CHAPTER 4 PSYCHOLINGUISTIC JUDGEMENT TASK
4.3 Results of the Object-Stimuli
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Additionally, the varied tendencies of the two adjectives for describing ‘inner’ or ‘outer beauty’ may not be the primary consideration for the participant’s judgments on the sentences.
In the following section, the participant’s use of pretty and charming in sentences describing object-related nouns is demonstrated.
4.3 Results of the Object-Stimuli
After showing the data of the participant’s acceptability of the person-stimuli, the next focus is to discuss the result of the object-stimuli. In this part of analysis, there are two types of controlled variables— ‘source of beauty’ (i.e., nature and artifact) and
‘patterns’ of collocates (pretty-only, charming-only, and common).
The Source of Beauty ⅹ Patterns Interaction
The analysis of Group A and B. When we included all the data of Group A
and B, the interaction effect of ‘source of beauty’ ⅹ ‘patterns’ was significant, F (2,703)
=5.400, p<.01. From observing the estimated marginal means, one could know that the participants may have different ratings in terms of the three ‘patterns’ used to describe different ‘sources of beauty’, as demonstrated in Figure 4-3.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 4-2 Overall Estimated Marginal Means of Acceptability of Sentences Describing Object-Stimuli
From Figure 4-2, it seemed that the participants showed that the object-related sentences within ‘pretty-only’ pattern were highly acceptable when they were describing the beauty of ‘nature’ (M=3.697, SD=0.112) and the beauty of ‘artifact’
(M=3.705, SD=0.112). The mean score of these sentences was about 3.7. As for
‘common’ pattern, all the participants agreed that both the adjectives were highly acceptable (common pattern: M=3.708, SD=0.079) to describe the beauty of the
‘artifact’—story; whereas, they showed that the sentences within ‘common’ pattern were the least acceptable sentences (M=3.198, SD=0.079) when they were depicting the beauty of ‘nature’—scenery. This results may suggest that the participants agreed with that pretty and charming were interchangeable when the adjectives were used to describe something created by a person (i.e., ‘artifact’). However, in terms of modifying the beauty of ‘nature’, the participants preferred to use pretty exclusively. They may be unfamiliar with the use of describing ‘nature’ in ‘common’ pattern.
To explore the differences of language use between Group A and Group B, in the
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
following parts, we analysis the two group separately.
The analysis of Group A. The two-way ANOVA was conducted and revealed that
the main effects of ‘source of beauty’ and ‘patterns’ were not significant (‘source of beauty’: F (1,344) =1.290, p=0.257; ‘patterns’: (2,344) =1.500, p=0.225). However, there was a significant interaction effect between ‘source of beauty’ and ‘patterns’, F (2,344) =3.833, p<.05. By looking at the estimated marginal means, shown in Figure 4-4 some interpretations can be made.
Figure 4-3 Group A’s Estimated Marginal Means of Acceptability of Sentences Describing Object-Stimuli
Figure 4-3 suggested that the participants showed higher acceptability of sentences describing the beauty of ‘nature’ in ‘pretty-only’ pattern (M=3.977, SD= 0.15) than those in ‘charming-only’ pattern (M=3.500, SD=0.150), and in ‘common’ pattern (M=3.477, SD=0.106). This is the reverse of the person-stimuli. More specifically, in terms of describing the beauty of ‘artifact’, it seemed that the participants were more likely to use pretty (M=3.721, SD=0.152) to describe the attractiveness than charming (M=3.674, SD=0.152). Furthermore, among the three ‘patterns’, the ‘common’ pattern
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
was regarded as the most acceptable pattern (M=3.943, SD=0.106) in describing the artifacts. The results were in line with the corpus analysis (see Chapter 3). Both pretty and charming were commonly used to describe something created by a person (i.e.,
‘artifact’). The beauty of ‘artifact’ was the most frequently type depicted by the two adjectives, accounting about 52% for pretty and 45% for charming shown in the COCA.
The Taiwanese EFL learners (i.e., Group A) had a similar preference of describing
‘artifact’. The sentences describing the ‘artifact’ in ‘common pattern’ were significantly more acceptable for the participants. However, in terms of describing the beauty of
‘nature’, sentences in ‘common’ pattern were the least acceptable for the participants.
They showed significant preferences of using pretty to describe the beauty of ‘nature’
(i.e., flower). From the results, we may suggest that the Taiwanese EFL learners were familiar with the use of both the two adjectives to describe ‘artifact’. Whereas, they seem to have a preference of describing the beauty of ‘nature’ by pretty exclusively.
The analysis of Group B. A two-way ANOVA was conducted and indicated that
the significant ‘source of beauty’ ⅹ ‘patterns’ interaction (F (2,353) =3.945, p<.05) was due to the participants’ higher degree of acceptability of the sentences describing the beauty of ‘artifact’ in pretty-only pattern (M=3.689, SD=0.160) than in the other two patterns. Among the two ‘sources of beauty’, the participants showed higher degree of acceptability toward the sentences describing the beauty of ‘nature’ in charming-only pattern (M=3.644, SD=0.160) than in other two patterns. In general, the participants in Group A and Group B had the similar tendency of describing the object-related nouns in ‘common’ pattern. They agreed with that using both the adjectives to describe
‘artifact’ were acceptable; whereas they tended to describe ‘nature’ in pretty-only pattern. The results can be easier demonstrated in Figure 4-4.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Figure 4-4 Group B’s Estimated Marginal Means of Acceptability of Sentences Describing Object-Stimuli
In addition, the main effect of ‘patterns’ was found significantly in Group B, F (2,353) = 4.641, p<.05, suggesting that the participants may have different degrees of acceptability with regard to different patterns. Overall, from the analysis, the results showed that the object-related sentences within ‘pretty-only’ pattern (M=3.556, SD=0.113) were rated higher than ‘charming-only’ pattern (M=3.533, SD=0.113), and
‘common’ pattern (M=3.200, SD=0.080).
To summarize what we found from the analysis of sentences focused on the object-related nouns, some points can be addressed. Firstly, the interaction effect between
‘source of beauty’ and ‘patterns’ seemed to serve as a vital factor that influence the judgements of both participants in Group A and Group B. However, the interaction effect performed differently in each group. The participants had different degrees of acceptability of the sentences describing something’s attractiveness in terms of the three
‘patterns’. For describing the beauty of ‘nature’, sentences within pretty-only pattern (i.e., pretty flower) was rated as the most acceptable description among the three
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
patterns in the Group A, while in the Group B, the participants showed that sentences within ‘charming-only’ (i.e., charming nature) were the most acceptable sentence followed by ‘pretty-only’ pattern and ‘common’ pattern. As for sentences depicting the beauty of ‘artifact’, the participants in Group A showed that story described by both pretty and charming (i.e., ‘common pattern’) were highly acceptable compared to the other two patterns. While, Group B had different opinions, they showed that sentences with ‘pretty-only’ pattern (i.e., pretty harbor) was more acceptable. Secondly, both Group A and Group B had similar tendency of describing ‘artifacts’ in ‘common pattern’. Overall, these results may suggest that the participants were not certain of the collocation use related to ‘artifact’ in both pretty-only and charming-only pattern. The two adjectives seemed to be regarded as an interchangeable pair of adjectives when describing ‘artifacts’.
Thirdly, we found that, regarding the three ‘patterns’, sentences within ‘pretty-only’
pattern were seen as the most acceptable when they were describing the beauty of
‘nature’. The results corresponded to the results of corpus analysis in the thesis. From the corpus analysis, it indicates that ‘nature’ was the second most frequent type of semantic categories that modified by pretty, accounting about 14% of all the instances in the study (see Chapter 3, section 3.2). One possible explanation is that all the learners were already familiar with the collocation use of ‘pretty-only’ used to modify ‘nature’.
However, the other two ‘patterns’ of collocates (i.e., ‘charming-only’ and ‘common’
pattern) may be less commonly used by the participants to describe ‘nature’. Clearly, the different semantic preferences of pretty and charming in describing object–related nouns could be further incorporated in the English language learning materials.
In the next chapter, the findings of both corpus analysis and the psycholinguistic judgement task would be discussed so as to solve the problems of distinguishing the similarities and differences between pretty and charming as well as clarifying the
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
factors that may affect the EFL learner’s acceptability of the sentences. It is expected that we could bring some teaching suggestions for the future.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
CHAPTER 5 DISCUSSION
There are two main foci of the thesis: the issues of (a) how to distinguish the differences and similarities of the two synonymous adjectives—pretty and charming, and (b) how EFL learners acquire them. This chapter is divided into two sections. Firstly, we would like to give a summary of the major findings in this thesis in section 5.1.
Then, in section 5.2, there will be some teaching suggestions for learning the set of near-synonyms—pretty and charming.
5.1 Major Findings in the Present Study
From the literature, the difficulty of the lexical choice between near-synonyms has been seen as one of the commonest language learning problems. Studies have tried to differentiate the differences between sets of near-synonyms through semantic roles of the words, collocational information, syntactic structure as well as their different performance in varied registers. Thus, in this thesis, we investigated all these information of the pair of near synonym—pretty and charming from the COCA. In order to explore the language learner’s use of the two adjectives, a psycholinguistic judgment task was adopted in the study as well. Additionally, we focused on the relation between gender and language for the reason that language use may reflect one’s social expectation or attitude toward different genders. Gender was used as a monitor variable in the psycholinguistic task to find out whether it would affect their language use. This integrated approach is different from those related studies because previous studies seemed to only focus on either the language use of native speakers or language learners.
In this section, the major findings of the thesis will be summarized.
First, regarding the semantic analysis of pretty and charming from the corpus, the results showed that both the adjectives had a similar tendency to describe the
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
attractiveness of a person or a thing. Both of them were more likely to describe person-related nouns compared to object-person-related nouns. The main differences between them were types of attractiveness they modified. For modifying a person, pretty tended to describe someone’s outer beauty; whereas, charming in most cases indicated someone having an attractive character. In terms of modifying an object, pretty tended to depict something created by people (i.e., ‘artifact’), such as dress, table, etc., followed by modifying the natural beauty (‘nature’), such as flower, tree, etc. Whereas, charming was mainly used to describe the attractiveness of an ‘artifact’ and a ‘place’, accounting for about 79% of the total instances. Furthermore, the use of the two adjectives seemed to correspond to the social expectation of females. Females were more likely to be pretty;
they were attractive in a delicate way (be slight, fragile, airy, and have beauty). Thus, from the semantic analysis of the pair of near-synonyms, we could know that the differences between pretty and charming could be identified from their preferences of describing varied types of attractiveness. Moreover, regarding the relation between the use of the two adjectives and gender, pretty tended to describe female’s outer beauty;
while, charming preferred to describe male’s inner beauty.
Second, although from the semantic analysis, few instances were found that males were pretty. According to the collocational information of person-related nouns, boy and boys were strong collocates with pretty. The results indicated that not only the sex roles but also the age of the modified person may be related to the discourse functions of pretty. Pretty can be used to express both compliment and insult. For charming woman (woman was one of overlapped noun collocates of the two target words within the top 20), we found that charming can be used to describe the woman’s attractiveness in both outer beauty and inner beauty. In terms of collocational analysis of object-related nouns, the results were in line with those in semantic analysis—pretty tended to describe outer beauty (e.g., pretty dress, pretty clothes); charming was apt to describe
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
inner beauty (e.g., charming book, charming town). Thus, from the collocational analysis, one could know that although the two adjectives had their own preferences to describe a particular gender of person (pretty for females; charming for males), they can be used to describe the two genders with both positive and negative discourse functions. In addition, since the two adjectives had a distinct preference for describing different sources of beauty (i.e., inner beauty and outer beauty), pretty tended to describe the outer beauty of a person or an object and charming preferred to describe the inner beauty, it seems that identifying the source of beauty of the modified nouns can be an effective way to differentiate the use of the two near-synonyms.
Third, from the investigation of the syntactic roles of the two adjectives, we found that both pretty and charming were more likely to appear in the attributive position.
However, what makes them different was the syntactic position which they applied when describing different categories of modified nouns. When describing person-related nouns, pretty did not show its preferences of syntactic positions (about 49% for both attributive and predicative position); whereas, charming showed a slight tendency to appear in predicative position (accounting about 57%). On the other hand, when modifying object-related nouns, both pretty and charming more frequently took the attributive position (about 55% for pretty and 69% for charming) than the predicative position.
Fourth, among the five registers showed in the corpus, pretty and charming commonly appeared in the fictional register. Their use in the academy register was the least. From the analysis, we could know the two adjectives were more likely to show in informal registers.
Fifth, following the findings in the corpus, we intended to explore what factors might influence EFL learners’ use of pretty and charming through a psycholinguistic judgment task. Three controlled variables were used to design the questionnaire; that is,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(a) the ‘gender’ of a person who has the qualities being described in the sentences (b)
‘patterns’ of collocates (i.e., pretty-only, charming-only, and common), and (c) ‘source of beauty’ (i.e., inner beauty and outer beauty). In addition, the combinations of the participant’s gender and person’s gender were labelled as a monitor variable in the analysis. Some results of psycholinguistic judgement task were in line with the corpus analysis. Firstly, when describing person-stimuli, charming was used significantly more frequently in describing someone’s inner beauty by the Taiwanese participants (i.e., Group A). Secondly, it seems that pretty and charming were regarded as interchangeable pair of adjectives to describe the attractiveness of ‘artifact’ by the participants in both Group A and Group B. The result was in agreement with corpus data since ‘artifact’ was the most commonly type of noun modified by the two target words. However, different from the corpus data, the combinations of genders were not a significant factor affecting the participant’s acceptability of the sentences. The language learners did not show a tendency of using the two adjective to describe male or female exclusively.
From the results, some possible dilemma of learning the set of near-synonyms were found and these findings may bring some teaching suggestions for the future.
Firstly, ‘patterns’ of collocates served as a significant factor affecting the participants’
use of the two adjectives. The sentences within the ‘pretty-only’ pattern were less acceptable for the most of the participants, with a mean score of 3.25. The results may suggest that the language learners in both Group A and Group B seemed to be uncertain about the use of collocates of pretty. Secondly, for describing the beauty of ‘nature’, the noun within ‘common’ pattern (i.e., pretty scenery, charming scenery) were less acceptable than the other two patterns. It appears that all the participants thought that pretty and charming were non-interchangeable in this context. Furthermore, the participants in Group A indicated that among the three patterns, pretty was the most
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
suitable adjective to describe the beauty of ‘nature’. However, the participants in Group B held a contrasting opinion—they thought that charming was the most appropriate word for describing ‘nature’. The findings may show that the participants had different tendencies of using the two adjectives in terms of depicting the beauty of ‘nature’. From these findings, some implications for pedagogical practices could be generated. In the next section, we will present one of possible ways of designing language learning task to provide learners with opportunities to notice the gap between their own language use and the target language.
5.2 Teaching Suggestions
Based on the results in the present study, we found that the language learners seem to be more unfamiliar with the collocational use of pretty and the different uses of pretty and charming in terms of describing the attractiveness of an object. One of possible ways to help the language learners to make a better lexical choice from this set of near-synonyms is to teach them to identify the main difference between the two adjectives—
pretty tended to describe the outer beauty of a person or an object. The attractiveness of outer beauty was like a visual stimulus which makes you feel instantly pleased. While, charming was apt to describe the inner beauty which needs the language user to think and to experience the attractiveness. It may take more time to find out the attractiveness of inner beauty.
The following language learning task was adapted from one of tasks mentioned in the Keck and Kim’s (2014, p. 114) Pedagogical Grammar. It is expected that the learning task can be used to raise the learners’ awareness of subtle differences between pretty and charming. The aim of the task is to help them notice the differences between the pair of near-synonyms from identifying the sources of beauty of noun collocates.
To begin with the language task, each student will receive one worksheet. In the task, eight noun collocates of the two adjectives were selected from the top 20 R1 noun
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
collocates from the COCA (see Chapter 3, Table 3-12).
Table 5- 1 The Design of the Language Learning Task
Types of Stimuli Sources of beauty Patterns Modified Nouns Items Person-stimuli Outer beauty pretty-only dresses (1)
face (2)
Person-stimuli Inner beauty charming-only accent (3) personality (4) Object-stimuli Outer beauty pretty-only flowers (5) pictures (6) Object-stimuli Inner beauty charming-only story (7)
book (8)
As demonstrated in Table 5-1, the eight noun collocates indicate the differences between pretty and charming in terms of their preferences of describing different sources of beauty. For pretty-only pattern, four noun collocates (i.e., dresses, face, flowers, and pictures) were selected as stimuli to indicate its preference of describing outer beauty. On the other hand, since the inner beauty was more likely to be described by charming, accent, personality, story, and book were chosen as the stimuli in the task.
Illustrations of the stimuli were provided to assist the learners in making judgements.
Illustrations of the stimuli were provided to assist the learners in making judgements.