Chapter 6 Antonym Sequences in Mandarin Chinese 130
6.3 Sequencing Preferences
It has been found in Section 6.1 that antonyms in Mandarin Chinese prefer a
particular sequence in text. In addition, as shown in Table 14, some antonym pairs show more stability in the preferred sequencing than others. For example, lao precedes its antonymous partner shao in as high as 98.7% of the database sentences featuring this pair, while ying precedes its antonymous counterpart shu in only 54.7%
of the database sentences featuring this pair. This section is intended to identify factors that contribute to the stabilization of antonym sequencing.
Note that the antonymous pairs in Table 14 are listed in order of the extent to which they adhere to their respective preferred sequence in text. At the top of Table 14 is the pair lao/shao, with lao preceding shao in almost every database sentence (98.7%). In addition to the pair lao/shao, another five antonymous pairs (i.e., dui/cuo, shi/fei, qianjin/houtui, nan/nu, and jin/tui) appear in their preferred sequence in more than 90% of the database sentences featuring these pairs. Moreover, another four antonymous pairs (i.e., ai/hen, yi/nan, jin/chu, and nansheng/nusheng) appear in the preferred sequence in more than 80% of the database sentences featuring these pairs.
Further down Table 14, five antonymous pairs show a bias between 70% and 80%, and seven antonymous pairs between 60% and 70%. Only three antonymous pairs score below 60%.
A tendency identified in English is that “the more positive the associations, the more marked the pattern tends to be” (Jones 2002:124). For instance, true, rich, and win, which carry positive connotations, are given priority over their antonymous partner (i.e., false, poor, and lose) in as high as 88.7%, 86.3%, and 84.5% of the database sentences respectively. On the other hand, the word innocence, which is negatively associated with naivety but remains more positive than its antonymous counterpart guilt, precedes guilt in a majority of sentences featuring this pair, but the preferred sequence is relatively less marked (65.9%).
The same tendency is observed in Mandarin Chinese. Those at the top of Table
14, such as dui (96.7%), shi (95.3%), qianjin (92.5%), and ai (86.7%), do bear strongly positive associations. On the other hand, some antonyms are positive, but their positive association is less strong than the above examples. Their sequence patterns are found to be less stabilized. For example, the word chengren ‘to confirm;
to admit’, which means strengthening a claim by conceding it as true, is semantically more positive than its antonymous counterpart fouren ‘to deny’, but its positive association is less strong than that of dui, shi, qianjin, and ai. As shown in Table 14, chengren precedes fouren in 68.7% of the database sentences featuring this pair, which is relatively low.
Intriguingly, most antonymous pairs in Mandarin Chinese do not pattern with their relatives in terms of how stabilized their preferred sequencing is. It is found that monosyllabic antonymous pairs adhere to the preferred sequencing more often than their disyllabic counterparts. As shown in Table 14, lao precedes shao in 98.7% of the database sentences featuring lao/shao while lao precedes nianqing in only 55.3% of the database sentences featuring lao/nianqing; dui precedes cuo and shi precedes fei in 96.7% and 95.3%, respectively, of the database sentences featuring the two pairs while zhengque precedes cuowu in only 60.0% of the database sentences featuring zhengque/cuowu. Since most monosyllabic words in Mandarin Chinese are inherited from Classical Chinese (Chao 1968; Li and Thompson 1981), most monosyllabic antonymous pairs in Mandarin Chinese are more stabilized than their disyllabic counterparts not only in terms of functions they exhibit and constructions in which they are embedded, but also in terms of the sequence in which their members co-occur.
As shown in Table 14, the monosyllabic antonymous pair nan/nu adheres to its preferred sequence in as high as 92.1% of the database sentences featuring this pair, and its disyllabic counterparts, i.e., nansheng/nusheng (82.0%), nanren/nuren (72.7%),
nanzi/nuzi (70.7%), and nanxing/nuxing (68.0%), show different degrees of stabilization in terms of antonym sequences.
In terms of antonym sequences, the monosyllabic antonymous pair yi/nan also shows a high degree of stabilization. As shown in Table 14, 86.7% of the database sentences featuring yi/nan adhere to the preferred sequence. On the other hand, the modern disyllabic counterpart, kunnan/rongyi, has not been conventionalized into the frame ‘positive antonyms preceding negative antonyms’ in Mandarin Chinese. That is, kunnan precedes rongyi (58.4%) more often than rongyi precedes kunnan (41.6%).
The small difference between the two figures can hardly confirm the sequencing preference either way.
A seemingly unusual antonymous pair is qianjin/houtui, which is disyllabic but adheres to its preferred sequencing more often (92.5%) than its monosyllabic counterparts jin/tui (91.3%) and qian/hou (70.0%) do. It is likely that the disyllabic pair qianjin/houtui is conventionalized into the frame ‘positive antonyms preceding negative antonyms’ in Mandarin Chinese, as its monosyllabic counterpart jin/tui is. As for the pair qian/hou, its ranking is relatively low in Table 14 although qian, like jin and qianjin, bears positive associations. The antonyms qian and hou often co-occur to define a period of time, with the latter preceding the former. Here is an example:
(150) 行政院及各部會首長，應於立法委員改選後、就任前，即行總辭，
Xingzheng yuan ji ge bu hui shouzhang, ying yu lifaweiyuan gaixuan hou, jiuren qian, ji xing zong ci,
administration institute and each organization director, should at legislator reelect after, inaugurate before, right do totally resign
‘All the officers at the Administration Yuan and related organizations should resign after legislators are reelected and before they are
inaugurated.’
In (150), gaixuan ‘to reelect’ and jiuren ‘to inaugurate’ co-occur in the lexico-syntactic frame …hou…qian ‘after…before…’ to delimit a period of time. It is suggested that frames of this kind have made the sequence qian/hou less stabilized than the sequences qianjin/houtui and jin/tui.
At the bottom of Table 14 is the pair ying/shu. While ying is highly positive, the sequence ying/shu accounts for only 54.7% of the database sentences featuring this pair. This pattern is incompatible with the finding in Jones (2002). In English, 84.5%
of the database sentences featuring the pair win/lose conform to the canonical sequence win/lose. In Mandarin Chinese, sheng is synonymous with ying and bai is synonymous with shu. Interestingly, jia dui da sheng yi dui ‘Team A won the game against Team B’ is synonymous with jia dui da bai yi dui ‘Team A defeated Team B’.
This suggests that the antonyms sheng and bai can be interpreted from different perspectives, with the same meaning yielded. It is likely that ying and shu is also more open to either perspective, so there is no apparent preferred sequencing. Another reason is that the pair ying/shu may have not been conventionalized into the frame
‘positive antonyms preceding negative antonyms’.
6.4 Summary
This chapter is intended to answer the following three questions. First, do antonym pairs in Mandarin Chinese manifest preferred sequencing in text? Second, what makes a given sequence preferred over the reverse order? Third, what contributes to the stabilization of a preferred sequencing?
The findings of this study support those in Jones (2002), the focus of which is on antonyms in English. First, most antonym pairs in Mandarin Chinese show a preferred
sequencing in text. Second, in terms of antonym sequences, almost all the antonymous pairs investigated in this study are compatible with their equivalents in English. The only difference is that lao precedes shao and nianqing more often while young precedes old more often. This discrepancy is speculated to have arisen from how youth and oldness are differently perceived in the two cultures.
In Section 6.2, it is found that the most important factor affecting antonym sequencing in Mandarin Chinese is the notion of positivity (Section 6.2.1). A positive word is usually given priority over its negative antonymous partner because the former is cognitively preferred. However, the context sometimes licenses perspective shifts so that the reverse antonym sequencing may show up.
As shown in Table 14, antonym sequencing preference varies from case to case.
Just as in English (Jones 2002:124), the more positive the first member of an antonymous pair is, the more preferred the canonical sequence tends to be (Section 6.3). However, other lexical properties, such as whether a given antonymous pair is a remnant from Classical Chinese, might override the notion of positivity, giving rise to the sequencing discrepancy observed between synonymous pairs such as lao/shao and lao/nianqing.
Chapter 7 Concluding Remarks
This study investigates how antonyms in Mandarin Chinese function in text by applying Jones’ (2002) functional framework of antonymy to data in Mandarin Chinese. In Chapter 1, three research questions are posed. First, to what extent is Jones’ (2002) functional framework of antonymy able to account for data in Mandarin Chinese? Second, what constructions can serve as lexico-syntactic frames for antonyms in Mandarin Chinese? Third, do antonyms favor a particular sequence in text? If so, then why? In Chapters 4, 5, and 6, the results of our analysis are presented and discussed. This chapter will summarize the findings (Section 7.1), pinpoint the implications of our findings (Section 7.2), and offer some suggestions for further research on antonymy in Mandarin Chinese (Section 7.3).
7.1 Findings of the Thesis
The present study takes a data-based perspective to investigate how antonyms in Mandarin Chinese function in text. The analysis of the study is based mainly on Jones’ (2002) functional framework of antonymy. Our Chinese data are found to manifest the eight textual functions of antonymy identified in Jones (2002), i.e., Ancillary Antonymy, Coordinated Antonymy, Comparative Antonymy, Distinguished Antonymy, Transitional Antonymy, Negated Antonymy, Extreme Antonymy, and Idiomatic Antonymy. In addition, three residual textual functions of antonymy identified by Jones (2002) also show up in our Chinese data, i.e., Specified Antonymy, Associative Antonymy, and Simultaneous/Equivalent Antonymy. Still, all the above textual functions cannot account for all the Chinese data. Two new textual functions of antonym co-occurrences are identified. The first one is Transitive Antonymy, i.e.,
antonyms co-occur in an agent-patient schema. The other is Negated Ancillary Antonymy, i.e., antonyms co-occur to signal a nearby contrast, and one of the members is negated. In total, thirteen different textual functions of antonymy have been identified in Mandarin Chinese. Among the thirteen textual functions, Coordinated Antonymy and Ancillary Antonymy are the most dominant, collectively accounting for 63.2% of the database sentences.
To establish a representative set of antonymous pairs in Mandarin Chinese, the present study includes antonymous pairs in different traditional categories of oppositeness. It is found that an antonymous pair’s traditional category of oppositeness affects its functional distribution. Typically, gradable pairs favor Transitional Antonymy, complementary pairs favor Specified Antonymy and Comparative Antonymy, relational pairs favor Transitional Antonymy and Transitive Antonymy, and directional pairs favor Negated Antonymy and Idiomatic Antonymy.
It is also found in the present study that an antonymous pair’s grammatical category affects how it functions in text. Typically, adjective pairs favor Ancillary Antonymy or Coordinated Antonymy, verbal pairs favor Coordinated Antonymy, and nominal pairs favor Coordinated Antonymy, Transitive Antonymy, Specified Antonymy, Comparative Antonymy, and Distinguished Antonymy.
Some antonymous pairs in Mandarin Chinese have a counterpart in Classical Chinese that is still used in Modern Mandarin. In the present study, an antonymous pair’s extent of modernization is taken into consideration. It is found that vernacular pairs (e.g., dui/cuo and lao/nianqing), in contrast with their counterparts in Classical Chinese (e.g., shi/fei and lao/shao), exhibit wider functional distribution or more flexibility in working with a variety of lexico-syntactic frames.
Due to the fact that Modern Chinese is evolving towards a situation where almost all content words contain at least two syllables to maintain a disyllabic
rhythmic pattern (Chao 1968; Li and Thompson 1981), an antonymous pair’s morpho-syllabic structure is also taken into consideration in the present study. In general, monosyllabic antonymous pairs prefer textual functions associated with four-character lexico-syntactic frames, such as Coordinated Antonymy and Transitional Antonymy.
In short, how antonyms in Mandarin Chinese function in text is influenced, to a certain extent, by their traditional category of oppositeness, grammatical category, extent of modernization, and morpho-syllabic structure. However, even though an antonymous pair’s functional distribution may be affected by various linguistic factors, it does not follow that how an antonymous pair functions in text is restricted by its these factors (Jones 2002:144).
A set of synonymous pairs, i.e., nan/nu, nanren/nuren, nansheng/nusheng, nanxing/nuxing, and nanzi/nuzi ‘man/woman’, show variations as they are used in text.
Coordinated Antonymy is strongly preferred by nan/nu and nanzi/nuzi, Specified Antonymy is favored by nansheng/nusheng and nanxing/nuxing, and Comparative Antonymy is preferred by nanxing/nuxing. The pair nanren/nuren, unlike its synonymous partners, illustrates the least skewed functional distribution. The findings suggest that distributional differences may reflect how near synonyms differ from each other in meaning.
The textual functions of antonymy in English and Mandarin Chinese are compared, with the focus on antonymous pairs selected both in Jones (2002) and in the present study. It is assumed that synonymous pairs in different languages (e.g., confirm/deny in English and chengren/fouren in Mandarin Chinese) perform similar textual functions. However, cross-linguistic discrepancies in how synonymous pairs function in text may arise from language-specific structural characteristics.
From a constructional perspective, each textual function of antonymy in
Mandarin Chinese has its typical lexico-syntactic frames. In other words, it is found that how antonyms function in text depends largely on the collocational environment.
In Mandarin Chinese, there are many four-character lexico-syntactic frames that allow antonym co-occurrences, as identified in the following:
(151) Coordinated Antonymy
•
…antonym X…antonym Y (e.g., shi dui shi cuo ‘be right be wrong’)•
monosyllabic interrogative + antonym X + monosyllabic interrogative + antonym Y (e.g., shei dui shei cuo ‘who right who wrong’)(152) Transitional Antonymy
•
xian + antonym X + hou + antonym Y ‘first…then…’•
cong/you + antonym X + dao + antonym Y ‘from…to…’(153) Negated Antonymy
•
(zhi ‘only’) + antonym X + bu ‘not’ + antonym Y ‘only…not…’(154) Comparative Antonymy
•
antonym X + monosyllabic stative verb + yu ‘than’ + antonym Y (155) Idiomatic Antonymy•
bu jin ze tui ‘either to make progress or recede’The above lexico-syntactic frames fit especially well with monosyllabic antonymous pairs, yielding a large number of four-character phrases.
Finally, it is found that antonym co-occurrences in Mandarin Chinese prefer a particular sequencing in text. In Mandarin Chinese, a positive word tends to precede its negative antonymous partner (e.g., dui/cuo ‘right/wrong’), and a word with a higher frequency tends to precede its antonymous partner (e.g., mai/mai ‘to buy/to sell’). The notion of positivity, which can be interpreted in psychological,
metaphorical, and socio-cultural terms, is the most dominant factor affecting antonym sequencing in Mandarin Chinese. In addition, the more positive the first member of an antonymous pair is, the more preferred the canonical sequence tends to be. However, when perspective shifts take place in text, the other sequencing order may show up.
7.2 Implications of the Thesis
Jones’ (2002) functional framework of antonymy is sufficient to describe how antonyms in English function in different modes of communication (Jones 2006, 2007;
Jones and Murphy 2005). The present study shows that Jones’ (2002) functional framework of antonymy is also useful in describing the textual functions of antonym co-occurrences in Mandarin Chinese, thus lending cross-linguistic support to this framework.
The present study pursues a corpus-based approach to exploring the textual functions of antonym co-occurrences in Mandarin Chinese by analyzing authentic data in a quantitative manner. The present study suggests that a collaboration between discourse analysis and corpus linguistics can complement traditional intuition-based approaches to language by probing into the actual use of language (Biber, Conrad and Reppen 1998; Halliday 1991, 1992; Stubbs 2001).
Following Jones (2002), this study takes a constructional perspective to establish a collocational profile for each textual function of antonymy in Mandarin Chinese. It is found that each textual function of antonymy has its own lexico-syntactic frames.
The syntagmatic aspect of antonymy, i.e., recurrent linguistic formulas for antonym co-occurrences, is revealed in co-text. This view complements the traditional approach, which emphasizes the paradigmatic aspect of antonymy (Stubbs 2001:39).
This study also touches on the cognitive facet of antonymy in Mandarin Chinese.
For example, the proximity effect explains mapping preferences in Ancillary
Antonymy (Section 5.2) and Specified Antonymy (Section 5.5). In addition, sequencing preferences in Mandarin Chinese reflect the notion of positivity in the mind.
To summarize, the present study of antonym co-occurrences in Mandarin Chinese has a number of methodological and theoretical implications. This study attempts a collaboration between discourse analysis and corpus linguistics, and lends cross-linguistic support to Jones’ (2002) functional framework of antonymy.
Moreover, this study takes a constructional perspective to explore the syntagmatic dimension of antonymy, identifying typical lexico-syntactic frames for antonym co-occurrences in Mandarin Chinese. Furthermore, this study offers a cognitive account on mapping preferences in Ancillary Antonymy and Specified Antonymy and antonym sequencing in Mandarin Chinese.
7.3 Suggestions for Further Research
The present study adopts Jones’ (2002) functional framework of antonymy to investigate antonym co-occurrences in Mandarin Chinese. In the following, several directions for further research on antonymy in Mandarin Chinese are briefly discussed.
First, the findings of the present study may need to be further verified with large-scale corpora of written language, for the corpus used in the present study, i.e., the Chinese Word Sketch Engine, only contains newspaper texts. As suggested in Kilgarriff (1997) and Stubbs (2001), different corpora have different sources and purposes, so a given research question may obtain different results from different corpora. While the Chinese Word Sketch Engine makes the present study comparable with Jones (2002) in terms of data sources (Section 3.1), it is fair and reasonable to ask whether antonym co-occurrences in Mandarin Chinese function similarly in
newspaper texts and, say, literary works.
Second, the present study focuses on the textual functions of antonym co-occurrences in written language. Further research may be conducted on how antonyms in Mandarin Chinese function in other modes of communication, such as spoken language. As proposed in Jones (2006, 2007), antonyms in English have different functional distributions in different modes of communication (Section 2.2.2.3). In Mandarin Chinese, how antonyms function might be affected by modes of communication as well. In functional linguistics, it is of interest and significance to the functional approach as to how a given language-internal phenomenon, such as antonymy, is affected by language-external factors, such as modes of communication (Biq 2000:378).
Third, the model adopted by the present study, i.e., Jones’ (2002) functional framework of antonymy, may have acquisitional applications. For example, one of the issues worthy of our further pursuit is how antonyms function in child-produced and child-directed speech. This issue has been addressed in English (Jones and Murphy 2005) (Section 2.2.2.3), and it should be addressed in Mandarin Chinese as well.
In fact, all of the three directions for further research on antonymy in Mandarin Chinese direct our attention to a multi-corpus, cross-linguistic approach to antonym co-occurrences. While the rapid development of corpus linguistics has made it possible to examine antonym co-occurrences from a usage-based perspective, no corpus is fairly representative of all the genres, styles, modes, topics, and sources (Kilgarriff 1997; Stubbs 2001). Meanwhile, for Jones’ (2002) functional framework of antonymy to gain more descriptive power, it needs to be examined in as many
In fact, all of the three directions for further research on antonymy in Mandarin Chinese direct our attention to a multi-corpus, cross-linguistic approach to antonym co-occurrences. While the rapid development of corpus linguistics has made it possible to examine antonym co-occurrences from a usage-based perspective, no corpus is fairly representative of all the genres, styles, modes, topics, and sources (Kilgarriff 1997; Stubbs 2001). Meanwhile, for Jones’ (2002) functional framework of antonymy to gain more descriptive power, it needs to be examined in as many