國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
CHAPTER 4
RESULTS AND DISCUSSION
This chapter reports the results of the Awareness Test and the Comprehension Test. The data are analyzed and discussed in order to examine the hypotheses stated in Chapter 2. In addition, the research questions raised in Chapter 1 are answered based on the findings of the experiments. The Awareness Test aims to measure the ability of the participants to recognize metaphors and metonymies. The finer-grained analyses regarding degrees of transparency and opaqueness of expressions are also reported.
The Comprehension Test aims to examine the ability to interpret
metaphoric/metonymic expressions fluently. In addition to quantitative analyses, qualitative analyses were also conducted, for that the participants‘ interpretations of metaphoric/metonymic expressions might be shown through their translations.
Results of the Awareness Test
A total of 115 participants participated in the experiments. There were 68 of them from a general university, and 47 from a four-year technological university.
Among them, 32 participants from a general university and 24 participants from a four-year technological university were grouped into the CM group, which received instruction on conceptual metaphors or conceptual metonymies. The rest 36
participants from general university and 23 participants from a four-year
technological university were put into the MM group, which received instruction on metaphoric mappings. Their performances on the pretest and the posttest were
calculated and analyzed according to the designs of the 48 test items in the Awareness Test.
‧
Ability to Recognize Metaphors and Metonymies
The participants were asked to judge whether each sentence contained
metaphoric or metonymic expressions as well as to decide on the certainty of their judgments on a five-level scale in order to examine their abilities to recognize metaphors and metonymies. Table 4.1 shows the mean performance scores of the participants. The mean scores show that the participants of both university and of both groups had overall improvements. The improvements of scores also indicate that the participants became more certain about their judgments of recognizing
metaphoric/metonymic expressions.
Table 4.1
Mean Performance Scores of Participants in the Awareness Test
Type
Participants of the general university Participants of the four-year technological university
CM group MM group CM group MM group
Pretest Posttest Pretest Posttest Pretest Posttest Pretest Posttest With
Note.k: number of items
Note. The mean scores were rounded off to two decimal places.
Note. Numbers in parentheses listed under mean scores were standard deviations.
The issue whether these improvements on mean scores reach statistic significances was then investigated. The differences between the participants‘
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
performance scores in the pretest and in the posttest would show their learning progress in terms of the awareness of metaphors and metonymies. Some external factors, such as the participants‘ scores in English on the JCEE or the TVE joint college entrance examination, their time spent on learning English by themselves outside of classes, and their study of relevant subjects during the week, were also calculated during the process of regression testing in order to exclude possible impacts of the participants‘ individual background differences. The constants of the regression models represent differences between the participants‘ responses to the sentences on the pretest and on the posttest.
Table 4.2 reports differences in participants‘ performances on sentences with metaphoric/metonymic expressions between pretest and posttest. Both the CM groups of the general university and of the four-year technological university showed no significant differences between the pretest and the posttest. However, the MM groups of both universities showed significant differences between two tests (=3.04, t=2.50, p<.01; =7.70, t=1.87, p<.05). The results indicate that, regardless of the background differences, the participants of the MM groups of both universities improved
significantly in the posttest when they were asked to identify metaphoric/metonymic expressions.
Table 4.3 reports differences in participants‘ performances on sentences without metaphoric/metonymic expressions between pretest and posttest. All four groups of the participants performed no significant differences between two tests. The results suggest that explicit instructions on metaphoric/metonymic expression indeed
enhanced learners‘ awareness of figurative language uses; however, the instruction on metaphoric mappings also improved their ability to recognize figurative expressions.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.2
Differences in Participants’ Performances on Sentences with Metaphoric/Metonymic Expressions between Pretest and Posttest
Participants of the general university Participants of the four-year technological university
CM group MM group CM group MM group
Variables t VIF t VIF t VIF t VIF
(Constant) 2.15 .98 3.04 2.50** .78 .76 7.70 1.87*
Scores -.12 -.75 1.04 -.20 -2.14+ 1.07 .01 .08 1.07 -.52 -1.75+ 1.31
Self-learning time -.03 -.47 1.03 .037 .64 1.05 .02 .14 1.11 .02 .32 1.24 Review or not .45 1.93+ 1.05 -.11 -.54 1.02 .08 .12 1.05 -.28 -.36 1.06
R2 .13 .13 .00 .22
Adj-R2 .03 .05 -.15 .10
F 1.35 1.60 .02 1.80
+ p < .05, one-tailed. * p < .05, two-tailed. ** p < .01, two-tailed.
Table 4.3
Differences in Participants’ Performances on Sentences without Metaphoric/Metonymic Expressions between Pretest and Posttest
Participants of the general university Participants of the four-year technological university
CM group MM group CM group MM group
Variables t VIF t VIF t VIF t VIF
(Constant) .10 .05 1.10 1.23 .79 1.30 2.82 .66
Scores .00 .00 1.03 -.05 -.72 1.07 -.03 -.69 1.07 -.23 -.74 1.31
Self-learning time .05 .86 1.05 -.08 -1.95+ 1.05 .03 .31 1.11 .03 .34 1.24 Review or not .28 1.20 1.04 -.07 -.47 1.02 -.22 -.58 1.05 -.16 -.24 1.06
R2 .09 .15 .04 .07
Adj-R2 -.01 .07 -.11 -.08
F .88 1.90 .25 .73
+ p < .05, one-tailed.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
The performances of the participants of the CM and the MM groups were compared in order to measure the effects of learning caused by each instruction. To exclude possible impacts of external factors and also to differentiate the two groups, the three individual background differences multiplied by the group variable (i.e., the CM group equals 0 and the MM group equals 1) were again calculated along with the participants‘ responding scores. The constants of the group variables in the regression models represent the results of comparisons of two groups.
Nevertheless, to confirm that the participants of the two groups did not differ significantly before receiving instruction, their performances on the pretest were compared first. Table 4.4 reports the results of comparing the participants‘
performances on evaluating sentences with metaphoric/metonymic expressions in the pretests. On the other hand, Table 4.5 reports the results of comparing the participants‘
performances on evaluating sentences without metaphoric/metonymic expressions in the pretests.
In the general university, the participants of both groups did not differ
significantly in the pretest, no matter in evaluating sentences with metaphoric and metonymic expressions or sentences without them (=.10, t=.79, p>.05; =.07, t=.61, p>.05). Since the external factors had been excluded from consideration, the results of Table 4.4 and Table 4.5 suggest that differences between two groups in the posttest could be attributed to different instructions they received.
However, in the four-year technological university, the participants‘
performances in the pretest were significantly different. In terms of evaluating sentences containing metaphoric/metonymic expressions, though the CM-group performances were better than the MM-group‘s, they did not reach significance level, (=-.07, t=-.40, p>.05). On the contrary, in terms of evaluating sentences without metaphoric/metonymic expressions, the MM group performed significantly better in
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.4
Differences in Performances on Sentences with Metaphoric/Metonymic Expressions in Pretest between the CM and the MM group
Participants of the general university Participants of the four-year technological university
Variables t VIF t VIF
(Constant) 1.55 .91 3.38 9.09***
Groupa .10 .79 1.09 -.07 -.40 1.16
Scores .14 1.17 4.12 .01 .44 1.18
Self-learning time .04 .79 2.08 -.05 -.82 6.00
Review or not -.27 -1.46 2.20 -.14 -.63 1.85
GMSCb -.11 -1.64 2.09 .10 .77 1.38
GMSLc .02 .15 3.98 .03 .42 6.04
GMRd .77 3.11++ 2.18 .42 1.17 1.83
R2 .22 .13
Adj-R2 .13 -.01
F 2.47 .81
aGroup: The CM group is coded as 0, and the MM group is coded as 1.
bGMSC: Means of the scores (MSC) multiplied by Group variable (G).
cGMSL: Means of the self-learning time (MSL) multiplied by Group variable (G).
dGMR: Means of answers to the review-or-not question (MR) multiplied by Group variable (G).
++ p < .01, one-tailed. *** p< .001, two-tailed.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.5
Differences in Performances on Sentences without Metaphoric/Metonymic Expressions in Pretest between the CM and the MM Group
Participants of the general university Participants of the four-year technological university
Variables t VIF t VIF
(Constant) -.20 -.13 1.70 3.75**
Groupa .07 .61 1.09 .58 2.88++ 1.16
Scores .15 1.35 4.12 .02 .56 1.18
Self-learning time .03 .68 2.08 -.05 -.62 6.00
Review or not -.23 -1.41 2.20 -.09 -.31 1.85
GMSCb .00 .05 2.09 -.09 -.54 1.38
GMSLc -.07 -.53 3.98 .018 .22 6.04
GMRd .49 2.18+ 2.18 .64 1.48 1.83
R2 .13 .26
Adj-R2 .03 .13
F 1.32 1.92
aGroup: The CM group is coded as 0, and the MM group is coded as 1.
bGMSC: Means of the scores (MSC) multiplied by Group variable (G).
cGMSL: Means of the self-learning time (MSL) multiplied by Group variable (G).
dGMR: Means of answers to the review-or-not question (MR) multiplied by Group variable (G).
+ p < .05, one-tailed. ++ p < .01, one-tailed. ** p < .01, two-tailed.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
(=.58, t=2.88, p<.01). The results suggest that the two groups of the four-year technological university were different from the beginning. Whether these differences may result in biased findings should be examined later.
Table 4.6 and Table 4.7 report the results of comparing the progress that the CM and the MM groups made in evaluating sentences in the posttest. In the general university, the CM group had better improvements than the MM groups did on evaluating both sentences with and without metaphoric/metonymic expressions (=-.29, t=-1.81, p>.05; =-.18, t=-1.25, p>.05). Similarly, in the four-year
technological university, the CM group also had better improvements than the MM groups on evaluating both types of sentences (=-.33, t=-.78, p>.05; =-.62, t=-1.86, p>.05). In summary, the CM groups in both universities had better performances in the posttest than the MM groups did, no matter in evaluating sentences with or
without metaphoric/metonymic expressions. However, none of the groups reached the significance level, meaning that the differences on the improvements between groups were not statistically significant.
To sum up, with regard to the question of whether the participants‘ awareness of metaphors and metonymies was raised after receiving instructions, the results of the posttest reported in Table 4.2 and Table 4.3 showed that awareness was raised in both CM and MM groups of both universities. In terms of evaluating sentences containing metaphoric and metonymic expressions, the CM-group participants did not make statistically significant progress. However, the MM-group participants made significant improvements. On the other hand, in terms of evaluating sentences
containing no metaphoric or metonymic expressions, the participants of both CM and MM groups made progress, even though such progress was not statistically significant.
The results suggest that instruction on metaphoric mappings is as helpful as
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.6
Differences in Participants’ Performances on Sentences with Metaphoric/Metonymic Expressions between the CM and the MM Group
Participants of the general university Participants of the four-year technological university
Variables t VIF t VIF
(Constant) 2.15 1.00 .78 .82
Groupa -.29 -1.81 1.09 -.33 -.78 1.16
Scores -.12 -.76 4.12 .01 .09 1.18
Self-learning time -.03 -.48 2.08 .02 .16 6.00
Review or not .45 1.98 2.20 .08 .13 1.85
GMSCb -.08 -.44 3.98 -.53 -1.58 1.38
GMSLc .07 .79 2.09 .00 -.01 6.04
GMRd -.57 -1.81 2.18 -.31 -.34 1.83
R2 .15 .12
Adj-R2 .06 -.04
F 1.56 .73
aGroup: The CM group is coded as 0, and the MM group is coded as 1.
bGMSC: Means of the scores (MSC) multiplied by Group variable (G).
cGMSL: Means of the self-learning time (MSL) multiplied by Group variable (G).
dGMR: Means of answers to the review-or-not question (MR) multiplied by Group variable (G).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.7
Differences in Participants’ Performances on Sentences without Metaphoric/Metonymic Expressions between the CM and the MM Group
Participants of the general university Participants of the four-year technological university
Variables t VIF t VIF
(Constant) .10 .06 .79 1.03
Groupa -.18 -1.25 1.09 -.62 -1.86 1.16
Scores .00 .01 4.12 -.03 -.55 1.18
Self-learning time .05 .99 2.08 .03 .25 6.00
Review or not .28 1.38 2.20 -.22 -.46 1.85
GMSCb -.05 -.32 3.98 -.20 -.73 1.38
GMSLc -.35 -1.28 2.18 -.01 -.04 6.04
GMRd -.13 -1.86 2.09 .06 .08 1.83
R2 .13 .14
Adj-R2 .03 -.02
F 1.31 .87
aGroup: The CM group is coded as 0, and the MM group is coded as 1.
bGMSC: Means of the scores (MSC) multiplied by Group variable (G).
cGMSL: Means of the self-learning time (MSL) multiplied by Group variable (G).
dGMR: Means of answers to the review-or-not question (MR) multiplied by Group variable (G).
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
instruction on conceptual metaphor in enhancing learners‘ abilities to recognize metaphors and metonymies.
However, when comparing the participants‘ performances of the two groups, the degree of progress the MM groups made was lower than the degree of progress the CM groups made in judging sentences with or those without metaphors/metonymies, as shown in Table 4.6 and Table 4.7. The results of the comparisons thus do not support the first hypothesis which hypothesizes instruction involving metaphoric mappings can be more helpful than instruction involving CM in raising learners‘
awareness. At the same time, the differences on the performances of the MM group participants from the four-year technological university at the default setting seemed to have no significant influence on their performances in the posttest.
Even though the MM groups did not achieve higher average scores than the CM groups, the results of the comparisons indicate that they did gain significantly higher overall improvements. In other words, the participants of the MM groups might have made more consistent progress in evaluating sentences with metaphoric/metonymic expressions on the posttest, and thus the variances of the changes resulted in
significance in Table 4.2. Contrarily, the participants of the CM groups might have performed inconsistently on the posttest, so they did not show significant overall changes. The results of the comparisons, therefore, support the sixth hypothesis:
instruction involving metaphoric mappings can lead learners toward longer-term effect on retention than instruction involving conceptual metaphors and metonymies can.
Moreover, in what specific ways could the instruction on metaphoric mappings facilitate steady learning still requires further investigation. Thus, the finer-grained analyses based on metaphor-metonymy continuum were conducted and reported in the following section.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Ability to Recognize Metaphors and Metonymies on the Continuum
The second hypothesis stated that metaphoric mappings might assist learners to establish image schemas of abstract concepts and thus facilitate learners‘ abilities to recognize more abstract expressions, that is, expressions located closer to the side of metaphor on the metonymy-metaphor continuum. To examine the hypothesis, the participants‘ performances in evaluating metaphoric and metonymic expressions belonging to different categories were analyzed.
Table 4.8 reports the mean performance scores in four categories of sentences with metaphoric or metonymic expressions. The participants received higher scores in all four categories in the posttest than in the pretest. These overall improvements in the posttest correspond to the previous findings that explicit instructions on figurative language can facilitate learners‘ awareness.
Nevertheless, among the four categories in all four groups of participants, Category Four (C4), which contained metaphoric expressions, had the lowest mean scores. The results indicate that metaphoric expressions, which were formed based on abstract cross-domain mapping processes, were the most difficult one for the
participants to recognize; the lowest mean scores also suggest that the participants were highly uncertain about their judgments.
A regression test was conducted in order to examine progress the participants made in the posttest. Table 4.9 shows the results of differences in performances on the four categories of sentences containing metaphoric/metonymic expressions between the pretest and the posttest. In terms of the participants of the general university, those in the CM group did not make significant progress on any of the four categories, though they received higher scores averagely in the posttest. However, the participants of the MM group made significant progress in Category Three (C3),
=2.86, t=2.28, p<.05, and Category Four (C4), = 4.49, t= 2.38, p<.05.
‧
Participants of the general university Participants of the four-year technological university
Note. k: number of items
Note. C1: Sentences Containing Metonymic Expressions. C2: Sentences Containing
Metaphoric-Metonymic Expressions. C3: Sentences Containing Metonymic-Metaphoric Expressions.
C4: Sentences Containing Metaphoric Expressions.
Note. The mean scores were rounded off to two decimal places.
Note. Numbers in parentheses listed under mean scores were standard deviations.
In terms of the participants in the four-year technological university, those of the CM group performed significantly better in evaluating sentences belonging to
Category One (C1), =1.20, t=1.87, p<.05. However, they did not make significant progress in the rest three categories. On the other hand, the participants of the MM group made no significant progress in any of the four categories.
The significant progress in C3 and C4 which the MM-group participants in the general university made suggest that the instruction on metaphoric mappings was helpful for learners in recognizing expressions referring to abstract concepts. On the contrary, the CM-group participants did not make significant progress in evaluating sentences in all four categories. Their different performances in the posttest thus
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
Table 4.9
Differences in Performances on Four Categories of Sentences Containing Metaphoric/Metonymic Expressions between the Pretest and the Posttest
Participants of the general university Participants of the four-year technological university
CM group MM group CM group MM group
Variables t VIF t VIF t VIF t VIF
C1 (Constant) -1.85 -.49 2.49 1.67 1.20 1.87* 3.80 1.31
Scores .17 .64 1.03 -.14 -1.26 1.05 -.02 -.48 1.07 -.26 -1.25 1.31
Self-learning time .03 .34 1.05 -.04 -.50 1.07 -.12 -1.10 1.11 -.06 -1.12 1.24
Review or not -.12 -.31 1.04 -.24 -.92 1.02 .21 .53 1.05 .29 .66 1.06
R2, Adj-R2, F R2=.02, Adj-R2= -.09, F=.17 R2= .09, Adj-R2= .01, F=1.06 R2= .09, Adj-R2= -.05, F= .64 R2= .10, Adj-R2= -.04, F=.71
C2 (Constant) 3.21 .96 2.30 1.35 .34 .49 5.39 1.35
Scores -.20 -.84 1.03 -.14 -1.10 1.05 .02 .45 1.07 -.36 -1.24 1.31
Self-learning time -.06 -.70 1.05 -.01 -.16 1.07 -.08 -.66 1.11 .04 .60 1.24
Review or not .90 2.53+ 1.04 -.11 -.39 1.02 .12 .27 1.05 -.41 -.68 1.06
R2, Adj-R2, F R2= .19, Adj-R2= .11 , F= 2.25 R2= .05, Adj-R2= -.04, F= .52 R2= .03, Adj-R2= -.12, F= .18 R2= .19, Adj-R2= .06, F=1.44
C3 (Constant) 2.40 1.26 2.86 2.28* -.40 -.59 2.68 .93
Scores -.13 -.96 1.03 -.19 -2.01 1.05 .05 .98 1.07 -.16 -.78 1.31
Self-learning time -.04 -.79 1.05 .02 .40 1.07 .11 1.03 1.11 -.01 -.12 1.24
Review or not .01 .05 1.04 -.01 -.02 1.02 -.06 -.15 1.05 -.35 -.80 1.06
R2, Adj-R2, F R2= .05 , Adj-R2= -.06, F=.46 R2= .11, Adj-R2= .03, F= 1.43 R2= .12, Adj-R2= -.01, F= .89 R2= .08, Adj-R2= -.06, F=.57
C4 (Constant) 4.86 1.46 4.49 2.38* .41 .57 3.51 1.10
Scores -.32 -1.32 1.03 -.31 -2.19+ 1.05 -.04 -.70 1.07 -.26 -1.14 1.31
Self-learning time -.04 -.48 1.05 .17 1.92 1.07 .13 1.06 1.11 .07 1.21 1.24
Review or not 1.03 2.90+ 1.04 -.10 -.31 1.02 -.11 -.25 1.05 .02 .04 1.06
R2, Adj-R2, F R2=.25 , Adj-R2= .17 , F=3.18+ R2=.18 , Adj-R2= .10, F=2.31 R2= .06, Adj-R2= -.08, F=.44 R2= .21, Adj-R2= .08, F=1.68 Note. C1: Sentences Containing Metonymic Expressions. C2: Sentences Containing Metaphoric-Metonymic Expressions. C3: Sentences Containing Metonymic- Metaphoric Expressions. C4:
Sentences Containing Metaphoric Expressions.
+ p < .05, one-tailed. * p < .05, two-tailed.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
support the second hypothesis in that instruction involving metaphoric mappings can facilitate noticing expressions locating closer to the metaphor end on the
metonymy-metaphor continuum than instruction involving conceptual metaphors.
However, the performances of the participants in the four-year technological university were different from those in the general university. Except for the
significant progress the CM-group participants made specifically in C1, no significant differences were shown between the pretest and the posttest in four categories. The results do not support the second hypothesis.
Since the metaphor-metonymy continuum is not a strictly-categorized line but a continuum with fuzzy boundaries, further fine-grained analyses are needed. It is hypothesized that human bodily experiences should be able to facilitate learners‘
recognition, especially experiences involving parts of the human body. Thus the metaphoric or metonymic expressions involving parts of the human body were examined.
Ability to Recognize Metaphors and Metonymies with Bodily Descriptions
The third hypothesis posited that metaphoric mappings could facilitate learners‘
awareness of more opaque expressions, the expressions which were formed with few body-related descriptions. Metonymic expressions, which are based on physiological reactions, were presumed to be more transparent semantically than metaphoric expressions, which involve abstract linking across conceptual domains. To examine this hypothesis as well as to further determine the effects of the two types of
instruction, the participants‘ performances on the tests were analyzed.
Among the 24 sentences containing the metaphoric/metonymic expressions, 12 sentences contained words referring to human body parts or images of bodily
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
descriptions, while the other 12 did not. Table 4.10 and Table 4.11 show the differences between the pretest and the posttest in performances on sentences with and without bodily descriptions respectively.
Table 4.10 shows that, regarding the 12 sentences with metaphoric or metonymic expressions comprising body-part descriptions, the participants of both the CM and the MM groups of the general university did not make significant progress in the posttest, and neither did the CM-group participants of the four-year technological university. However, the MM-group participants in the four-year technological university performed significantly worse in the posttest than in the pretest,=-4.14, t= -2.60, p< .01.
Table 4.11 shows that, regarding the other 12 sentences with metaphoric or metonymic expressions which did not contain bodily descriptions, the participants of the general university did not make significant differences between the pretest and the posttest, and neither did the participants of the MM group of the four-year
technological university. Nevertheless, the CM-group participants of the four-year technological university performed significantly worse in the posttest than in the
technological university. Nevertheless, the CM-group participants of the four-year technological university performed significantly worse in the posttest than in the