30 More than thirty Activity verbs were listed in the learners’ textbooks Practical Audio-Visual Chinese I and Practical Audio-Visual Chinese II (Lesson 1), while only twelve Semelfactive verbs were found.
All of the Semelfactive verbs can be seen in our test questions (ke ‘cough’, ti ‘kick’, pai ‘pat’, reng
‘throw’, shuai ‘shatter’, tiao ‘jump’, gua ‘scratch’, ci ‘poke’, da ‘hit’, kan ‘cut’, qiao ‘knock’ and diu
‘throw’).
31 The morpheme da ‘hit’ was taught earlier than all the other Semelfactives (Lesson 22, Book I). But, it was not introduced as an independent verb. Rather, it was combined with the morpheme qiou ‘ball’ to form a verb phrase da-qiou ‘hit-ball’, meaning playing ball.
This section primarily concerns the L1 influence on the L2 acquisition of RVCs.
The first language influence, also known as the cross-linguistic influence or language transfer has been a central issue in the field of second language acquisition. SLA researchers investigating the role of learners’ native language are concerned with the question as to whether or not the presence of an L2 structure patterned like the L1 in the development of L2 proficiency is the result of L1 transfer. Results of studies along this line (e.g., Lado 1957; Ard and Homburg 1992; Odlin 1994) indicate that the transference of one linguistic label from the L1 to the L2 has the involvement of an obvious correspondence between the learner’s first language and the second language.
In this study, the acquisition of three RVC types was examined. In Chapter Two, the three types of RVCs and the English verbs that correspond to them have been compared and contrasted in terms of their semantic meaning and grammatical properties. As summarized in Tables 2-2 and 2-3, there are cross-linguistic variations between Act-/Sem-R RVCs and English Accomplishments, though their semantic components denote the action-result semantic relation. The cross-linguistic comparison shows that like English Achievements, such RVCs consist of only a result in their meaning. Moreover, RVCs resemble Achievements in the following ways.
First, similar to Achievements, RVCs are incompatible with durative linguistic labels.
Second, RVC sentences with the adverbs chayidianr ‘almost’ only have the result
interpretation. In light of the similarities and differences between Action-Result RVCs, it is predicted that the English learners are liable to transfer what they have perceived in the English verbs to the new forms: there will be a negative transfer if they appeal to Accomplishments when learning the grammatical properties of the action-result RVCs; a positive transfer will take place if they treat such RVCs as Achievements.
In the present study, we designed two experimental tasks to collect pertinent data—the grammaticality judgment (GJ) and sentence interpretation (SI) tasks. The GJ task, on the one hand, aimed to examine whether or not the subjects had the grammatical knowledge that RVCs are not allowed to go with durative linguistic labels and the second set of questions in the SI task, on the other hand, was to investigate how a sentence that had both the RVC and adverb chayidianr ‘almost’ was interpreted by the English learners. The result and discussion of the GJ task will be presented first, followed by those of the SI task. There were 18 test items in the GJ task. The 18 test items were divided into three groups according to the three durative markers— zhengzai, kaishi and tingzhi, 6 items for each marker. The 6 items were regrouped into three categories according to the three RVC types—Act-R, Sem-R and Sta-R RVCs. Therefore, the highest mean score for each test question that had one RVC and one durative linguistic label was 2.
The mid and high groups’ mean scores on Act-R and Sem-R RVCs are shown in
Table 4-5 and Table 4-6, respectively.
Table 4-5. The mid group’s mean scores on Act-R and Sem-R RVCs Durative
Table 4-6. The high group’s mean scores on Act-R and Sem-R RVCs Durative
Tables 4-5 and 4-6 indicate that on the whole, the mean scores of the subjects’ correct responses to Sem-R RVCs were statistically lower than those to Act-R RVCs. They had more correct responses in questions with Act-R RVCs than with Sem-R RVCs and one-way ANOVA confirmed this (p=.006*). In the previous discussion, we pointed out that subjects’ poor performance on Sem-R RVCs could be ascribed to their unfamiliarity with this type of RVC.
In addition to the reason of unfamiliarity with the meaning of Sem-R RVCs, we can also account for the significant difference existed between subjects’ performances on Act-R and on Sem-R RVCs with the L1 influence. That is, as mentioned already,
most learners were able to detect the incompatibility between the meaning of the Act-R RVC and a durative grammatical structure, but unable to notice such a feature in the Sem-R RVCs. A further examination shows that the learners seemed to understand that the result component constitutes the semantic focus of RVCs;
consequently, they treated these RVCs as Achievements instead of Accomplishments:
Achievements consist of only a result, whereas Accomplishments include both a process and a result as parts of their meaning. It should also be noted that Act-R RVCs are treated by the L2 learners as Achievements without preliminary stages and Sem-R RVCs as Achievements with preliminary processes. Given these, the English learners may rely heavily on the meaning of Achievements when making judgments on the combinatory infelicity between the meaning of the RVC and a durative expression.
From the above analyses, L1 transfer can be found. Learners treated Act-R RVCs as typical Achievements without preliminary stages. Typical Achievements cannot occur with grammatical structures that involve the feature of duration; therefore, they performed better on Act-R RVCs. On the other hand, the learners did worst on Sem-R RVCs because they treated these RVCs as Achievements with preliminary processes.
Recall that some Achievements are allowed to go with the English progressive, and the imperfective aspect spans the preliminary stage of Achievements. If Sem-R RVCs were conceptualized as Achievements with preliminary stages by the learners, then
their performance may reflect interference from their L1. In other words, the compatibility between the meaning of the Achievement and the English progressive could mislead the learners into considering that Sem-R RVCs can appear with the Mandarin imperfective aspect marker zhengzai. Such assumption conformed to the results presented in Tables 4-5 and 4-6.
Tables 4-5 and 4-6 show that in questions with zhengzai, the accuracy mean score for Sem-R was exceptionally low (M=0.65 for the high group and M=0.40 for the mid group). In test questions with zhengzai, Sem-R RVCs had the lowest accuracy mean score, which suggests that the English learners, even those with higher Chinese proficiency, tended to perceive that the Sem-R RVC is compatible with the imperfective aspect marker zhengzai. It should also be noted that the one-way ANOVA indicates that the learners did not make any progress on these questions (p=.559). However, as the results suggest, the L2 learners had the grammatical knowledge that Act-R RVCs are not allowed to go with zhengzai (M=1.45 for the mid group and M=1.05 for the high group). And, as their Chinese proficiency increased, they showed progress on this type of question (p=.045). As the findings suggest, the English learners relied on Achievements when making judgment on the incompatibility between the Action-Result RVC and durative linguistic label. It is worth mentioning that though both Sem-R and Act-R RVCs were conceptualized as
Achievements, the learners treated the two types of RVCs as Achievements with and without the preliminary stage, respectively. Thus, their performance on Sem-R RVC reflected the L1 interference, but their performance on Act-R RVCs did not. The findings corresponded to our prediction that the learners are likely to transfer what they have perceived in the English verbs, Accomplishments, or Achievements, but did not fully conform to the prediction that (only) a positive transfer would occur if they treat such RVCs as Achievements.
Now let us examine the learners’ performance on the SI task to see if the learners would rely on Achievements when making decision on the center predication of sentences that had both the RVC (i.e., Action-Result RVC) and the adverb chayidianr
‘almost’. As mentioned in Chapter Three, there were 12 test questions in the SI task.
The 12 questions were divided into two groups according to whether or not the test items contained the adverb chayidianr, 6 for RVC sentences without the adverb and 6 for RVC sentences with the adverb. It was also noted that the two set of questions were designed for different purposes: while questions without ‘chayidianr’ was investigate whether or not the result part of an event is less salient to English-speaking people than to Chinese-speaking people, questions with ‘chayidianr’ aimed to find out if the English learners would appeal to Accomplishments/Achievements when making decision on the center predication of RVC sentences with the adverb chayidianr
‘almost’. To examine the possible L1 influence, we present the two sets of sentences in the discussion. Presented in Table 4-7 are the frequencies of the result interpretation response of questions without the adverb chayidianr ‘almost’. Such frequencies for questions with the adverb are shown in Table 4-8. As noted already, the 12 test questions in the SI task were grouped into two categories, 6 test items for each category. The six test items were further divided into three groups according to the RVC types, 2 for each type of RVC. Hence, the highest frequency of the result interpretation response for each type of RVC is 40 and that for each group, comprising
20 people, is 40.
Table 4-7. Frequency of the result interpretation for test items with Act-R, Sem-R and Sta-R RVCs in the SI task
Group
Type of RVC Mid High Control
Activity-Result RVC 17 (42.5%) 28 (70%) 40 (100%)
Semelfactive-Result RVCs 19 (47.5%) 21 (52.5%) 38 (95%)
State-Result RVCs 36 (90%) 36 (90%) 37 (92.5%)
Table 4-8. Frequency of the result interpretation response for the items containing both the RVC and chayidianr ‘almost’ in the SI task Group
Type of RVC Mid High Control
Activity-Result RVC 26 (65%) 32 (80%) 40 (100%)
Semelfactive-Result RVC 24 (60%) 30 (75%) 40 (100%)
State-Result RVC 22 (55%) 31 (77.5%) 40 (100%) As is evident form Tables 4-7 and 4-8, test items with the adverb chayidianr
‘almost’ elicited a contradictory result to those without the adverb. The frequency totals displayed in Table 4-7 indicate that the performances of the English learners and Chinese controls on Sta-R RVCs were nearly equivalent. In the Sta-R RVCs, we obtained 37 result responses out of a maximum of 40 from the control group. The frequency of the result interpretation response received by either the mid or high group was 36 out of 40. The Chi-square confirmed that the frequencies obtained from the experimental groups were not statistically different from that of the control group (Chi-Square-value=.624, p=.732). However, with respect to the result interpretation response to Sta-R RVCs, the Chi-square showed a significant difference between the experimental and control groups (Chi-Square-value= 16.064, p=.003*). A further exploration was conducted to find out why the learners had the knowledge that the result part constitutes the semantic focus of the Sta-R RVC, but they did not sense that an RVC sentence with the adverb chayidianr has the result interpretation.
The exploration showed that the learners’ performance on State-Result RVCs reflected L1 interference. As noted already, the learners’ responses to the Sta-R RVCs in test items without the adverb chayidianr ‘almost’ were more close to the native speakers’ when compared with their responses to Sta-R RVCs in questions without
chayidianr. That is, the L2 learners, especially those in the mid group, were not so
aware that when a Sta-R RVC appears with the adverb chayidianr in a sentence, only the result reading is deducible from the RVC sentence. One possible explanation is that the learners had difficulty dealing with the property that the Sta-R RVC, which presents a state inducing an endpoint, can occur with the adverb chayidianr ‘almost’
since the adverb almost in English is only allowed to go with verbs representing events. In Chapter Two, we mentioned that the test with the adverb almost is used to distinguish English Accomplishments, which include the non-detachable process in their meaning from Achievements, some of which have a detachable process (Smith 1997). Hence, it is very possible that the English learners were influenced by their L1 knowledge that the adverb almost cannot occur with verbs naming a state and thus confused about the test questions that contain both a Sta-R RVC and the adverb chayidianr ‘almost’. Moreover, in light of the fact that the learners performed well on
the test items with Sta-R RVCs in questions that do not contain the adverb chayidianr, we can conclude that their fewer result interpretation responses to these test questions is likely to result from the L1 influence (i.e., the L1 interference).
Moreover, a negative L1 transfer could also be found in the learners’
performance on Action-Result RVCs. To better understand how these RVC sentences were interpreted by the English learners, the frequencies and percentages for Act-R
and Sem-R RVCs collected from the two sets of questions in the SI task were compared and contrasted and are shown in Tables 4-9 and 4-10:
Table 4-9. Frequency and percentage for Act-R RVCs with and without adverb chayidianr ‘almost’ enjoyed high frequency of the result responses. When such clues are removed, the result response decreases in frequency. On the basis of the result, we can state that the adverb ‘almost’ assisted the learners to make correct decision on the semantic focus of an action-result RVC. The findings were consistent
with the prediction that the learners relied on L1 knowledge of Achievements/Accomplishments when learning the semantic property of Action-Result RVCs. Moreover, they also accorded with the prediction that there would be a positive transfer if the learners transferred the L1 knowledge of Achievements to the L2 learning of Action-Result RVCs.