General Discussion - Acquisition of Literal and Non-literal Types

4.1 Acquisition of Literal and Non-literal Types

4.1.2 General Discussion

The major finding showed that to acquire the non-literal meaning of lao posed more challenges than the literal meaning of lao. Figure 4-1 shows a significant

outperformance on the literal type over the non-literal type for all the age groups, which is exactly what we have expected. The literal meaning is generally easier to acquire than the non-literal meaning due to the fact that children‟s general cognitive structure reflects the property of intellectual concreteness and realism; that is, only referential conception of language is likely to be processed (Levorato and Cacciari 1995). For instance, the children inclined to interpret the meaning of lao in lao jingcha literally as „an old police officer‟ rather than non-literally as „an experienced police officer,‟ since they generally made use of the referential feature represented by lao such as wrinkles or gray hair to interpret the meaning. On the other hand, the non-literal meaning „experienced‟ denoting the result of the aging process involves a less concrete property which is hardly referential from the physical appearance. As a result, to interpret the non-literal meaning of lao, children need to notice that there is a discrepancy between what is said and what is meant (Levorato and Cacciari 2002).

This result evidently supports the universal principle that explains the order of language development, the Markedness Theory (Jakobson 1971, Eckman 1977, Goodluck 1991). According to the Markedness Theory, once the marked form is acquired, the unmarked form is consequently acquired since the marked form refers to a less frequent or less intuitive language property (i.e., non-literal meaning). In other words, the literal meaning of lao was an unmarked form which was used more frequently and acquired easily. The results showed that the literal meaning of lao was easier to acquire than the non-literal one, supporting the Markedness Theory.

Regarding the between-group comparison, the older children (G3 and G4) acquired the literal meaning similarly to the control group, and they performed significantly better than G2. It was found that G5 did not significantly outperform the younger children (G1 to G4). To figure out the possible reasons for this, we

re-examined the answers of G5 to the literal type and noticed that the literal item lao muqin „old mother‟ was misleading for some subjects. To be specific, in the Interpretation Task, some children interpreted lao as other adjectives such as „nice‟ or

„amiably‟ instead of the original meaning „aged‟ as stated in (1) and (2); as for the Multiple-choice Task, they chose one of the options „diligent‟ rather than the correct one, „aged.‟

(1) cisiang de mama „nice mother‟ (taken from G5S1) (2) heaikeqin de mama „amiable mother‟ (taken from G5S6) However, other literal items like lao xiansheng „old man,‟ or lao furen „old woman‟

were not misleading because the noun phrases xiansheng „a man‟ and furen „a woman‟

were frequently modified by lao. It was concluded that part of children in G5 regarded the meaning of lao as showing respects when lao occurs with the kinship terms muqin „mother.‟

Regarding children‟s performances on the non-literal meaning, a steady and developmental curve was presented. Generally speaking, the correct responses to the non-literal meaning of lao increased with age. The older children (G4 and G5) significantly outperformed the younger group (G1 to G3), and in particular G2 and G3 performed significantly better than G1. The children at age 11 (G5) obtained the highest scores among all the experimental groups and their scores between literal meaning and the non-literal meaning were close (F (1, 28) = 5.468, p < .05). In addition, G4 performed similarly to G5, indicating that the non-literal meaning was not challenging for children beyond ten years old (G4). Since lao is a polysemous word which involves the core meaning and meanings beyond the literal level, it was found that to acquire words that stand for many meanings would result in difficulties.

Moreover, the children generally acquired the core meaning of polysemous words

first, and the secondary meanings were acquired gradually in subsequent years (Durkin and colleagues 1985). This evidence supports the developmental pattern of acquisition for meanings of lao. However, although older children achieved a significant difference with the younger children, the control group significantly outperformed all the children, indicating that the older children still did not fully acquire the non-literal meanings of lao. To figure out the possible reasons for this finding, we re-examined our subjects‟ responses and found that lao-phrases which were metaphorically opaque such as lao youtiao „a sophisticated person‟ and lao jianghu „a sophisticated person‟ posed great challenges on the Interpretation Task particularly. Therefore, the degree of markedness (cf. Jakobson 1971) for these items was fairly higher than other marked non-literal meanings and raised the difficulties of non-literal meanings for the older children in our study.

The results demonstrated that the age of acquisition of non-literal expressions basically started with ten years old, which is consistent with the results of previous studies (Gardner and Winner 1978, Gibbs 1987, Hsieh 2008, Levorato and Cacciari 2002, Prinz 1983). The developmental trend emerging from the result of the present study is supported by Levorato and Cacciari‟s (2002) Global Elaboration Model (GEM), which consists of five developmental phases: Phase 1 refers to children who are up to age seven, in which children elaborate meanings piece-by-piece in a primitive or literal way (i.e., G1 in the study). Children aged at eight to nine (i.e., G2 and G3 in this study) are at Phase 2 in which they start to acquire the sensitivity to contextual clues and make use of world knowledge to infer non-literal meanings. At Phase 3, children at ages ten to twelve (i.e., G4 and G5) are aware of the existing discrepancy between what is said and what is intended and that the literality can be violated on the basis of communication purposes. Although children at ages 10 and 11

in the present study did not fully acquire non-literal meanings of lao, the evidence that the older children in our study indeed performed significantly better than those at ages 7 to 9 implied that the children at age 10 developed a remarkable meta-linguistic ability to process the meanings beyond the literal sense (Levorato and Cacciari 2002).

Moreover, Nippold (2004) states that children at age 10 are able to read proficiently and thus are capable of acquiring new vocabularies by employing the meta-linguistic strategies such as inferring the meanings from linguistic context. These evidence give accounts for why the age 10 is a cutting-point for better acquisition of figurative meanings. However, it was indicated that with the subjects‟ age increased, their ability to acquire non-literal expressions continued developing up to adolescents and finally to the adulthood (Hsieh 2008). The GEM also showed that the adult-like competence in figurative expressions was not acquired after the age of fifteen.

4.2 Acquisition of Transparent and Opaque Types

In addition to exploring children‟s ability of literal and non-literal meanings of lao, the present study further examines the acquisition difficulty of non-literal meanings: transparent and opaque. As mentioned in Chapter Two, the non-literal meanings of lao were re-classified into two types: (1) the transparent non-literal meaning, which is closely related to the literal meaning of lao, and (2) the opaque non-literal meaning, which is not apparently related to the core meaning of lao. The second research question in the study aims to examine the difficulty between transparent and opaque types of non-literal meanings of lao. Section 4.2.1 reports the overall findings of the two types and Section 4.1.2 further discusses the results of the subjects‟ responses.

4.2.1 A Comparison between Transparent and Opaque Non-literal Meanings As shown in Table 4-3, the overall scores for the two types of non-literal meanings of lao, transparent and opaque were significantly different (F (1, 119) = 80.198, p < .001), and the post hoc comparison showed that generally the transparent type was acquired better than the opaque type.

Table 4-3 Subjects’ Overall Performances on the Transparent and Opaque Meanings of Lao

Type M SD F p-value

Transparent 0.82 0.19 80.198 .000*

Opaque 0.69 0.21

Figure 4-2 presents the mean scores of the subjects‟ responses to the two types of non-literal meaning: Transparent and Opaque.

Figure 4-2 Overall Performance of Each Group on Transparent and Opaque Meanings of Lao

= 0.86 > 0.79, G5: M = 0.9 > 0.78, Control: M = 0.99 > 0.95). Table 4-4 further shows the within-group differences for the performance of each age group between two types.

Table 4-4 The p-values for the within-group Differences between the Transparent and the Opaque Meanings

G1 G2 G3 G4 G5 Control

F 20.775 21.486 26.201 5.431 15.915 8.598

p .000* .000* .000* .031* .001* .009*

A significant difference between the transparent and opaque meanings was found in G1 (F (1, 28) = 20.775, p < .001), G2 (F (1, 28) = 21.486, p < .001), G3 (F (1, 28) = 26.201, p < .001), G4 (1, 28) = 5.431, p < .05), G5 (F (1, 28) = 15.915, p < .01), and the control group (F (1, 28) = 8.598, p < .01). Therefore, it has been found that the opaque meaning was more difficult to acquire than the transparent meaning.

Concerning the between-group differences in the two types, one-way ANOVA indicated that there was a significant difference among the six age groups in the two sub-types of the non-literal meaning of lao (transparent: F (5, 114) = 12.661, p < .001;

opaque: F (5,114) = 48.687, p < .001). The post hoc comparison showed that for the transparent meaning, G2, G3, and G4 significantly outperformed G1 (p < .01, p < .01, and p < .001 respectively). G5 significantly outperformed G1 (p < .001), G2 (p < .01), and G3 (p < .05). G5 was not significantly different with G4 (p > .05). However, the control group performed significantly better than the experimental group (p < .001) except for G5 (p > .05). As for the opaque meaning, G2 performed significantly better than G1 (p < .05), G3 performed significantly better than G1 (p < .001), and G4 and G5 significantly outperformed G1, G2, and G3 (p < .001). However, the control group significantly outperformed all the children (p < .001). Based on the finding, although

G5 significantly outperformed the younger groups on both types of the non-literal meaning, they still had a significant difference from the control group. However, it was concluded that even G5 had not fully acquired the opaque meaning of lao and found the opaque meaning especially difficult.

4.2.2 General Discussion

In addition to the finding that the literal meaning of lao was easier to acquire than the non-literal meaning, we further explored the difficulty between the two types of the non-literal meaning, transparent and opaque. Previous studies have noted that metaphoric transparency was found to influence the developmental process of children‟s acquisition of figurative expressions (Gibbs 1987). The major finding illustrated that the transparent meaning was easier to acquire than the opaque meaning as we had expected. In other words, the meaning which is not closely related to the core meaning of lao as in lao niao „an expert‟ was more challenging for the children to acquire than that which is apparently related to the core meaning of lao like lao pengyou „a friend that one has made for a long time.‟

As mentioned in Chapter Two, lao functions as an adjective and later is grammaticalized as a frequency and degree adverb. Liu (2007) argued that the grammaticalization of lao is undergoing the mechanism of metaphor. Specifically, the grammaticalization is formed by means of the similarity between the cognitive domain (i.e., to understand the abstract concept based on the concrete concept). Heine and his colleagues (1991) propose a scale representing the route of grammaticalization motivated by metaphor as represented in (3).

(3) PERSON > OBJECT > ACTIVITY > SPACE > TIME > QUALITY

The scale accounts for the domain of conceptualization in terms of the degree of metaphorical “abstraction” (i.e., left: most concrete; right: most abstract). That is, any one of the concepts on the scale may serve to conceptualize any other categories to its right. Accordingly, Liu (2007) stated that the grammaticalization of the adverb of lao is transferred from the PERSON domain (i.e., the origins of the core meaning of lao

„a seventy-year-old elder‟) onto the TIME domain (i.e., „the duration of action‟ as in lao zai shuijiao meaning „always sleeps‟), following the route from the most concrete toward the abstract concept. Along this line, the transparent meaning of lao is transferred from the PERSON domain (i.e., the origins of the core meaning of lao „a seventy-year-old elder‟ onto the TIME domain (i.e., „the duration of time‟ as in laopengyou „someone who existed from a long time ago.‟) In other words, lao originally denotes the “maturation (growth of age),” in which “the long-standing of age in the lifespan” is used to express the temporal concept about the continuity of interpersonal relationship metaphorically. Furthermore, the opaque meaning of lao as in laoshou „an experienced person‟ is conceptualized from the TIME domain (i.e., „a prolonged period of time‟) into the QUALITY¹ domain (i.e., „an experienced or sophisticated person.‟ The concept in the TIME domain „a prolonged period of time‟

is used to express the consequence or characteristic of someone who undergoes a prolonged period of time. As a result, an explanation for why the transparent meaning of lao is easier to process may be due to the fact that according to the scale of conceptualization for human experiences, the transparent meaning is conceptualized by concreteness (i.e., PERSON → TIME), whereas the opaque meaning of lao poses more challenging for acquisition since it is conceptualized by a relatively abstract concept (i.e., TIME → QUALITY). According to Ma (2002), the diverse meanings

1 Heine and his colleagues (1991) defined the QUALITY concept as „the catchall for a number of quite divergent conceptualization.‟

of lao comprise a three-level semantic field. The first level refers to the core meaning

„old‟ and the feature „a prolonged period of time‟ which is derived from the core meaning constructs the second level. The second level consists of lao pengyou

„someone who existed from a long time ago,‟ and another meaning as in lao shou „an experienced person.‟ The latter one (i.e., the opaque meaning) is claimed to be indirectly related with the former meaning as illustrated in Figure 4-3. It has been implied that on the scale of metaphorical concepts, there seems to be a cognitive activity that illustrates an egocentric distance proceeding from the concept closest to human experience (i.e., PERSON) to the one that is the most remote (i.e., QUALITY) (Heine and his colleagues 1991). In other words, the opaque meaning is remote from the children‟s personal experience and they need to bear greater cognitive load in processing the metaphorical opaque meaning.

Figure 4-3 The Match between the Semantic Field of Lao (Ma 2002) and the Scale for the Grammaticalization Route

The results of the study are consistent with the previous findings of Gibbs (1987), First level (Core meaning)

laoxiansheng „an old man‟

Second level (Non-literal meaning) laopengyou „someone who existed

from a long time ago‟

laoshou „an experienced person‟

Time

Quality Person

and Nippold and Rudzinski (1993) that their children more easily comprehended and explained the transparent meanings than opaque meanings. Gibbs (1987) states that when processing transparent idioms the children tended to metaphorically extend the literal meanings to other contexts and create figurative interpretations; while processing opaque idioms, the children encountered more difficulties since the obscure historical nature for the meanings of opaque idioms was not easy to recognize.

In the case of lao, the opaque meaning „an experienced person‟ denoting the state as the consequence of being through „a prolong period of time,‟ which is obscure and involves historical origins. For instance, lao jianghu originated from the Chinese novel Ershinian mudu zhe guaisiansiang² (Vol. 50) in Qing Dynasty, in which one described himself as a lao jianghu „ a man of long experience‟ and claimed that how a man of long experience fell for your tricks. For that reason, the historical nature of the opaque meaning of lao was hardly recognizable for children.

With regard to the between-group comparison in the transparent meaning, the fifth graders were found to perform significantly better than the younger groups (G1 to G3) and similarly to the control group. It was found that the fifth graders gained the ability to acquire the transparent meanings as the adults did. Although G4 performed similarly with G5, they did not achieve the adult-like performance. In addition, G2, G3, and G4 were not significantly different from one another, indicating that the younger children did not acquire the ability of the transparent meaning of lao until the age of eleven (i.e., G5).

However, for the opaque meanings of lao, the older children (G4 and G5) exhibited a better performance than the younger children (G1 to G3), indicating that the children at ten have started developing the ability to infer the non-literal meaning

2 ErshiNianMuDuZhiGuaiXianZhuang was written by Wujianren and the original quote was Shuzhi wo ye shige laojianghu, qi ken shang ni de dang.

that is not closely related to the literal meaning of lao. This is somewhat consistent with the finding in Gibbs (1987) that five- to nine-year-olds (i.e., G3 in the study) have not fully acquired the opaque idioms. However, this was found in examining the acquisition of the opaque idioms in the without-context condition (Gibbs 1987).

Although the older children performed significantly better than the younger children, the acquisition was still in the progress since the control group significantly outperformed the children. Generally speaking, the results indicated that Grade 5 has acquired the transparent meaning of lao; however, they still have not acquired the opaque meanings as in lao niao „an expert.‟

4.3 Acquisition of the Animacy Effect of Lao

The third research question addressed in this study aimed to examine whether the animate feature of the NP modified by lao is likely to affect the interpretation of lao.

With respect to the animacy effect on the non-literal meanings of lao, it is hypothesized that lao with an animate noun is easier to acquire than lao with an inanimate noun.

4.3.1 A Comparison between Animate and Inanimate NPs Modified by Lao Table 4-5 shows that the overall scores for lao-phrases with animate and inanimate NPs achieved a significant difference (F (1, 119) = 153.276, p < .001), and the scores for animate NPs gained a higher score (Mean = 0.83) than those for inanimate NPs (Mean = 0.68).

Table 4-5 Subjects’ Overall Performances on the Animate and Inanimate Features

To further identify the animacy effect in each age group, Table 4-5 compares the within-group difference for the scores on the animate and inanimate noun phrases of lao.

Table 4-6 The p-values for the within-group Differences between the Animate and Inanimate NPs of Lao

G1 G2 G3 G4 G5 Control

F 41.514 61.059 9.417 64.006 45.38 16.54

p .000* .000* .006* .000* .000* .001*

A significant difference was found between the scores for the animate and inanimate NPs of lao in all the groups: G1 (F (1, 28) = 41.514, p < .001), G2 (F (1, 28)

= 61.059, p < .001), G3 (F (1, 28) = 9.417, p < .05), G4 (1, 28) = 64.006, p < .001), G5 (F (1, 28) = 45.38, p < .001), and the control group (F (1, 28) = 16.54, p < .01).

Generally speaking, the differences were highly significant (p < .001) for all the children except G3 (p < .01). Moreover, the post hoc comparison showed that the scores for animate NPs were significantly higher than those of inanimate NPs. It was concluded that the lao-phrases with inanimate NPs were more difficult to interpret than those with animate NPs. Figure 4-4 presents the mean scores of each group for the animate and inanimate NPs of lao.

Type M SD F p-value

Animate 0.83 0.18 153.276 .000*

Inanimate 0.68 0.21

Figure 4-4 Overall Performance of Each Group on the Animate NPs and Inanimate NPs of Lao

As shown above, the scores for the animate NPs of lao for all the subjects were high (Mean = 0.64, 0.77, 0.76, 0.91, 0.91, 1 for G1 to G5 and the control group, respectively). On the contrary, the scores for the inanimate NPs of lao for all the groups were relatively low (Mean = 0.44, 0.53, 0.63, 0.74, 0.77, and 0.95 for G1 to G5 and the control group, respectively).

Figure 4-5 Subjects’ Overall Performances of the Animacy Effect on the Transparent and Opaque Meanings of Lao

G1 G2 G3 G4 G5 Control

Specifically, it has been found that in the transparent meanings, lao with animate NPs (Mean = 0.9) was acquired significantly better than lao with inanimate NPs (Mean = 0.74; F (1,119) = 62.995, p < .001). In the opaque meanings, lao with animate NPs (Mean = 0.76) was acquired significantly better than lao with inanimate NPs (Mean = 0.61; F (1,119) = 80.073, p < .001).

With regard to the between-group differences in animate and inanimate NPs of lao, one-way ANOVA indicated that there was a significant difference among the six age groups (animate: F (5, 114) = 20.071, p < .001; inanimate: F (5,114) = 36.789, p

< .001). The post hoc comparison demonstrated that for the animate NPs of lao, G3 significantly outperformed G1 (p < .01). The older children (G4 and G5) significantly outperformed the younger group (G1 (p < .001), G2 (p < .01), and G3 (p < .001)). The control group significantly outperformed G1 to G3 (p < .001), and G4 (p < .05). As

在文檔中中文「老」字之第一語言習得研究 (頁 68-0)