Metaphor And Metonymy In Comparison And Contras...
The book elaborates one of Roman Jakobson's many brilliant ideas, i.e. his insight that the two cognitive strategies of the metaphoric and the metonymic are the end-points on a continuum of conceptualization processes. This elaboration is achieved on the background of Lakoff and Johnson's twodomain approach, i.e. the mapping of a source onto a target domain of conceptualization. Further approaches dwell on different stretches of this metaphor-metonymy continuum. Still other papers probe into the specialized conceptual division of labor associated with both modes of thought. Two new breakthroughs in the cognitive linguistics approach to metaphor and metonymy have recently been developed: one is the three-domain approach, which concentrates on the new blends that become possible after the integration or the blending of source and target domain elements; the other is the approach in terms of primary scenes and subscenes which often determine the way source and target domains interact.
Metaphor and Metonymy in Comparison and Contras...
Metaphor (drawing a similarity between two things) and metonymy (drawing a contiguity between two things) are two fundamental opposite poles along which a discourse with human language is developed.[1] It has been argued that the two poles of similarity and contiguity are fundamental ones along which the human mind is structured; in the study of human language the two poles have been called metaphor and metonymy, while in the study of the unconscious they have been called condensation and displacement.[2] In linguistics, they are connected to the paradigmatic and syntagmatic poles.[3]
The couple metaphor-metonymy had a prominent role in the renewal of the field of rhetoric in the 1960s. In his 1956 essay, "The Metaphoric and Metonymic Poles", Roman Jakobson describes the couple as representing the possibilities of linguistic selection (metaphor) and combination (metonymy); Jakobson's work became important for such French structuralists as Claude Lévi-Strauss and Roland Barthes.[3] In his essay, Jakobson also argues that metaphor is the basis for poetry, especially as seen in literary Romanticism and Symbolism, whereas metonymy forms the basis for Realism in literature.[4]
In 1957, Jacques Lacan, inspired by an article by linguist Roman Jakobson, argued that the unconscious has the same structure of a language, and that condensation and displacement are equivalent to the poetic functions of metaphor and metonymy.[1][2][12]
Dirven, R. (2002). Metonymy and metaphor: Different mental strategies of conceptualisation. In René Dirven and Ralph Pörings (Ed.) Metaphor and metonymy in comparison and contrast, (pp.75-111). Berlin: Mouton de Gruyter.
Nerlich, B. (2011). The role of metaphor scenarios in disease management discourses: Foot and mouth disease and avian influenza. Windows to the mind: Metaphor, metonymy and conceptual blending, 115-142.
Nerlich, B. (2011). The role of metaphor scenarios in disease management discourses: Foot and mouth disease and avian influenza. In: Handl, S. and Schmid, H.-J. (eds.). Windows to the Mind: Metaphor, metonymy and conceptual blending (pp. 115-142). De Gruyter Mouton.
Many experimental studies of both types of polysemy demonstrated that because of their greater relatedness to the literal sense, metonymy-based senses exhibited evidence of being stored together with the literal sense in the mental lexicon, in contrast to less related metaphors that were stored separately (Klepousniotou and Baum, 2007; MacGregor et al., 2015; Lopukhina et al., 2018). However, the existing experimental studies on polysemy do not reflect the complexity of this phenomenon in its entirety. First, they mainly concentrate on nouns and sometimes on verbs, while leaving aside adjectival polysemy, which, with the few exceptions (Lopukhina et al., 2018; Weiland-Breckle and Schumacher, 2018) has not received experimental attention. Yet adjectives are an important and linguistically well-represented category among content words, with unique patterns of polysemy. Second, previous studies consider only coarse-grained distinctions between metonymy and metaphor. However, each of these two semantic extensions has more fine-grained subtypes which might conceivably be stored differently in the mental lexicon. Our research aims at partially bridging this gap.
Klepousniotou and Baum (2007), employing a timed lexical decision task, found that metonymically polysemous words (such as bottle for container or its content) were processed faster than homonyms, whereas metaphorically polysemous words (such as star for celestial body or person) did not show such an advantage. Similarly, Klepousniotou et al. (2008), showed that in a timed sensicality judgment task, words with highly overlapping senses (metonymy) demonstrated reduced effects of context and sense dominance as compared with words with moderately or low overlapping senses (metaphorical polysemy and homonymy). The authors suggested that the comprehension of ambiguous words was mediated by the semantic overlap of alternative senses.
Contrary to the majority of results and based on two lexical decision and semantic categorization experiments, Jager and Cleland (2015) found that metonymy-based senses were stored in separate representations, whereas metaphorical senses were stored together with the literal sense. The authors explained these findings by suggesting that the particular instances of metonymy and metaphor, namely, animal-product metonymy (as in She wore her silver fox) and animal-person metaphor (as in He is a sly fox) represented a counterexample to the generally accepted claim that metonymy is more sense-related to the literal sense than the metaphor. Therefore, the result by Jager and Cleland (2015) may be due to the bias in the stimuli and may indicate that the pattern of sense storage depends on semantic properties of words.
We suggest the following hypotheses regarding the mental representation of different senses in polysemous adjectives: (1) proximal metonymy should have a greater overlap with the literal sense than distal metonymy; (2) distal metonymy should have a greater overlap with the literal sense than metaphor; (3) proximal and distal metonymy should have a greater overlap with each other than any type of metonymy with metaphor. In order to estimate which types of senses in polysemous adjectives overlap and which do not, we apply a semantic clustering paradigm suggested in Lopukhina et al. (2018). We ask participants to cluster short contexts expressing literal, proximal metonymic, distal metonymic, or metaphorical senses of an adjective together into virtual baskets, with an unlimited potential number of categorization groups. We measure how well the participants group together sentences with the same sense and what misclassifications they might have. The semantic clustering approach with multiple options solves both the problem of forced choice and the problem of restricted context, which was typical for previously used categorization paradigms (Klein and Murphy, 2002; Jager and Cleland, 2015). We assume that the offline clustering paradigm implies deep semantic processing of word senses in context because the participants can spend as much time as they need analyzing subtle differences in sense relatedness and deciding about each context. For example, using the offline categorization task Klein and Murphy (2002) showed that the participants sometimes group together different senses of one word (wrapping paper and liberal paper) that was not detected in the online lexical decision task (Klein and Murphy, 2002). Therefore, we suppose that the clustering task with multiple options provides favorable conditions to investigate semantic representations of polysemous words. In the present study, we expect participants to confuse literal senses with proximal metonymic senses more often than with distal metonymic senses. Distal metonymic senses will be confused with literal senses more often than metaphors. Proximal and distal metonymies will be confused with each other more than with metaphor.
We found that in Type 1, literal senses were placed in separate baskets significantly more often than either of the two proximal metonymies (p
Second, we analyzed the misclassifications made by the participants, see Table 3 for the proportions of incorrect classifications within each Type and Table 4 for the Adjusted Rand Index information. We found that most often the participants incorrectly placed two proximal metonymies in one basket in Type 1 (ARI = 0.292, lower ARI indicates worse discrimination of senses). Least often the participants misgrouped literal senses and metaphors in Type 3 (ARI = 0.835, higher ARI indicates better discrimination of senses) and Type 4 (ARI = 0.812). The participants also rarely confused metonymies and metaphors: ARI for proximal metonymy and metaphor was 0.636 and for distal metonymy and metaphor 0.671, this difference was significant (p = 0.008). Finally, for the Type 2, where we could compare the misclassifications of literal senses, proximal metonymies and distal metonymies, we found that literal senses were grouped with proximal metonymies more often than with distal metonymies (ARIs were 0.447 and 0.574, respectively, p 041b061a72