Nội dung

OVER AGAIN: POTENTIAL NOVEL PERSPECTIVES FROM LEXICAL CONCEPTS & COGNITIVE MODELS THEORY Do Tuan Long* Department of Language Training and Professional Development, VNU University of Languages and International Studies, Pham Van Dong, Cau Giay, Hanoi, Vietnam Received 31 May 2018 Revised 27 July 2018; Accepted 31 July 2018 Abstract: This article aims at analyzing the approaches by different authors to the English preposition over, showing their strengths and weaknesses. We then try to apply the theoretical constraints of Lexical concepts & Cognitive Models (hereafter: LCCM) to treat over from a novel perspective. Seth (2009) can describe the pragmatics of over in stances of use, but he fails to present the distinct senses of the word. Though Lakoff’s framework of Full-Specification could figure out the distinct senses of over, the vast proliferation of senses and a lack of methodological constraints make the approach inappropriate in certain cases. Other works by Kreitser (1997), Tyler & Evans (2003), Deane (2005) and Maria Brenda (2014) could, to certain extent, fill the gaps of Lakoff, but they have failed to address the issue of combining both the linguistic (parametric) and the non-linguistic (analogue) representations in analyzing the semantics of over. It is hypothetically proved that as a linguistic vehicle, the preposition over encodes 17 lexical concepts with unique semantic and formal selectional tendencies. Keywords: over, distinct, LCCM, lexical representation 1. Introduction: The challenge of over 1 English prepositions were once neglected and linguists never seemed to take them seriously (Jackendoff, 1983: 345) but prepositions turn out to be appealing to cognitive linguists. Perhaps the English preposition over is the most special one as it has different syntactic functions and has received a great deal of attention from numerous researchers (Brugman, 1981; Boers, 1996; Deane, 2005; Dewell, 1994; Kreitzer, 1997; Lakoff, 1987: 416–461; Tyler and Evans, 2003:64-106; Yoon, 2004; Set, 2009; Roussel, 2013; Maria Brenda, 2014). In general, the approaches of above-mentioned authors could be classified into three main * Tel.: 84-985227867 Email: tuanlongcfl@gmail.com trends: The descriptivist approach (Seth, 2009); the Cognitive spatial division approach (Roussel, 2013) and the Cognitive semantics image-schemas approach (Brugman, 1981; Boers, 1996; Deane, 2005; Dewell, 1994; Kreitzer, 1997; Lakoff, 1987: 416–461; Tyler and Evans, 2003:64-106; Maria Brenda, 2014; Yoon, 2004). 1.1. The descriptivist approach The preposition over, together with above, is introduced and analyzed in each instance of use by Seth Lindstromberg through the book English Prepositions Explained published in 2010. From the very beginning of chapter 9, Seth (2010:109) made a reference to Coventry et al (2008) to suppose that above has no or little functional meaning so its usages are far less varied as those of over. 84 D.T. Long / VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 Seth Lindstromberg shows the usage of over and above in great details, and the difference between over and above was made clearly in each case of use from spatial senses to non-spatial ones. When denoting spatial meanings, over is neutral about geometrical separation while in every sense of above, the subject and landmark are separated. The pragmatics of over and above were shown through three levels: basic spatial meanings, variations from basic applications, and additional metaphorical usages. In chapter 12, Seth Lindstromberg clarifies the usage of under and below. The preposition under is the approximate opposite of over while below is the opposite of above. Here are the four basic spatial configurations that the four prepositions denote according to Seth Lindstromberg. Figure 1. Spatial configuration1 denoted by above (Seth, 2010: 110) Figure 2. Spatial configuration denoted by over (Seth, 2010: 110) Figure 3. Spatial configurations denoted by under & below (Seth, 2010: 158) What Seth was trying to do is to describe the real usages of the four prepositions and distinguish between two pairs of synonymous prepositions overabove, under-below. The approach by Seth 1 1 The dark sphere represents the Subject (or Trajectory), the rectangle represents the Landmark. is similar to that of a dictionary with indepth explanation and examples, which is of importance to learners of English. However, Seth fails to explain how extended senses of the four prepositions arise and the relationships between those senses. VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 1.2. The cognitive spatial division approach This approach is advocated by Roussel (2013) when he analyzes the structure A over B. Particularly, he uses the componential analysis of the inanimate and animate referents of B with the total number of 196 instances of use. The main findings of the study show that the use of over in the structure A over B, as written by Roussel, is: Over exhibits properties combining visual limits and motion, which seems to be a sign for perceptive calculations governing the linguistic establishment of spatial relations. The preposition appears to prioritize the expression of the observer’s visual assessment of their own relation to the surrounding entities. Therefore, the geometrical patterns and the various degrees of measure, place, position, height, passage or judgment may not reveal the meaning of the preposition, or the interpretation of the speaker, but they may convey the final stage of perceptive calculations which, subsequently, operate aspectually or pragmatically in the utterance…. More generally, the linguistic analysis demonstrates that the system of English prepositions seems not only to encode space relations egocentrically [DEANE 2005], but also allocentrically (by reference to the external world), as shown both by the body parts and the semantic features selected in the referents of A and B. (Roussel, 2013: 223) Although Roussel realizes the tendency of space limits of over, he fails to explain if the preposition in some cases carries or encodes distinct senses. The author is concerned with the “perception in the brain prior to the encoding of the extra linguistic world in new schematic forms”, but Roussel does not analyze the speaker’s meaning in each instance of use, or at least to generalize the speaker’s meaning when using the preposition over. 85 1.3. The cognitive semantics image-schemas approach This is the approach that a number of scholars make use of when analyzing the semantics of over. In fact, Lakoff with his case study of over laid a foundation for other discussions (Brugman, 1981; Boers, 1996; Kreitzer, 1997; Tyler and Evans, 2001, 2003; Yoon, 2004; Deane, 2005; Dewell, 1994; Maria Brenda, 2014). The first two approaches, namely, Full-specification Approach and Partial-specification Approach, were critically analyzed by Tyler and Evans (2001) who then developed an approach termed Principled Polysemy. Maria Brenda (2014) exploited Langacker’s model of cognitive grammar (2000) to analyze the semantic and syntactic structure of over. The approach by Maria Brenda, to certain extent, can be described as Extended Principled Polysemy. The notion of image schema was developed by Johnson (1987) and Lakoff (1987), and in their opinion, physical domain is the most significant level of human interaction. The two authors suppose that conceptual representations in human mind are arising from embodied experience. The definition of image schemas, proposed by Gibbs and Colston (1995: 347), are experiential gestalts and they come from different sensorimotor of human beings. Evans and Green (2006: 180190) synthesized that image schemas possess eight properties: (i) Image schemas are preconceptual in origin; (ii) An image schema can give rise to more specific concepts; (iii) Image schemas derive from interaction with and observation of the world; (iv) Image schemas are inherently meaningful; (v) Image schemas are analogue representations; (vi) Image schemas are not the same as mental images; (vii) Image schemas are subject to transformations and; (viii) Image schemas can occur in clusters. In short, Lakoff supposes 86 D.T. Long / VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 that words are seen as radial categories, and the following figure represents the diagram to illustrate the radial categories. Figure 4. Modelling the radial categories (Evans & Green, 2006: 332) with over are structured with respect to this image-schema which provides the category with its prototype structure. In sum, Lakoff claims that the schemas which are different from the central schema are considered to represent distinct senses associated with over. According to this model of word meaning, the central schema for over has at least six distinct and closely related variants (see Figure 5), each of which is stored in semantic memory. The following part represents the above approaches with critical comments. 1.3.1. Full-specification approach Lakoff took over as a case study in English prepositions and his analysis is sometimes described as the full-specification approach to lexical semantics. The core point in Lakovian theory is that the senses associated with prepositions like over, which are grounded in spatial experience, are structured in terms of image-schemas. According to Lakoff, an image schema combining elements of both ABOVE and ACROSS is the prototypical sense of over. The distinct senses associated Figure 5. Central image schema (adopted from Lakoff, 1987: 423) Beside the ABOVE-ACROSS sense, over denotes a number of other senses summarized in Table 1. As can be seen, this model results in a potentially vast proliferation of senses for each lexical item. Table 1. Schemas proposed by Lakoff (1987) for over besides the central schema (Evans & Green, 2006: 337) Schema type ABOVE schema Basic meaning The TR is located above the LM. COVERING schema The TR is covering the LM Examples The helicopter is hovering over the hill. The board is over the hole. The TR is reflexive: the TR is simultaneously the REFLEXIVE schema TR and the LM. The final location of the TR is The fence fell over. understood with respect to its starting position EXCESS schema REPETITION schema When over is employed as a prefix it can indicate The bath overflowed. ‘excess’ of TR relative to LM After receiving a poor grade, Over is used as an adverb to indicate a process the student started the that is repeated. assignment over (again). VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 Consequently, over has, at the very least, several dozen distinct senses. Tyler and Evans (2001, 2003) show two main problems of Lakoff: (1) a failure to distinguish between polysemy and vagueness, (2) unconstrained methodology. They indicated that Lakoff denied the role of context in meaning altogether. Particularly, Tyler and Evans (2003) argue that the examples in (1) do not represent distinct senses of over (one specifying contact and one specifying lack of contact): (1a). The bird flew over the wall. (b). Sam climbed over the wall. Instead, Tyler and Evans suppose that the interpretation of over with respect to contact or lack of contact derives from the integration of over with the other elements in the sentence. Our knowledge about birds (they can fly) and people (they cannot) provides us with the inference that birds do not come into contact with walls when crossing over them while people do. In other words, the linguistic context together with encyclopedic knowledge provides the details relating to the presence or absence of contact. According to Tyler and Evans, over here is vague with respect to contact. Tyler and Evans argue that while Lakovian position on polysemy as a conceptual phenomenon is correct, it is also important to take into account the crucial role of context in word meaning. Lakovian approach has also been blamed for a lack of methodological constraints. In other words, Lakoff provides no principled criteria for determining what counts as a distinct sense. This means that the polysemy account presented for over (or whatever lexical item we might apply the approach to) results purely from the intuitions (and perhaps also the imagination) of the analyst rather than actually representing the way a particular category is represented in the mind of the language user. This problem has been discussed in some detail 87 by Sandra and Rice (1995) and by Sandra (1998) [cited in Evans, 2006: 342]. 1.3.2. Partial-specification approach Kreitzer (1997) made use of the works by Lakoff (1987) and Talmy (1983) to modify Lakovian framework to semantically analyze over. Kreitzer posits that there are three distinct levels of schematization inherent in the conceptualization of a spatial scene: the component level, the relational level, and the integrative level. Evans (2001) criticized Kreitzer for failing to decide which sense is the primary sense of over and neglecting senses presented by Lakoff. Hence, there is a lack of how senses of over are distinguished. 1.3.3. Principled Polysemy The framework’s aim is to analyze the meanings of certain English prepositions and present them in semantic networks. In fact, the framework is built upon works by Lakoff and Claudia Brugman as part of cognitive lexical semantics and provides a theoretical constraint how a sense is counted as distinct. Founders of the framework provided two criteria: (1) for a sense to count as distinct, it must involve a meaning that is not purely spatial in nature, and/or a spatial configuration holding between the TR and LM that is distinct from the other senses conventionally associated with that preposition; and (2) there must also be instances of the sense that are contextindependent: instances in which the distinct sense could not be inferred from another sense and the context in which it occurs. Tyler and Evans (2003:64-106) took over as a case study to shed light on the analysis of other prepositions. They provided a semantic network for over with one central meaning and fifteen extended meanings. The framework Principled Polysemy could successfully explain “how new 88 D.T. Long / VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 meanings develop from established ones on the basis of experiential correlations” (Thora, 2004). However, to the best of our knowledge, Tyler and Evans’ network may not help us trace the meaning of over in certain cases. Consider over in the following sentence, Normally, over may denote a spatialphysical configuration between a LM and a TR in the above sentence; but in fact there is no such configuration. So in this case, over denotes a non-spatial meaning. In order to understand the non-spatial meaning of over, we consulted several dictionaries2 ; however, the consultation yields no answers. The semantic network for over proposed by Tyler and Evans did not give us any clues to trace the meaning of over in this case. Yet, reading through the article, we can see that it is the joke that makes the ambassador himself in trouble. Tyler & Evans may argue that the use of over in (2) is constructed online and perhaps the prototypical spatial function of over can help readers guess the meaning of over joke. This is true to certain extent; however, we do not suppose that the polysemous use of over is sufficient in the following figure. (2). British Ambassador in hot water over joke. 1 The above example is a headline on BBC, and there is no verb. The complete sentence, as understood by readers, is “British Ambassador is in hot water over joke”. We can analyze the structure of the sentence like in the following table. “Over joke” is treated as something new because it needs analyzing to understand the meaning of the whole sentence. Table 2. Analysis of “British Ambassador in hot water over joke” British Ambassador is in hot water Subject Verb Adverbial over joke. Something new Figure 6. The semantic network for over (Tyler and Evans, 2003:80) Yoon (2004) reanalyzed the semantics of over in the light of Principled Polysemy and proposed a revised semantic network for the preposition as follows: 1 2 Oxford Dictionary Online, Cambridge Dictionary Online, Merriam Webster Dictionary. VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 89 Figure 7. Revised semantic network for over by So Yeon Yoon (2004) Additionally, recent research studies show that polysemy is not as simple as it is assumed, so Evans (2014) himself admitted errors in analyzing the phenomenon of polysemy: In the final analysis, what this reveals is that polysemy is a complex and multifaceted phenomenon. It is probably overly simplistic to assume, as has sometimes been done (e.g., Tyler and Evans, 2001, 2003) that discussions of polysemy boil down to the polemic of monosemy on the one hand, versus the multiple distinct sense-units of the principled polysemy approach that I espoused with Andrea Tyler in our 2003 book. This bifurcation is too neat, and consequently ignores some of the very phenomena that I have been addressing in this paper. While polysemy as viewed through the eyes of Charles Ruhl (1989) is surely empirically flawed, it is fair to say that the view of polysemy developed in Tyler and Evans (2003) is probably also too simplistic. There we argued for neat semantic networks, where word-senses constituted clearly demarcated, discrete nodes within a radiating lattice of semantic memory, which we thought, ultimately, would be locatable in the brain. (Evans, 2014:122) Hence, Principled Polysemy seems to be a rather simplistic view of over. 1.3.4. Extended Principled Polysemy Inspired by Langacker’s work in 2000 and Tyler and Evans’ work in 2003, Maria Brenda wrote a monograph, published in 2014, on the cognitive perspective of the polysemy of over. We term the framework that Maria Brenda exploited Extended Principled Polysemy. Besides presenting what Tyler and Evans (2003) had done, Maria added new meanings of over by analyzing the syntactic structures of her corpus while taking context into consideration. She also succeeded in presenting the historic use of preposition over over the centuries. However, as Evans admitted (2015), the starting point of Maria Brenda’s work is rather, in Evans’ term, “simplistic”. In addition, the data collected were from various dictionaries, so it is de facto that usage-based model was not exploited. 1.3.5. Other works Boers (1996) made a reference to Lakovian notion of image-schema to treat over in different metaphors. In general, what was semantically found by Boers was “grateful” to results of Lakoff and Brugman. However, Dewell (1994) proposed the six image-schema transformations more seriously to address over in each distinct sense. As a result, how extended senses of over arose from the prototypical sense was rigorously explained. Deane (2005) proposed a multimodality spatial representation to analyze over. He proved that the polysemy of over derives from the prototypical sense on the basis of preference rules. 90 D.T. Long / VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 Personally, we see that the works presented by above-mentioned scholars are of great significance and have their own values. In fact, The case of over by Lakoff laid a foundation for other theoretical framework to be built upon. However, there are two precautions that must be taken when treating English prepositions. 1. Linguists have proved that prepositions carry meanings, or in other words, encode certain senses. It is true that spatial configuration denoted by prepositions facilitates meanings in utterances. Non-spatial meanings are originally rooted from spatial ones through different mapping processes. 2. Evans showed that three prepositions in, at, on denote three types of polysemy: conceptual polysemy, lexical polysemy and inter-lexical polysemy. Then, he successfully proved the point of view that the polysemy of the three prepositions emerges from different sources and in different ways. Evans supposes that an account of the nature of semantic structure - a representational format unique to language and conceptual structure - a representational format that is wholly non-linguistic in nature is needed (Evans, 2014:122). Additionally, in reference to what Evans (2014) advocated, the works by Lakoff (1987), Brugman (1988), Kreitzer (1997), Tyler and Evans (2001; 2003), Maria Brenda (2014) and others have failed to address the issue of combining both the linguistic (parametric) and the non-linguistic (analogue) representations in analyzing the semantics of over. Hence, a more comprehensive framework would provide a more proper perspective for the polysemy of over. 2. A reanalysis of OVER Inspired by Evans’ works (2005, 2009, 2014), we suppose that frameworks that analyze the semantics of over must follow the premises and assumptions below (Navarro, 1998; Evans, 2009): 1. A prepositional vehicle3 is always meaningful, and always contributes to meaning construction no matter what syntactic construction where it occurs. 1 2. The meanings of all the uses of a preposition should be explained by virtue of a single coherent semantic structure. 3. The semantic structure should represent the polysemy4 of a preposition with primary cognitive model and secondary cognitive model. 2 4. All the senses of the semantic structure should be linked with no gap in the chain. 5. Metaphorical and abstract uses should be derivable from senses based on bodily experience. 6. The semantic structure should make apparent the mechanisms and patterns of meaning elaboration and extension. These mechanisms and patterns should explain how the semantic category extends and how it could possibly extend in the future, but they do not predict the exact way in which this will happen, or if it will happen at all. 7. The interface between language, communication, and cognition constitutes the process of meaning construction which is influenced by usage. 8. The meaning of the prepositions under supervision is attached to a distinct syntactic and semantic structure. In reference to the presented hypothesis together with the above-analysis of previous approaches to the English preposition over, we come to the proposal that the theory of LCCM is the most potential to account for the 3 4 Evans’ term, in his book entitled “How Words Mean” published by Oxford University Press in 2009. Basically, there are two contrasting approaches to the semantics of prepositions: polysemy and monosemy. Recent studies have shown the appropriateness of the polysemy approach, especially in the light of cognitive linguistics (Tyler and Evans, 2003). VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 semantics and meaning transference of over. In the following part, we will first explain why LCCM is chosen as the theoretical framework, and then introduce how LCCM works, i.e. the semantics of over in the light of LCCM. 2.1. LCCM as the conceptual framework It is seen that a new framework which can provide both the lexical representation and semantic compositionality of over is needed. In other words, that framework has to deal with the preposition’s conceptual structure and semantic structure - two distinct concepts which are discussed by Evans (2013:15) as follows: Conceptual structure is a level of nonlinguistic representation that derives from sensory-motor, proprioceptive and subjective experience. Semantic structure is a language-specific level of representation encoded at the semantic pole associated with words and other multiword constructions. These two levels are modelled by the theoretical constructs that give the theory its name: the lexical concept and the cognitive model. (Evans, 2013: 15) The novelty of LCCM is that “lexical concepts – units of language-specific semantic structure – facilitate access to units of conceptual structure – cognitive models” (Evans, 2013: 15). Additionally, LCCM also provides a detailed explanation 91 for the linguistically instantiated processes of integration (Evans, 2013: 21). That is the reason why LCCM is made use of. As the name of the framework denotes, there are two important notions: lexical concepts and cognitive models. In Evan’s view, a lexical concept is part of the linguistic knowledge that conveys various types of highly schematic linguistic content. Specifically, linguistic content includes information relating to the selectional tendencies associated with a given lexical concept - the range of collocational and collostructional behaviour of a given lexical concept. Evans supposes that because the lexical concept of an open-class word gives access to numerous association areas within the concep­ tual system, it also guides to access to numerous cognitive models. A cognitive model profile of a lexical concept is the range of cognitive models to which it facilitates direct access, and the range of additional cognitive models to which it therefore facilitates indirect access. In fact, the framework is the result of Evans’ continuous works (Evans 2006, 2007, 2009a, 2009b, 2010a, 2010b) which were based on works of Langacker (2002), Goldberg (2006); but the difference lies in the fact that it provides a methodological framework for conducting semantic analysis of lexical concepts (Evans, 2013:25). Here is the architect of the framework: Figure 8. The architect of LCCM Theory (Evans, 2009: 76) 92 D.T. Long / VNU Journal of Foreign Studies, Vol.34, No.4 (2018) 83-103 According to Evans, there are two mechanisms of linguistically mediated usage events namely lexical concept selection and fusion in the semantic compositionality. Lexical concept selection serves to identify the most appropriate lexi­ cal concept associated with a given form during the processing of an utterance. Fusion is the integrative process and results in the construction of a conception. This is achieved by recourse to two sorts of knowledge: linguistic content and conceptual con­tent. Fusion is itself made up of two constituent processes: lexical concept integration and interpretation (see figure 9). Figure 9. Processes of semantic composition in LCCM (Evans, 2009: 219) One more issue addressed by LCCM is figurative language (metaphor and metonymy). Evans proves that literal meaning of an utterance is interpreted within the default or primary cognitive model profile while the non-literal meaning must be understood in the secondary cognitive model profile. The distinction between metaphor and metonymy is due to the emergence of alignment between what were termed figurative target and figurative vehicle. In case of metaphor, there is divergence between the two while in case of metonymy, there is alignment. 2.2. The methodology to account for the lexical representation of a linguistic unit in the light of LCCM In reference to LCCM, we can make a methodological procedure to identify the lexical concepts of the preposition over as follows. A lexical concept shows selectional tendencies made up of two types of information. The first kind, termed the lexical concept’s formal selectional tendencies, relates to the vehicle types that can encode the lexical concept. The second type concerns the semantic arguments that make up the argument-structure lexical concept: its semantic selectional tendencies. A lexical profile, in nature, refers to the two types of selectional tendency, and it is assumed to be unique for any given lexical concept. Hence, a given lexical concept will exhibit a range of formal and semantic selectional tendencies that, in principle, should be sufficient for identifying a distinct lexical concept. Look at the following sentences from Oxford Dictionary5: 1 (3). I went over and asked her name. (4). He’s gone over to the enemy. It is rather easy to schematize the spatial scenes denoted in sentences (3), (4) and there is a shift in movement towards the other side of something or somebody, which can be represented by the following figure: 5 We advocate the usage-based thesis of Langacker (2000) but due to the role of this paper as a proposal, the data were collected from various dictionaries namely Oxford, Cambridge, Collin COBUILD, etc. and some examples are ours which are reviewed by native speakers of English.