Comparison of General Accuracy Averages between the Two Groups
Figure 2 displays the accuracy averages of the 50 sentences between the two proficiency groups. The curves of G1 and G2 overlapped greatly, indicating that the degree of difficulty of questions for the two groups is similar despite the varied duration of learning English. The overall averages of G1 and G2 are 2.6985 and 2.6509, showing only a slight advancement of G1 over G2. Although, in some questions, the accuracy averages of G1 are greater than those of G2, such as in sentences 14-17, 22 etc., there are also questions in which G2 outperformed G1 and showed greater scores, such as in sentences 8-13, 40, 47 etc. The similar accuracy rates give the
Figure 2
The Accuracy Averages between the Two Proficiency Groups
110
impression that G1 improved little in the learning of English aspect structures relative to G2, despite their 5.3 more years (11.8-6.5) of learning English.
Acquisition Development of English Aspectual System in Verb Types
The accuracy means of all verbs and of each of four situational types of verbs––state, activity, achievement and accomplishment––
are displayed in Table 3. The overall means of the two proficiency groups (G1: 2.6985; G2: 2.6509) show little difference, as shown in the cell ‘all verbs’ in Table 3, but with further scrutiny into each situational type, the means of state and accomplishment verbs reveal a disparity between the two groups, especially in accomplishment verbs.
An observation of the means among the four types indicates that, between the two groups, the activity verb has the least difference (t = 0.987, p > .05) and the accomplishment verb has the greatest difference (t = 4.352, p < .01), meaning that the activity verb is least problematic in accommodation with aspect, and the accomplishment verb, the most problematic. Between the two extremes, the state verbs develop rapidly, as shown in the greatest mean 3.0544 of G1, and thus generate a significant difference between the two proficiency groups (t = 2.778, p < .01). In contrast, achievement verbs reveal a slow development, as shown in the small means 2.2524 and 2.0513 in G1 and G2 respectively, and show no significant distinction between the two groups (t = 1.421, p > .05). Accomplishment verbs are difficult for both groups as shown in the small means of both groups.
Although G1 did not show great advancement in accomplishments, this improvement made it significantly distinct from G2.
111
In sum, G1 advanced greatly in state and activity verbs, but lagged in achievement and accomplishment verbs; only the differential performance in knowing how states and accomplishments can combine with grammatical aspects is significant between the two groups.
Table 3
Means Comparison of Verb Types Between Two Proficiency Groups
To probe further, we compared the means of each verb type to discern whether there exists differences in acquisition among them.
Table 4 shows the mean differences of verbs of each type in comparison with the other three types. Tables 5 and 6 show the mean differences of verbs of each type in comparison with the other three types in G1 and G2, respectively.
According to these means, the general order of development of verbs of the four types is activity > state > achievement >
accomplishment. On examination, the means in Table 4 roughly
112
Table 4
Mean Comparison among Four Verb Types in All Subjects
Mean (SD) Activity
divide the four verb types into two categories: state and activity, as the better developed verbs with means 2.8211 and 2.8818, and achievement and accomplishment as the slowly developed verbs with means 2.1381 and 2.1404. The significant differences also provide strong evidence between the two major divisions of verbs of the four types; that is, state is significantly better than achievement (t = 8.214, p < .01) and accomplishment (t = 5.297, p < .01), respectively, and likewise activity (achievement: t = 4.196, p < .01; accomplishment: t
= 4.357, p < .01); whereas no significant difference is observed between state and activity (t = 0.631, p > .05), nor between achievement and accomplishment (t = -0.896, p < .05). This result indicates that Taiwanese EFL learners have difficulty in learning verbs of achievement and accomplishment types in aspect forms. This result concurs with J. Lin’s (2004) assertion that Chinese has only two primitive verbs, i.e., state and activity, and achievement and accomplishment are derived from the primitive verbs via interaction with the grammatical aspect or with other verbs.
113
Table 5
Mean Comparison among Four Verb Types in G1
Mean (SD) Activity
Achievement 2.2524 (0.6559) t = -0.896
p > .05 Accomplishment 2.3800 (0.5128)
Table 6
Mean Comparison among Four Verb Types in G2
Mean (SD) Activity
Achievement 2.0513 (0.7035) t = 0.783
p > .05 Accomplishment 1.9585 (0.4302)
Tables 5 and 6 reveal the same pattern as in Table 4: state and activity verbs are better developed whereas achievement and accomplishment verbs are difficult and developed much more slowly in aspect sentences. The only difference is that the development order for G1 is state > activity > accomplishment > achievement, whereas the development order for G2 is activity > state > achievement >
accomplishment. Within the two rapid-slow major categories, the order of the two verbs becomes switched, which seems to indicate
114
that in the early acquisition stage the activity verb is easier, but the state verb develops rapidly and outperforms the activity verb in the advanced stage. For the two slowly developing verbs, the accomplishment verb also develops rapidly and outperforms achievement verb in the advanced stage, but neither difference in the two pairs of verbs attains a significant level.
Acquisition Development of English Grammatical Aspects
Comparison of progressive and perfective aspects. Table 7 compares the means of progressive and perfective sentences in the entire group of subjects and within G1 and G2. The two grammatical aspect forms, i.e., progressive and perfective, demonstrate sharp and significant differences across all three comparisons (all: t = 15.355, p
< .01; G1: t = 10.443, p < .01; G2: t = 11.191, p < .01), indicating the degree of ease that these two aspects might differ greatly. In form, the two aspects seem equally heavy in construction. The progressive constitutes a verb to be and a gerundive verb, (Be + Ving), and the perfective constitutes an auxiliary have and a past participle, (have + PP). In concept, the perfective aspect is much more complicated than the progressive aspect. The progressive simply means an action is in progress at the reference time. The perfective can mean several indications: a past experience, a past ended event, an event lasting from the past to the present, etc. To complicate further, the aspect form together with the temporal phrases it contains must agree with the lexical aspect of the verb so as to make the sentences grammatically correct, as indicated in the literature review section.
A further comparison between the two proficiency groups, as shown in Table 8, reveals that a significant difference exists in each
115
Table 7
Mean Comparisons between Progressive and Perfective Aspects
Mean (SD) t sig.
All progressive 2.8936 (0.4006)
perfective 2.0938 (0.3885) 15.355 p < .01 G1 progressive 3.0612 (0.3106)
perfective 2.2578 (0.3933) 10.443 p < .01 G2 progressive 2.7663 (0.4166)
perfective 1.9693 (0.3382) 11.191 p < .01
individual grammatical aspect form (progressive: t = 3.495, p < .01;
perfective: t = 3.838, p < .01), indicating that the two aspects are difficult and require protracted learning. Although the greater means in the progressive aspect indicate more rapid improvement than the perfective aspect, the great distinction between G1 and G2 implies there is still much to acquire in the progressive aspect. The small means in the perfective aspect in both groups signifies that the perfective aspect is a major obstacle in learning English aspects.
Table 8
Mean Comparisons between Progressive and Perfective Aspects
Mean (SD) t sig.
Progressive G1 3.0612 (0.3106)
G2 2.7663 (0.4166) 3.495 p < .01
Perfective G1 2.2578 (0.3933)
G2 1.9693 (0.3382) 3.838 p < .01
Group comparison of verb types in the progressive aspect. To probe the origin of difficulty, we further examined the means of verb
116
types in the progressive aspect between the two groups. Table 9 shows a comparison of means of verbs of each individual type.
Statistical significance is observed in state (t = 2.070, p < .01), activity (t = 3.180, p < .01) and accomplishment (t = 2.074, p < .05) verbs between the two proficiency groups, implying that the progressive aspect improves in verbs of these three types during the additional 5.3 years of English learning between the two groups.
Table 9
Group Mean Comparison of Progressives among Four Verb Types
Mean (SD) t sig.
No difference is detected in the achievement verb (t = 0.436, p
> .05), signifying that verbs of this type in the progressive form of sentence improves little in the 5.3 extra years of learning, and is the source of difficulty in progressive aspect learning. The small means in both G1 and G2 also indicate that the achievement verb in the progressive form is conceptually difficult for Taiwan EFL learners.
The difficulty of learning achievement verbs in progressive form might derive from the English conceptual incompatibility of instantaneous achievements with a durative progressive aspect. If the proposal that Chinese has only state and activity verbs and that
117
achievement verbs are derived from state verb plus a termination aspect le is true, the lack of achievement verbs in Chinese would be a possible source of difficulty. Next, building the conceptual incompatibility between achievements and the progressive aspect requires not only that learners’ English proficiency reach a certain level, but also that their sensitivity to subtle aspectual features of verbs, inspiration from linguistic input, or the directions of teachers do so as well. All these factors might contribute to the little or no improvement in achievement verbs in English progressive aspect.
Group comparison of verb types in perfective aspect. Table 10 displays a comparison of the G1-G2 means in perfective aspect among verbs of the four types. Whereas the means of state and accomplishment verbs in perfective aspect show significant differences between the two proficiency groups (state: t = 2.321, p
< .05; accomplish: t = 3.103, p < .01), those of activity and achievement verbs do not (activity: t = -0.811, p > .05; achievement: t
= 1.707, p > .05). As activity verbs are the type that develop first and most rapidly, as shown in the overall means of G2 in Table 3, it is natural to see G2 perform equally as well as or even better than G1, as shown in Table 10 (G2: 2.9954 > G1: 2.8354). No significant distinction is observed in achievement verbs, and both groups display small means here (G1: 2.0122; G2: 1.6852), indicating that little improvement exists in verbs of this type in perfective aspect form.
Similar to the progressive aspect, achievement verbs are again a major obstacle in perfective aspect learning. Accomplishment verbs are also difficult to learn in the perfective aspect, as shown in the small means of both groups (G1: 1.4146; G2: 0.8611), and are to be discussed in detail in section 4.4.
118
One possible reason for the difficulty with achievement verbs might arise from the English-Chinese contrast in showing the perfectiveness with achievement verbs. In English, the instantaneous nature of the achievement verb is inherent in the lexical property of the verb, whereas in Chinese the achievement verb is formed externally, comprising a state verb plus a perfective aspect marker le or a verb plus an achievement verb (J. Lin, 2004). The inherent property of English achievement verbs would be consequently difficult for Taiwanese EFL learners to acquire because of the contrast of perfective formation in the linguistic systems.
Table 10
Group Comparison of Perfectives among Four Verb Types
Mean (SD) t sig.
G1 2.7622 (0.9730) State
G2 2.3333 (0.8256) 2.321 p < .05 G1 2.8354 (1.0210)
Activity
G2 2.9954 (0.8961) -0.811 0.419 G1 2.0122 (0.9337)
Achievement
G2 1.6852 (0.9178) 1.707 0.091 G1 1.4146 (0.8725)
Accomplishment
G2 0.8611 (0.8393) 3.103 p < .01
Secondly, the instantaneous nature of the English achievement verb is incompatible with the durative phrase in the perfective aspect, exemplified in (23a). In contrast, in Chinese the achievement verb can co-exist with the perfective aspect using the mechanism of sentential subject and temporal predicate (Li & Thompson, 1981; Teng, 1975), as shown in (23b). The little improvement of achievement verbs
119
(from G2: 1.6852 to G1: 2.0122) might thus be a result of negative L1 transfer from Chinese.
23. a. *Sarah’s father has died for more than 10 years.
b. Sala de fu-qin si-le shi duo nian le Sarah poss. father die-ASP ten more year LE
It has been more than ten years since Sarah’s father died.
Durativity and Telicity in the Perfective Aspect
As expected, the addition of a for-phrase and an in-phrase in the perfective aspect increases the complication of grammatical aspect in the sentence, which is reflected in the unexpected variability of accuracy means of G1 and G2 in Table 11 below. The accuracy means of activity verbs are greater than those of accomplishment and achievement verbs in sentences with both an in-phrase and a for-phrase. As only a for-phrase in the perfective aspect is taught in junior high school,5 the untaught in-phrase in the perfective naturally appears more difficult for both groups as indicated in the smaller accuracy means of both groups, and significant divergence is observed in this area (t = 2.091, p < .05), likely because of the extra 5.3 years of English learning. For achievement verbs, the acquisition situation is similar, with for-phrase sentences outperforming in-phrase
5 The grammar of perfective aspect with for NP is mentioned early in all textbooks of junior high schools in Taiwan. Most books elaborate the aspect with the durative phrase “for +NP,” as indicated in the following publishers in Taiwan, but no perfective sentence has ever mentioned the completive “in + NP” in the perfective aspect.
Kanghsiung: Book 4, Lesson 8 Taiwan Pason: Book 4, Lesson 5 Hanlin: Book 5, Lesson 1 120
Table 11
Group Comparison of Perfectives with For-Phrase and In-Phrase
G1-G2
for NP 2.8354 (1.0210) 2.9954 (0.8961) -0.811 0.419 in NP 1.7439 (0.8487) 1.8472 (0.9827) -0.538 0.592
sentences, but no significant difference is observed between the two groups. The in-phrase sentences of G2 has an unexpectedly greater accuracy mean than that of G1 and also than that of its own for-phrase sentences. The same phenomenon occurs in accomplishment verbs.
For accomplishment verbs, the means are the least of the three verb types, indicating this to be the most confusing type of verb. In for-phrase perfective sentences, the G1-G2 group comparison shows a significant difference (t = 3.130, p < .01). In in-phrase perfective sentences, G2 performed even better than G1, although no significance is observed. This condition indicates there might exist a period of confusion with both accomplishment and achievement verbs as the in-phrase sentences had greater means than the for-phrase sentences, contrary to the normal development of verbs of other types.
G2 might have answered sentences correctly for the wrong reasons, and hence surpassed G1 in achievement on in-phrase sentences, and the inverse could be true for G1 with respect to in-phrase and for-phrase sentences.
121
Durativity and Telicity in Interaction with Achievement Verbs As durativity (for-phrase) and telicity (in-phrase) in the grammatical aspect generally disagree with the lexical aspect of achievement verbs, we probe further to discover what acquisition order of this type of verb displays. Table 12 presents the accuracy means of achievement verbs in three types of perfective patterns in the experiment: perfectives with no phrasal modifier, perfectives with for-NP, and perfectives with in-NP.
Table 12
Mean Comparison of Achievement Verbs in Three Perfective Aspect Patterns
Mean (SD) + no phrase + for NP G1
+ no phrase 2.1298 (0.9859)
+ for NP 2.0122 (0.9337) t = 0.485
+ in NP 1.7439 (0.8487) t = 2.155 t = 1.153 G2
+ no phrase 2.1169 (0.9044)
+ for NP 1.6852 (0.9178) t = 2.129
+ in NP 1.8472 (0.9827) t = 2.220 t = -0.725
Although achievement verbs reveal no distinction in durative perfectives (i.e., perfective + for NP) and telic perfectives (i.e., perfective + in-NP) or between G1-G2 groups (cf. Tables 10 and 11), an internal within-group scrutiny shows different results. According to Table 12, G1 showed a significant difference between pure (i.e., no-phrase) perfectives and in-NP perfectives (t = 2.155, p < .05) whereas G2 showed a significant difference not only between no-phrase perfectives and in-NP perfectives (t = 2.220, p < .05) but also
122
between no-phrase perfectives and for-NP perfectives (t = 2.129, p
< .05). The means indicated that the pure perfective was the easiest type, followed by perfectives with for-phrase, which is again followed by perfectives with an in-phrase. The two significant distinctions in G2 indicate that the intermediate group is effective with only pure perfectives and had difficulty with perfectives added with other phrasal features. G1, the advanced group, improved in acquiring the for-phrase and therefore showed no more distinction, but still struggled with the in-phrase perfectives; the significant difference thus remained. For the two temporal phrases, the fact that the for-phrase is often taught together with perfectives in Taiwan whereas the in-phrase is overlooked in formal English education makes the acquisition of perfectives with these two phrases unequal in acquisition development. This result supports Buczowska and Weist’s (1991) relevance of pedagogical practices to the foreign language learning process.
CONCLUSION
We investigated the acquisition progress of Taiwanese EFL learners of the lexical aspect of verbs in interaction with grammatical aspects. For research question 1, English aspect constructions are found to be difficult for most learners; two groups with 5.3 years difference in learning English still patterned similarly and with no significant difference observed in judging the grammaticality of English sentences. For research question 2, progressive, in the two major types of aspect, is significantly easier than perfective between the two groups, probably because of the conceptual simplicity and the
123
lack of interference from cross-linguistic difference, i.e., the incompatibility of state verbs in progressive aspect in both languages.
However, Chinese learners made errors in using achievements in the progressive aspect due to lack of primitive achievements in the lexicon and, thus, were insensitive to the compatibility requirement.
In terms of acquisition order of verbs of the four lexical types in research question 3, states and activities advanced much more rapidly than achievements and accomplishments in both proficiency groups, in agreement with J. Lin’s (2004) assertion that Chinese has only two primitive verbs, i.e., state and activity, and achievement and accomplishment are derived from the primitive verbs in interaction with grammatical aspect or with other verbs. Positive L1 transfer of state and activity verbs was observed here.
The perfective aspect was difficult for Taiwanese learners because, in addition to the complicated form of grammatical structure, i.e., have +PP., there was an internal agreement of lexical aspect of verbs with the concept of grammatical aspect. The hindrance was shown in the delayed development of achievement and accomplishment verbs in English aspect structures. The punctuality feature of achievements, incompatible with the durative phrase and occasionally with the terminative phrase, was acquired late only by advanced learners. Negative L1 transfers were typically seen in mistakenly regarding achievements in the perfective aspect as the Chinese sentential clause with a temporal predicate. Accomplishment verbs had no problem in the progressive aspect but appeared to be difficult in the perfective aspect, especially with durative and terminative phrases. There was even a period of confusion, during which the advanced learners performed worse than the intermediate
124
learners. The complication derived from the reason that the telicity feature of accomplishment verbs disagreed with the durative phrase but agreed with the terminative phrase. Hence, for research question 4, due to unawareness or insufficient knowledge about the agreement of lexical aspect in interaction with the grammatical aspect, Chinese EFL learners, often had prolonged laborious experiences in acquiring English aspect structures.
For research question 5, the addition of a durative for-phrase and a terminative in-phrase in the perfective aspect indeed increased the complication in the agreement of lexical-grammatical aspect in the sentence, which was reflected in the significantly low accuracy means in the two structures in comparison with the pure perfective structure. The in-phrase perfective was even more difficult than the for-phrase perfective since the former was rarely taught in school, whereas the latter was usually taught together with perfective structure in Taiwan. This result supports the effect of formal instruction in L2 acquisition proposed by Buczowska and Weist (1991).
Chinese monosyllabic verbs are only of two types: state and activity. Achievement and accomplishment verbs are morphologically derived by making verb compounds or syntactically derived by adding aspect marker le, or a terminative object NP. Hence, with the lexical aspect mingling with grammatical aspect, Chinese verbs have less lexical property division than English verbs, and matching is less strict between the lexical aspect and the grammatical aspect than in
Chinese monosyllabic verbs are only of two types: state and activity. Achievement and accomplishment verbs are morphologically derived by making verb compounds or syntactically derived by adding aspect marker le, or a terminative object NP. Hence, with the lexical aspect mingling with grammatical aspect, Chinese verbs have less lexical property division than English verbs, and matching is less strict between the lexical aspect and the grammatical aspect than in