回 “return”, zou ⾛走 “leave”, dao 到 “arrive”, dao 倒 “fall”, qi 起 “rise”, etc. The full list of verb types in conversation can be found in Appendix I.
According to Log Poisson Regression model, it was presented that an increase of types of Manner verbs and Path verbs was not significant (b = 0.35, Z = 1.93, p>.
05). In other words, the developmental changes in verb type usage was not obvious.
However, the overall results of verb type implied that children’s motion verb types did present language-specific patterns at an early stage in conversation, which is that the type of Manner verbs were more than Path verbs.
In addition to verb types, verb token was also examined. Each motion verb in each motion clause was coded. The results of the number of tokens of verbs and the percentage of motion verbs in conversation are presented in Table 4. Three-year-old children chose to mention M, P and D with similar frequency, which is between
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
27.6% and 37.3%. According Ch-square test, there were no significant differences between M, P, and D in three-year-old group, X2 (2, N = 217) = 3.46, p < .05; four-year-old children also used M, P and D with similar frequency, between 29.8% and 35.9%; there was no significant differences between M, P, and D, X2 (2, N = 198) = 2.17, p < .05. In five-year-old group, the preference of the motion verbs in frequency was similar to that of the other age groups. The frequency of the three motion verbs was between 29.8% and 36.8%; there were no significant differences between M, P, and D in the age five group, X2 (2, N = 285) = 2.17, p < .05. In sum, within each age group, the choices of the motion verbs were not significantly different. It seems that children in each age group had no particular preferences for M,P, or D.
The developmental curves of the usage of each category of verb also showed little differences. The curves of the developmental changes are presented in Figure 1.
No significant differences were found between 3-, 4-, and 5-year-old groups in the choice of M, X2 (2, N = 229) = 0.51, p < .05. No significant differences were found between 3-, 4-, and 5-year-old groups in the choice of P, X2 (2, N = 236) = 1.61, p
< .05. Last, no significant differences were found between 3-, 4-, and 5-year-olds in the choice of D, X2 (2, N = 235) = 0.88, p < .05.
In sum, children’s verb token usage did not seem to behave like the adults’.
According to Chen (2005), Manner frequency and Path frequency were 65% and 26% in adults’ utterances. However, children in the current study used only 32.7% of Manner, but 61.3% of Path and Deixis in their total motion encoding in conversation.
Furthermore, the developmental changes of verb token were not obvious between the ages of 3 to 5.
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
4.1.3 Motion constructions
In this section of motion constructions, the number of verbs in the motion constructions and the tokens of each construction are presented. The number of verbs in motion constructions was first presented in Table 5. The results showed that children’s patterns were similar throughout all three age groups. Three-year-olds used one-verb constructions in over half of the motion constructions (51.2%), followed by three- and two-verb constructions (25.6% and 23.2%). Four-year-olds also used one-verb constructions the most (47.0%), followed by three- and two-verb constructions (28.7% and 24.3%). Five-year-olds still favored one-verb constructions the most (47.8%), followed by two- and three-verb constructions (26.7% and 25.5%).
It shows that one-verb constructions were always the most frequently used construction in children’s speech in conversation (between 47.0% and 51.2%) compared to two- and three-verb constructions (between 23.2% and 26.7%) across age groups. The results coincide with Allen el al. (2003), who also found in his study that three-year-old children used one-verb constructions more.
Table 5. Token and percentage of number of verbs in motion verb in conversion
No. of
verbs One Two Three Total
Age token % token % token % token %
3 64 51.2% 29 23.2% 32 25.6% 125 100%
4 54 47.0% 28 24.3% 33 28.7% 115 100%
5 77 47.8% 43 26.7% 41 25.5% 161 100%
Total 195 48.6% 100 24.9% 106 26.4% 401 100%
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 2 presents the developmental changes of each construction. It seems that the changes remained mild. No significant differences were found in one-verb constructions across age 3, 4 and 5, X2 (2, N = 195) = 0.21, p < .05. No significant differences were found in two-verb constructions across age 3, 4 and 5, X2 (2, N = 100) = 0.26, p < .05. No significant differences were found in three-verb constructions across age 3, 4 and 5, X2 (2, N = 106) = 0.18, p < .05.
Figure 2. The percentage of number of verbs in motion construction in conversation
10%
20%
30%
40%
50%
60%
3 4 5
One Two Three
‧
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
The findings of each motion constructions are presented in Table 6. The top three motion constructions that the three-year-olds opted for were MPD, D and M. A Chi-square test presented that there was no significantly difference among the three most used constructions, X2 (2, N = 125) = 0.27, p>.05. Example (1) is when the three-year-old child YIN used MPD construction. The child was playing at home with her mother. The child pretended to cook something on the toy oven, so she pretended to turn on the switch of the oven and talked about the motion at the same time.
Table 6. Token and percentage of motion construction in conversation
Age D M P MD MP PD PP MPD PPD Total
*YIN: 然後 再 轉過去.
ranhou zai zhuanguoqu then again turn-pass-go
“then switch it over.”
%act: pretending to turn on the switch of the toy oven
The construction which the four-year-olds preferred the most was MPD, significantly more then other constructions, X2(1, N = 115) = 4.36, p <.05. The constructions followed MPD construction were D, M, and MP, without significant difference between them, X2 (2, N = 125) = 0.27, p>.05. For instance, in Example (2), the four-year-old child KUO used MPD construction when she was doing a jigsaw puzzle of the cartoon Aladdin. She mumbled about the motion she was doing when she was putting together the puzzle pieces.
Example (2) [4-year-old]
%act: mumbling while putting together the pieces of puzzle
‧
The five-year-olds still used the most MPD, followed by P and D. Example (3) illustrates the five-year-old QIN using MPD. QIN was playing with a puppet frog.
He put the puppet frog on a toy car, and described that he was able to perform the motion “to put it up on the toy car” to the Figure “the puppet”.
Example (3) [5-year-old]
*QIN: 也 是 當成 他 +…
ye shi dangchen ta also is pretend -as he
“pretend this also to be his…”
*QIN: /ei/?
INT
%act: putting the puppet frog on the toy car
*QIN: +, 他的 廚房 然後 +//. age samples, X2(2, N=104) = 1.2, p>.05. The token of MPD suggests that children are capable of constructing motions that are semantically condensed by adding different categories of motion verbs in serial verbs. The frequent use of MPD is also found in adults’ motion expressions (Guo & Chen 2009; Lin 2006). Therefore, it is possible that children had learned this language-specific construction from at least age three and widely used this construction in conversation.
‧
In conclusion, the results of motion constructions in conversation of children from age 3 to 5 shared a similarity to adults’ choices: MPD was a very productive construction. However, our results of number of verbs in motion constructions implied that although children were capable of encoding motion events with various motion verbs, their optimal construction of encoding motion events in conversation was still the simple ones. In this respect, children started to use MPD construction frequently, but not as frequent as the adults did. They still preferred one-verb constructions more in conversation.
In the section of 4.1, children’s motion expressions in conversation were studied, and the results answer the first research question of how Mandarin-speaking children ages 3, 4 and 5 encode motion events in natural conversation. It showed that, first, children used more type of Manner verbs than Path verbs just like the adults do starting from age three. Also, Manner types increased in four- and five-year-old groups. Second, children of all age groups used Manner, Path, and Deixis token with similar frequency, which was different from the Manner prominence in adults’ motion encoding. Last, children of all age groups used one-verb construction the most (including D, M, and P), but MPD alone was still the most common construction in children’s speech, which was similar to adults’ usage. The results of verb type and verb construction were similar to what was expected, while verb toke was not.