A diachronic analysis of the FIRE character

3.2.1 Step 1: The semantic structure of the FIRE character

After reorganizing the data, we identified 27 senses (see Table 2 for details). Based on their lexical meanings, we analyzed the potential mechanisms of semantic extension at the origin of the different senses as well as the semantic relations that hold between them. This yields the semantic structure of the FIRE character displayed in Figure 2. Figure 2 reveals that the semantic structure of the FIRE character appears to be a radial network. In other words, all the senses are mutually connected in the network. In addition, the structure radiates from the inner circle to the outer area, with senses developing from the core to the periphery. All the senses are listed together with one of the most typical quotations they occurred in. The selected quotations are illustrated with the assistance of The Leipzig Glossing Rules.^[9]

Figure 2:

The semantic structure of the FIRE character.

In this quotation, FIRE is described from two perspectives, i.e. what the literal fire is and in consequence, what FIRE means semantically as a character and a radical. At the literal level, FIRE is a complicated natural phenomenon – a burning process that simultaneously releases byproducts such as light, heat, flames, etc. At the semantic level, FIRE is a pictographic character that can also be used as a radical in composite characters, which are semantically related to its literal meaning. In this respect, we assume that the original sense describes not only the literal fire but also the prototypical sense of FIRE, since it is an experiential phenomenon. Therefore, the FIRE character is a pictographic character which in turn confirms the prototypical nature of FIRE. The core of the semantic structure of the FIRE character is literal FIRE, or more specifically, the complicated burning process with multiple facets such as light, heat, and flames mentioned in quotation (1). If we think of FIRE, different people may bring up different facets, since the experience of fire is individual. Hence, it is not so strange that the dictionary may not list all the characteristic facets of the burning process. For instance, FIRE is also typically reddish and produces energy, so we argue that RED and ENERGY should also be included as characteristic facets. The prototypical core is displayed in Figure 3. Since all the other facets are released during the burning process, BURNING is found to be the central facet. Thus, the other facets are found to be connected to BURNING through metonymy. Moreover, the overall semantic structure is a complicated radial network where all facets have related clusters of extensions. Therefore, the analysis is introduced from facet to facet followed by their extended clusters. Since BURNING is the central facet, we now start from the related senses of BURNING.

Figure 3:

The core of the semantic structure of the FIRE character.

Here, senses (2) ‘burn out’ and (3) ‘to refer in particular to burning fire as a natural disaster’ are related to the facet BURNING in the two dictionaries. First, sense (3) seems to be closely connected to sense (2) in that they may both be interpreted as concrete specializations of the more general sense ‘burning’ by focusing on different aspects of the process. For instance, ‘burn out’ may be seen as a specialization of the burning intensity (i.e. a thorough burning) whereas ‘burning fire as a natural disaster’ is regarded as a type of burning. Furthermore, we also find metonymical relations between both senses. Given that FIRE ‘burns things down and destroys things in it,’ ‘burn out’ may also be interpreted as the result of a complete combustion whereas ‘burning fire as a natural disaster’ presents the consequence of the burning process. In this respect, both senses may be seen as metonymical nuances of the more general sense ‘burning.’ Yet, if ‘burn out’ is considered as a thorough burning, ‘burning fire as a natural disaster’ may be seen as the consequence of sense (2). Then, sense (3) may also be seen as a metonymical extension of sense (2). In addition, the metonymical relation between sense (3) and ENERGY may hold in both directions, since FIRE can be seen as a kind of strong ENERGY, which can be a disaster when it is out of control.

Here in (4), we find a simple and straightforward extension of the literal FIRE. The burning fire is used as a tool or method to ‘cook food thoroughly.’ Therefore, the notion of using fire to ‘cook food thoroughly’ is metonymically connected to the literal FIRE, or more specifically to the facet BURNING. Following this line of extension, senses (5)–(7) also appear to be related to this cluster. In particular, quotation (5) shows that ‘the roasting process of tea leaves’ is a clear specialization of the notion ‘to cook food thoroughly by fire.’ What’s more, sense (6) is also found to be related to (4) ‘to cook food thoroughly by fire,’ in which FIRE refers to the literal fire used to cook in the army. In this case, it has a metonymical interpretation due to the eating habits in the army, i.e. soldiers set up fire to cook and eat in a unit of 10 people in the army. Furthermore, sense (7), which appears chronologically later, may be seen as a generalization from sense (6) to a larger notion GROUP/UNIT. Figure 4 summarizes the relations of senses (1)–(7).

Figure 4:

Semantic relations of senses (1)–(7).

Now let us move on to the extended clusters that are linked to the facet HEAT. For example, the notion of ‘internal heat’ described in quotation (8) as the sensory feeling of getting warm is indirectly linked to FIRE through HEAT, given that sense (8) is a metonymical extension of the facet HEAT. Moreover, based on the same physiological experience of feeling warm internally, FIRE can also be metaphorized from the literal FIRE to refer to sense (9) ‘physiological fire produced by yang qi and providing energy to the whole body’ in medical contexts. In contrast to the simple sensory feeling described in sense (8), the ‘physiological fire’ described in (9) is a metaphorized medical term based on the real physiological experience.

In senses (10) and (11), FIRE is also used to refer to medical terms. In (10), FIRE can be ‘one of the six causes of disease’ if the ‘physiological fire’ is dysfunctional or disrupted in the body, which indicates a metonymical development from (9) to (10). Sense (11) may also be seen as a metonymy that developed from the ‘physiological fire’ since ‘fever’ in quotation (11) may be seen as a pathological performance caused by the disrupted ‘physiological fire’.

However, the reason that FIRE or INTERNAL HEAT can be interpreted in traditional Chinese medicine terms is also related to sense (12) ‘FIRE is one of the five elements.’ Generally, the five elements doctrine is a philosophical theory descending from the classic Daoist theory of yin-yang.^[20]

According to the yin-yang doctrine, all phenomena and things in the universe, the human body included, are governed by the law of unity between opposites represented by two Chinese words: 阴 yin ‘feminine; negative’ and 阳 yang ‘masculine; positive’. This ultimate law states that yin and yang are both opposed to and dependent on each other, and can even turn into each other (Yu 2009: 105).

In quotation (12), the notion of ‘FIRE as one of the five elements’ is introduced in its philosophical sense. That is, the five elements are not concrete substances but abstract concepts that illustrate how the universe is dynamically equilibrated according to the tradition of Daoism. More specifically, each element is generated after another, while each is restrained by another in sequence. In particular, the outer arrowheads indicate the direction of mutual promotion, whereas the inner arrowheads indicate the direction of mutual restraint between five elements (Yu 2009: 106). The mutually promoting and restraining relations are exhibited in Figure 5 (from Chen 1989: 1000, translated by Yu 2009: 106).

Figure 5:

Relation of mutual promotion and restraint among the five elements (from Chen 1989: 1000, translated by Yu 2009: 106).

In terms of the five elements theory, FIRE represents the concepts that are closely related to the literal FIRE, viz. concepts such as HOT, UP, and LIGHT. Therefore, anything that shares similar features with the literal FIRE is classified in the FIRE element category. In this regard, the sense ‘FIRE as one of the five elements’ is a specialization of the literal FIRE. Moreover, the five elements are not only important concepts but also symbols of a deeply rooted philosophy and ideology in the community. In consequence, the five elements theory has been applied and interpreted metonymically in a wide variety of ways. For instance, senses (9) ‘physiological fire,’ (10) ‘one of the six causes of disease,’ and (11) ‘pathological hyperfunctions’ above are the extensions interpreted in medical contexts, whereas senses (13) ‘Mars’ and (14) ‘Antares’ below are the extensions interpreted in the context of constellations.

In constellation contexts, ‘Mars’ is metonymically linked to the literal FIRE through the feature RED, because it has a seemingly reddish color in the sky, in particular, ‘as reddish as fire.’^[21] Moreover, ‘Mars’ and ‘Antares’ are related to FIRE in that FIRE ‘represents the direction south.’ Indeed, each element is assigned a direction in the five elements theory, with FIRE representing the south by virtue of the HIGH TEMPERATURE of southern areas, which typically enjoy a warmer climate than northern areas. Furthermore, the idea of ‘Antares’ appearance’ implying the fifth month ‘May’ in quotation (14) also shows that summer is a season of HIGH TEMPERATURE. So senses (13) and (14) are metonymies related to the five elements theory of the literal FIRE.

Senses (15)–(17), which describe ‘strong feelings/desires/emotions,’ can be interpreted in two ways. If we consider FIRE as a kind of strong ENERGY, namely ‘FIRE burns and destroys everything in it,’ then ‘strong feelings/desires/emotions’ may be regarded as a direct metaphor of FIRE. If we consider ‘strong feelings/desires/emotions’ as ‘pathological hyperfunctions,’ they may be considered as metaphoric extension of (12) ‘FIRE as one of the five elements.’ Also, sense (15) is related to sense (16) in that they are actually the nominal and the verbal interpretations of the same concept, ANGER, respectively. As can be seen from quotations (15) and (16), the different syntactic functions of the FIRE character in the contexts are the result of their different word class.

There is one more extension related to ENERGY, namely sense (18) ‘urgent or pressing.’ The notion of ‘urgent or pressing’ is extrapolated from the sense of a FIRE compound in its oldest quotation (18a), in which FIRE refers to its literal meaning. Concretely, the compound 火驰 huǒ-chí denotes ‘fire-speedy’ in its literal sense whereas the overall reading of the compound conveys ‘as fast as fire,’ which is a metonymical description for urgency. To a certain extent, FIRE may be seen as a redundant intensifier to strengthen the sense ‘speedy’ in such compounds. However, we also find quotation (18b), in which the notion of ‘urgent or pressing’ is directly denoted by FIRE itself. In this sense, ‘urgent or pressing’ may be seen as a metaphorical extension of the literal FIRE given that FIRE is a violent and intensifying phenomenon so that when the burning fire is out of control, it is an ‘urgent’ or ‘pressing’ situation that requires immediate attention.

Similar to the two ways ENERGY extends, the FIRE character also refers to the notion of RED in two ways. That is, the FIRE character either refers to the color red directly by itself or it is used in its literal sense as a compositional part in compounds describing the ‘reddish’ sense as a whole. For instance, the notion RED is expressed by the FIRE character directly in the compound 火旗 huǒ-qí ‘red flag’ in quotation (19). In compounds such as 火红 huǒ-hóng ‘red (fiery-red),’ which as a whole is denoted ‘as red as fire,’ the FIRE character may act as a redundant intensifier to express its literal meaning. Figure 6 illustrates the semantic relations we found in senses (8)–(19).

Figure 6:

Semantic relations of senses (8)–(19).

The oldest quotation of ‘guns and ammunition’ occurs in the compound 火器 huǒ-qì^[27] ‘firearms.’ In the context of quotation (20), specifically, FIRE is used as part of the weapon, namely to ignite the gun powder of weapons. In this respect, FIRE is interpreted in its literal sense and the notion of ‘firearms’ is a metonymical extension of the literal sense.

The next sense centers on the notion that FIRE is used as a temporal marker. The ‘Cold Food Day or Festival’ in (21) is an old traditional holiday which originated from southern China. Its name ‘cold food’ comes from the tradition of avoiding any uses of fire, including no cooking with fire. However, the main reason for this ‘no fire’ practice lies in seasonal change for the amount of daily fire consumption in the south. That is, people will adjust the amount of fire to be burned for daily uses in the warmer seasons. Particularly on this holiday, people are forbidden to use any kind of fire before changing to the new amount of fire for daily uses. Consequently, FIRE here is to be understood in its literal sense and is further extended as the temporal marker, as in quotation (21) “火前 huǒ-qián ‘before the fire.’” As quotation (22) points out, “there are four alternations of fire in one year, so people in Xixiang, Qiongzhou call one year ‘one huǒ.’” Therefore, the use of ‘temporal quantifier, synonym of year’ is a direct generalization of sense (21), in which FIRE is used as temporal marker to refer to the actual cycles of adjusting fire in the real world.

Sense (23) ‘flame-shaped patterns on ancient clothes’ is found to be the only metonymy that directly developed from the facet FLAMES. LIGHT seems to be a more productive facet, as it is the core of a cluster of extensions, including the metonymical extensions (24)–(26) and the metaphorical extension (27). As for the metonymical extensions of LIGHT, sense (24) and sense (25) are clearly related to each other in that the literal FIRE is used as ILLUMINATION in both quotations. However, they differ in their syntactic function of the character FIRE. In particular, quotation (24) displays a nominal interpretation, as the character FIRE is used to refer to ‘illumination tools (torch or candle).’ In contrast with (24), the character FIRE is used as a transitive verb in quotation (25), where its contextual sense ‘checking things by firelight (methods)’ plays an adverbial role. In addition, ‘thunder and lightning in the rain’ is metonymized as ‘fire in the water’ in quotation (26). As for the metaphorical extension of LIGHT, quotation (27) typically occurs in idioms and proverbs. At the literal level, quotation (27) as a whole conveys the meaning ‘when you see burning light, you know there is fire,’ which is then metaphorized as ‘when you see or recognize the features of things, you get the insights or knowledge to understand the things.’ Meanwhile, each constituent part of quotation (27) is also metaphorized individually, more specifically ‘the burning light’ is metaphorized as ‘features of things’ and ‘fire’ is metaphorized as ‘insight or knowledge.’ Hence, ‘light (n.); brightness’ is linked to the figurative meaning of LIGHT. The semantic relations of senses (20)–(27) are portrayed in Figure 7.

Figure 7:

Semantic relations of senses (20)–(27).

To sum up, it is now clear how Figure 2 summarizes the findings of the first step of the analysis. To make it easier to follow the process of extensions, Figures 3, 4, 6, and 7 build up Figure 2 in a step-wise sequence: Figure 3 illustrates how FIRE facets are connected in the core circle of the prototypical FIRE; Figure 4 mainly presents the extension cluster that centers around the sense ‘to cook food thoroughly by fire’; Figures 6 and 7 zoom in on extensions that are related to the facets of the prototypical FIRE.

3.2.2 Step 2: Diachronic pathways

A diachronic investigation of the semantic development of the FIRE character is needed for several reasons. First, we need to check whether our analysis of the directions of the extensions and mechanisms that drive semantic change are valid by looking at the chronological appearance of each sense. For example, we hypothesized that sense (7) ‘companion; group made by companionship, company’ is an extension of sense (6) ‘ancient unit of military system, a military unit made by 10 people’ through generalization. However, if our diachronic data shows that sense (7) appears earlier than sense (6), the relation between the senses will be changed in two ways. On the one hand, the direction of the extension will be changed whereby sense (6) is the extension of sense (7). On the other hand, this also means that the mechanism of semantic change changes: sense (7) is no longer a generalization from sense (6); sense (6) is a specialization of sense (7). Moreover, differences between the panchronic analysis and the diachronic analysis may reveal interesting cases for future research. What’s more, a diachronic investigation of semantic change provides more aspects of meaning than the synchronous aspects given by panchronic denotations in the dictionaries.

To determine the diachronic pathways of the senses, we dated the earliest appearance of each sense in the dictionaries. A corpus-based approach is usually a very good option for collecting the diachronic information by virtue of its authentic reflection of actual language use (e.g. Gablasova et al. 2019: 130; Gilquin and Gries 2009). However, the earliest appearances of the senses are mainly found in Ancient Chinese, where lexical items are very often ambiguous and polysemous in meaning. As a result, subjective interpretation of corpus attestations by the authors could be misleading. More importantly, a corpus-based approach is also inconvenient for a practical reason. After a pilot search in the Corpus of Historical Chinese,^[34] we gathered 53,044 and 159,348 concordances of the FIRE character in Ancient Chinese and Modern Chinese, respectively. Given this large amount of data for this single case study, it is possible that the amount of the data we get for follow-up case studies will be even larger. A common practice in dealing with such large amounts of corpus data is to select samples to represent the whole population (e.g. McEnery et al. 2006: 13–21). However, this method of sampling is not viable in this case because we have to manually go through all concordances one by one in order to identify the earliest appearance of each sense. Sampling a subset of the entire data will potentially leave out the concordances in which the earliest appearances of the senses occur. That is to say, it is impractical to deal with such large amount of data for this single case study as well as the coming case studies. As an alternative to a corpus-based approach, this study adopts a dictionary-based approach. A dictionary-based approach is legitimate for several reasons. First, as mentioned earlier, the abundant data of the GCDC and the GDC covers a long period of time. The oldest quotations can be located before the Qin Dynasty (221–207 B.C.), while the more recent quotations appear after the Qing Dynasty (1636–1912). Second, despite the fact that the dictionaries provide only approximate diachronic data, all the quotations are representative instances from texts of diverse genres, such as poems, ancient dictionaries, encyclopedic collections, and historical records. Moreover, the quotations are usually listed chronologically, which helps determine the diachronic appearances of the senses.

To date the oldest quotation of each sense in the dictionaries, we could simply have checked the publication dates of the texts in which the oldest quotations occurred. However, most of the oldest quotations come from early philosophical classics, and therefore, the publication dates are very often found within a certain time span. In this case, the biographical information of the authors of the texts may be helpful, as this is typically better documented than the publication dates. However, the same problems applies to the information regarding the author: their life is also a time span, and sometimes it is unknown. Hence, the majority of the publication dates and the authors’ biographical information are not available as exact dates. For simplicity and consistency, the end of the time spans was used to date the oldest quotation of each sense because we can be certain that the oldest quotations occurred at the latest by the end of the publication dates or the death of the authors. To determine the earliest appearances of the senses, we adopted the following three-step process.

1. When the publication date of the text in which the oldest quotation occurred is known, the exact date was used in the analysis. For example, the oldest quotation of ‘get angry’ is from the drama Ri Chu, ‘Sunrise,’ Scene 1, which was published in 1936.

2. When the author is unknown or the author’s lifetime is undocumented, we used the end time of the publication date.^[35] For example, sense (23) ‘flame-shaped patterns on ancient clothes’ is from Shu, ‘Book of history,’ whose author is unknown. So the end time of the publication date was used in the analysis.

3. When both the publication date and the author’s biography are time spans, we compared the end of the publication date time span and the death of the author of the text and used the earlier of the two.^[36]

Table 2 presents the chronological appearances of the senses. Figure 8 incorporates the diachronic information into the panchronic structure of FIRE from step one of the analysis, which yields the diachronic development of the character FIRE. Each sense in Figure 8 is accompanied by its order of appearance according to Table 2. More specifically, the senses are numbered from old to new. Senses that show up around the same time are marked with the same number. For instance, senses ‘burn out’ and ‘fire as a natural disaster’ show up around the same period of time as the earliest senses of the FIRE character, so they are both marked with number 1. The senses ‘light (n.); brightness,’ ‘to cook food thoroughly by fire,’ and ‘illumination tools, such as torch, candle, etc.’ that appear respectively around 376, 348, and 286 B.C. are marked as numbers 2, 3, and 4. The ordering ends with the sense ‘get angry’, which is first seen in 1936 and is ranked as number 20.

Figure 8:

Diachronic structure of the FIRE character.

The overall diachronic development of the FIRE character in Figure 8 generally supports the panchronic analysis. That is, the semantic development of FIRE radiates from the core to the periphery. On the one hand, most of the senses in the core circle are among the most ancient senses, as most of them show up before 5 B.C., whereas most of the senses outside the core circle mostly appear after the year 513. On the other hand, if we zoom in on individual clusters of extensions, senses are also chronologically developing from the core to the periphery in their individual development lines in two ways: (i) each line of extensions starts from the core; and (ii) the panchronic analysis of how one sense develops from the other generally fits the chronological appearances of the senses. For example, if we look at the sense ‘to cook food thoroughly by fire,’ it is a metonymical extension of the facet BURNING in the core. Therefore, the cluster of meanings extended from ‘to cook food thoroughly by fire’ is also seen as extensions from the core. Moreover, regarding the cluster centering around the sense ‘to cook food thoroughly by fire,’ the chronological order supports the direction of the extensions of this line of development: the chronological order from BURNING to ‘tea roasting’ is 1, 3, 14, whereas the chronological order from BURNING to ‘companion; group made by companionship, company’ is 1, 3, 8, 9. The same results can also be found in other individual development lines of clusters of extensions in Figure 8.

Yet, Figure 8 reveals that certain figurative extensions said to develop from the facets HEAT, FLAMES, LIGHT, and RED appear much earlier than the facets themselves, which seems to contradict the panchronic analysis. For example, LIGHT’s metonymical and metaphorical extensions, ‘light (n.); brightness,’ ‘illumination tools, such as candle and torch,’ and ‘thunder and lightning’ seem to show up around 376 and 286 B.C. respectively, whereas LIGHT starts to appear around 5 B.C. How can this finding be explained provided we posit that all senses develop from the center of the prototypical FIRE to the peripheral extensions? To answer this question, we need to separate the two meanings of the facets. If we consider fire as a natural phenomenon, the facets are the byproducts such as light, flames, heat, etc., which are automatically present when there is fire. As parts of the complex phenomenon of burning, these byproducts are inseparable from fire. In consequence, if we consider the burning fire as the literal and the prototypical meaning of FIRE, the existence of these facets is naturally presupposed regardless of whether they are recognized as meanings or senses in the dictionaries. Not to mention that it may also be the case that it takes longer for dictionaries to explicitly identify these facets as separate meanings, especially since they are implied by the definition of the literal FIRE.