4.2 Results of the Semantic Feature Analysis
4.2.1 The Comparison of semantic features between in and on
4.2.1.3 Between figure nouns of metaphorical expressions
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
68
container from the perspective of image schema theory, this may be one of the possible reasons for the difference between the prominent features of ground noun in in and on construction. Both prepositions can be used to refer to location, but the location in the former constructions should be able to include the figure entity while in the latter the figure should remain on the surface of ground after located on it, as the examples shown in (4.5).
(4.5) a. The protests were smaller than those of March 1989, when police opened fire on marchers and martial law was imposed. (BNC_CRC 868)
b. It was richly covered with trees, glinting in the sunlight, a land hidden in mountains, completely cut off. (BNC_HGK 841)
In (4.5a), instead of stating *opened fire in marchers, on is used to refer to the location of the occurrence of fire, which is in contact with the outer side of the marchers and does not really go into the body of them. In (4.5b), we can also see the contrast between in and on: it is not possible to say *a land hidden on mountains since from the semantic point of view, only mountains implies the meaning of surrounding or inclusion.
After examining the result of literal prepositional phrases, the cases in
metaphorical phrase will be discussed in the following two sections to investigate if the result may exactly conform to the hypotheses and help illustrate the construction of metaphorical phrases.
4.2.1.3 Between Figure Nouns of Metaphorical Expressions
Table 4.7 shows the distribution of semantic features of figure nouns in
‧
metaphorical preposition expressions in BNC. Metaphorical expression is a
construction which may not be comprehended from its literal meaning. For example, in situation in control, dependence on oil, the literal meanings of in ―inclusion‖ and on ―in contact with or supported by‖ cannot be applied to explain these two phrases, and extended meanings are needed to fully explain them. For metaphorical data, a statistic chi-square test was also run to compare if there is any significant difference between in and on in the semantic features.
Table 4.7 The Comparison of Semantic Feature of Figure Nouns in Metaphorical Preposition Expressions in BNC
feature in on χ 2 Df
Table 4.7 shows that significant differences between in and on may be found in
‗concrete‘, ‗animate‘, ‗human‘, ‗mobile‘, ‗partitive‘, ‗countable‘, ‗solid‘, and
‗activity‘ features. This indicates that when English native speakers use preposition in and on in their metaphorical meanings, more figure nouns with semantic features of animate, human, mobile, partitive, and solid may be found in in, or these features appear to be less in on. On the other hand, more concrete, countable, solid, and
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
70
activity features may be found in the figure nouns of on, or these features appear to be less in in, as discussed in the following passages.
For metaphorical in, the kind of figure nouns is like those in literal expression which involved people‘s names, references to people, as well as body terms (e.g. skull, face, blood). As these figure nouns are part of an animate being, undoubtedly most of them are active in moving (e.g. doctors, Tom). Extending this result to the distribution of sense in those metaphorical prominent categories (including Senses 2, 3, 4, 12, and 16, which appear to have higher frequency in metaphorical construction in Table 4.3), Figure 4.4 shows the graphic representation of how this set of semantic features may help represent the figures in metaphorical constructions..
NA
Vantage point is interior
Figure 4.4 The Graphic Representation of Figures in Metaphorical Prominent Senses of in in BNC
In prepositional expressions categorized into Sense 2, which is used to describe
―means, medium, or instrumentality‖, the figures tend to be inanimate noun phrases.
Cluster Location
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
71
Sense 3, which expresses ―limitation, qualification, or circumstances‖, tends to describe concrete things, especially mobile items, such as human beings, objects, or other things that are closely related to human activity (e.g. stock market, language).
Comparing this result with the hypothesis postulated in Table 4.2, the hypothesis cannot be proven in the ―means sense‖ within the LOCATION cluster, in which inanimate nouns appear to be more frequent in the data.
Sense 4, showing the ―purpose‖ of the figure, tends to include figure nouns that are more concrete, mobile, countable, and solid (e.g. mouth, gene), though non-solid items are also very common in the figures of this sense (e.g. vote, course). The data in Sense 12 display great consistency in the sort of features they contain; most of them are concrete, animate, human, mobile, countable, and solid. This result might be unique in all senses as it intends to show the ―position of participant, insider, or officeholder‖, so the figure surely needs to be human beings. Sense 16 mainly contains figures in vogue or season, which may not necessarily be concrete as they may vary from concrete objects, activities, attitude, to organizations.
Compared with in-construction, a lot of concrete nouns can also be found in the metaphorical on-construction. Evaluating the result based on the hypothesis presented in Table 4.2, we can see that some of the metaphorical prominent categories, that is, senses that appear to have higher frequency in metaphorical construction in Table 4.4 do not have their corresponding semantic clusters in the network in Figure 4.3, including Senses 21, 22, 25, 26. For Senses 6, 17, 18 which are grouped under the SUPPORT cluster, the result does not correspond to the distribution predicted in hypothesis. In this cluster which includes the senses of ―physical support‖, ―the means of conveyance‖, and ―basis‖, the current researcher hypothesized that concrete,
animate, human, and mobile feature may be more prominent in the metaphorical data;
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
72
however, the result shows that, in addition to concrete feature, countable, solid, and activity features also appear in high frequency. The senses under the SUPPORT cluster actually display great variation among them, as the basis sense usually involve an [-animate] figure (e.g. movie in This movie is based on a true story.), while the
―physical support‖ and ―the means of conveyance‖ senses tend to be constructed by human beings as the figure noun (e.g. children, student). Since most of the data identified within the senses of this cluster belong to the basis one, less animate nouns can be found in this set of data, leading to the inconsistency with the hypothesis.
As for Senses 13 and 14 belonging to the STATE cluster, the result basically conforms to the hypothesis with more activity semantic feature found in these nouns (e.g. car collection, treatment). Extending this result to the distribution of sense in those metaphorical prominent categories (including Sense 6, 13, 14, 17, 18, 21, 22, 25, and 26), Figure 4.5 shows the graphic representation of how this set of semantic
features may help represent the figures in metaphorical constructions.
Support State Support NA
Figure 4.5 The Graphic Representation of Figures in Metaphorical Prominent Senses of on in BNC
Cluster
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
73
From Figure 4.5, we can observe that there are some features that are more prominent in particular senses. For example, the figures in the construction of Sense 21 are purely inanimate and non-human as these nouns are nearly all negative descriptions of incidents which may cause ―harm‖ to the grounds, as in (4.6).
(4.6) It is significant that one which was encountered more than once was a “lack of follow-up support after the course and a lack of procedures for evaluating impact on participants‟ performance as opposed to the quality of the training experience”. (BNC_HPX 1396)
In (4.6), the impact symbolizes detriment to the ground performance. Other nouns that may appear in the position of figure for Sense 21 includes patenting, duty,
hardship, loss, trick, etc., which are all non-human in feature. In Sense 6, no partitive, measure, and temporal feature can be observed in the data. In Sense 13, concrete and solid feature appear extremely high in frequency. Both can also be explained through looking into the description of their senses, as the former refers to a source of
dependence which has to be whole entity, while the latter directs to an entity being involved in a condition. The examples of these two senses are given in (4.7).
(4.7) a. Affirmation depends on negation: white is valued at the expense of black;
youth acquires status through the devaluation of ageing. (BNC_C9S 202) b. Kwik Save also prosecuted for offering Danish blue cheese which was unfit
because of its bacteria count, and having food on sale at temperatures of up to 20 degrees which should have been stored at a maximum of eight degrees
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
74
centigrade. (BNC_K97 2038)
The tendency in the distribution of semantic feature in different senses within the semantic clusters provides evidence that prepositional phrases of a certain sense can only be constructed with a particular type of nouns. After examining the distinctive features in the figures of metaphorical phrases and the correlation of distinctive features with particular senses, the result of grounds in metaphorical phrase is probed into to see if there is any significant difference in the feature of these grounds.