Altaic influences on Beijing dialect: the Manchu case.
Chinese(1) has been characterized as being resistant to outside influence. It has been assumed that if there is any borrowing between Chinese and neighboring languages, the direction is nearly always from Chinese to the other languages, rather than vice-versa. A statement on word order made by Li and Thompson (1974: 206) typifies what has been the majority opinion: "It is to be noted that the Chinese language is particularly suited to the investigation of the principles and directions of word order changes because of the overwhelming dominance of the Chinese civilization in pre-twentieth-century Asia. Such cultural dominance precludes the possibility of any external influence on Chinese in its word order development. Any change observed in Chinese word order must be originated internally."(2)
THE IDEA OF LANGUAGE DIFFUSION
More recently, however, a few scholars have challenged this idea of Chinese being a closed system. One has gone so far as to characterize Chinese as a language "mosaic."(3) Mantaro Hashimoto (1976a; 1976b; 1978; 1980; 1986) has done, perhaps, the most in advocating the study of areal features to gain information on the development of Chinese. He studied the Chinese dialects of northern China and noticed the further north one traveled in China, the more the Chinese dialects began to resemble the Altaic languages that bordered them. Conversely, as one traveled south in China, the Chinese languages began to resemble Austro-asiatic languages that bordered them in the south. He theorized that the Chinese languages had been heavily influenced by the non-Chinese languages on their periphery. Going even further with respect to the situation in northern China, he theorized that during the Qing (Manchu) dynasty (1644-1911), the language of the capital was not Chinese but rather a pidgin made up of Manchu and Chinese elements, as well as a few elements from Mongolian and other minor languages. Furthermore, he suggested that modern Beijing dialect was a descendant of that pidgin (Hashimoto 1986). Though this theory is impressionistically persuasive, the major obstacle in accepting it is the lack of solid empirical evidence. It might be expected that little or nothing would be written in this hypothesized "mixed" language. Pidgins rarely, if ever, produce a literature. Hashimoto (1986: 92-93) pointed to several texts of zidi shu (a type of "drum song" popular in northern China during the Qing dynasty(4)) that were written in a mixture of Manchu and Chinese as a proof that the mixed language he hypothesized actually existed. Charles Li (forthcoming) has argued convincingly that these texts were purposeful manipulations of the two languages, much in the same way that macaronic verse in the West was often simply a manipulation of Latin and a vernacular language to produce a comic idiom for the enjoyment of those who had to study Latin. Indeed, it would be unusual for a pidgin to assume the form of the language in zidi shu, since most pidgins rely on essentially a single language for their lexicon, rather than being a mixture of lexical elements (Whinnom 1971). The fact that there are so few texts in this form also argues against them being a representation of the spoken language of time - although more and more texts of this type are coming to light (see Tulli 1992).
BORROWING VS. LANGUAGE SHIFT
The evidence given us by the existence of zidi shu should not be abandoned entirely, however. The fact that the texts were produced tells us something about the language situation in northern China. First, there must have been some, even much, bilingualism among the inhabitants of Beijing then, otherwise there would be no audience for the pieces. Ji Yonghai (1993) has in fact presented evidence that the Manchus went through a period of bilingualism before abandoning their language for Chinese. Second, although the mixed-language zidi shu appear to be deliberate manipulations of language for comic effect, one may assume that part of the comedy is that the language parodies an actual language situation. As Whinnom (1971) has pointed out, target-language speakers often parody the language of non-native speakers of the language, not in the way they actually speak, but rather how they are perceived to speak. Thus they may produce evidence of a mixed language even though they do not produce the actual mixed language. In the zidi shu text Chaguan ("Inspecting the Pass") the Manchu language is put in the mouth of the chou (clown),(5) a method that exactly matches genres of tanci (chantefable) in the south, wherein the major characters speak in Mandarin (guanhua) but the chou speaks in the local dialect. From that single text it would appear that Manchu was considered the local language, in contrast to Mandarin, which was the lingua franca.
Let us return for a moment to the areal situation in Chinese. Despite the lack of specific evidence to support Hashimoto's theory of a Manchu-Chinese pidgin, it remains a fact that when one travels north or south in China, the Chinese dialects one encounters tend to resemble to some degree the non-Chinese languages that border them. In the Fujian dialects, Jerry Norman (1976) discovered that some features of the language did not go back to Middle Chinese cognates and postulated a substratum language from the Austroasiatic family. Charles Li (forthcoming) has recently done work on Chinese and Tibetan dialects in western China and has discovered mutual influences, sometimes to an advanced degree, between the two languages. But since Standard Chinese (a standard based on modern Beijing dialect) has only a handful of words that can be identified as loans from other languages, these for the most part being Buddhist religious terms from Sanskrit and technical and scientific terms from English, most still think of Chinese as proceeding in an almost pure genetic line from Old Chinese through Middle Chinese and on to Modern Chinese. The pivotal argument supporting this idea is the observation that in situations of language contact and borrowing, lexicon is generally most easily borrowed, while phonology, syntax and morphology is, in that order, increasingly more difficult to borrow. Some, in fact, would argue that morphology cannot be borrowed from one language to another, and that the borrowing of syntax is hardly more possible. Given that many Chinese dialects, and most notably the standard dialect, seemed to have borrowed very little vocabulary from any source, some have dismissed the phonological, syntactic and, in cases where it may apply, morphological similarities of Chinese dialects with non-Chinese neighbors as either coincidental or involving Chinese influence on their neighbors. In the case of Manchu contact with Chinese, the above assumption appears to hold true. The Manchu found in Qing-dynasty texts borrows heavily from the Chinese lexicon, while at the same time, other than a handful of obscure Manchu loans into Chinese, there is little evidence of lexical borrowing from Manchu into Chinese.(6)
New thinking on situations of language contact, however, may put this evidence into a new light. Kaufman and Thomason (1988) have divided contact-induced change in language into two types. One type is "borrowing," where the pattern indicated above is followed: lexicon is borrowed first, then phonology, and if the borrowing continues, perhaps syntax and finally morphology. In these cases one would expect extensive lexical borrowing before one encounters any sort of phonological borrowing. Syntactic and morphological borrowing come, if at all, after the lexicon and phonology of the target language have been almost completely adopted.
The second type of contact-induced change is what they term "language shift." In these cases a significantly large group acquires as a group - and in a short period of time - the target language and begins to use it as their own. But there are varying degrees of competency among these groups learning the target language, depending on such things as availability of the target language, social factors, perceptions of prestige, etc. As a result the target language is often learned "imperfectly," thereby producing a mixed language of sorts. This "imperfect" language could consequently be imitated by the original target-language population, resulting in a contact-induced language change for the entire target-language population. The crucial factor in this scenario is that the influence of the first language on the second, or target language, occurs in precisely the opposite order in terms of language features "borrowed" in the first case of borrowing described above. The first-language speakers are following the regular pattern in acquiring the target language: they borrow lexicon first, then phonology, etc. Morphology, being the most difficult item to borrow, is the first casualty as the group acquires the target language and makes it their own. That is, they retain the morphology and/or syntax of their original native language or simplify the morphology/syntax of the target language. When they eventually abandon their original native language and become absorbed into the target language, as new "native speakers" of the target language, they begin to exert influence on "their" language. As a result, the native morphology, syntax, and even sometimes phonology of the target language are, in that order, replaced or eliminated. Its lexicon however, with the exception of a few scattered, obscure words for which there are no counterparts in the target language, is retained.
CHINESE AS A MOSAIC OF LANGUAGES
Looking at Chinese language and culture as ancient, homogeneous, and little-changing has obscured the complicated ethnic diversity that makes up the land of China. Though China has a unifying culture, it is made up of a wide variety of Asian ethnic groups that found themselves under this common cultural umbrella at different times and places, each with its own cultural and linguistic peculiarities. It is probably no mistake to say that "Chinese," in whatever linguistic form it assumed throughout the centuries, was indeed the prestige language of East Asia. But far from making it thus impervious to foreign influence, in fact, it is possible that, as masses of non-Chinese language speakers began to adopt Chinese as their language, they altered the language according to the rules of their previous native languages. It would make sense that Chinese is resistant to borrowed lexicon, because most instances would be situations of language shift. It would also make sense that Chinese has almost no morphology since it is one of the first casualties in cases of massive language shift, much in the same way morphology disappears in cases of pidgin or creole languages.(7)
EXTERNAL VS. INTERNAL INFLUENCE
A general rule followed by historical linguists in mapping linguistic change is that internal factors (rather than outside influence) must always be considered the main, if not the sole motivation for change. External factors are looked for only when all internal possibilities are exhausted (Kaufman and Thomason 1988: 57 [in reference to Martinet (1955) and Polome (1980)]). But, Kaufman and Thomason have shown (1988: 54-64) that this rule is not necessarily valid. Language change is often induced by external influence, and, even if there are internal factors that justify a change, there is no need to dismiss external factors. This is particularly true of simplifications.
INFLUENCE FROM ALTAIC LANGUAGES
We have rejected, for the time being, Hashimoto's claim that a pidgin language was spoken in the Manchu capital, due to a lack of hard supporting evidence. However, this is not to say that there has not been any Altaic (and possibly Manchu) influence on the northern Chinese dialects. Hashimoto has claimed a number of features in northern Chinese are due to Altaic influence. Some of these are: the relatively few tonal distinctions in northern dialects compared to the complex tonal systems of the southern dialects; the replacement of specific classifiers with the general classifier ge; the loss of all final consonants in syllables except "n" and "ng"; the phonotactic rules of northern vs. southern dialects, wherein the syllable can be the terminal level of analysis in southern dialects but the phoneme must be the terminal level in northern dialects; the change from modified-modifier word order in the south to modifier-modified in the north, and concurrently, the regular placement of the verb after the object in the north, rather than before the object, as in the south; the greater number of polysyllabic words in the north vs. the larger number of monosyllabic words in the south (Hashimoto 1976b). Unfortunately, many of these features may be considered simplifications of the language. That would be what one would expect in a language shift situation, but following the standard rule above, simplifications are generally explained as being internally motivated.
Jerry Norman (1982: 245-46) first commented on a parallel feature between Manchu and Chinese that involves a marked morpho-syntactic structure.(8) That is, the use in northern Chinese of the same markers for both passive and causative constructions (i.e., jiao, rang). Manchu also uses only one marker for both causative and passive constructions (-bu). This feature appears to have come into the Chinese language quite late and is typical of features of the language that are borrowed in "language shift" situations. The fact that the northern Chinese dialects use native lexical markers rather than borrowing, say, the Manchu bu, actually strengthens the argument that the feature appeared through language shift. The shifting group (here, the Manchus) acquires the target-language lexicon but reinterprets the grammatical functions of those words according to what they consider as equivalent items in their original language (Thomason and Kaufman 1988: 138).
THE MANCHU CASE
Manchu is a term designating the peoples led by the brilliant organizer and strategist Nurhaci. Nurhaci, during the late sixteenth and into the seventeenth century, organized an army, first from his own people, the Jurchen tribes living in what is now northeast China. These people were descendants of the Jurchens who ruled the area of northern China during the Jin dynasty (1115-1234), and Manchu is a direct descendant of this earlier Jurchen language. The term Manchu was not used until 1635 when Nurhaci's son Hong Taiji, abandoned the term Jurchen, allegedly because it was associated with subservience vis-a-vis China (Sun 1991: 9). Nurhaci's army came to consist of a number of groups, including Mongolians and Chinese, from the area. When his son and successor, Hong Taiji, used this army to invade Ming-dynasty China, it was estimated to contain two hundred thousand troops, of which approximately half were Manchu and a quarter each were Chinese and Mongolian, respectively (Ji 1993: 39). Ji Yonghai (1993) has given evidence to show that when this army attacked the Ming, the troops (including Chinese and Mongolians) probably spoke a form of Manchu. Ji goes on to say that, due to a number of factors, after the Manchus took over from the Ming they went from speaking only Manchu, through a period of widespread bilingualism, and finally reached the point at which Chinese became their native language. The question becomes: as this large group of people switched from Manchu to speaking Chinese what effect, if any, did it have on their new native language (i.e., northern Chinese), particularly within the Beijing dialect, which subsequently became the model for the standard language in all of China.
In looking at the possibilities involved in the contact between Chinese and Manchu, an exhaustive lexical, phonological, syntactical, and morphological comparison of the two languages would have to be made. The history of each language would have to be looked at to determine if changes in either language could be the result of contact with the other. Unfortunately, work on the historical development of Chinese is only just beginning and hardly any work has been done on Manchu. In terms of lexicon, Schmidt (1931) has a substantial monograph on Chinese loanwords into Manchu in which he alleges that over twenty percent of the Manchu lexicon is derived from Chinese. Little has been done on Manchu loans into Chinese (see note 5), indeed, some say nothing need be done, since there was no borrowing from Manchu into Chinese.
MANCHU LOANS INTO CHINESE
The difference between the number of Manchu loans into Chinese and Chinese loans into Manchu is enormous. Schmidt (1931) lists about two thousand Chinese loanwords into Manchu. A few of these have later been shown not to have been loans, but a large majority are unquestioned.
The Chinese lexicon, on the other hand, appears to have very few Manchu intrusions. A look at the Beijing fangyan cidian (Dictionary of the Beijing Dialect) and the Hanyu wailaic cidian (Dictionary of Foreign Words in Chinese) yields a list of approximately two hundred possible loanwords. It is difficult to determine how extensively these loanwords were incorporated into the dialects they entered. Chinese lexicography has tended to shy away from non-standard and dialectal terms, and that is precisely where loan items would appear. As a result, the Guoyu cidian (Dictionary of the National Language), a dictionary produced during the first years of the Chinese republic, contains only a couple of Manchu loanwords in its pages,(9) and one of these is missing in the recently published, allegedly more comprehensive and much larger, Hanyu da cidian (Great Dictionary of Chinese). Many of the loans found in the Beijing fangyan cidian and the Hanyu wailaici cidian are military and governmental terms; there are also a few kinship terms and some words representing plants and animals that are native to the northeast. All of which is what one would expect in a case of language shift. An examination of the local gazetteers of the northeast(10) has yielded a number of previously unknown loans. This is an area that warrants more detailed examination.
There are a number of features of the northern Chinese dialects that suggest that they were influenced by Altaic languages that bordered them on the north. Thomason's and Kaufman's theory of language shift, though not applied by them to the situation in China, appears able to account for the patterns of influence that can be observed, in ways that previous theories of language contact and borrowing could not. Of course, other explanations are also possible, including internal change. More investigation of the historical development of Chinese and the Altaic languages that border the Chinese language areas needs to be done before we can resolve the question of language influence between these two groups.
STEPHEN A. WADLEY PORTLAND STATE UNIVERSITY
1 The term "Chinese" is used throughout this paper as if it referred to a single language entity. This is a fiction of convenience, less true for the Chinese language family than any other major language. The Chinese dialects, as has been often pointed out, are separate languages if one uses the linguistic criterion of "mutual intelligibility." Even within the seven major dialect groups posited by Li Fang-kuei, and generally accepted without comment by scholars, are found many mutually unintelligible languages. Dialect studies in China are still at a rudimentary stage, and generalizations about the Chinese dialects will certainly have to be scrutinized and revised as our knowledge of them increases (see, for example, Norman [1988: 181-89]). Nevertheless, imperial identity, and more recently, national identity, as well as a mutually shared writing system has created the illusion, if not the actuality of a "national language." The interplay between a local northern language and the perceived "national" language is the major focus of this paper.
2 In fairness to Professors Li and Thompson, it should be mentioned that their later work indicates that they no longer subscribe to this extreme position.
3 Hashimoto (1986: 77) quotes Ballard, who apparently formulated the idea of Chinese being a linguistic "mosaic," in part influenced by Triestman (1968; 1970; 1972), who considered Chinese culture a "cultural mosaic."
4 Zidi shu are a form of amateur singing entertainment popular in northern China during the Qing dynasty. Most scholars now agree that the entertainment originated with the unemployed Manchu bannerman (called zidi 'boys') in the capital. It was generally performed by two men, one played the san xian 'three-stringed guitar' and one sang. Most of the written manuscripts of the genre that remain are in Chinese, though one manuscript entirely in Manchu and several in a combination of Manchu and Chinese have survived. For information on the genre, as well as examples of the mixed language texts, see Guan and Zhou (1984), Wadley (1991), and Tulli (1992).
5 In nearly all forms of Chinese dramatic entertainment, one finds a set of specific, conventional roles. Besides the male and female leads, there is often a secondary character called the chou 'clown.' The clown's role is often much more lively than the leads' and his language often has a much more vernacular flavor. In some dramatic forms, the clown will speak in a local dialect while the leads will speak in a formal, almost literary, language. For more information on these conventions see West (1986).
6 Jiao Jie (1993), in a recent article in Zhongguo yuwen, presented thirty-three words he considered loans from Manchu into Chinese and developed a number of rules about how this borrowing process occurred, only to be challenged on nearly all of his alleged loans by two scholars, writing independently of each other, in articles in a later issue of the periodical (Zhou Yimin, Zhu Jiansong ). About the only word accepted nearly universally as being a Chinese loan from Manchu is saqima, a type of confectionery. There were, however, a number of words borrowed into the Beijing dialect which appear in Qing-dynasty texts, then subsequently disappear when the dynasty ends (see section "Manchu loans into Chinese," below). The intriguing question remains: how can Chinese be, as we are discovering, so amenable to loan words on a local level and yet so resistant to them in the standard language?
7 It has in fact been noticed that Chinese aligns itself typologically with many of the world's creole languages. The problem with arguing that this is due to its role as a lingua franca for many disparate Asian peoples as they entered the Chinese cultural "umbrella" is that the language, even from the earliest documents, has never shown much evidence of morphological processes. There is evidence of what appears to be an early case system (e.g., the difference in Classical Chinese between nominative and objective case first-person pronouns), but early Chinese otherwise is almost a perfect example of an isolating language. This does not negate the possibility that it was "creolized" at a very early date, but makes the supposition highly speculative. The lack of morphology in Chinese makes the study of possible Manchu influence all the more problematic.
8 Marked structures are considered stronger evidence of borrowing since they are seldom the result of simplification (see Thomason and Kaufman (1988)).
9 Saqima 'a type of pastry' from Manchu sacima, and henduo 'to upbraid, to reproach' from Manchu hendu 'to speak,' which is missing in the Hanyu da cidian. The Hanyu da cidian, however, contains wadan 'a type of bag' from Manchu wadan, which is missing in the Guoyu cidian. Variations in the characters used to represent words borrowed from other languages makes it difficult to look them up in dictionaries using the traditional format of listing characters.
10 The Qinding shengjing tongzhi (The Imperially Ordered Gazetteer of the Capital), the Jilin tongzhi (Gazetteer of Jilin Province), and the Heilongjiang zhi gao (Manuscript Gazetteer of Heilongjiang Province); the Qinding shengjing tongzhi was produced during the final years of the Qing dynasty, the other two during the first years of the Republic.
Ch'en Chieh-hsien. 1976. The Decline of the Manchu Language in China during the Ch'ing Period (1644-1911). In Altaic Collecta, pp. 137-54. Wiesbaden: Otto Harrassowitz.
Guan Dedong and Zhou Zhongming, ed. 1984. Zidi shu cong chao. Shanghai: Shanghai guji.
Hashimoto, Mantaro. 1976a. Language Diffusion on the Asian Continent. Computational Analyses of Asian and African Languages 3:49-64.
-----. 1976b. The Agrarian and the Pastoral Diffusion of Language. In Genetic Relationship, Diffusion and Typological Similarities of East and Southeast Asian Languages. Pp. 1-14. Tokyo: Japan Society for the Promotion of Sciences.
-----. 1980. Typogeography of Phonotactics and Suprasegmentals in the Languages of the East Asian Continent. Computational Analyses of Asian and African Languages 13:153-64.
-----. 1986. The Altaicization of Northern Chinese. In Contributions to Sino-Tibetan Studies. Pp. 76-97. Leiden: E. J. Brill.
Ji Yonghai. 1993. Manzu zhuan yong Hanyu de licheng yu tedian. Minzu yuwen 1993.6: 38-46.
Li, Charles N. (forthcoming). Language Contact in China: Is Mandarin Chinese Derived from a Pidgin?
Li, Charles N., and Sandra A. Thompson. 1974. An Explanation of Word Order Change SVO-SOV. Foundations of Language 12:201-14.
Norman, Jerry. 1978. A Concise Manchu-English Lexicon. Seattle: Univ. of Washington Press.
-----. 1982. Four Notes on Chinese-Altaic Linguistic Contacts. Tsing Hua Journal of Chinese Studies 1982.1-2:243-47.
Schmidt, Von P. 1931. Chinesische Elemente im Mandschu. Asia Major 7:573-628.
Sun Bang. 1991. Jilin Manzu. Changchun: Jilin renmin.
Thomason, Sarah Grey and Terrence Kaufman. 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley and Los Angeles: Univ. of California Press.
Tulli, Antonella. 1992. Analisi metrico-formale di una raccolta di poesie ibride sino-mancesi. In Aetas Manchuica, ed. Giovanni Stary, no. 3 (1992): 222-89.
Wadley, Stephen A. 1991. Two Mixed-language Verses from the Manchu Dynasty in China, Papers on Inner Asia, no. 16. Bloomington: Research Institute for Inner Asian Studies.
West, Stephen H. 1986. Drama. In The Indiana Companion to Traditional Chinese Literature, ed. William H. Nienhauser. Pp. 13-30. Bloomington: Indiana Univ. Press.
Whinnom, Keith. 1971. Linguistic Hybridization and the 'Special Case' of Pidgins and Creoles. In Pidginization and creolization of languages, ed. Lyle Hymas. Pp. 91-115. Cambridge: Cambridge Univ. Press.
Zhao Jie. 1993. Beijing hua zhong de Man Han ronghe tanwei. Zhongguo yuwen 1993.4: 281-87.
Zhou Yimin and Zhu Jiansong. 1994. Guanyu Beijing hua de Manyu ci (1) (2). Zhongguo yuwen 1994.3: 201-9.
|Printer friendly Cite/link Email Feedback|
|Author:||Wadley, Stephen A.|
|Publication:||The Journal of the American Oriental Society|
|Date:||Jan 1, 1996|
|Previous Article:||Glosses and Commentaries on Aristotelian Logical Texts: The Syriac, Arabic and Medieval Latin Traditions.|
|Next Article:||Prosody or pharyngealization in old Chinese? The origin of the distinction between Type A and Type B syllables.|