Data Retrieval I: Identification of Crossmodal Metaphors and Metonymies

We retrieve crossmodal metaphors and metonymies by undertaking an identification process of crossmodal mapping. Since crossmodal mapping involves the concepts of perception (from the senses of sight, hearing, touch, smell, or taste) as the target concepts, and an asymmetrical relation of metaphoric expression, we adopt the Crossmodal Metaphor Identification Procedure (CMMIP) adapted from Ullmann (1959), Williams (1976), Marks (1996), Group (2007), and Shen and Gadir (2009). With regards to the semantics of the target discourse, the Chinese WordNet (Huang & Hsieh, 2010 ; Huang et al., 2010) is employed to distinguish the perceptual modalities that are applied within the semantics network of certain lexicons.

Originated from a project sponsored by the National Science Council, Chinese WordNet (CWN) is a lexical knowledge base of detailed Chinese lexical-semantic analyses. Involving researches of lexical and cognitive semantics, CWN aims to serve as a fundamental reference for linguistic investigation and “an indispensable infrastructure in application to Chinese natural language processing and ontology engineering” (Huang & Hsieh, 2010; Huang et al., 2010). CWN is chosen for use due to its solid framework of lexical semantics and ontology.

There has long been a lack of improvements to Since Corpus. The foremost authorized corpus

of Modern Chinese, CWN offers a refreshed perspective on the construction of contemporary Chinese lexicons. Furthermore, CWN offers a network of words by visualizing the lexical-semantic relations in a net-like image (shown in Figure 3.3). Thus, the searched lexicons can be viewed as synchronic data representative of contemporary usage.

As we investigate the interactions of different perceptual modalities in terms of metonymies and metaphors using the data of coffee cupping practices, we identify the metonymies and metaphors based on the belief that crossmodal metaphors do not exist within conceptual metaphors only. They are not simply forms of multimodal metaphors, as Forceville suggested; they are a kind of creative metaphor based on cross-sensual interactions in language.

Unlike individual intuition, this method of identification offers a comparably strict and well-knit procedure for comprehending metaphorical expressions. Accordingly, the CMMIP requires the following steps:

1. Scan the complete text-discourse to gain a full understanding of its semantics.

2. Segment the whole text-discourse into lexical units.

3. Determine the contextual meaning of each lexical unit through Chinese WordNet (CWN), contemplate whether the text-discourse contains at least two lexical units possessing distinctive expressions or descriptions regarding the five basic perceptions (i.e., sight, hearing, touch, smell, and taste).

4. Determine whether the two (or more) lexical units refer to the same event or situation.

5. Determine whether the entailed distinctive perceptions follow the directional mapping from one perception in the target domain to the other perception(s) (different from that in the target domain) in the source domain.

6. If these two (or more) lexical units denoting the same event or situation can be comprehended through mutual comparison, then the expression can be considered a crossmodal metaphor.

As for the actual application, we take the following example (3.1) to illustrate how CMMIP is applied to analyze the crossmodal metaphorical expressions made during coffee cupping.

(3.1)

(濃縮咖啡)雖然說它有一點點微酸微苦，但後面的餘韻蠻長的，那喝起 來也有淡茶的口感，那它的酸則是有淡柑橘的酸，亮亮的酸這樣子，那 也有一些焦糖的味道，整體的味道則是比較偏向慢慢地展開。

“Although it (this espresso) is a little bit sour and bitter, its aftertaste stays quite long, and it has the mouthfeel of weak tea. Its acidity is that of tangerine, and it is a bright acidity. There is also the flavor of caramel. The overall flavor tends to unfold slowly in the mouth.”

The steps are as follows:

1. Reading the two lines, we understand that the overall meaning concerns a speaker describing the flavor emerging from tasting a first cup of espresso.

2. We segment the lines according to the units of lexicons by inserting slashes between each unit as follows:

“（濃縮咖啡）雖然說/它/有/一點點/微/酸/微/苦/，/但/後面 的/餘韻/蠻/長的/，/那/喝起來/也/有/淡茶的/口感/，/那/它 的/酸/則/是/有/淡/柑橘的/酸/，/亮亮的/酸/這樣子/，/那/也 /有/一些/焦糖的/味道/，/整體的/味道/則/是/比較/偏向/慢慢 地/展開/。”

3. We examine the contextual meaning of each lexical unit using CWN’s visualization of word meanings. For instance, the network image of liang (亮) is shown in Figure 3.3. Based on CWN, we consider that

the text-discourse contains two lexical units (i.e., 亮亮的 and 酸), possessing distinctive expressions featuring the two basic perceptions of sight and taste.

4. We determine that these two lexical units, “bright” (亮亮的) and “acidity”

(酸), refer to the same event of coffee tasting.

5. The directional mapping of the entailed distinctive perceptions from the semantics network of the two lexical units of “bright” (亮亮的) and

“acidity” (酸) extends from taste, as the target domain, to sight, as the source domain.

6. We analyze two lexical/phrasal units containing crossmodally metaphorical similarities. In the lines, “…那它的酸則是有淡柑橘的酸，亮亮的 酸這樣子，那也有一些焦糖的味道…” (…Its acidity is that of

caramel…), the crossmodal metaphor of ACIDITY IS LIGHT can be identified.

Figure 3.3 The Chinese WordNet of Liang

In the Chinese data, we underline the letters to indicate the presence of crossmodally metaphorical similarities, since the italics form does not seem clear in Chinese.