國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
105
prepositional phrases of particular senses. Moreover, in order to facilitate the learning of English prepositions, these findings are also extended to probe into the learners‘
tendency and problems in producing prepositional phrases.
5.2 Discussion of the Results
Traditionally, studies on the sense of prepositions concern mostly on nouns going after a preposition as the factor of influencing the meanings of it. For example, from the cognitive perspective, in Wierzbicka‘s (1993) paper, the distinction in the meanings of temporal prepositions in, on, and at was inspected in terms of the length of time indicated from nouns going after these prepositions. The use of at implies sameness of time in the word following it (e.g. at breakfast, at 10 o‟clock), in implies that there is a period for the time of the event (e.g. in summer, in the 19th century), while on implies that there must be a clear identification of time which is an indivisible and clearly defined part of a period (e.g. on Thursday, on the last day of each month). However, from the findings above, we discovered that all the elements, including the prepositions, figure and ground nouns, in constructing the prepositional expressions, must be comprehended as a whole and have tremendous influence on the sense and usage of the chosen preposition. Through understanding how these
elements can be fitted into the construction, the meaning of this expression can then be generated. Previous studies have also shown that the relative location and
locomotion of figure and ground are able to generate various senses in the semantic network (cf. Lakoff, 1987; Evans and Tyler, 2004), while this thesis shows how the meanings of a particular preposition vary according to its co-text. Although there have been studies focusing on the semantics of preposition by viewing different meanings of the preposition as generation from the primary sense instead of distinctive sense
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
106
categories, how the meanings vary according to different linguistic contexts is rarely explored in a systematic way. In fact, as shown in this work in a prepositional phrase formed by a particular meaning, there is a range of semantic preferences in their co-text. In other words, the meaning of a preposition could be coerced by the lexical words it is surrounded by (cf. Pustejovsky, 1995), but not merely the word that goes after that preposition. The co-occurrence of the preposition and the figure and ground nouns thus form a semantic unit, which cannot be comprehended without viewing them as a whole. To summarize what we have emphasized so far, in the sentence I am on the stage with I as the figure and stage as the ground, it can be interpreted both
with its locational meaning as well as the extended ‗an involvement in an activity‘
meaning. However, in the sentence the chair is on the stage with the chair as the figure and a similar ground, the chair may be seen as the property that is put on the stage. There are many instances like this pair that are similar in construction but differ in meaning(s). Our study on the semantic features of the figure and ground nouns thus contributes to finding out the differences of the pair(s). One cannot comprehend these sentences without considering the features of the figure and ground nouns. Figures 5.2 and 5.3 summarize our findings from the analysis of BNC, displaying the semantic preferences of the nouns that function as the figure and ground in the prepositional construction and how the senses of in and on might be influenced by these nouns.
‧
Figure 5.2 The Diagram of Figure and Ground Nouns in the co-text of Preposition in in Influencing the Senses of Prepositions
Figure 5.2 shows how the distribution of semantic features might differ between literal and metaphorical constructions of in, and how these features can be further distinguished among different semantic meanings under the category of metaphorical construction (The literal construction usually has one meaning—a ‗location‘ meaning).
The features listed in the grid of ‗metaphorical construction‘ are the shared features identified in these metaphorical semantic meanings, and the grids below are the extended metaphorical semantic meanings with their own unique prominent semantic
Ground
‧
features. However, owing to the distribution of data in the senses, not all the semantic meanings can be identified with their prominent semantic features. Therefore, those shown in Figure 5.2 only include the meanings with higher distribution in the sense analysis. This also shows how each sense of in is distributed in BNC. The diagram for the preposition on is also shown in Figure 5.3.
Figure 5.3 The Diagram of Figure and Ground Nouns in the co-text of Preposition on Influencing the Senses of Prepositions
ON
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
109
Similarly in Figure 5.3, the diagram only shows the partial representation of the semantic meanings of on since the shortage of corpora data identified for all the senses. For the features in the overall metaphorical constructions, there is no shared feature identified in the metaphorical semantic meanings, which indicates that the types of nouns constructing the metaphorical expressions possess greater diversity than those appeared in the metaphorical expressions of in.
From Figures 5.2 and 5.3 we can observe that it is the varied nouns with different features that decide which semantic meanings should the construction belongs to.
They show how the sense of a preposition would be greatly influenced by different sets of semantic preference that restrict particular types of figure and ground. The interaction among the figure and ground noun phrases and the sense(s) of prepositions in and on observed in the native speaker data may be represented through the model shown in Figure 5.4.
Figure 5.4 The Model of the Interaction among the Figure and Ground Nouns and the Prepositions In and On
This model displays the interaction among the noun phrases and the preposition, as Senses of in/on
Noun Phrase 1 Figure
Noun Phrase 2 Ground
in / on
Prepositional Phrase
Literal Construction Metaphorical Construction
(a) (a)
(b) (b)
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
110
well as the role of the noun phrases in this construction. In sense construction of a prepositional phrase, different types of figure and ground nouns may have some influences on the senses of the preposition, generating either literal or metaphorical meanings. As the production of metaphorical expressions is more varied in sense, the constructions are shown in dotted lines (b) in this model. This indicates that a great variety of nouns may result in different metaphorical meanings of a particular preposition. Moreover, as the figure and ground are important elements in the construction of prepositional phrases, the change of the figure or ground nouns may also shift the sense denoted in this construction. For example, in the clause she is in medicine with she as the figure and medicine as the ground, in refers to the activity done by the figure. When the figure is changed to some other nouns with inanimate feature, such as statistics, careers, etc., these nouns may affect the whole construction to become a different interpretation. In this way, the ground medicine in statistics in medicine refers to a particular field of study. The ground medicine is taken as working on something in careers in medicine. Thus, the establishment of meaning goes beyond the locomotion or the relative location of figure and ground, and this can be
compensated through the semantic preferences of a particular sense and how the preferred nouns in this sense work to generate the meaning under a certain structure.
This model reflects some of the concept of coercion proposed by Pustejovsky (1995), who claimed that nouns surrounding the preposition may also influence the
prepositional senses denoted in a particular sentence. In this study, the nouns are further identified according to their semantic features for discussing how they may appear in forming constructions of different senses, and we also focus on the interaction between the figure and ground nouns in contributing to the alteration of senses. The above are some of the important contributions of this thesis.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
111
Another finding of this work is that the selection of the figure and ground nouns seems to be bounded by the preposition, and there is no strict relation between these two nouns. Therefore, the interaction shown on this model reveals that the influence does not merely concern how the figure and ground nouns pose changes in meaning, but also how the preposition decides the types of noun it should go with in that expressions with particular senses. For example, we hypothesized that most of the literal expressions of in should contain the feature ―figure being smaller than the ground (F<G)‖ since in the image schema theory reviewed in Chapter 2, the
proto-scene is pictured as the ground being the container where the figure is placed in.
Therefore, in the phrase the boy is in the house, where the ground noun the house is a concrete noun that denotes a location, in which the figure noun the boy situates, the figure boy must be smaller than the ground house, and the opposite case is not preferred in the construction of the locational expression of in. In this way, the relative location or locomotion are revealed from the semantic features used to describe them. However, if we replace in with on, the relative size of the boy and the house in the boy is on the house is not restricted as long as the boy is in contact with and is supported by the top or a part of the house. In this way, the contact between the figure and the ground becomes an essential feature in this sentence.
The discussion so far has focused on native data; we now examine the data
produced by L2 learners. As the sense is greatly influenced by the words in the co-text, the types of the nouns used by the learners were also analyzed. The findings indicate that in the learner data, the range of semantic preference tends to restrict to specific semantic features in a particular set of data in both literal and metaphorical
constructions. For example, learners used in as a locational preposition to refer to a region or location, in which the figure tends to be human beings who stay in a
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
112
particular place that acts as the ground. In metaphorical instances, the restricted set of semantic preferences seems more obvious, as fewer instances were found to be activity nouns (e.g. battle, driving) in the grounds of in-constructions, which is a prominent characteristic in native speaker data. Learners‘ metaphorical usages of in were constructed more frequently with body part (e.g. mouth, heart) and measure and temporal nouns (e.g. era, minute), implying that learners may learn prepositional phrases by the semantic fields these nouns belong to. In this way, learners‘ production of other senses of the preposition in is thus limited. Therefore, in language teaching, the teacher could attend to the variety of the elements in the prepositional
constructions, which will be discussed further in Section 6.2.
For the case of the preposition on, learners‘ problem in using this preposition is slightly different from that of in. The above-mentioned findings indicate that the sense of a particular preposition is greatly influenced by its co-text and we found that
learners‘ production of nouns is restricted to nouns of certain semantic fields. When acquiring the usages of English prepositions, learners tend to perceive the
prepositional phrases as multi-word units or even fixed expressions and use the prepositions particularly with a set of words. In this way, we may find a lot of set usages in the learner data of on (e.g. spend some time on doing something, keep an eye on something) which may result in the restrictive types of nouns in producing prepositional expressions of on since these nouns tend to come from the same semantic field. In language teaching, different methods should be employed in inspiring productions of different combinations.
Investigating into the errors found in all of the learner data, the current researcher observed that most of the errors occur with concrete nouns, especially locations (e.g.
*on [in] 7-eleven, *on [at] the bus stop). If we look into the corpora, we may not
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
113
know whether this is due to learners‘ avoidance or learners‘ good command in prepositional expressions other than locational ones. This is a question which may be answered in other elicitation experiment, but the above-mentioned points still remind language teachers that different facets of attention toward the complex meanings and usages of English prepositions should not be overlooked.
Compared with the previous study conducted by De Vega et al. (2002) that utilized the semantic features to analyze the semantics of prepositions, the findings of this thesis proceeded to the exploration of metaphorical constructions. In De Vega et al.‘s study, they claimed that the integration hypothesis can be used to explain the locative construction of prepositions. This hypothesis posited that a
multiple-constraint process should be activated to comprehend locative senses, through which the sensory-motor features of the objects are retrieved and fitted into a simulation that facilitates comprehension. However, for constructions beyond the locative ones, the present study found that the integration hypothesis can solely be used to ―describe‖ the features of figures and grounds instead of precisely predicting which preposition or sense may be used with particular figures and grounds in a unique way. Therefore, the model in Figure 5.4 is presented above in order to provide a more comprehensive explanation for both literal and metaphorical constructions of English prepositions.
5.3 Summary of the Chapter
To sum up, in this study the feature of figures and grounds do not show such consistent differences between in and on, since the contrast goes beyond the
locational level to more metaphorical ones that are diverse in meaning. This research also presents possible bi-directional interactions of all the important elements (figure,
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
114
ground, and the preposition) in constructing a meaningful prepositional phrase, through which the words appearing in the co-text are found to have influenced meaning change to both the target preposition and the semantic preferences in deciding the meaningful words in forming a prepositional phrase.
In Chapter 6, the main points of this thesis will be summarized, and limitations of this study and suggestions for future study will also be addressed. Pedagogical suggestions will also be discussed so as to provide language teachers and learners with a different direction regarding the teaching and learning of English prepositions.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
115
CHAPTER 6
CONCLUSION, LIMITATION, AND PEDAGOGICAL SUGGESTIONS
This chapter concludes with a short summary of this thesis (6.1), followed by some pedagogical suggestions on teaching and learning English prepositions (6.2).
Limitations and suggestions for future related studies are addressed in (6.3).