國立台灣師範大學翻譯研究所碩士論文 A Thesis Submitted to Graduate Institute of and Interpretation National Normal University

語言變異的(不)可譯性：以中文的文白並陳為例 (Un-) Translatability of Linguistic Variation: Chinese Diglossic Situation as Case in Point

指導教授：賴慈芸 博士 Thesis Advisor: Dr Sharon Lai

研究生： 牟傳門 Advisee: Todd Klaiman

中華民國一O二年一月 January 2013

國立台灣師範大學翻譯研究所碩士論文 A Thesis Submitted to Graduate Institute of Translation and Interpretation National Taiwan Normal University

語言變異的(不)可譯性：以中文的文白並陳為例 (Un-) Translatability of Linguistic Variation: Chinese Diglossic Situation as Case in Point

指導教授：賴慈芸 博士 Thesis Advisor: Dr Sharon Lai

研究生： 牟傳門 Advisee: Todd Klaiman

中華民國一O二年一月 January 2013

Dedications

I would like to dedicate this work to Sunny, Cecibel Martinez and the memory of Sheila Gray. Their love, patience, and support made it possible for me to complete my studies and produce the present thesis.

i Acknowledgments

I would like to thank my thesis advisor Dr Sharon Lai for all her help and support in the preparation and writing of the present work. Also, I would like to thank the staff and faculty of the Graduate Institute of Translation and Interpretation for their support and guidance throughout my studies at the National Taiwan Normal University.

ii Abstract

The present paper addresses (un‐) translatability of linguistic variation as it pertains to literary translation of diglossia from Chinese to English. The paper investigates issues of potentiality and importance of translation of diglossic linguistic variation. Discussion of potentiality is carried out through an extensive review and comparison of major trends in the Translation Studies literature on linguistic variation. The discussion brings into contrast discourse‐ restrictive, yet ubiquitous, “translatability of dialect” issues with a rare, yet discourse‐liberating, notion of “syntagmatic alteration of distance” put forth by Anthony Pym (2000) as basis for translation of linguistic variation. A brief example from ’s (魯迅, 1881‐1936) novella The True Story of Ah Q (阿Q正 傳, A Q Zhengzhuan) (1921) provides an argument for importance. The example focuses on parody and linguistic variation as they function together in the language and qualities of the novella’s main antagonist and his biographer, the narrator, as they are translated out from the historicized Chinese referential frame of the original text into the English‐speaking‐world of the translation. Analysis is carried out across four English‐language : Wang Chi‐ chen’s Ah Q and others: Selected stories of Lusin (1941); Xianyi and ’s The true story of Ah Q (1956); William A. Lyell’s Diary of a madman and other stories (1990); and Julia Lovell’s The real story of Ah­Q and other tales of (2009).

Keywords: Diglossia, historiography, linguistic variation, standard‐with‐dialects, Lu Xun

iii Table of Contents

1. Introduction…………………………………………………………………………………...………. 1 2. Translation Studies and Linguistic Variation 2.1 Early Considerations……………………..……………………………………………. 6 2.2 Minority Languages and Power Relations…………………………………... 10 2.3 Reflexivity and the (Un‐) Translatability of Linguistic Variation….. 10 2.4 Reconsideration of Textual Function…………………………………………. 16 2.5 Beyond Standards‐with‐Dialects……………………………………………….. 18 3. Diglossia 3.1 Definition and Ferguson’s 9 Variables….………………………….. 22 3.2 Classical and Extended Diglossia………………………………….…. 31 3.3 Diglossia and Bilingualism……………………………………………… 33 3.4 Diglossia and Standards‐with‐dialects…………………………….. 36 3.5 Diglossia and Code‐switching………………………………………..... 37 4. Chinese Language Situation 4.1 Chinese Diglossia…………………………………………………...………. 40 5. Parody, Translation, and Chinese Diglossia 5.1 Chinese Historiography and Language Use………………..…….. 45 5.2 Ah Q Zheng Zhuan……...…………………………………………..……….. 47 5.3 Translators and Translations …………………………………………. 51 5.4 Text Excerpt: Analysis and Discussion…………………….………. 60 5.4.1 Text Excerpt (Original)…………………………………...... 60 5.4.2 Text Excerpt (Wang)……………………………………….... 63 5.4.3 Text Excerpt (Yangs)……………………………………...... 64 5.4.4 Text Excerpt (Lyell)………………………………………….. 67 5.4.5 Text Excerpt (Lovell)………………………………...... 69 6. Conclusion………………………………………………………………………………………...... 72 Bibliography………………………………………………………………………………………...... 76

iv 1. Introduction

Within Translation Studies literature, the issue of the (un‐) translation of linguistic variation has received a certain degree of attention (Catford 1965; Newmark 1988; Hatim and Mason 1990; Brisset 1996; Sanchez 1999; Sanchez 2009). The attention, however, has been quite unflinchingly focused upon (un)‐translatability of dialect — in particular, socio‐ and geographic dialect — and the problems associated with it. The present paper would like to suggest that the limited scope of the discussion has caused research into various other aspects of linguistic variation — bilingualism, multilingualism, and diglossia,1 to name a few — to be left underdeveloped. The call for increased awareness and broadened scope for discourse related to translatability of linguistic variation is not new. The issue has been raised before and even with particular reference to diglossia (Muhammad Raji Zughoul & Mohammed El‐Badarien 2004; Meylaerts 2006; Anderman 2007). However, based upon research carried out for the present paper, it seems that the Chinese diglossic situation, in particular, has never before been considered within the Translation Studies Literature. It will be dealt with in the present paper. It remains to be seen whether or not a call to broaden the scope of the discussion will be answered.

A simple example of the limited scope of recent research upon the topic of translatability of linguistic variation may be observed in the limited treatment of the subject in two of Translations Studies’ most general, but comprehensive, resource texts — Jeremy Munday’s (2009) The Routledge Companion to Translation Studies and Mona Baker’s (2011) The Routledge Encyclopedia of Translation Studies. Munday’s text, for example, provides an extensive treatment — written by Basil Hatim (2009:36‐53)— of register‐related issues that pertain, in particular, to register and discourse analytic models imported from applied linguistics and to

1 Diglossia, the focal point of the present paper, will be discussed in further detail in Chapter 3. The term generally describes a language contact situation in which two variants of the same language co-exist in an asymmetrical relationship — across variables of function, prestige, literary heritage, acquisition, standardization, stability, grammar, lexicon, and phonology — within a society (Ferguson 1959:325-340).

1 translation quality assessment models such as those developed by Julian House. With respect to translation of other types of linguistic variation, the book offers a brief discussion of dialect — ideolect and sociolect — in audiovisual translation (2009:158‐159) and a short definition of ‘dialect translation’ as a practice characterized by general adherence to the norm of ‘homogenizing’, or standardizing, the target text language (2009:181). Diglossia as a form of linguistic variation is not mentioned. In Baker’s (2011) The Routledge Encyclopedia of Translation Studies, dialect, register, and style are all presented and dealt with extensively. Diglossia, however, receives only a single, obscure reference in a section of the book dealing with ideology and translation. In particular, diglossia is mentioned in reference to a promising, yet sluggish, increase in awareness of alternative perspectives on translation in non‐Western cultures (2011:140)

The aim of the present paper is to further address through presentation and explication of the Chinese diglossic situation the lacuna that exists within Translation Studies and the discourse on translatability of linguistic variation as it pertains to diglossia. In particular, the present paper will investigate both the potentiality and importance of translation of diglossia as a form of linguistic variation.

In order to address issues relating to the potentiality of translation of diglossia as linguistic variation, the present paper will compare and contrast popular variety‐ for‐variety, equivalence‐fidelity‐based approaches to translation of dialect with more general, register‐based, relative‐distance styled approaches taken from Catford (1965) and Pym (2000) readings of the problem. A comprehensive review and discussion of Translation Studies literature as presented in the subsequent section of the present paper will function to identify an overall trend in the field toward discourse that focuses upon translation of linguistic variation issues that pertain to structuralist notions of equivalence and fidelity to source text. The trend, considered to be out of fashion amongst post‐structuralist translation theorists, will serve as comparison to infrequently referenced, relativist notions based upon the

2 author’s reading of Catford (1965) and Pym’s (2000) treatment of the issue of translatability of linguistic variation. Finally, analysis of a representative instance of literary linguistic variation and its translation will serve to highlight the potentiality of a more register‐based, relative‐distance styled approach to the translation of variation.

The importance of translation of diglossic variation will be addressed by way of analysis of a representative piece of Chinese fiction and its translation into English. Lu Xun’s (魯迅, 1881‐1936) novella Ah Q Zhengzhuan (阿Q正傳, The True Story of Ah Q) (1921) will serve the purpose.2 In her examination of the translingual reinvention of the national character myth in China and its association with the May Fourth literary discourse, Lydia Liu (1995) takes as her point of focus Lu Xun’s Ah Q Zhengzhuan. Liu approaches Ah Q Zhengzhuan, the climactic event of the May Fourth discourse (1995:47), as a means to explore the complexity of 20th century China’s intellectual battle with the seemingly contradictory paths of Chinese tradition and Western modernity. The present paper makes use of Lu Xun’s influential novella for a similar reason. The burden of the Chinese intellectual, as Liu refers to it (ibid.), is tied quite closely to the Chinese diglossic language contact situation.

Early 20th century Chinese intellectuals found themselves positioned precariously at the precipice of modernization. The situation was made all the more urgent by debris from peaks of ancient Chinese history and tradition that had begun to crumble to the ground all around. In “Blossoms in the Snow: Lu Xun and the Dilemma of Modern Chinese Literature,” Theodore Huters explains the scenario, “traditional Chinese thought underwent such severe distortion in the years after 1900 that it could not (or would not) recognize itself” (1984: 75). During first two decades of the 1900s, amidst the influences of Western modernization and strong nationalist concerns, Chinese intellectuals began a drive for the advancement within

2 The writer known by the pen name Lu Xun was born Zhou Zhangshou (周樟壽) in Shaoxing, Zhejiang province, China. Although he also used the name Zhou Shuren (周樹人), he is generally referred to in Western literature and research by his pen name, Lu Xun.

3 China of democracy, science, and literary change (Chen 1999:72). In hopes of thrusting China into the new modernity, intellectuals championed the abandonment of the traditional Chinese diglossic scenario in which the “standard written language (wenyan) was completely divorced from actual speech” (ibid.) — thus, exacerbating the already widespread issue of illiteracy, for a more accessible, vernacular standard (baihua).

It is against the backdrop of linguistic, cultural, and intellectual revolution that Lu Xun, the narrator of Ah Q Zhengzhuan, and the story’s protagonist Ah Q, all co‐exist. As Liu points out, the asymmetric power relation between Ah Q and his narrator functions to foreground a “vast chasm existing between them as members of two different classes” (1995:75). Moreover, according to Liu, “the narrator’s criticisms of Ah Q and condescension, sympathy, and even ambivalence toward him are conditioned by his elevated status as a writer and by his exclusive access to knowledge” (ibid.). In the present paper, Lu Xun’s portrayal in the text of the written‐vernacular diglossic language situation, referred to by Liu above, will form the base upon which an argument for the consideration of Pym’s ‘syntagmatic alteration of distance’ approach to variation translatability will be built.

In the section that follows, Chapter 2, I will present a thorough review and discussion of the Translation Studies literature pertaining to the translatability of linguistic variation. The literature review will help to demonstrate the lack of attention given to translation of diglossia and the Chinese diglossic situation. Also, the discussion presented in Chapter 2 will provide opportunity for further explication of Pym’s concept of syntagmatic alteration of distance.

Chapter 3 will provide review and discussion of the literature pertaining to diglossia. Definitions of the term ‘diglossia’ will be presented and they will also be compared and contrasted with other language contact situations — namely, bilingualism, stadards‐with‐dialects, and code‐switching.

4 Chapter 4 will present further discussion of the term diglossia and its manifestation in the Chinese language contact situation in Mainland China. Unique characteristics of the Chinese diglossic situation will be considered.

Chapter 5 provides a detailed description of Lu Xun’s Ah Q Zhengzhuan, the four English translations, and relevant issues pertaining to Chinese Historiography and language use. Also, Chapter 5 comprises the main analysis and discussion of the text excerpt used to present the Chinese diglossic case. In particular, the chapter will address issues relating to narrative function, parody, Lu Xun’s representation of diglossia, and translation of diglossia.

Chapter 6, the final section of the main body of the present paper, presents conclusions and suggestions for future research into diglossia and the translatability of linguistic variation.

5 2. Translation Studies and linguistic variation

2.1 Early Considerations

Early writings on the subject of translation of linguistic variation, J C Catford’s (1965) A Linguistic Theory of Translation, for example, were concerned with the possibility of finding an equivalent target‐side representation to match the variation found in the source text. Catford’s unique notion of equivalence with respect to language variation is concomitant to his claim that “all varieties of a language have features in common,” but that they also have “features which are peculiar” to each individual variety (1965:86). These peculiar features [phonetic, grammatical, lexical, phonological, graphological, et cetera] “serve as formal (and sometimes substantial) criteria or markers of the variety in question” (ibid.). As such, a distinction can be made between standard, unmarked dialect, which Catford points out, “shows little variation (in its written form at least) from one locality to another, (ibid.)” and marked, non‐standard dialects such as idiolect, geographical dialect, and social dialect. These are the only three forms of non‐standard dialect discussed by Catford. This marked‐unmarked distinction makes it possible for Catford to assert that “equivalence is set up between varieties [of dialects],” without the need for equivalence to occur across homologous features. A unique feature of Catford’s views on the (un‐) translatability of linguistic variation — dialect, register and style are all classified by Catford as forms of language variation — is that “equivalence must be set up between the varieties as such, and the specific markers may be different in the SL and TL texts” (1965:91). This is an important point that I will return to in my discussion of Anthony Pym’s (2000) claims about translation of linguistic variation and syntagmatic alteration of distance.

Eugene Nida (1976) categorizes linguistic variation as an aspect of form (as opposed to content). Linguistic variation is reduced to an issue of style. Nida accepts that both conceptual and formal features can create and influence meaning in the text, but maintains that choices relating to (non‐) translation of formal features should

6 depend upon their relevance to the text and to the text receiver. In particular, he sees dialect as “not rhetorical, but rather [are] role related” (1976:55). According to Nida, dialect gives information about “the roles of those who participate in the discourse or about the settings in which the discourse takes place” (ibid.). I would suggest that this view of linguistic variation, dialect in particular, has persisted in Translation Studies literature at least until the early 1990s and in the practice of translation right up to present day (See Lane‐Mercier 1997, below). In spite of his view of dialect as not related to rhetorical function in the text, or perhaps because of it, Nida touches upon two issues that to‐date have remained central to the (un‐) translatability of linguistic variation debate; namely, authenticity and risk of meaning creation (See Lane‐Mercier 1997, below). In his treatment of the subject a decade earlier, Catford (1965) did not make any mention of either of these two issues. My own contention is that Catford’s notion of equivalence as it relates to the maintenance of relative distance between marked and unmarked dialect enables him to avoid dealing with many of the issues that have come to define the linguistic variation debate.

Peter Newmark (1988) provides an extremely short discussion of linguistic variation. He does, however, introduce a few novel points. Newmark refers to a case in which dialect appears meta‐lingually. It is transferred to the target text as is then translated into neutral language and accompanied by clarification — in the form of footnote, presumably — of its function in the text (1988:195). Neither Catford nor Nida provide any discussion of the possibility of clarification of dialect through the translator’s use of footnotes. Perhaps, this can be attributed to the fact that such visible involvement of a translator into a text is something that did not receive much consideration until well into the 1990s. As Susan Bassnett points out in the preface to the third edition of Translation Studies, “so important has research into the visibility of the translator became in the 1990s, that it can be seen as a distinct line of development within the subject as a whole” (2002:9). Of course, the visibility Bassnett refers to goes well beyond the mere introduction of footnotes by the

7 translator, but such simple exchanges between translator and reader are essential aspects of research into the (in‐) visibility of the translator.

A second novel notion Newmark introduces to the conversation is the issue of the translator’s (un‐) familiarity with a given dialect. Unfamiliarity with dialect(s) is not a problem that is necessarily so difficult to resolve, however. Both Catford (1965) and Pym (2000) point to the fact that in the majority of cases, dialect as it is reproduced in literature is not an exhaustive cataloguing of a real‐world dialect markers, but rather a caricature produced through an exaggeration of commonly acknowledged stereotypes. As such, it is not perfect knowledge of dialect that is important, but rather understanding of stereotypes of dialects. I will return to this point in my discussion of Pym’s (2000) claims regarding the distinction between authenticity and parody of dialect. One final point that seems to set Newmark apart from both Catford (1965) and Pym (2000) is his view of dialect as “a self‐contained variety of language not a deviation from standard language” (1988:195). Such a view does not allow for relative translation approaches such as Catford’s “equivalence across varieties” or Pym’s syntagmatic alteration of distance.

Based on models of register analysis proposed by Halliday, McIntosh, and Strevens (1964) and Corder (1973), Basil Hatim and Ian Mason (1990) focus their discussion of language variation upon geographic dialect, temporal dialect, social dialect, standard dialect, and idiolect (1990:39‐45). As with Newmark, Hatim and Mason, too, bring up new concerns regarding (un‐) translatability of dialect, some of which include: issues of aesthetics, comprehension, ideology, politics, and social implications. Of the various forms of variation and their associated problems of (un‐) translatability mentioned, geographic and social dialect and issues of ideology and politics seem to have dominated the discussion within the Translation Studies literature.

André Lefevere, in relation to theories and practices of literary translation, briefly discusses sociolect, idiolect, and language variants (1992a:64‐70). Lefevere points

8 out that “distinction between sociolect and idiolect is not always easy to make and less easy to maintain” (1992a:67). This is true because, as Lefevere explains, each of us is not only a solitary individual, but also a member of a social group. Language tendencies within a social group affect language tendencies of individuals within the group and, of course, language tendencies of individuals within the group also affect tendencies of the group as a whole. As a result, categorizations of sociolect and idiolect may best be seen as dynamic points along a continuum instead of static, poles of a dichotomy.

The only mention in Lefevere’s Translating Literature: practice and theory in a comparative literature contex of strategy for translation of language variation is “to merely replace one sociolect by another that is perceived to play a similar part in the receiving culture” (1992a:66). In a discussion of what he refers to as language variants (seems to still be talking about sociolect) that follows the sociolect/ideolect section, however, Lefevere does bring up issues relating to comprehensibility, authenticity, and social implication. Moreover, in Translation, Rewriting, and the Manipulation of Literary Fame — also published in 1992 — Lefevere chastises translators of American and Scottish versions of Aristophanes’s Lysistrata for choosing to use Texan and Scots to render “Dorian, a variant of Greek [spoken by the Spartans], which appears to the Athenians as a funny kind of sociolect” (1992a:65). He explains, “neither translator stops to consider either the ‘validity’ or the stereotypes, cultural mechanisms to ‘affirm’ the superiority of one subgroup over another, or the possible anachronistic effect of the use of Scotch and Texan in classical Athens” (1992b:49). It is in a similar vein — concern over possible social implications — that much of the research of the remainder of the decade continues. Although geographic and social dialect remain the main objects of research, issues relating to language‐dialect distinctions begin to arise and bring to the foreground power and nationalist‐related struggles.

9 2.2 Minority Languages and Power Relations

Michael Cronin (1995), Annie Brisset (1996), and John Corbett (1997) explore translation into and out from Irish, Québécois, and Scottish, respectively. Their treatment of the subject presents the three as languages in their own right — minority languages — instead of mere geographical variations of a preferred, standard dialect. Brisset (1996), in particular, highlights the post‐colonial, colonizer‐ colonized struggle concomitant with minority languages and their translation. In a similar vein, Lawerence Venuti (1996) infuses the debate with the idea of the existence of hierarchical power struggles occurring between centre and periphery language variants. Venuti sees hierarchical power relationships existing between standard and non‐standard dialects. He explains, “language is a continuum of dialects, registers, styles, and discourses positioned in a hierarchical arrangement and developing at different speeds and in different ways” (1996:109). As a result of the supposed pervasiveness of these power struggles, Venuti sees any variation away from the norm (standard dialect) as an important means of subversion of “the major form by revealing it to be socially and historically situated” (1996:91).

2.3 Reflexivity and the (Un­) Translatability of Linguistic Variation

If we take Nida’s (1976) search for an equivalent target‐side dialect and Venuti’s (1996) subverting of imperialist norms as two extremes upon a continuum that makes up two decades of discussion of (un‐) translatability of language variation, Gillian Lane‐Mercier’s (1997) treatment of the subject may be seen as a middle ground attempting to reveal the shared inadequacy inherent in these two extremes. To the continuum of views (translator’s responsibility at the semantic, aesthetic, ideological, and political level), Lane‐Mercier’s work adds the idea of responsibility at the ethical level. “Ethical” responsibility, here, refers to “the translator’s own ethical code, his or her responsibility and engagement with respect to the choices for which he or she opts and the aesthetic, ideological and political meanings these

10 choices generate” (Lane‐Mercier’s, 1997:63). Lane‐Mercier seems to propose a sort of ethical subjectivism of translation practices. Stated simply, if we accept the notion — championed by post‐structuralist translation theorists — that translator agency cannot be removed from the act of translation — nor from the translational product, itself — then we must also accept that each individual translator has the right and, perhaps, the responsibility to translate in any way he or she sees fit. As a result, much of the discussion that fills the literature about the (un‐) translatability of linguistic variation becomes irrelevant. In contrast to this expectation for ethical responsibility, Lane‐Mercier suggests that in relation to translation of dialect issues, translation scholars have tended to put forth arguments that reactivate “the very assumptions of fidelity to and transparency of source‐text meaning that post‐ structuralist approaches to translation have sought to combat” (1997:49). In order to support this claim, Lane‐Mercier neatly summarizes the main issues of the sociolect (un)‐translatability debate; namely, risk of meaning creation or loss, risk of ethnocentricity, risk of unauthenticity, and risk of conservatism or radicalism. These all‐too‐ubiquitous arguments do seem to endlessly revolve around questions of equivalence, fidelity, and original meaning.

In a 1999 paper titled “Translation as a(n) (Im)possible Task: Dialect in Literature”, Maria Sanchez takes translation of literary dialect as a case in point to argue for the inclusion of both translatability and untranslatability under a single definition of the term “translation”. Sanchez begins with a brief presentation of the varying views found within Translation Studies literature that take translation as either a possible or impossible proposition. Again, much of the disparity inherent in the varying views can be found to revolve around questions of the aforementioned issues of equivalence, fidelity, and original meaning (Sanchez 1999:301‐302). Based on the notion that translation be perceived as an act of communication, albeit imperfect at times, Sanchez sees translation as “necessary and unavoidable in spite of all its limitations” (1999:301). In other words, although translation may be impossible in theory, the fact that translations have been carried out throughout history and continue to be created today is an irrefutable fact. The aim of Sanchez’s paper is to

11 suggest that a complete definition of the term “translation” must incorporate both the theoretical impossibility and the empirical evidence of its existence. She chooses to present her argument by way of discussion of translation of literary dialect.

In her paper, Sanchez hits upon several important points that continually surface within Translation Studies literature on translation of linguistic variation. Firstly, by way of Sumner Ives’ (1950) discussion of dialect, Sanchez refers to the approximate representation of dialect in literature. As with both Catford (1965) and Pym (2000), Sanchez, with the help of Ives, alludes to the fact that literary dialect is merely a representation and not word‐for‐word reproduction of dialect. In contrast to Catford and Pym, however, Sanchez sees the imprecise nature of literary representation of dialect as a problem for the author/translator to overcome rather than a simplification to the task of representation. For whatever reason, it seems that Sanchez misinterprets Ives’ discussion of dialect. She takes what Ives describes as the problem of dialect to the linguist — “every variation from the conventional system of writing the language is a problem for the linguist” (1950:138 quoted in Sanchez 1999:304) — and incorrectly extends it to include writer and translator — “From Ives' comments we can deduce that, from the start, any form of language which deviates from the norm is a problem for the linguist, the writer and the translator” (Sanchez 1999:304). As can be seen from the original quotation,3 however, Ives did not wish to extend the problematic to either the writer or the translator. On the contrary, in fact, Ives points out that the author’s — and presumably the translator’s — incomplete rendering of dialect is deliberate and artistic. As with Catford (1965) and Pym (2000), Ives’ description of literary dialect as a mere representation of reality offers both writers and translators, alike, a certain degree of freedom with regard to choice of characteristics to be represented

3 “Nearly all examples of literary dialect are deliberately incomplete; the author is an artist, not a linguist or a sociologist, and his purpose is literary rather than scientific. In working out his compromise between art and linguistics, each author has made his own decision as to how many of the peculiarities in his character's speech he can profitably represent” (Ives 1950:138).

12 in the text. This point will be discussed further in the present paper’s section on Pym’s (2000) work.

The second point brought up in Sanchez’s paper is one mentioned above in relation to Hatim and Mason’s (1990) work — namely, ideology. In contrast to Hatim and Mason, however, Sanchez’s mention of ideology as a problem in translation of literary dialect is merely done in passing — “usually, but not always, the language they speak [dialect] carries some connotations of inferiority” (Sanchez 1999:304)— and is only developed slightly further in order to applaud a Spanish translator’s use of Andalusian to represent Pygmalion character Eliza Doolittle’s Cockney dialect. Sanchez’s discussion, in this case, seems to side‐step the issue of creation of meaning — referred to in the present paper with reference to Nida (1976) and Lane‐Mercier (1997). Although dialects from different languages or cultures may appear to have “similar sociolinguistic connotations” (Sanchez 1999:308), in many cases, dissimilar connotations also will exist. The creation of meaning produced through dissimilar sociological connotation is the main issue at the root of the ideological question of translation of dialect.

Two final points discussed by Sanchez include the translator’s degree of familiarity with source language (SL) dialect and the degree of importance of dialectal features for the conveyance of overall effect of the source text. Familiarity with SL dialect is an issue discussed earlier in the present paper with regard to Newmark’s (1988) work. As mentioned previously, I suggest it is not perfect knowledge of dialect that is important, but rather an appreciation of stereotypes and characteristics used by authors in the representation of dialects. With regard to the issue, however, Sanchez does bring to the forefront the importance of a basic understanding of SL dialect in order that translators may avoid semantic mistranslation (Sanchez 1999:305). Sanchez presents examples from two Spanish‐language translations of Wuthering Heights in order to emphasize the point. In both cases the translator misinterpreted the meaning of certain terms in the source text dialect and, therefore, represented them incorrectly in the target text. The basic issue of source text comprehension is

13 one that individual translators must deal with as they see fit. One obvious solution to this problem is for translators to seek out advice from an expert (this could take any number of forms — linguist, native speaker, author, and so on) in the form of dialect being used. It is not, however, an issue that is of concern for the translation of dialect alone, but rather something to be considered in many other kinds of translation scenarios — translation of field‐specific jargon, slang, and cultural terms, for example.

Finally, with respect to degree of importance of dialectal features for the conveyance of overall effect of source text, Sanchez falls back upon something similar to what Pym (2000) calls the easy answer: Often use of target‐side standard language to replace source‐side dialect is a choice in favor of ease and economy (1999:306). As Pym points out, however, the crux of the issue is for the translator to “know what [linguistic] varieties are doing in cultural products in the first place” (2000:69). Perhaps, discussion of dialect in the depth necessary to tease out such subtleties is beyond the scope of Sanchez’s paper. After all, her investigation of issues relating to translation of dialect is meant to highlight the assumption that translation is at times possible, but in other instances impossible. As mentioned earlier, Sanchez’s paper takes translation of dialect as the obvious starting point from which to add weight to her claim, “any definition of translation should never exclude the possibility or impossibility of language transfer, but should include both aspects as unavoidable and complementary” (1999:309).

One final point I would like to add is in her 2009 book, “The problems of literary translation: a study of the theory and practice of translation from English into Spanish”, Sanchez does make a slightly more thorough study of dialect and its translation. Again, however, her investigation of specific examples — Shaw’s Pygmalion; Brontë’s Wuthering Heights; and Lawrence’s Lady Chatterley’s Lover (2009:205‐228) — is aimed toward the goal of providing support for her claim that any definition of translation must include aspects of both possibility and

14 impossibility. Sanchez’s work does introduce into the translation of dialect debate a couple of novel issues.

Firstly, in a section dedicated to discussion of the “use of non‐standard language for literary purposes” (2009:197), Sanchez — by way of a quote taken from Altano (1988) — presents the notion, “when we speak of a literary work written in non‐ standard – or dialectal – language, we usually mean that the non‐standard variant has been used for direct speech, while the standard norm has been used in the narrative passages” (ibid.). Sanchez goes on to suggest universal applicability (ibid.) of Altano’s claim that Italian dialectal novels almost always use standard language for narrative voice and dialect “for the scope of characterization” (1988:152). This seems like quite an oversimplification since, for example, obvious counter‐examples in English — Twain’s Adventures of Huckleberry Finn and Irvine Welsh’s Trainspotting — are readily provided. I would suggest the universality of Altano’s distinction remains to be shown.

Secondly, as in her 1999 paper, Sanchez quotes Ives (1950) reference to the inauthenticity of literary dialect. Although in her 2009 book she does not make the same erroneous extrapolation to apply Ives quote to translator and author as she does in her paper, she does still, however, seem to miss the point that literary dialect is deliberately created as a caricature of the real‐world language variant. This is a point that will be developed further in the present paper’s discussion of Pym’s (2000) claims regarding the distinction between authenticity and parody of dialect. As mentioned earlier, appreciation of the function of linguistic variation in a literary text is of utmost importance with regard to any meaningful discussion of the translation of variety (Pym 2000:69). Although this is not a novel issue, it does seem that, as it is with Sanchez, a great deal of Translation Studies literature on the subject still posits faithful replication of reality as one of the goals of the author who uses literary dialect in his or her work. As a result, anything short of “faithful replication of reality” — whether it be source‐side or target‐side reality — by the translator is considered, as Sanchez suggests, proof of an impossibility of translation.

15 The issue seems to bring us back to Lane‐Mercier’s suggestion that in relation to translation of dialect issues, translation scholars have tended to put forth arguments that reactivate “the very assumptions of fidelity to and transparence of source‐text meaning that post‐structuralist approaches to translation have sought to combat” (1997:49). The twelve years between Lane‐Mercier and Sanchez’s publications do not seem to have moved Translation Studies much distance away from issues of faithfulness and fidelity.

2.4 Reconsideration of Textual Function

If Lane‐Mercier’s discussion of sociolect can be said to have been the pinnacle of the linguistic variation (un)‐translatability debate’s swing away from the semantic moving toward the aesthetic to the ideological, the political, and the ethical, then Ineke Wallaert’s (2001) discussion of the illocutionary force inherent in sociolect brings us full‐circle back to the textual. Although Wallaert does not make light of ideological and political concerns referred to in research that precedes her work, she does feel that “considerations of the textual functions of literary sociolects should precede questions regarding the translator’s ideological responsibility (2001:171). Wallaert returns the conversation to the question of the function of sociolect in the text. In particular, she points to the self‐referential, illocutionary nature of sociolect to “say something about language by using a particular form of language” (2001:177). Wallaert argues that although “the way literary sociolects are translated does indicate the translator’s ideological stance” (2001:171), it is equally, if not more, important to recognize that literary sociolects function in texts as meaning‐creating and foregrounding devices. It is in this line of argument that Wallaert brings up the notion of the illocutionary effect of literary sociolects. In her article, Wallaert focuses the discussion entirely on self‐reflexive texts, ones in which the language of the text is used to point directly to, and overtly make comments on, the nature of language itself. For example, Wallaert looks at Baudelaire’s translation of Edgar Allen Poe’s The Gold Bug in order to raise questions about the validity of

16 ascribing all sociolect‐related translation choices to ideological considerations. Wallaert suggests Baudelaire’s decision not to translate Poe’s sociolect may have been based purely upon issues related to the translator’s “poetical standards” or his assessment of the difficulty of finding “a suitable equivalent” (2001:181). Oddly, both of these reasons seem obviously ideological in nature. Although this aspect of Wallaert’s argument is not convincing, I would suggest her assertion that such assessments can only be addressed through careful examination of both textual and meta‐textual analysis (ibid.) is not only valid, but also a valuable counterbalance to the ideologically‐politically‐heavy research trend of the past two decades.

Anthony Pym’s (2000) discussion of linguistic variation was published a year before Wallaert’s research, but I present it slightly out of order here because of its uniqueness in comparison to the work that precedes it, excluding, perhaps, Catford’s (1965). Pym’s perspective is reminiscent of Catford’s in that they both emphasize the distinction of dialect as a functional representation of variety and as a shift away from an established norm (Catford 1965:83‐86; Pym 2000:70‐71). The main premise of Pym’s argument is that linguistic variation is usually represented in literature in a way that approaches either parodied language at one extreme or authentic language at another. Pym’s use of the term “parody” refers to the fact, as mentioned previously, that dialect as it is reproduced in literature is generally just a caricature of a real world variation accomplished through exaggeration of commonly acknowledged stereotypes. According to Pym, authentic language, on the other hand, “is the multiplication of variations beyond anything that the popular imagination can identify” (2000:70). In other words, a given variety is represented in such detail that the receiver, overwhelmed by the range and depth of the reproduction, accepts it as authentic without necessarily understanding it completely. In regard to the two extremes of linguistic variation described thus, Pym explains that their creation within literature depends upon a “rapid shift away from an established norm,” and that this norm is one that is established within “the particular cultural product in question” (2000:71). Based upon these two points — dialect as a functional representation and as a shift away from a norm — as Catford

17 did nearly four decades before him, Pym argues that it is the “relative deviation from the norm” that should be rendered in translation, not a given source‐text variety (2000:72). Pym refers to this as syntagmatic alteration of distance. In order to investigate the potentiality of translation of linguistic variation, the present paper takes as its focal point syntagmatic alteration of distance as translation strategy. In particular, I will explore the effectiveness of this strategy to deal with parodic function as created through manipulation of the Chinese diglossic language scenario.

2.5 Beyond Standards­with­Dialects

As can be seen in the review of literature presented thus far, discussion of (un)‐ translatability of linguistic variation has focused mainly upon geographical and social dialect and the issues relating to their translation. A variety of issues — semantic, aesthetic, ideological, political, and ethical — have been addressed, but the discussion has been confined to an unrealistic, single‐user‐single‐variant context. By “single‐user‐single‐variant”, I refer to the status of the presumed speaker of a geographical or social dialect taken as theoretical object upon which most debate has been focused. The dichotomy created in much literature and, as such, found to pervade the (un)‐translatability of linguistic variation debate is that of a mono‐ lingual/mono‐variant speaker of one dialect — standard or non‐standard; majority language or minority language, central or periphery, and so on — placed in contrast to a mono‐lingual/mono‐variant speaker of another. Although the community in which the interaction takes place may be multilingual, the actors within the community are usually taken to be mono‐lingual/mono‐variant. The contrast between mono‐lingual/mono‐variant speakers of opposing dialects is then played out upon any number of levels including, but not limited to, semantic, aesthetic, ideological, political, and hierarchical. It seems that the mono‐lingual/mono‐variant speaker is an oversimplified point upon which to base any discussion of (un)‐ translatability of linguistic variation because: Firstly, mono‐lingual/mono‐variant agents are not the only ones represented in works of literature; and secondly, many

18 language communities are not composed of only mono‐lingual/mono‐variant speaking individuals.

Within the Translation Studies literature, there have been a small number of scholars who have explored linguistic variation from a point of entry other than the usual standard‐with‐dialects or minority‐majority dichotomies. I will discuss briefly four: Ben‐Z Shek (1977, 1988), Muhammad Raji Zughoul and Mohammed El‐ Badarien (2004), Jane Wilkinson (2005), and Reine Meylaerts (2006). Shek is one of the first to investigate the effects of diglossia on translation. Shek’s research takes up the issue of imbalance in literary translation in Canada (French translated into English much more highly represented than English into French). As way of explanation for the imbalance, Shek proposes the diglossic historical relationship between the nation’s two official languages — French and English. It is important to point out that Shek explores, as he puts it, “Quebec’s socio‐cultural evolution” and not just a static moment, for example, in present‐day Canadian society. This is an important distinction because it is not entirely clear whether or not modern‐day Quebec (post‐1974, when French was made the only official language) meets the requirements to be categorized as a diglossic community. However, viewed over the full course of its history, Quebec would probably make an interesting example of what Ferguson (1959) refers to as an evolution away from diglossia toward a standard‐with‐dialect or, perhaps, a majority‐minority situation. Regardless, Shek’s work helps to balance an otherwise monolingual‐heavy discussion of (un)‐ translatability of linguistic variation.

In contrast to the questionable categorization of Quebec as a diglossic community, research on translation and the diglossic nature of Arabic carried out by Muhammad Raji Zughoul and Mohammed El‐Badarien (2004) can without a doubt be viewed as related to diglossia in the classic sense of the word. In response to their view that “treatment of variation has always been restricted to “dialect” and has not encompassed the notion of diglossia” (2004:447), Zughoul and El‐Badarien investigate equivalence of variety to context as it pertains to translation into the

19 Arabic diglossic situation. Although it is quite prescriptive in nature, their conclusion — “use of the wrong variety in translating a text not only fails to transfer the intended meaning but also distorts the message” (2004:454) — might still be instructive in that it refers to a diglossic community that Ferguson, himself, described as belonging to a category in which prestige of the high language is such that “H alone is regarded as real and L is reported ‘not to exist’” (1959:29). Of course, not all members of a community will hold such a view, but it surely will affect both meaning and reception at various levels of society.

Finally, both Jane Wilkinson (2005) and Reine Meylaerts (2006) return the focus back to issues related to national identity. Wilkinson explores theatre translation in German‐speaking Switzerland. In particular, she discusses translation from H to L variety within a language community that meets an extended definition of diglossia as opposed to a classical one. The distinction between extended and classical diglossia is that Ferguson’s original categorization did not include genetically unrelated languages while later research in both psychology and sociology — John Gumperz (1961, 1962) and Joshua Fishman (1967) — helped to extend the definition of diglossia to allow H and L to be represented by genetically unrelated or historically distinct languages. Wilkinson’s argument, with respect to such a diglossic community — in particular, modern‐day German‐speaking Switzerland — holds that translation choice often reflects an L‐side desire to promote local and national identity and to resist H‐side culture (2005).

In a similar vein, Reine Meylaerts (2006) delves into some of the issues relating to the struggle for equality — or, perhaps, dominance — that often arises between H and L language varieties in a diglossic setting. Meylaerts explores translation of Flemish novels into French within the heteroglossic context of Belgium during the 1920s and 30s. The unique heteroglossic nature of the context is such that monolingual, H‐variety‐French speakers comprise the target audience for which multilingual Flemish‐French speaker‐produced L‐variety Flemish novels are translated. Much like Shek’s research into “Quebec’s socio‐cultural evolution,”

20 Meylaerts work investigates within a diglossic setting an L‐variety language in the early stages of resisting H‐language‐variety dominance. The study of translation — in all its permutations — as it exists across an evolving — both in space and time — diaglossic community is an interesting area that to‐date has not received much attention in the field of Translation Studies. For the time being, however, the present paper will deal with just one instance of translation from a diglossic community — Mainland China — to a non‐diglossic one — North America and the UK.

21 3. Diglossia

3.1 Definition and Ferguson’s 9 Variables

In order to expand the focus of the discussion, the present paper looks at a contrasting — but, perhaps, more realistic — scenario in which speakers have access to more than one dialect at a time — namely, diglossia. Linguist Charles Ferguson (1959) coined the term diglossia in order to refer to a situation in which “two varieties of language exist side by side throughout the community, with each having a definite role to play” (1959:25). A more detailed definition is as follows:

A relatively stable language situation in which, in addition to the primary dialects of the language (which may include a standard or regional standards), there is a very divergent, highly codified (often grammatically more complex) superposed variety, the vehicle of a large and respected body of written literature either of an earlier period or in an other speech community which is learned largely by formal education and used for most written and formal spoken purposes but not used by any sector of the community for ordinary conversation (Ferguson 1959:34‐35).

In his investigation of this phenomenon, Ferguson focuses upon four diglossic language communities: namely, Arabic, Modern Greek, Swiss German, and Haitian Creole (1959:26). According to Ferguson’s classification system, separate varieties in a diglossic community are called either “H” for the high variety — the superposed dialect, or “L” for the low variety — standard or regional dialects. In his 1959 paper, Ferguson describes 9 variables — function, prestige, literary heritage, acquisition, standardization, stability, grammar, lexicon, and phonology — inherent to a diglossic language community.

First, Ferguson outlines issues relating to the function variable. According to Ferguson, H and L varieties of diglossic languages can be differentiated on the basis of a disparate specialization of function. As Ferguson, himself, puts it “in one set of situations only H is appropriate and in another only L” (1959:29). It is important to note that Ferguson admits there is a certain degree of overlap between the two sets

22 of functions, but he notes that it is minimal. In a 1991 paper revisiting the concept of diglossia, Ferguson clarifies his position on the issue of function, ‘‘If we assume that there are two basic dimensions of variation in language, dialect variation correlating with the place of the speaker in the community and register variation correlating with occasions of use, then the H and L varieties of diglossia are register variants, not dialect variants’’ (222). The uncompromising distinction between diglossia as strictly use‐related and other forms of variation — standards‐with‐dialects and societal bilingualism — as user‐related has met with both support and criticism.

Alan Hudson, who has provided one of the most exhaustive bibliographies of diglossia to date (1992),4 proposes distinction between use‐specific and user‐ specific variation as a means by which to differentiate diglossia in the classic — Fergusonian — sense of the term from other sociolinguistic phenomena that have come to be associated with it (Hudson 2002:2). According to Hudson, “diglossia, in its ideal form, may be conceived of as the quintessential example of linguistic variation where linguistic realization as opposed to language acquisition — here, grossly oversimplified, the use of H or L — is a function solely of social context, and not of social identity of the speaker” (Hudson 2002:6). It has been argued, however, such a view is based upon an unrealistic separation of social context and social identity. Henry and Renée Kahane, for example, regard the diglossic matrix of H and L varieties as a “linguistic representation of a class system” (1979: 183).

The Kahanes argue that H — the prestige language — is exclusively used in and by a privileged sector of society distinguishable from the rest of the language community by its advantage in power, education, manners, and heritage (ibid.). Hudson seems to acknowledge a class‐based bias present in acquisition and opportunity for use of H within diglossic language communities. He recognizes that “whereas social access to the informal situations in which L is appropriate is universal, access to those

4 Hudson’s bibliographic review of diglossia contains 1,092 entries spanning from the time of publication of Ferguson’s first article in 1959 to publication of Hudson’s bibliography in 1992. The following year, Mauro Fernandez (1993) published a bibliography of diglossia that contains 2900 entries.

23 more formal situations in which H is appropriate is asymmetrically distributed in favor of those educationally privileged, literate, or otherwise specialized classes in society most likely to have had the opportunity to acquire H formally” (Hudson 2002:5‐6). However, Hudson goes on to make a distinction between opportunity for H‐L acquisition and use, on the one hand, and actual use of H‐L, on the other. According to Hudson, “social‐class stratification of the use of H and L (or its absence) must be distinguished from social stratification of the opportunity to acquire H, and from social stratification of access to those social contexts in which H is appropriate” (Hudson 2002:5). I would agree that H‐L use is not directly influenced by social‐ class stratification, but it seems that apart from an idealized case in which every member of a community knows all varieties equally well all other cases would exhibit obvious indirect social‐class influence at work. Although Ferguson (1996) has reiterated his original claim that H and L varieties of diglossia are use‐related variants rather than user‐related, he has gone on to suggest the importance of considering class difference and power differential issues (1996:60). In particular, he asks, “how widespread is competence in the H variety and which social groups have greater access to it” (ibid.)?

In reference to the question of power differentials, Ferguson alludes to an example — of special interest to the present paper — that may be described as ‘literary diglossia’. Ferguson cites the case of early twentieth century Arab playwright Mikhail Nu’aimeh who used literary diglossia in his plays. Ferguson explains, “Mikhail Nu’aimeh had the educated people speak H and the less educated people speak L, although in real life everyone spoke L in the situations presented in his dramas’’ (Ferguson 1996:61). In another example, literary diglossia is discussed in relation to classical Sanskrit. Researchers (Hock and Pandharipande 1978; Lee 1986; Deshpande 1991) have described classical Indian drama in which Sanskrit is spoken by the educated, upper class, while various other L varieties are spoken by the uneducated, lower class. In reference to Deshpande’s (1991) paper, Ferguson explains:

24 In an attempt to remain loyal to their literary tradition and yet portray realistic dialogue, Sanskrit dramatists chose to have kings and priests speak in the H variety and the people of lower status speak in L varieties, although of course in real life everyone spoke in one or another L variety and used H only for religious, literary, and official documents and certain public occasions (Ferguson 1996:61).

Similar to Sanskrit dramatists’ juxtaposition of H and L to contrast upper and lower classes in their plays, in the case of English, for example, Shakespeare used L variety to portray certain characters as “rustic, comical, uneducated,” and so on (Schiffman 1998:207). Although Hudson dismisses literary portrayal of a social‐class stratified diglossia as nothing more than a “dramatic device that does not accord with contemporary social reality” (Hudson 2002:5), representation through literature of an artistic, characterization of linguistic variation is precisely the parodic treatment Pym (2000) points to in his description of literary linguistic variation as syntagmatic alteration of distance. Thus, in the case of the present paper, I feel it justifiable to move on with Ferguson’s variables and to leave as a stalemate the discussion of diglossia as horizontal, situational stratification of H and L versus diglossia as vertical, social‐class stratification of H and L (Hudson 2002:5).

Ferguson’s second area of discussion in relation to diglossia is prestige. According to Ferguson, in a diglossic scenario, the H variant of the language in question is regarded by all speakers to be superior to the L variant “in a number of respects” (1959:30). In particular, Ferguson notes that H is often believed to be “more beautiful, more logical, better able to express important thoughts, and the like” (1959:31). To a certain extent, the prestige of the H variety is closely linked to the value ascribed to the domain of H. ‘Domain’ refers to the linguistic behaviour for which either H or L varieties are accepted as appropriate as prescribed by the language community norms. As Schiffman explains, “these domains are usually ranked in a kind of hierarchy, from highly valued (H) to less valued (L)” (1998:205‐ 206). In his discussion of the weaknesses of his original paper on diglossia, Ferguson notes that he has been criticized for not making a clearer distinction between people’s behaviour with respect to H‐L variety use and their attitudes toward them

25 (1996:59‐60). People’s attitudes toward the two varieties are an important factor defining prestige. According to Ferguson, in many diglossic language communities, “people will say that the L variety has no structure, no grammar, no rules, and it is only chaos” (1996:59). Of course, attitudes will differ both across and between speech communities. In many cases, one can conceive of fairly politicized attitudes developing toward H and L — H‐favouring bourgeois and L‐favouring proletariat sentiment, for example. However, as Ferguson shows with Greek and Arabic — L favoured by Greek left wing groups, whereas Arab left wing groups favour H — attitudes toward H and L may exhibit case specificity and should not be assumed to be generalizable over all diglossic scenarios (1996:60). Most importantly, questions about attitude — “Which comes first, the language variety or the feelings toward the variety?”, “How did the attitude get set in place?”, and “How did the variation get set in place?” (ibid.) — may help to untangle issues related to use, value, and prestige.

Literary heritage is the third variable discussed by Ferguson. As with all of Ferguson’s variables, there is obvious overlap and interconnectivity between literary heritage and prestige. According to Ferguson, in each of the four defining languages explored in his work, not only is there “a sizable body of written literature in H which is held in high esteem by the speech community,” but also, “contemporary literary production in H by members of the community is felt to be part of this otherwise existing literature” (1959:31). Both Ferguson (1959) and Schiffman (1998) emphasize literary traditions written in the L variety can and do exist — dialect poetry and folk literature, for example, but such works are usually held in much lower esteem than H variety counterparts. Also, Ferguson does point out that although “all clearly documented instances [of diglossia]” known to him are found in “literate communities,” it is possible that diglossia could develop in a non‐ literate language community (1959:41, note 18). In the case of a non‐literate diglossic community, a body of oral literature would function as the community’s literary heritage. Role played by writing systems and literacy in development, maintenance, and disappearance of diglossia has been an issue much discussed in Sociolinguistic literature (Ferguson 1959; Coulmas 1987, 2005; Walters 1996).

26

Ferguson cites both literary heritage — specifically, a body of literature both in a language related to “the natural language of the community” and that embodies “fundamental values of the community” — and restricted literacy as conditions likely to lead to the establishment of diglossia (1959:37). As Schiffman points out, in the case of restricted literacy, although not all members of the community attain mastery of the H variety, it is still expected that H will be used in all H‐variety domains. Schiffman explains, “this does not mean that illiterates have the option of using the L variety in H‐variety domains; rather, the expectation is that they will remain silent rather than exhibit inappropriate linguistic norms” (1998:211). It is clear that under such circumstances, class, social, and power‐related differences between members of the community competently literate in H and those not would continue to accumulate. Walters describes the restricted literacy upon which diglossia is both established and maintained as “involving the written variety of a language that becomes increasingly distant (and therefore distinct) from the native variety of language spoken in a speech community that is overwhelmingly illiterate’’ (1996: 161–162). In light of such descriptions of diglossia and restricted literacy, it is to be expected that diglossia would come to be seen as a language scenario that not only establishes but also maintains social‐class disparity among members of a language community. According to Coulmas, however, although it is true, “diglossia has been characterized as a language situation that not only coincides with but causes and helps to maintain high levels of illiteracy, thus reinforcing social inequality,” it is also the case that in certain parts of the world today, diglossia continues to exist even in communities in which widespread literacy is the norm (2005:128). Examples of diglossia existing in contemporary communities where illiteracy has ceased to be a social problem include: Switzerland (Ferguson 1959; Keller 1982) and Hong Kong (Snow 2010a; Snow 2010b). Snow’s discussion of diglossia in China will be dealt with further in Section 4.1 — Chinese Diglossia — of the present paper.

27 Acquisition is the fourth of Ferguson’s variables of diglossia. In Ferguson’s view, in a diglossic community, L language varieties are learned naturally, through informal interaction with one’s parents, peers, and local community. H‐varieties, on the other hand, are learnt through formal schooling and by way of grammar rules and norms for imitation. The unique case of acquisition in the diglossic language scenario has been shown to be one aspect that can be used to differentiate diglossia from bilingualism, for example. As Timm explains, “acquisition of the L variety by all members of the speech community prior to the acquisition of the H variety is not pertinent in bilingual settings” (1981:362). In some cases, bilinguals acquire H first. Moreover, in bilingual language communities, since H is a native vernacular for some members, not everyone in the community will necessarily learn L. Differences between diglossia and bilingualism will be discussed further in Section 3.3 — Diglossia and Bilingualism — of the present paper.

In much of the Sociolinguistic literature, strong connectivity has been suggested between function, acquisition, and stability (Britto 1991; Hudson 2002). Britto, for example, describes function and acquisition as the two most important defining principles of diglossia (1991:60‐84). According to Hudson, the conceptual unity found in descriptions of diglossia by Ferguson and others is to be found in “a quite specific set of relationships between functional compartmentalization of codes, the lack of opportunity for the acquisition of H as a native variety, the resulting absence of native speakers of H, and the stability in the use of L for vernacular purposes” (2002:40). Simply stated, both echoing and elaborating upon Britto’s claim, Hudson sees two variables — namely, difference in method of acquisition of H as the superposed variety and L as the native variety and differentiated of H and L in terms of specialization of function — as not only constituting the diglossic scenario but also leading to its maintenance and stability. In addition to the two important variables of function and acquisition, I would like to add Ferguson’s fifth variable, standardization.

28 Ferguson’s fifth variable of diglossia, standardization, describes the establishment within a diglossic language community of a strong tradition of grammars, dictionaries, canonical texts, academic fields of enquiry, and the like, related to the H language variety. Schiffman adds, “L is rarely standardized in the traditional sense, or if grammars exist, they are written by outsiders” (1998:207). In this sense, standardization is closely linked to acquisition. According to Coulmas, diglossia may arise in language communities within which the written language (H) is “cultivated by a small caste of scribes and not allowed to change in the direction of the language of the people” (1989: 13). The vernacular language (L) of the people, Coulmas continues, is “not influenced by writing as they are largely illiterate” (ibid.). Lacking in a standardized, explicitly stated grammar, L is bound to its position as vernacular, unsuited to institutions of prestige — namely, learning, literature, religion, and politics. Standardization leads to further partitioning and compartmentalization of function and mode of acquisition of H and L; this, in turn, leads to stability — Ferguson’s sixth variable.

Stability, Ferguson’s sixth variable, follows quite logically from the previous five variables. With functions of H variety, for example, encompassing a language community’s most important, influential, and furthest reaching forms of communication — religious liturgy and sermons, political documents and speeches, pedagogic materials, news media, literature, and so on, L varieties lack significant opportunity to displace it. Prestige and literary heritage only add to the stability of the H variety’s position of superiority over L within the hierarchy of the diglossic scenario. Finally, both formal acquisition and standardization of H serve to ensure the consistency, stability, and effective transmission of it over time. According to Ferguson, “diglossia typically persists at least several centuries,” and he adds, “evidence in some cases seems to show that it can last well over a thousand years” (1959:33).

The final 3 variables — grammar, lexicon, and phonology — differ from the previous 6 variables in that they are demonstrably linguistic in nature, rather than

29 sociological or sociolinguistic. Hudson classifies Ferguson’s 9 variables of diglossia as follows: “Function, prestige, acquisition, literary heritage, and standardization are all in one way or another contextual aspects of diglossia,” and he adds, “grammar, lexicon, and phonology are linguistic aspects” (2002:9). According to Hudson “stability, clearly, refers to the temporal axis along which the various contextual and linguistic variables may vary” (ibid.). The notion of stability as the axis along which the other 8 variables vary helps to locate Ferguson’s final 3 variables — grammar, lexicon, and phonology — within the idea of connectivity I briefly outlined above in relation to stability and its relationship to the first 5 variables.

In his discussion of the grammar variable, Ferguson presents two possible levels of distinction between H and L. First, he suggests a conservative, general statement explaining, “in diglossia there are always extensive differences between the grammatical structures of H and L” (1959:33). I refer to this statement as conservative because it — quite wisely, perhaps — avoids any sort of value judgement as to the nature of the differences between the two grammatical structures. Before moving on to a second, less conservative statement about the distinction between H and L grammars, Ferguson acknowledges, “it is always risky to hazard generalizations about grammatical complexity” (ibid.). With the salutary caveat in place, he ventures to suggest, “in at least three of the defining languages [Arabic, Greek, and Haitian Creole], the grammatical structure of any given L variety is simpler than that of its corresponding H” (1959:34). Ferguson’s assessment of grammatical complexity is based upon four factors: morphology, gender, agreement, and syntax. Validity of assessment of complexity based on these four factors may be called into question since the criteria seem to be assumed to be somehow applicable universally across all languages. In fact, depending upon the language there could be a variety of factors differentiating H from L in terms of structure. Obviously, each case must be considered individually.

Finally, with respect to Ferguson’s variables of lexicon and phonology there seems to be much less clarity of distinction between H and L variants. According to Ferguson,

30 H and L share a great deal of vocabulary (1959:34). However, since the two variants differ quite extensively in function within the diglossic community, each necessarily adapts its own specialized vocabulary not present in the counter‐variant. As Ferguson points out, an interesting feature of diglossia is the existence of a disparate pairing of H and L lexical structures that refer to a single common item (ibid.). The point, taken alone, may be seen to be nothing more than a difference of register such as that found in English words such as trip/excursion and discover/ascertain, for example. Ferguson makes clear, however, differences in register are quite different from H and L lexicon pairing disparities because possibility for overlap in use in the case of the former is much greater than in the case of the latter (ibid.).

As it is with lexicon, the variable of phonology within Ferguson’s definition of diglossia presents a parameter that is difficult to generalize. Ferguson offers a description of two possible scenarios. First, H and L share the same phonological elements, but H may have divergent features that are more complicated morphophonemics of L (1959:35). Second, H may contain phonemes not found in L, but L phonemes may be used to substitute in oral use of H (1959:36). Regardless into which scenario a diglossic community may be grouped, it is obvious that phonological differences alone may not be a discrete enough distinction upon which to base H and L categorization. As such, it is important to consider together all 9 of Ferguson’s variables of diglossia.

3.2 Classical and Extended Diglossia

Thus far, discussion in the present paper has taken as its point of focus Ferguson’s original definition of diglossia and its 9 variables. Much of the research found in the two most extensive bibliographies of diglossia (Hudson 1992; Fernandez 1993) to date, however, has been concerned with refinement and extension of Ferguson’s original formulation. The most well known extension came from the work of Joshua A. Fishman (1967). According to Timm, Fishman’s reworking of Ferguson’s original

31 definition has lead to two unfortunate outcomes: Firstly, it has created an entire body of work, from various authors, that has, without reference to Ferguson’s original work, depended entirely upon Fishman’s characterization of diglossia (1981:356). Since Timm (1981) questions the validity of Fishman’s extended definition of diglossia, it is quite obvious that he should criticize the work of authors who take Fishman instead of Ferguson as their starting point. Secondly, Timm sees Fishman’s work as creating “conceptual fuzziness” about the definition of diglossia and about the “criteria” by which diglossia is meant to be recognized (1981:356). Timm is not the one to criticize Fishman’s attempts to extend Ferguson’s original notion of diglossia. Schiffman, for example, claims that since diglossia, as a “gradient, variable phenomenon,” was never meant to be “boxed into an either‐or, binary system of categorization,” it is neither useful, nor desirable, to try to “`refine' or `extend' diglossia, or to discern whether such and such is or is not a case of diglossia” (1997:208). Schiffman’s criticism is quite obviously leveled at the work of Fishman and his supporters. Schiffman follows the aforementioned statement with an explanation of Fishman’s extended definition of diglossia and a discussion of the differences between classical — Fergusonian — and extended diglossia. Moreover, it does seem that Fishman’s rendering of diglossia to include bilingualism gave rise to a great deal of research (Rubin 1968; Gumperz 1971; Fishman 1971; Denison 1971; Mkilifi 1972) aimed at, as Schiffman put it, discerning “whether such and such is or is not a case of diglossia” (1997:208).

As Fishman points out in the introduction to his 1967 paper, “Bilingualism with and without Diglossia,” his main concern is “to relate diglossia to psychologically pertinent considerations such as compound and co‐ordinate bilingualism” (30). As Timm understands it, “the main purpose of Fishman' s 1967 article was to set forth systematically the logically possible relationships between diglossia and bilingualism and to indicate speech communities thought to represent the different combinations of bilingualism and diglossia” (1981:357). Fishman’s reworking of diglossia begins with his restatement of Ferguson’s original definition. According to Fishman, diglossia was initially considered in relation to “a society that used two or

32 more languages for internal (intrasociety) communication” (1967:29). As Timm makes clear in his criticism of Fishman, the extension of the definition of diglossia to include a language situation in which two or more languages, rather than two variants of a single language, in a society goes beyond the scope of Ferguson’s original intention for the term. As Ferguson himself clearly states in reference to his original paper, “no attempt is made in this paper to examine the analogous situation where two distinct (related or unrelated) languages are used side by side throughout a speech community, each with a clearly defined role" (1959:40, note 2). In other words, Ferguson did not intend for diglossia to be applied to the kind of bilingual language situation to which Fishman attempts to extend it. Finally, Fishman’s retelling of Ferguson’s original concept goes further astray in that he reduces to one — function — Ferguson’s original 9 variables. As Timm explains, “Fishman mistook this [domain specialization] as the sole criterion defining diglossia” (1981:357). Some researchers (Gilbert 1975; Timm 1981; Hawkins 1983; Kaye 2001; Hudson 2002) have criticized Fishman’s extension as weakening the term to the point of uselessness. In other words, as Kaye explains, categorization of societal bilingual language situations as a form of “extended diglossia,” as Fishman would have it, “in essence, leads to labeling all speech communities diglossic” and, therefore, causes the term to lose any significance of differentiation (2001:121).

3.3 Diglossia and Bilingualism

Since Fishman’s (1967) discussion of diglossia that included within the scope of the term — extended diglossia — novel notions regarding bilingualism and diglossic speech communities comprised of both genetically and non‐genetically related languages, researchers (Hawkins 1983; Hudson 2002; Coulmas 2005; Marfany 2010) have attempted to clarify the relationship between diglossia — and/or extended diglossia — and bilingualism. In many cases, the two are seen to differ in regard to their relation to Ferguson’s variables of function and acquisition.

33 According to Hawkins, diglossia is “not a uniquely monolingual phenomenon, but has affinities with bilingual and multilingual situations” (1983:3). It is important to note that although Hawkins claims diglossia does not describe a purely monolingual scenario, he does not go as far as to classify it as entirely bilingual either. He merely professes that diglossia possess some similar characteristics to bilingual or multilingual scenarios. The distinction is important because it means that Hawkins, unlike Fishman, maintains Ferguson’s original distinction of diglossia as referring to variants of a single language rather than two or more languages in use in a speech community. In his explanation of the difference between bilingual and diglossic situations, Hawkins emphasises the point that “in a bilingual speech community both the codes, H and L, may be fully specifiable, independent languages, natively‐ spoken with descriptively‐based grammars” (1983:18, note 12). It is clear that Hawkins description of the bilingual scenario is presented in such a way that it may be contrasted not only to diglossia as variation within a single language but also across Ferguson’s variables of acquisition and standardization. In contrast to the bilingual scenario in which both H and L have standardized grammars and are acquired as ‘native’ languages, the diglossic scenario as described by Ferguson comprises a language community in which only H has standardized grammar and only L is acquired as a ‘native’ language.

Both Hawkins and Ferguson suggest the existence of quite clear and distinctly delimited speech communities that fit precisely into either a category described as diglossic or one described as bilingual — or, perhaps, multilingual. Hudson, however, helps to complicate and, perhaps, add a certain actuality to the issue. As Hudson points out, classifications of diglossic, bilingual, and standards‐with‐dialects situations tend to be less categorical that one might hope (2002:2). Obviously, in many cases, there will be areas of overlap and obscurity. As Hudson puts it, “there are examples of societal bilingualism that bear some resemblance to diglossic situations, just as there are, or have been, examples of diglossia that, in certain aspects of their social evolution, resemble societal bilingualism”(ibid.). However, Hudson argues, diglossia and societal bilingualism differ in their “origins,

34 evolutionary courses of development, and resolutions over the long term” (ibid.). In terms of origin, development, and resolution, Hudson is mainly concerned with issues relating to Ferguson’s variables of acquisition, function and, stability. For example, Hudson explains that distinction between the terms ‘diglossia’ and ‘societal bilingualism’ helps to tease apart issues relating to ‘functional distribution’ of codes and code ‘stability’ or ‘displacement’, on the one hand, and ‘linguistic form’ and ‘language function’, on the other hand (2002:2‐3). Obviously, ‘functional distribution’ and ‘language function’ refer to Ferguson’s variables of prestige and function, while ‘stability’ and ‘displacement’ refer to stability, and ‘linguistic form’ refers to grammar, lexicon, and phonology. Although Hudson seems to claim that distinction between these two terms helps to distinguish more clearly issues relating to as many as 6 of Ferguson’s 9 variables, it is clear throughout the paper that he feels both function and acquisition remain the critical points of distinction. However, connections between the 2 variables and the other 7 variables are easily made. For example, Hudson refers to the “critical distinction” of a presence of a ‘prestige group of H‐speakers’ in the bilingual speech communities and an absence of such a group in the diglossic community (2002:21). Apart from just relating to acquisition, the distinction also implicates Ferguson’s variable of prestige. Moreover, Hudson makes it clear that he is not only interested in the dichotomous notion of presence or absence of a prestige group of H‐speakers in these communities but also issues relating to stability that include facilitation or obstruction of the origin, development, and extinction of codes (ibid.). As a final point of contrast between the two language community scenarios — a point that is mentioned by Fishman (1967) and echoed in both Coulmas (2005) and Marfany (2010), Hudson refers to Ferguson’s variable of function. As Hudson explains, bilingualism refers to a type of “individual linguistic versatility” that is based upon an individual language user decision for any number of reasons — habit, prestige, language ability, and so on — to switch variants or languages (2002:43). Diglossia, on the other hand, refers to a scenario in which “societally held norms” direct “differential functional allocation of codes” (ibid.). As a most general form of distinction between the two, function — a speaker in a diglossic situation chooses varieties in order to suit different “contexts

35 and social domains”, while a speaker in a bilingual situation chooses varieties in order to accord with “strictly personal circumstances” — seems to be the point most often referred to in the literature (Coulmas 2005:135; Marfany 2010:12).

3.4 Diglossia and Standards­with­Dialects

In his argument against Gumperz (1971) and Fishman’s (1971) extension of Ferguson’s classic definition of ‘diglossia’, Hawkins (1983) offers several useful points upon which the diglossic language situation may be distinguished from the standards‐with‐dialects scenario. Although Hawkins admits similarities amongst various types of speech community — diglossic, bilingual, multilingual, and dialectal, for example — he still insists there are “important differences” between them (1983:14).

As with the aforementioned differences between the bilingual and diglossic situation, differences between diglossic and standards‐with‐dialects scenarios begin with the distinction of diglossia as a more or less monolingual case. As mentioned before, Ferguson’s original formulation of the idea of ‘diglossia’ referred exclusively to a language community in which different, genetically‐related variants of a single language exist distinctively — across 9 variables — over time. As Hawkins explains, “the critical differences are to be found in the relationship between H and L” (ibid.). Hawkins refers to a relationship in which “H varieties are derived from L by the process of purification” (ibid.). Hawkins view of H as having been derived from L offers a clear statement as to the monolingual, genealogical relationship between the two. Moreover, built into Hawkins description of the relationship between H and L within the diglossic situation is an obvious reference to the importance of Ferguson’s variables. For example, Hawkins’ description of the process as ‘purification’ points to an asymmetrical relationship similar to the one described by Ferguson in regard to variables of prestige, grammar, and phonology. In a statement that confirms an asymmetrical relationship between H and L with respect to

36 grammar, Hawkins also brings up Ferguson’s all‐important variable of acquisition, “while descriptive grammars can be written for L, grammars of H must be based on prescriptive rules for purification ‐ there is no natively spoken code” (ibid.). In other words, Hawkins’ model describes H as a variant of the language that depends upon L for its existence. The H‐L relationship of dependency, however, does not imply power imbalance in favor of L. Hawkins makes it quite clear that the purification process proceeds from L to H “in the direction of a more highly classicized representation” of the language (ibid.).

In contrast to his model of diglossia, Hawkins description of dialect highlights the reversed order of the relationship between H and L. As Hawkins points out, in the case of the standards‐with‐dialects speech community scenario, H has an “independent existence”, while varieties of L are “described by reference to” H (ibid.). In other words, dialect (L) — also referred to as non‐standard language — is usually considered to be a corruption — as opposed to a purification — of standard language (H). Description of the process that moves from H to L — in the dialectal scenario — as corruption and the movement from L to H — in the diglossic scenario — as purification makes clear an obvious reversal of order with respect to process. However, the asymmetrical relationship between H and L with regard to Ferguson’s variables of prestige, grammar, and phonology, for example, does not change between the two language situations. In other words, in both cases, H is the purer, more complex variant of the language. The asymmetrical relationship between H and L is preserved between the two disparate scenarios because, as Hawkins explains, “the difference [between the two situations] lies ultimately in the nature and relationship of the codes” and not in “the uses of the codes” (1983:15). In fact, Hawkins cites Ferguson’s misguided formulation of diglossia based upon code use, rather than code nature and relationship, as the main weakness preventing the classical definition of diglossia from being able to distinguish true diglossia — Greek, Arabic, and Chinese types — from situations of standards‐with‐dialects (ibid.).

37 3.5 Diglossia and Code­switching

As Marfany (2010) notes, the concept of ‘code‐switching’ as form of language contact has received a great deal of attention in the field of sociolinguistics, in general, and in relation to language contact concepts such as bilingual (Weinreich 1953; Vogt 1954; Gumperz 1964), standards‐with‐dialects (Gumperz 1958; Gumperz 1961), and diglossia (Fishman 1972; Müller and Ball 2005; Gafaranga 2009; Marfany 2010), in particular. In a paper reviewing sociolinguistics and linguistic anthropology literature on code‐switching, Nilep offers a general definition of the term as follows:

Code switching is defined as the practice of selecting or altering linguistic elements so as to contextualize talk in interaction. This contextualization may relate to local discourse practices, such as turn selection or various forms of bracketing, or it may make relevant information beyond the current exchange, including knowledge of society and diverse identities. (2006:1)

Based upon his obviously thorough — the paper discusses in depth much of the research done on the subject across a number of fields — review of the literature, Nilep, himself, is forced to admit that although a great deal of work has been done with regard to the concept of code‐switching, scholars remain in disagreement upon a consistent definition for the term (ibid.). For example, in contrast to Nilep’s somewhat vague object of alteration, ‘linguistic elements’, Müller and Ball, in their definition of the term, specify ‘elements’ as part of “at least two linguistic systems” (2005:49). ‘Linguistic systems’ is explained further as “separate languages, or distinguishable varieties of one language” (ibid.). Moreover, whereas Nilep’s definition frames the speech situation within the terms of overly obscure jargon such as ‘contextualize talk in interaction’, Müller and Ball state quite clearly that the speech situation in which alteration of linguistic systems takes place includes “participants, their motivations and agenda, the constraints on language use imposed by various factors, as well as the talk (or writing) that is produced by the participants” (2005:50). Müller and Ball’s clarification with respect to the object of alteration and the speech situation described by the term ‘code‐switching’ helps to

38 bring into focus some distinguishable similarities and differences — disparate languages versus varieties of one language and individually motivation versus socially motivated motivation — between the term and Ferguson’s ‘diglossia’.

Marfany (2010) takes an unforgiving and sharply defined approach to the question of defining language contact as belonging to the category of either ‘code‐switching’ or ‘diglossia’. In Marfany’s view, description of the existence of linguistic variation may be reduced to three possible scenarios. According to Marfany, “people switch or mix languages because either they try to speak in a language they do not know well enough and keep lapsing into the one they know, or they are diglossic, or they are bilingual” (2010:17‐18). Marfany’s system of categorization is, perhaps, an oversimplification, but it is presented from the sociological or historical perspective as relief from the obscurity of what Marfany calls, with respect to ‘code‐switching’, sociolinguistics’ “unstoppable generation of increasingly esoteric subtopics and debates within debates, uncontrolled proliferation of technical vocabulary, corresponding increasing opacity, and equally increasing self‐referentiality” (2010:13). In fact, Marfany’s view of language contact brings us full circle back to Ferguson’s (1991) endorsement of diglossia as belonging to one of the two basic dimensions of language variation — namely, register or use‐related, and not the other — dialect or user‐related. In other words, situations of language contact such as bilingualism, standards‐with‐dialects, and code‐switching — psychologically motivated alteration of codes — may be categorized as user‐related instances of language variation, while diglossia may be categorized as use‐related.

39 4. Chinese Language Situation

4.1 Chinese Diglossia

Although Chinese is not one of the defining languages described in his classic paper on diglossia, Ferguson does note, “it [Chinese] probably represents diglossia on the largest scale of any attested instance” (1959:36). In reference to what he also calls “the most extensive case of diglossia in history”, Don Snow explains, “in pre‐modern times, Classical Chinese functioned as the high (H) language variety in not only China, but also Korea, Japan, and Vietnam” (2010a:124).5 In the case of the Chinese diglossic situation in China, Classical Chinese (文言; wenyan) can be considered to function as the H variant of the language, and a variety of vernacular (白話; baihua) forms of Chinese can be regarded as L.

In “Modern Written Chinese in Development,” Ping Chen explains that wenyan functioned as H in China for approximately 2,000 years, until the language reform associated with the May 4th Movement of 1919 led to its abandonment as the standard written language (1993:506‐507). However, as Snow’s paper helps to clarify, it wasn’t until the 6th or 7th century that the Chinese language contact situation developed into a “full‐fledged diglossic pattern” (2010a:127). Moreover, it was around the same time — Tang (618‐907 CE) dynasty — that baihua as a written vernacular began to develop (Chen 1993:507). Although development of a written L form is not necessary for the establishment of diglossia — it may even hinder establishment in some instances, it is a unique characteristic of the evolution of diglossia in the pre‐modern Chinese language community. Baihua as written vernacular will be referred to again in a subsequent discussion of Ferguson’s prestige variable and its manifestation in the Chinese diglossia situation.

Consistent with defining characteristics of Ferguson’s variable of function, wenyan (H) and baihua (L) were clearly differentiated with respect to functional domain. As

5 The term ‘pre-modern China’ usually refers to the historical period before formation of the Republic of China in 1912.

40 Chen points out in Modern Chinese: Its history and sociolinguistics, for most of China’s history, wenyan has played the role of H variant as “the classical standard written language for literary, scholarly, and official purposes” (1999:68). Baihua, on the other hand, has “served all low‐culture functions such as transcriptions of Buddhist admonitions, scripts for folk stories, and plays” (Chen 1993:507). Moreover, as will be discussed in a subsequent section — 5.1 Chinese Historiography and Language Use — of the present paper, throughout most of China’s history, there have existed clear conventions of language use and domain — with respect to wenyan, in particular.

Distinct conventions of language use and domain tend to facilitate establishment of an obvious system of prestige, as described by Ferguson’s second variable. The classic diglossic asymmetrical relationship of prestige between H and L can be observed quite clearly in Chen’s description of the Chinese language contact situation in which “wenyan was considered refined and elegant, thus ideal for high‐ culture functions, while baihua was despised as coarse and vulgar, suitable only for low‐culture functions” (1999:69). As Snow points out, much of the value placed upon wenyan as the prestigious variant arises due to its link to a system of written examinations (科舉; keju) establishment during the Han dynasty (202 BCE‐220 CE) for the purpose of selection of government officials (2010a:126). The importance of the link between wenyan and its use in imperial examinations is highlighted further in Chen’s astute observation that it is no coincidence that “the abandonment of wenyan as the standard written language occurred in less that two decades after the abolition of the state examination system” (1999:68).67

The use of wenyan as the language of examination in the Imperial exams, and the importance of the system of examinations for the appointment of government office

6 In The Reform and Abolition of the Traditional Chinese Examination System, Wolfgang Franke dates the official abolition of the system of Imperial Examinations to approximately 1905 (1972:69-71). 7 Chen connects the ‘abandonment of wenyan as the standard written language’ with events occurring during and directly after the May 4th Movement (1999:74-75).

41 obviously served to institutionalize wenyan as the H variant. Of course, quite often, in a diglossic situation, variables of standardization and acquisition are concomitants of the institutionalization process. Promotion during the Qin dynasty (221 BCE‐206 BCE) of a unified, standardized Chinese script and an unrelenting sponsorship throughout the history of pre‐modern China of early — first millennium BCE — wenyan texts such as the Analects and Mencius as prototypical resources for Imperial examination preparation led to increased grammatical, lexical, and phonological distance between wenyan and baihua (Chen 1993:506‐509; Chen 1999:67‐68; Snow 2010a:126). Increased distance between the H and L variants fixes differences in acquisition, as well. As what Snow calls “the first mechanism by which diglossia was created in East Asia,” the ever‐widening gap between wenyan and vernacular Chinese through a process of isolation and fossilization of the former and continual evolution and transmutation of the latter solidified the diglossic disparate relationship of acquisition in which L is assimilated naturally in the home, while H must be learned formally at school (2010a:126‐127).

The distance between wenyan and baihua in China was made even greater by the fact that as a logographic — as opposed to phonographic — writing system wenyan was almost entirely divorced from the phonetic details of vernacular speech (Chen 1999:67). As Chen notes, the logographic nature — lack of direct association between sound and graphic forms — of wenyan gave it, “a degree of accessibility across space and time” and “insulated it from changes in the vernacular language” (1999:68). As a result, Chen explains, wenyan was able to “serve as the medium whereby Chinese literary heritage was preserved and continued, and information could be spread across a land of great dialectal diversity” (ibid.). Consistent with Ferguson’s description of the diglossic variable of literary heritage, the unique characteristics of wenyan as a logographic writing system and its place in the Imperial examinations and selection of government office process have enabled it to reside at the heart of “a sizable body of written literature” that is “held in high esteem by the speech community” (Ferguson 1959:31).

42 There is a final point of interest that is, perhaps, unique to the diglossic situation in China. It is an issue that relates to Ferguson’s variables of function and prestige and one that seems to highlight an interesting divergence between the diglossic language communities — Arabic, Modern Greek, and Haitian Creole — studied by Ferguson and the Chinese case. In his investigation of the three aforementioned language communities, Ferguson found that often “the superiority of H is connected with religion” (1959:31). He offers examples for Greek, Haitian Creole, and Arabic, showing that the H variant is used for all written records of scripture — the Bible and the Koran, respectively. In his discussion of the Chinese diglossic language scenario, however, Chen notes, “baihua served all low‐culture functions such as transcriptions of Buddhist admonitions, and scripts for folk stories and plays” (1999:69). This observation is echoed in Victor Mair’s “Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages” in which Mair explains that in contrast to the majority of pre‐modern Chinese literature which was written in Literary Sinitic — wenyan — “beginning in the medieval period, however, an undercurrent of written Vernacular Sinitic [baihuawen] started to develop” (1994:707).

According to Mair, “the earliest instances of written VS [baihua] occur almost exclusively, certainly with absolute and unmistakable predominance, in Buddhist contexts” (1994:709). It seems the Chinese diglossic scenario, at least in the instance of the language of religious texts, is quite the opposite to Ferguson’s description of the Greek, Haitian Creole, and Arabic situations. Obviously, as Mair’s paper sets out to establish, the use of baihua in early Chinese Buddhist texts may be attributed to several different factors. One contributing factor which is of greatest relevance to the field of Translation Studies is the recognition that since early Chinese Buddhist texts — approximately 1st to 6th Century CE — were translated from Indic originals, it is likely that use of certain baihua elements in these translation “was in direct response to the linguistic features of the Indic (and perhaps Iranian and Tocharian) prototypes” (Mair 1994:710).

43 Although it is beyond both the aim and the scope of the present paper, it may prove fruitful in future research to investigate the influence of translation upon the establishment, maintenance, and resolution of diglossic language communities. In fact, Mair hints at a possible direction for such research in his assertion that:

Implantation of Buddhism into the Chinese sociolinguistic body also served to elicit in an active way vernacular, colloquial, and dialectical elements that belonged properly to spoken Sinitic languages but that had been rejected by the indigenous textual tradition as vulgarisms (Mair 1994:710)

As Harold Schiffman points out in his discussion of diglossia as sociolinguistic situation, belief systems about language that are present within speech communities — beliefs about origin myths, faults and virtues of given language varieties, and language taboos, for example — constitute one example of a social institution that may affect maintenance and transmission of language (1998:211). In other words, it seems translation, in the case of the language used in Chinese Buddhist texts, created an opportunity for baihua, an early L variety of Chinese,8 to establish a written function for itself within the H‐L diglossic continuum. However, the stability of the baihua religious text functional niche was vulnerable to stress exerted upon it through its subordinate relationship to wenyan within the asymmetrical diglossic system of prestige. According to Mair, Chinese Buddhist writing eventually evolved to more closely approximate H‐variety style. Interestingly, Mair adds that this is particularly true for “texts that were composed by native authors and were not translated from non‐Sinitic languages” (1994:712). Again, translation seems to play a unique role in the establishment, development, maintenance, and resolution of diglossia in China.9

8 Gurevich (1985) has shown that baihua found in early Buddhist texts contained a significant amount of wenyan elements and, thus, cannot be considered a pure baihua form (quoted in Mair 1994:712). Mair refers to the early form of baihua as a ‘Buddhist Hybrid Sinitic’ (1994:712).

9 Connections between translation and establishment, development, and maintenance of diglossia in China have been alluded to in the present section’s discussion of baihua use in early Buddhist texts. For the case of translation’s role in the resolution of diglossia in China, Chen (1993) presents a detailed discussion of the influence of translation of Western languages on the shaping of norms of modern written Chinese.

44 5. Parody, Translation, and Chinese Diglossia

5.1 Chinese Historiography and Language Use

In a seminal work on the topic, Chinese Traditional Historiography, Charles Gardner explains, “historical composition in Chinese has evolved a considerable body of conventional diction which requires special study for complete comprehension” (1961:80). In Ah Q Zhengzhuan, Lu Xun devotes most of the first two chapters of the book to the narrator’s first‐person narrative that outlines the myriad difficulties he encountered trying to fit Ah Q’s story and personal details to the conventions of traditional historiography. In “The Inescapable Predicament: The Narrator and His Discourse in ‘The True Story of Ah Q’,” Martin Huang explains, “In the introduction, Lu Xun is consciously playing the conventions of traditional historiography against those of traditional vernacular fiction” (1990:433). In other words, Lu Xun offers his readers a critical investigation of the internal workings of the relationship between H (wenyan) and L (baihua) in the diglossic language contact situation. In a conventional, condescending tone, Lu Xun’s narrator discusses in detail the problems of finding within the historiographical tradition a proper title for Ah Q’s story. As Huang points out, by meticulously explaining “why none of the traditional categories of biography fits his own ‘biography’ of Ah Q,” the narrator (and Lu Xun) “ridicules traditional historiography” (ibid.). In the end, according to Huang, the narrator “finds a ‘new’ variety of historiography by which to identify, hence to authenticate, his own narrative” (ibid.). Throughout the process, the asymmetry of the H‐L relationships of function, prestige, and literary heritage is foregrounded, and ultimately challenged.

It is important to remember that significant changes in Chinese historiography did not occur until the turn of the 20th century. Prior to that time, traditional Chinese historiography wrote political and military history in an annals­biographic form (Wang 2001:16). This enabled historians to focus, across various biographies, upon

45 the deeds of important individuals (Wang 2001:16). The traditional method, however, did not allow for a detailed telling of the stories of the Chinese people, but rather painted them all into a nondescript backdrop upon which the lives of heroic figures could be played out. Moreover, histories were written in the literary language of wenyan rather than in the vernacular baihua, the language of the people. It wasn’t until the revolutionary period of the early 20th century that China began to evaluate limitations associated with traditional practices of education, literature, science, social science, and language. As Chen points out, during the first two decades of the 1900s and China’s push toward modernization, the Chinese language, blamed for widespread illiteracy, “was picked as one of the most important targets for reform” (1993:505). Under the influence of imported Western — and Eastern in the form of Japanese — notions of modernization and democracy, 20th century Chinese thinkers such as Lu Xun began to push for changes that would lead to greater accessibility to knowledge through the replacement of wenyan by a language variant closer in structure to vernacular forms used and understood by the masses (Chen 1993:509).

One of the main scholars to lead the revolt against traditional Chinese historiographical practice was Qing dynasty reformist Liang Qichao (梁啟超, 1873‐ 1929). According to Edward Wang, Liang felt that the “main problem in the traditional practice of historical writing was its failure to acknowledge the role of the people and to foster a national awareness” (2001:16). Through his exposure to Japanese and Western ideas while studying abroad in Meiji Japan, Liang began to give rise to notions of reforming the traditional Chinese system of historiography. Based upon Japanese scholar Fukuzawa Yukichi’s (1835‐1901) ideas about a “history of civilization” (bummeishiron), Liang would come to propose a new, modernized system through which to understand China’s historical past (Wang 2001:44). It is in this context that the creation by Ah Q’s narrator of a “new variety of historiography” becomes an important piece of the parodic puzzle presented in the novel.

46

Ah Q’s classically trained, tradition‐bound narrator – he adheres unquestioningly to Chinese traditional historiographic conventions of naming and register – attempts in vain to figure the story of Ah Q, an unimportant peasant unworthy of the attention of history, into the restrictive forms and categorizations of traditional historiography. Not only does the narrator failure to find a place for Ah Q within the hegemony of the traditional form, but also he is unable or, perhaps, unwilling to break from tradition even after he has established its restrictiveness. The narrator provides a clear parallel to Lu Xun’s own struggle within the context of his time. Lu Xun stood in a precarious position pinned between the precipice of modernity and the long, looming shadow cast by the cliffs of tradition.

In his discussion of the “sense of predicament” of the turn‐of‐the‐century Chinese intellectual, Huang explains, “Lu Xun even questioned whether, as a cultivated intellectual, he could ever free himself completely from the bondage of the ‘ancient’ tradition which he criticized so harshly” (1990: 431‐432). This is the context under which Lu Xun’s parody of traditional historiography comes into full view. Subsequent section — 5.4 Text Excerpt: Analysis and Discussion — will focus upon presentation, analysis, and discussion of narrative style, diglossia, and linguistic variation as they interact to create a context in which Lu Xun’s parody was created in the original and might be relived in the English translation.

5.2 Ah Q Zhengzhuan

Ah Q Zhengzhuan (阿Q正傳; The True Story of Ah Q) (1921) is a short episodic novella written by Lu Xun (魯迅, 1881‐1936). The story traces the adventures of Ah Q, an idler and odd‐jobber living in the fictitious village of Weizhuang (未莊; weizhuang). The story is set in China during the time that leads up to the Revolution of 1911. Zongxin Feng, in a paper that deals with the interplay of historical fact and fictional narrative in the novella from a Cultural Studies perspective, explains that although the 1911 Revolution may be said to make up the story’s central theme, Lu

47 Xun focuses the plot upon the life of the innocuous peasant who is concerned more with personal interest than with issues of the nation (2008:192). In telling the story of the peasant Ah Q against the overall grand theme of the Revolution, Lu Xun attempts to draw attention to the plight of the rural people, crumpling under the heavy burden of living in a nation in the midst of political upheaval.

Ah Q, a member of the rural peasant class, has little education, no special training, and no real understanding of the wider world around him. One talent that Ah Q is especially well known for is his ability to claim a “spiritual victory” (read “self‐ deceptive victory”) whenever faced with defeat or humiliation. Ah Q uses this talent to perfection early in the story, for example, when he gets into an altercation with Mr Zhao, an honored landlord of Weizhuang. After receiving a beating from Zhao, Ah Q consoles himself with the idea that he must, indeed, be an important person if someone as well respected as Mr Zhao has taken the time to beat him up. Apart from being bullied, Ah Q is also a bully himself. He harasses the weak and the less fortunate while, at the same time, remains fearful of anyone above him in rank, strength, or power. As Feng describes him, Ah Q is a character who “does not realize the reality and hence lives free of worries, with arrogance and vanity in pursuit of spiritual triumphs” (2008:192). Ah Q, Feng continues, “never sees his weaknesses but always takes pride in his past family glory and wishful glory of his future generations” (ibid.). Many of Ah Q’s personality traits have been taken as representative of certain aspects of the Chinese national character. As Jon Solomon explains, “Ah Q is understood as a composite figure, supposedly bringing together all the negative traits specific to Chinese culture” (1993:248). Although, Solomon’s own position focuses on the indeterminateness of Ah Q’s character,10 questions as to the specific identity of the archetype upon which the repulsive character was based have arisen.

10 Solomon emphasizes the generalization of the Ah Q character to include the entire social collective rather than a single, concrete individual or group (1993:248-249).

48 Ah Q’s story was first published in the Morning News as a serial between December 4, 1921 and February 12, 1922. The piece is generally held to be a masterpiece of modern Chinese literature, since it captured for the first time in vernacular Chinese the struggles of the Chinese nation as it teetered at the expansive crevasse between tradition and modernity (Luo 2004:84). As Jon Kowallis points out, Lu Xun “was really striving to remake baihua wen into a new written language for a new literature”(1994:283). As has been discussed in the present paper, Lu Xun used the juxtaposition of hitherto irreconcilable language phenomena — H and L languages and the entire complement of variables and ideologies associated with them — to bring into question the practicality of traditionalists’ uncompromising adherence to defunct, traditional language practices even in the face of China’s seeming demise at the hands of the modern West. One aspect of Lu Xun’s technique, as Kowallis explains it, uses irony created “by juxtaposing ideals expressed in the classical language against the harsh realities of the present day” (1994:283‐284). In order to address issues relating to the Chinese diglossic language contact situation and the translation of linguistic variation, the present paper will provide a detailed investigation of Lu Xun’s juxtaposing of language and ideals in his 1921 novella Ah Q Zhengzhuan.

As mentioned above, there has been a great deal of discussion about the significance of Lu Xun’s character Ah Q. As Gloria Davies points out, for example, at the time of publication, Ah Q Zhengzhuan already had begun to cause readers to be “intrigued by the question of whether the portrayal of Ah Q was based on a real person” (1991:58). Even Lu Xun has hinted at the possible significance of the Ah Q character as a “portrait of himself [the reader] as well as all other Chinese” (cited in Lee 1985a:132). According to Feng, Ah Q portrayed as Everyman provided Lu Xun an avenue through which to “criticize the incompleteness of the Revolution and its sad fate of failure, highlighting peasant issues in the democratic revolution in early twentieth‐century China” (2008:190).

49 Ah Q Zhengzhuan begins with a preface in which the story’s narrator confesses, in first‐person narrative, the various difficulties that he has been faced with in his attempt to write Ah Q’s biography. These difficulties are, in fact, problems that the narrator, who Jeremy Tambling calls a fussy Confucian, traditionalist (2007:59), has encountered while trying to write a conventional biography about an individual who, as the narrator himself puts it, ‘is obviously not one of those whose name is preserved on bamboo tablets and silk’ (Yang 1956:79). The narrator is attempting to write a “history” of a person who has been completely ignored by historians. The situation foregrounds what Tambling calls, “a parodying of the possibility of writing history” (2007:59).

In the preface, Lu Xun creates a situation in which the narrator, in his unwavering attachment to the conventions of the Chinese traditional historiographic system, actually lays bare the flawed nature of the system, itself, and its inability to deal with the story of the common person. In his article “The Inescapable Predicament: The Narrator and His Discourse in The True Story of Ah Q,” Martin Huang explains that “the life of Ah Q, according to accepted historiographical conventions, is not a subject worthy of the ‘elegant’ discourse used to tell it” (1990:435). Huang adds, “At the same time, the elegant discourse itself becomes ridiculous and awkward when applied to the life of Ah Q. Thus the conventions themselves are seriously questioned” (1990:435).

Apart from the preface, which spans most of the first two chapters of the book and is told mostly in first‐person narrative, the rest of Ah Q Zhengzhuan — telling of the misadventures of Ah Q, the town’s folk, the revolution, and the eventual execution of Ah Q as scapegoat for crimes committed in the name of the revolution — is told entirely by an omniscient, third person narrator. In his book Madmen and Other Survivors: Reading Lu Xun's Fiction, Jeremy Tambling describes Ah Q’s narrator thus:

What sort is the narrator, then? The answer is that he is a mass of contradictions, like Ah Q himself, a mixture of pedantries and obscure

50 traditions, which take over the prose and his thoughts, and someone who without realizing it, by writing the life of Ah Q, shows that history lies in the documents that historians have discarded. Hence everything in the narrator is digressive, the opposite of what is expected from a true story (60).

The parodic discourse that arises from the narrator’s ‘mass of contradictions’, then, will play a central role in the present paper’s discussion of register and narrative style as the two pertain to Lu Xun’s original text as well as to the four English translations that will be described in detail in a subsequent section — 5.3 Translations and Translators. In particular, the recreation in English translation of Lu Xun’s juxtaposition of classical literary language with the vernacular will be analyzed in order to investigate various options with regard to translation of literary linguistic variation.

An important aspect of Lu Xun’s narrative style in the preface, and throughout the book for that matter, is the use of Classical Chinese syntax, referred to above by Huang as the elegant discourse, placed in juxtaposition to, as the narrator, himself, calls it, “the debased vulgarity of [the story’s] content and characters” (Lovell 2009:80). It is this juxtaposition of ‘high intellect’ with ‘low class’ that draws attention to the narrator’s inability to break free from traditional historiographical conventions (Lovell 2009:81). It is this juxtaposition of form and content that drives the parody present in Lu Xun’s original work. In order to place Lu Xun’s parodying of historiography into the context of the present paper’s discussion of Ferguson’s diglossia, it is essential to understand something more about traditional Chinese historiography and language use in China.

5.3 Translators and Translations

According to Donald Gibbs and Yun‐chen Li’s 1975 A bibliography of studies and translations of modern Chinese literature, 1918­1942, the first translations of Lu Xun were done in 1926 by George K. Leung (梁社乾, Liang Sheqian) (1898‐?) (105). In

51 1926, the Commercial Press published Leung’s English translation of Lu Xun’s Ah Q Zhengzhuan in Shanghai under the title, The True Story of Ah Q.11 In his review of William Lyell’s translation of Lu Xun in Diary of a Madman and Other Stories (1990), Jon Kowalli explains, “the first attempt at introducing Lu Xun to readers of English, made by American huaqiao Liang Sheqian (George K. Leung) in Shanghai in 1925, was aimed primarily at the treaty‐port readership” (1994:283). In an explication of an entry in Lu Xun’s diary referring to his reception of a letter from Leung in April 1925, Dong Dazhong (董大中) explains that Leung was writing to Lu Xun in order to discuss translation of Ah Q Zhengzhuan (2007:145). According to Dong, Leung was born and schooled in New Jersey, studied theatre and music in New York, and eventually moved to China, living in Beijing, Guangzhou, and Hangzhou (ibid.). Leung’s 1925 translation was followed by E. Mills’ translation — published in the UK in 1930 and in the US in 1931 — of J.B. Kin Yn Yu’s (1926) French version. Reviews of the early translations of Lu Xun are quite rare (Eber 1985:249). Journalist and Sinologist Edgar Snow, one of the first Americans to translate Lu Xun,12 offers a brief evaluation of the work of Leung and Mills, “two English translations have already been published, but alas, both of them are full of arbitrary insertions and deletions, as well as crude mistakes" (Zhelokhovtsev 1982:68).13 Kowalli, too, criticizes the “halting English” of the two earliest translations of Lu Xun (1994:283). In light of the fact that Leung’s translation is filled with elementary linguistic errors,14 Mills’ translation is taken from the French rather then the original Chinese, and the aim of the present paper does not include an exhaustive comparison of Lu Xun’s work in English translation, the aforementioned two translations will not be used for analysis in the present discussion.

11 Opened in 1897, in Shanghai, The Commercial was the first modern publishing house in China. 12 Snow published English translations of Lu Xun in Asia (1935; 1936), a US journal, and in his book Living China: modern Chinese short stories (1937). 13 Zhelokhovtsev translates the quote from Ge Baoquan’s (戈寶權) The True Story of Ah Q Abroad《<阿Q 正傳>在國外》(1981:28). 14 In Cosmopolitan publics: Anglophone print culture in semi-colonial Shanghai, Shuang Shen (2009) also expresses concern over the amount of errors and deletions found in Mills’ translation (107).

52 The next entry in Gibbs and Li’s bibliography refers to Wang Chi‐chen’s (王際真 ) (1899–2000) translation of Ah Q Zhengzhuan. Wang’s translation was first published between 1935 and 1936 in the New York based periodical China Today (1933‐ 1937). In a comprehensive transnational investigation of the influence of Western‐ educated intellectuals upon China’s modernization process in Shanghai in the late 1920s, Shuang Shen explains that China Today was published and edited in New York by an organization called the American Friends of the Chinese People (2009:108). According to Shen, the international periodical was closely connected to Communist parties in both China and the US and presented an obvious pro‐ Communist, anti‐imperialist position (2009:108‐115). Whether or not the political position advanced by the periodical held any influence over the choice to publish Wang’s translation remains open to debate. However, Shen points out that Wang’s translations — “A Hermit at Large” and “Remorse”, for example — of Lu Xun’s work were published at around the same time in other English‐language periodicals in China not associated with the Communist Party, nor exhibiting leftist‐leaning tendencies whatsoever (2009:116). According to Shen, it may be argued that Lu Xun, in fact, was reflected in the minds of his commentators as possessing two distinct personae (ibid.). The representation of such a view is observable, as Shen explains, in the fact that “Lu Xun’s short stories were accepted by both the left and the non‐left and perceived as less political,” while “his essays were published only in leftist magazines and thus had a different connection with politics” (ibid.). The present paper argues that in light of Shen’s point about the apolitical‐nature of contemporary English‐language publication of Lu Xun’s short stories and the fact that the only other translations of Ah Q Zhengzhuan available at the time were the less than ideal versions produced by Leung and Mills, the potential connection between the political views associated with China Today and its choice to publish Wang’s translation is, at the very least, contentious and, perhaps, entirely impossible to substantiate. As a result, the present paper will reserve judgement of the possibility of Communist party influence over translation choice for a more obvious

53 and more verifiable case — namely, Yang Xianyi and Gladys Yang’s 1956 translation published by the state sponsored Foreign Languages Press in Beijing.

In 1941, Columbia University Press published Wang’s translation of Ah Q Zhengzhuan along with his translations of ten other works by Lu Xun in a collection titled Ah Q and Others: Selected stories of Lusin. Wang was born in 1899 in Shandong Province, China. He was the first Chinese translator to translate Dream of the Red Mansion (紅樓夢, Hongloumeng) into English.15 He graduated from Tsinghua University (清華大學) in Beijing and travelled to the US as an exchange student in 1922. He received a bachelor's degree in Politics and Journalism from Columbia University, New York. While in the US, Wang also held faculty posts at both Columbia University and the University of Wisconsin. According to Zhang Jie (张杰), author of a comprehensive Chinese‐language study of the transmission and reception of Lu Xun’s work in the West, during the Second Sino‐Japanese War (1937‐1945), the Chinese Nationalist Party sent “right‐wing sympathizer” Wang on cultural exchange duties to the US (2002:284). During his time in the US, as Zhang explains, Wang’s 1941 Ah Q and Others: Selected stories of Lusin became the first book‐form publication in the US relating to the translation and study of Lu Xun (2002:285). Irene Eber, translator of the Hebrew version of Ah Q Zhengzhuan, argues that reception and wide publicity enjoyed by Wang’s 1941 publication was “associated with the American efforts to develop popular support for China’s wartime role” (1985:244). However, it wasn’t until the 1950s, Eber notes, that English‐language critical works on Lu Xun began to appear (1985:247). Although Wang’s Ah Q and Others: Selected stories of Lusin received some attention contemporaneously (Pritchard 1942; Kao 1942), his work seems to have been left out of later discussions on the topic of Lu Xun.

15 Although Wang did not produce a complete translation of the novel, as Xu points out, his version was praised for concise, fluent style (2011:326).

54 In his thorough treatment of Lu Xun’s short stories, Madmen and Other Survivors: Reading Lu Xun's Fiction, Jeremy Tambling only briefly alludes to select passages from Wang’s translation. In his chapter‐length discussion of Lu Xun’s Ah Q Zhengzhuan, Tambling focuses his analysis upon two later translations — namely, Yang Xianyi and Gladys Yang’s The true story of Ah Q (1956), and William A. Lyell’s Diary of a madman and other stories (1990). As Tambling explains, most translations of Lu Xun have remained somewhat in the shadow of the Yangs’ 1956 work (1990:6).

In 1956, the Foreign Languages Press in Beijing published Yang Xianyi (1915–2009) and his wife Gladys Yang’s (1919–1999) translation of Ah Q Zhengzhuan. The work is titled, The True Story of Ah Q. As has been alluded to above, the Yangs’ translation is probably the most well known English version to date and has been republished several times over the years (1960; 1972; 1980; 1990; 2002; 2006). Aside from paratextual differences — covers, forewords, presence or absence of bilingual presentation, and so — the only appreciable difference between the 1956 text and later publications is the system of romanization used for Chinese place and person names. As Kowallis explains, “the 1980 edition of Yang Xianyi and Gladys Yang’s Lu Xun differs from the 1956‐1960 edition in that all of the romanization has been redone into ” (1994:153). The 1990 Cheng & Tsui publishing company edition of the Yangs’ translation will be used for the purpose of analysis in the present paper. Again, apart from changes to the romanization practice mentioned by Kowallis, this edition is, for all intents and purposes, equivalent to the 1956 Foreign Languages Press edition.

As translators, Yang Xianyi and Gladys Yang make a unique team. Yang Xianyi was born in 1915 in , Northeastern China. He was a well‐known Chinese translator who has translated a great deal of Chinese classical literature into English. Yang was also the first to render both Homer's and Virgil's Eclogues into Chinese using the ancient Greek and ancient Latin originals respectively (Yang 2002:215). Yang’s knowledge of Greek and Latin came from his time at Merton

55 College, Oxford, in which he did a degree in Classics. After meeting his future wife Gladys Taylor in the UK, Yang and Taylor moved to China where they translated exclusively for the state‐funded Foreign Languages Press in Beijing. Interestingly, Taylor was actually born in Beijing but moved with her family back to England as a child. She attended school in the UK and graduated from the University of Oxford with a degree in Chinese before returning to live in China.

As can be expected from a team made up of two well‐educated, well‐cultured translators, the Yangs’ translations are generally regarded as accurate and well written. In a review article dedicated to the discussion of contemporary Chinese literature in translation, Leo Ou‐fan Lee explains, “there is no question of the "correctness" of translation…when it comes from a learned team such as the Yangs” (1985b:565). As with any creative undertaking, however, there have been aspects of the Yangs’ work that have been prone to criticism. In general, the criticism has taken two forms.

Firstly, there have been claims that the Yangs’ translations tend to be overly truncated. In the aforementioned article, Lee complains that much of the translation of traditional Chinese literature done by the Yangs is severely abridged and, thus, “gives a partial and often erroneous impression of the original work” (ibid.). Not all the blame for excessive truncation is placed upon the translators, however. Lee also holds accountable the two publishing companies the Yangs have been associated with over the years, the Foreign Language Press (FLP) and Panda Books (ibid.). As Lee explains, “For an academic audience, irresponsible truncation is one of the liabilities of the Panda series (and for that matter, a liability of the earlier literary translations from the FLP in general)” (ibid.).

A second criticism often brought against the Yangs’ translations is that it is overly British in style and slightly archaic. Lee points out, “from the American angle, the Yangs' ‘limitations’ may be their use of Anglicisms and occasional expressions or sentences that sound slightly ‘archaic’” (1985b:565‐566). He does add, however,

56 that, “one also finds a ring of informality that lends itself to easy readability” (ibid.). Gladys Yang, in the preface to her translation of Gu Hua’s (古華) (b. 1942) Furong zhen (芙蓉镇), alludes to the circumstances surrounding the archaic style of her writing: "Owing to the limitations of my English, now out of date after over forty years in China, I have failed to convey the raciness and earthiness of Gu Hua's language, which heavily draws on Chenzhou colloquialisms" (1983:11).

In addition to concern over Yang’s out‐dated English, Tambling adds, with respect to the Yangs’ 1956 The True Story of Ah Q, that the translation was not only “criticized for its rather British English,” but also for its “failure to register the different modes in which Lu Xun writes literature in the vernacular” (2007:5). Lee, too, qualifies his statement about the ‘easy readability’ of the Yangs’ work by adding, “[it] does not necessarily capture the cadence of the original works, especially in traditional Chinese fiction” (1985b:566). Obviously, the question of capturing the mode and cadence of the original is something that will be considered in the present paper’s analysis of the four English translations of Lu Xun’s Ah Q Zhengzhuan.

The third translator to be discussed in the present paper is American‐born Sinologist William A. Lyell (1930–2005). Lyell is the only one of the translators thus far mentioned to have pursued scholarly research in the field of Sinology, in general, and Lu Xun Studies, in particular. Lyell’s Lu Hsün's vision of reality (1976) was one of the first comprehensive English‐language works on Lu Xun (Lin 1979:365). In the book, Lyell offers detailed technical and formal analysis of many of Lu Xun’s works of fiction and presents extensive research relating to the author’s life and historical context.

Lyell was born in New Jersey in 1930. He did an undergraduate degree in French and English literature before going on to study Chinese at the Air Force Language Program at Yale. After his time in the forces working as Chinese interpreter during the Korean War, Lyell went on to spend three years studying Chinese philosophy

57 and literature in Taiwan at the National Taiwan University. He earned a doctorate in Chinese studies at the University of Chicago and was an associate professor emeritus of Chinese at Stanford and Ohio State University. Some of his most well known works include: Lu Xun: Diary of a Madman and Other Stories (1990); Cat Country: A Satirical Novel of China in the 1930's (1970); and Shanghai Express: A Thirties Novel (1997).

In 1990, in Honolulu, Hawaii, the University of Hawaii Press published Lyell’s translation of Ah Q Zhengzhuan under the title Lu Xun: Diary of a Madman and Other Stories. The work includes a brief biographical sketch of Lu Xun, an introduction to the works to be presented, a translator’s preface, and a collection of Lyell’s translations of 25 of Lu Xun’s works of fiction spanning the period from 1918‐1925.

Comparing Lyell’s translation of Ah Q Zhengzhuan to the Yangs’ translation, Tambling claims that although the latter is more neutral, it is also duller (2007:5). Tambling points out that, “William Lyell’s enthusiasm is marked, and it makes for a text which includes a number of different registers, even different typographies on the page – capital letters, italics – to convey the different modes in which narrative takes place” (ibid.). Also willing to offer criticism of the work, Tambling explains that although Lyell’s translation captures the cadence and varied register of the original, his use of vernacular “occasionally looks like paraphrase, and the style is American, racy and slangy, as though that was an equivalent for Lu Xun’s China” (ibid.). Analysis in the present paper will help to foreground the racy, slangy American style featured in Lyell’s translation. The concern of the analysis will be to assess how such a style works to represent parody and diglossia‐based linguistic variation as seen in Lu Xun’s original work.

The final translation to be discussed in the present paper is, in fact, written in a fashion that may be referred to as the polar opposite of Lyell’s slangy, American style. It is the lofty, British‐styled translation produced by UK‐born translator and Sinologist Julia Lovell. In 2009, in London, Penguin published, Lovell’s translation of

58 Ah Q Zhengzhuan under the title The Real Story of Ah Q and Other Tales of China. The book includes a detailed chronology of Lu Xun’s life, an introduction to Lu Xun, his work, and his ideas, suggestions on further reading about Lu Xun and his work in translation, a brief note on the translations presented in the book, Lovell’s translations of the full collection of Lu Xun’s fiction, and an afterword written by Beijing‐born, American novelist Yiyun Li.

Lovell was born in the UK in 1975 and has spent extended periods of time in China. She now lectures in Chinese history and literature at the University of Cambridge. She also writes on China for The Guardian, The Times, The Economist and The Times Literary Supplement (Lovell [Interview] 2010). Lovell has translated works from 20th century Chinese authors such as Yan Lianke (閻連科) — Serve the People (2007), Eileen Chang (張愛玲) — Lust, Caution (2007), and Han Shaogong (韓少功) — Dictionary of Maqiao (2003). She also has produced original, scholarly works such as The Politics of Cultural Capital: China's Quest for a Nobel Prize in Literature (2006) and The Great Wall: China Against the World 1000 BC­AD 2000 (2006).

In an interview regarding the publication of The Real Story of Ah Q and Other Tales of China, Lovell explained her reasons for choosing Lu Xun’s work as translation subject:

[Lu Xun] is a sharp stylist, with a command of tone (surrealism, irony, black humour) that gives him an appeal beyond China specialists. Anyone who works on modern Chinese culture encounters Lu Xun – he’s kind of James Joyce and Dickens rolled into one. And I would suggest that anyone who wants to get a handle on modern Chinese literature and culture － and particularly on the sense of crisis that gripped 20th‐century writers and thinkers – can’t do better than start with Lu Xun (2010; Interview).

With respect to the analysis carried out in the present paper, it is important to note that in the interview Lovell exhibits a clear grasp, and appreciation, of, firstly, the ‘irony’ and ‘black humour’ present in Lu Xun’s work, and, secondly, the work’s

59 embodiment of ‘the crisis that gripped 20th century writers and thinker’ in China. As has already been mentioned, the two issues play an important role in the present paper’s discussion of register, linguistic variation, and narrative style as it pertains to the narrator of Ah Q Zhengzhuan. Moreover, in the interview, Lovell, briefly describes her method for portraying in English Lu Xun’s juxtaposition in the original of literary and vernacular language. Lovell explains, “Where Lu Xun used classical Chinese to make a contrast with the vernacular elsewhere, my translation style for these excerpts became less comfortable, more stilted” (2010; Interview). Lovell goes on to make it clear that she is also very aware of the parodic narrative style of the original. Lovell tells the interviewer, “Lu Xun’s doing something similar in the facetious opening to “The Real Story of Ah‐Q”, poking fun at the flatulence of Confucian literary convention” (ibid.). Analysis presented in the subsequent section will discuss in more detail Lovell’s attempt to recreate through juxtaposition of literary and vernacular language Lu Xun’s parodic narrative style in Ah Q Zhengzhuan.

5.4 Text Excerpt: Analysis & Discussion

5.4.1 Text Excerpt (Original)

The following excerpt is taken from the second chapter of the novella. In the passage, the narrator begins telling of the exploits of the main character Ah Q. After a brief discussion of Ah Q’s utterly unremarkable existence, the narrator starts right in with quite a scathing description of the character’s exaggerated sense of pride:

阿Ｑ又很自尊，所有未莊的居民，全不在他眼睛裡，甚而至於對於兩位 “文童”也有以為不值一笑的神情。夫文童者，將來恐怕要變秀才者也； 趙太爺錢太爺大受居民的尊敬，除有錢之外，就因為都是文童的爹爹， 而阿Ｑ在精神上獨不表格外的崇奉，他想：我的兒子會闊得多啦！ (Lu Xun 2002:3)

Ah Q also [is] very proud. The Weizhuang villagers are not worth his notice. Toward the two “young scholars”, in particular, he feels them not

60 even worth a smile. It is said, one who is a young scholar, it seems, will become a talented literati. Mr Chao and Mr Qian enjoy the villagers’ respect, not only because they have money but also, because they are the dads of the two young scholars. However, Ah Q is alone in not being in the spirit of showing any particular worship toward them. He thinks, “My children gonna be much more richer!” (my translation)

This particular passage corresponds to an example cited in Huang’s (1990) article “The Inescapable Predicament”. It will serve to direct the present paper’s initial analysis of Lu Xun’s original text. Based upon discussion of the passage, a detailed analysis of the four English translations — Wang (1941); the Yangs (1956); Lyell (1990); and Lovell (2009) — will be carried out in order to further investigate issues pertaining to the (un‐) translatability of linguistic variation as it pertains to Lu Xun’s use of parody and the unique characteristics of the Chinese diglossic language contact situation.

According to Huang, an important feature of the particular section of text is, “the narrator introduces the discourses of ‘others’ without any apparent acknowledgement (such as typographical indication)” (1990:437). In other words, although the main voice of the passage is that of the narrator, there is also, in the text, a second voice, namely that of the people of Weizhuang. Huang suggests that in the underlined sentence above, and in Lu Xun’s choice of the informal ‘爹爹’ (‘dad’), the narrator is actually narrating “from the perspective of the villagers” (ibid.). Huang also notes that this change of voice is important because it points to the narrator’s shifting of satirical subject — from Ah Q to the villagers.

Finally, Huang explains that the success of the satire lies in the awkward hybrid construction of the villagers’ voice: literary language (‘夫文童者’; ‘young scholar’ and ‘秀才者也’; ‘talented literati’) mixed with a dash (‘將來恐怕要變’; ‘it seems will be’) of vernacular (ibid.). In terms of Ferguson’s diglossia this means H variety used inappropriately and combined, again, inappropriately with L variety. According to Huang, “this ‘hybrid construction’ captures perfectly the typical villager's

61 combination of snobbery and envy: they are trying awkwardly to imitate or repeat what the rich and ‘educated’ say” (ibid.).

I also would like to add that not only does Lu Xun juxtapose literary language with vernacular within the villagers’ voice, but he also juxtaposes H and L across the entire passage, itself. As Mair explains, literary wenyan and vernacular baihua demonstrate a stark disparity across grammatical structures (1994:709). In particular, Mair points to the difference in the grammatical function of the Chinese word shi (是). As Mair explains, under the literary wenyan system of grammar, shi has a demonstrative use, while under the baihua system it is used as a copulative verb (1994:710). According Mair, “This distinctive characteristic of VS [baihua] (A shih B {“A is B”}) which is so apparent even up to present day, is utterly different from LS [wenyan], which lacks a copulative verb altogether” (ibid.). “Instead,” Mair continues, “LS [wenyan] employs the nominative sentence structure A B yeh (也) (“A {is} B”)” (ibid.).16 Based upon Mair’s explanation, a clear juxtaposition of H and L can be seen in Lu Xun’s employment of both nominative and copulative verb function in the passage. The underlined phrase: 將來恐怕要變秀才者也 (one who is a young scholar, it seems, will become a talented literati), employs the wenyan nominative yeh (也), while the subsequent phrase: 就因為都是文童的爹爹 (because they are the dads of the two young scholars), clearly employs the baihua copulative verb function of shi (是).

Finally, in addition to the points outlined in Huang’s article, I would also like to suggest that the final sentence of this passage, ‘我的兒子會闊得多啦’; ‘my children gonna be much more richer’ (in bold above), offers insight into Lu Xun’s treatment of the voice of the character Ah Q, and, therefore, will prove useful in our analysis of

16 Due to the fact that he employs a previously popular romanization system — Wade Giles Pinyin — that differs from the Hanyu Pinyin system favored presently, Mair uses the romanized shih and yeh to represent the Chinese words 是 and 也, respectively, instead of the Hanyu Pinyin shi and ye.

62 the two English translations. The voice of Ah Q will be discussed further in relation to analysis carried out upon the four English translations that follows.

In summary, analysis of the four English translations will focus upon representation of the following four aspects: firstly, the change in voice at the phrase ‘夫文童者，將 來恐怕要變秀才者也’; ‘It is said, one who is a young scholar, it seems, will be one who becomes a talented literati’; secondly, the hybridity of the villagers’ voice as it is presented in this particular phrase and in the use of the informal ‘爹爹’; ‘dads’; thirdly, variation between narrator and character voice as it exists in the final sentence of the passage; and finally, contrast in nature between the narrative discourse and the actual subject matter being narrated so as to present a parody of traditional, Chinese historiographical conventions.

5.4.2 Text Excerpt (Wang 1941)

Wang’s 1941 translation will be the first to be discussed. As analysis will demonstrate, the translation lacks any variation of register whatever and obscures somewhat the change in voice at the sentence in question:17

阿Ｑ又很自尊，所有未莊的居民，全不在他眼睛裡，甚而至於對於兩位 “文童”也有以為不值一笑的神情。夫文童者，將來恐怕要變秀才者也； 趙太爺錢太爺大受居民的尊敬，除有錢之外，就因為都是文童的爹爹， 而阿Ｑ在精神上獨不表格外的崇奉，他想：我的兒子會闊得多啦！ (Lu Xun 2002:3)

Ah Q was very proud and held all the inhabitants of Wei in contempt, even to the extent of sneering at the two students. Now a student might one day pass his examination and become a licentiate. The reason Their Honors Chao and Chien were so esteemed by the villagers was that, besides their wealth, they were fathers of students. But in spirit Ah Q had no special regard for them. “My son would be much better than they,” he would assure himself. (Wang 1941: 82)

17 I have included the original text before each translated text in order to facilitate comparison.

63

As can be seen in the excerpt, whereas Lu Xun’s original demonstrated — through a change of register — a clear change of voice at the phrase, ‘夫文童者，將來恐怕要 變秀才者也,’ Wang’s version (in bold above) does not show any variation of register. Moreover, it is unclear who could possibly be expressing the belief that, ‘a student might one day pass his examination and become a licentiate’. To be fair, since the phrase is rendered in a declarative form it is clearly offset from the rest of the narrative. The syntax of the statement, however, does not offer the reader any clue as to whom could have spoken it, nor does there exist any hybridity within the statement itself. Wang, quite uninterestingly, renders the informal ‘爹爹’ as ‘fathers’ (underlined above).

Wang’s translation really misses the mark when it comes to recreating the linguistic variation, the change in voice, and the diglossic nature represented in the original. When Ah Q quite articulately thinks to himself, “My son would be much better than they,” (in bold and underlined above) the reader is left to wonder how it is that Ah Q, an uneducated peasant, is seemingly as well spoken as the narrator, an classically trained historian. Also unfortunate for the reader of Wang’s translation, the unremarkable, average‐sounding voice of the narrator makes it difficult for the work to re‐create any of the parodic narrative discourse used in the original to poke fun at Chinese traditional historiographical conventions.

5.4.3 Text Excerpt (the Yangs 1956)

The second translation to be discussed is Yang Xianyi and Gladys Yang’s translation. Published in 1956 by the Foreign Languages Press as The True Story of Ah Q, the translation reads as follows:

阿Ｑ又很自尊，所有未莊的居民，全不在他眼睛裡，甚而至於對於兩位 “文童”也有以為不值一笑的神情。夫文童者，將來恐怕要變秀才者也；

64 趙太爺錢太爺大受居民的尊敬，除有錢之外，就因為都是文童的爹爹， 而阿Ｑ在精神上獨不表格外的崇奉，他想：我的兒子會闊得多啦！ (Lu Xun 2002:3)

Ah Q, again, had a very high opinion of himself. He looked down on all the inhabitants of Weichuang, thinking even the two young "scholars" not worth a smile, though most young scholars were likely to pass the official examinations. Mr. Chao and Mr. Chien were held in great respect by the villagers, for in addition to being rich they were both the fathers of young scholars. Ah Q alone showed them no exceptional deference, thinking to himself, "My sons may be much greater!" (Yang 1956:82)

The voice of the villagers in this passage is entirely undistinguishable from the voice of the narrator. The only hint that, perhaps, the sentence taken to be the voice of the villagers, highlighted in bold above, is in any way different from the rest of the text is that it is offset by a comma; but, in fact, it is difficult to discern whether the utterance belongs to the narrator, the villagers, or, perhaps, even Ah Q, himself. Moreover, Lu Xun’s informal ‘爹爹’; ‘dads’, as well as the mix of literary and vernacular in the villagers’ voice, which worked so well to highlight the ‘awkward hybrid construction’ of their worldview, are entirely absent in the Yangs’ version. It may be argued that the villagers’ voice juxtaposes a vernacular ‘though most young scholars were likely to pass the official examinations’ (in bold above) with the more colloquial ‘not worth a smile’ (underlined above). However, I would suggest that the relative distance of register between the two does not seem to befit Catford’s “equivalence across varieties” or Pym’s syntagmatic alteration of distance. Moreover, the language of the Yangs’ narrator is not entirely different from that of Ah Q. This is an issue that Jeremy Tambling refers to in his claim that the Yangs’ translation of The True Story of Ah Q fails to ‘register the different modes in which Lu Xun writes literature in the vernacular’ (2007:5). As a result, contrast between narrative discourse and subject matter as used in the original to set up a parody of traditional, Chinese historiographical conventions fails to materialize in this sampling of the Yangs’ translation.

65 In order to avoid sounding prescriptive, I hope that perhaps an example of the criticism levelled against the Yangs’ work will help to suggest a possible reason for their choice to disregard the linguistic variation of the original. As mentioned previously, Leo Ou‐fan Lee has criticized the Yangs for overly abridging and, thus, providing “a partial and often erroneous impression of the original work” (1985b:565). Lee also has blamed the Foreign Language Press (FLP) and Panda Books — the two main publishers for whom the Yangs worked throughout their careers — for “irresponsible truncation” (ibid.). In a paper dealing with the issue of state commissioned publishing and translation, Bonnie McDougall explains that the Beijing FLP was modelled after the Foreign Languages Publishing House established in Moscow in 1931 (2009:3). As with its counterpart in the USSR, the Beijing FLP was a government‐funded and government‐run publishers commissioned to translate and publish in a number of foreign languages national literature, political (communist party) literature and books on all subjects Chinese (ibid.). Established in Beijing in 1952, as McDougall describes it, “the public mission of the Bureau [FLP] was always directed by the political line adopted by the CCP [Chinese Communist Party]” (2009:5). McDougall claims that translation under the direction of the FLP “was neither source‐oriented nor reader‐oriented,” but “served the self‐defined short‐term interests of the state as producer (2009:38). In his autobiography, White Tiger: An Autobiography of Yang Xianyi, Yang Xianyi explains that Liu Zunqi, the first head of the FLP and a veteran Party member, recruited him in 1952 to act as head translation consultant to the FLP (2002:184-185). The Yangs employment by the FLP and, in particular, Yang Xianyi’s position within the ranks of the state-run agency would more than likely have had some influence upon translation choices made by the pair.

With particular regard to the Yangs’ choice not to translate the linguistic variation of the original, I also would like to cite Merle Goldman’s (1982) discussion of the CCP’s political use of Lu Xun and his work. According to Goldman, the CCP carried out a double‐edged — simultaneously praising the man while condemning his work — deification of Lu Xun as an attempt to win over Lu Xun’s readership while continuing to promote Party ideology (1982:446‐447). Mao Zedong eulogized Lu Xun as “the

66 chief commander of China's ,” and claimed him to be “not only a great man of letters, but a great thinker and revolutionary” (ibid.). However, according to Goldman, the popular Party line with respect to Lu Xun’s work — in particular, his satirical style — was that it was inappropriate to the times and harmful to the establishment of communist ideals (1982:447). Of particular interest to the present paper’s suggestion that Translation Studies take a more realistic — diglossic or heteroglossic instead of monolingual — view of linguistic variation is Goldman’s description of the CCP’s position with respect to literature: “no longer was literature to reflect life as it is or as the individual saw it as exemplified in Lu Xun's work, but as it will be and as the Party and Mao saw it” (ibid.). In other words, the heterogeneity of voice and language within Lu Xun’s writing was to be replaced by the homogeneity of the Party’s ‘monoglossic,’ if you will, interpretation of the world.

5.4.4 Text Excerpt (Lyell 1990)

Moving the focus back to analysis of the texts, the next excerpt comes from Lyell’s 1990 translation published by the University of Hawaii Press as Diary of a madman and other stories. Lyell’s translation is quite different to the previous two translations in that it reflects a discernable effort on the part of the translator to represent the linguistic variation present in the original. As is demonstrated in the discussion presented above, comparison between the various translations is not meant to be prescriptive or critical of the work, but, instead, is offered as a means through which to highlight differences in translator choice and ideology.

阿Ｑ又很自尊，所有未莊的居民，全不在他眼睛裡，甚而至於對於兩位 “文童”也有以為不值一笑的神情。夫文童者，將來恐怕要變秀才者也； 趙太爺錢太爺大受居民的尊敬，除有錢之外，就因為都是文童的爹爹， 而阿Ｑ在精神上獨不表格外的崇奉，他想：我的兒子會闊得多啦！ (Lu Xun 2002:3)

67 Since he thought so well of himself, Ah Q considered the other villagers simply beneath his notice. He went so far with this that he even looked down his nose at the village’s two Young Literati. He didn’t realize, of course, that up there in the rarefied world of scholar‐officialdom those whom one doth Young Literati name can darn well get to be those whom one must Budding Talents proclaim – if you don’t keep an eye on them. That’s why Old Master Qian and Old Master Zhao were so all‐fired respected in the village: they were daddies to those two Young Literati – and rich to boot. Ah Q, however, was less than impressed. “My son’s gonna be a lot richer.” (Lyell 1990:108)

Analyzing the passage in question, we immediately observe a hybridity in the voice of the villagers. Speaking in the voice of the villagers of Weizhuang, Lyell’s narrator proclaims in an overtly literary tone, “those whom one doth Young Literati name (can darn well get to be) those whom one must Budding Talents proclaim.” In the middle of this lofty utterance, which Lyell has offset explicitly with his use of italics (as above), the villager’s ‘slangy, American’18 vernacular resurfaces (in bold and underlined above), further foregrounding the juxtaposition of language varieties within the single voice. Not only is the relative distance of register between these two varieties quite well defined, but also the literary styled variety used may be regarded as a quasi‐H variety of English that has not been used in spoken English since well before the time — early 20th century — in which the novel was set; thus, maintaining Ferguson’s claim of disparity of function between H and L varieties.

As mentioned in Huang’s “The Inescapable Predicament,” Lyell’s narrator speaks, throughout the novella, in a colloquial, slangy, American tone as well (1990:5). In the case of the passage above, this makes it difficult to discern the narrator’s voice from one that, potentially, could represent either the voice of Ah Q or of the villagers. That having been said, however, we may, in Lyell’s defence, suggest that there is variation in degree of ‘colloquiallity’ between what we know as the narrator’s voice and what we have seen in the original is the hybridized voice of the villagers. The problem is that this assessment of degree of colloquiallity, in this case, is confounded by the fact that Lyell’s narrator has such a propensity for the

18 As mentioned earlier, Tambling calls Lyell’s vernacular “American, racy and slangy” (2007:5).

68 vernacular. Case in point, the following phrases underlined in the passage above: ‘went so far’; ‘looked down his nose at’; ‘keep an eye on them’; ‘so all‐fired respected’; and ‘rich to boot’. It is difficult to judge whether or not these colloquialisms are actually appreciably different in register to the two utterances, ‘can darn well get to be’ and ‘daddies’ — presumably, meant to represent the L‐ variety aspect of the villagers’ voice.

Finally, although the voice of Lyell’s narrator is presented in the vernacular, his representation of Ah Q’s voice is such that the reader is, at least, still able to recognize the difference between the two. The use of ‘gonna’ in the representation of Ah Q’s mental discourse (underlined above) is, in fact, a level of colloquialism that Lyell’s narrator does not ‘lower’ himself to at any point in the novella. In this sense, Lyell’s translation strategy can be said to work on a level of relative difference — although it is overly complicated — that is consistent with Pym’s syntagmatic alteration of distance and Catford’s “equivalence across varieties”. However, my own feeling is that Lyell’s translation choice also may be complicated by a desire to represent Lu Xun’s vernacular style through easily recognizable, stereotypical markers associated with popular “American” vernacular of the early part of the 20th century — close to the time in which the novel was set.

5.4.5 Text Excerpt (Lovell 2009)

In clear contrast to the previous paragraph’s discussion of Lyell’s slangy, American style, the present section will now focus upon the lofty, British style of Julia Lovell’s narrator. In 2009, Penguin Books published Lovell’s The Real Story of Ah-Q and Other Tales of China. It is the most recent translation of Ah Q Zhengzhuan. Lovell’s translation is as follows:

阿Ｑ又很自尊，所有未莊的居民，全不在他眼睛裡，甚而至於對於兩位 “文童”也有以為不值一笑的神情。夫文童者，將來恐怕要變秀才者也；

69 趙太爺錢太爺大受居民的尊敬，除有錢之外，就因為都是文童的爹爹， 而阿Ｑ在精神上獨不表格外的崇奉，他想：我的兒子會闊得多啦！ (Lu Xun 2002:3)

Ah‐Q had a robust sense of his own self‐worth, placing the rest of Weizhuang far beneath him in the social scale. Even the village’s two aspiring young scholars – the Zhao and the Qian sons – he considered with haughty contempt. In time, they could both reasonably be expected to get through at least the lowest rung of the official examinations – the path to power and riches. Their fathers, the venerable Mr Zhao and Mr Qian, therefore received the village’s craven respect not just for their personal wealth, but also for their son’s academic prospects. Only Ah‐Q remained invulnerable to the glamour of their future promise: My son will be much richer than them! He thought to himself. (Lovell 2009: 84)

As with Lyell’s translation, Lovell’s version is clearly working to represent the linguistic variation found in the original. Unlike Lyell, however, Lovell’s problem is not an overly vernacularly styled narrator but an overly literary vernacular. Lovell’s narrator expresses himself in a similarly pompous and lofty style to that of Lu Xun’s narrator. This can be seen in the following phrases underlined in the passage above: ‘robust sense of his own self‐worth’; ‘considered with haughty contempt’; ‘the village’s craven respect’; ‘academic prospects’; and ‘remained invulnerable to’. The highly literary register of Lovell’s narrator helps to represent the parodic juxtaposition of, as Huang (1990) calls it, a biographer using the ‘elegant’ discourse and a story of an entirely ‘insignificant person’. However, Lovell has also coupled this with a vernacular voice that is, perhaps, overly high in register. As a result, the unique diglossic relationship that exists between H and L varieties of Chinese does not clearly emerge. In the case of the villagers’ voice, for example, Lovell seems to have chosen to present the first part of the statement – ‘they could both reasonably be expected to get through at least the lowest rung of the official examinations’ – in the narrator’s voice. She then offsets what is presumably the villagers’ view – ‘the path to power and riches’ – from the rest of the sentence by way of hyphen.

Unfortunately, because the utterance is not entirely different in register from that of the narrator, the reader is, again, unsure of whether this is the view of the narrator

70 or of someone else. Moreover, the hybridity of the villagers’ voice present in the original is in no way re‐created here in Lovell’s translation. Perhaps, if Lovell were to present the entire sentence in more of a declarative form, similar to what we saw in Wang’s version, she could then lower slightly the register of ‘power and riches’ to make it more clear that not only does the entire utterance belong to the voice of the villagers, but, also, this voice is, as in the original, a hybridization of H and L variants.

Although the passage does not clearly demonstrate how Lovell treats the voice of Ah Q, throughout the rest of the novella, she does give the character a distinct register and style that is fairly obviously different from that of the narrator. Again, however, I would suggest that the difference in register could be made greater still and, thus, more obvious.

71 6. Conclusion

Through presentation and discussion of major Translation Studies research trends — namely, an unrealistic, disproportionately mono‐lingual/mono‐variant treatment of dialect — relating to (un‐) translatability of linguistic variation, the present paper has tried to set a context within which alternative perspectives of variation — namely, diglossia and heteroglossia — may be afforded some attention.

In particular, analysis of four English‐language translations — Wang Chi‐chen’s Ah Q and others: Selected stories of Lusin (1941); Yang Xianyi and Gladys Yang’s The true story of Ah Q (1956); William A. Lyell’s Diary of a madman and other stories (1990); and Julia Lovell’s The real story of Ah­Q and other tales of China (2009) — has provided an example of translation from a diglossic language — Chinese — into a non‐diglossic one — English. Moreover, the diglossic scenario presented in the original has been shown to be an important feature of the parody of traditional Chinese historiography functioning within the source text, in particular, and the historicized Chinese referential frame, in general.

Discussion of the four translations also has helped to highlight differences between a variety‐for‐variety, equivalence‐fidelity‐based view of dialect quite common within the (un‐) translatability of linguistic variation literature and an ‘alteration of relative distance’ view that the present paper claims is unique to the work of J C Catfords (1965) and Anthony Pym (2000). It is my hope that the present paper may be one of many research projects to come that will extend the scope of the (un‐) translatability of linguistic variation discussion to include investigation of translation into, out of, and within diglossic, heteroglossic, and multilingual language communities and language contact situations.

This paper has tried to paint a clear picture of the condition of the narrator of Lu Xun’s novella The True Story of Ah Q as he is recreated in the English‐speaking world. This has been carried out through a detailed analysis of four separate English‐

72 language translations of Lu Xun’s novella, namely Wang Chi‐chen’s Ah Q and others; Yang Xianyi and Gladys Yang’s The True Story Ah Q; William A. Lyell’s Diary of a Madman and Other Stories; and Julia Lovell’s The Real Story of Ah Q and Other Tales of China. Although analysis was carried out on only one passage, the clear representation of register variation and heteroglossia in this particular passage make it an important point of comparison between original and translation. Moreover, throughout the novella, the characteristics of the voice of the narrator, the villagers and of Ah Q as demonstrated by the passage analyzed do not change in regard to register variation or heteroglossia. Also, the reproducibility of the analysis presented here makes it quite simple to investigate further passages of the text if necessary.

With respect to the analysis presented in the present paper, the focus has been places upon (non‐) representation within the four translations of diglossia, heteroglossia, and linguistic variation as presented in Lu Xun’s Ah Q Zhengzhuan. In particular, discussion has been focused upon representation of linguistic variation within the story’s narrative as it functions to foreground a parodic treatment of Chinese traditional historiography and the diglossic language contact situation in Pre‐modern China.

In summary, analysis has produced the following results:

Firstly, with respect to the representation of heteroglossia in the four English translations, only Lyell’s work succeeds fully. Although Lovell does present a heteroglossic narrative, in the particular case of the sampling of text used for analysis in this paper, her narrative is not clearly differentiated enough to produce the effect present in Lu Xun’s original and, therefore, is only marginally successful.

Secondly, none of the four translations were able to fully re‐create the hybridity within voice that we saw in the voice of the villagers in the sampling of text from the original. Perhaps, this characteristic of Lu Xun’s narrative was difficult to reproduce

73 because it not only involves the representation of two distinct registers within the single voice of the villagers but, also, needs something to offset both of those registers from that of the narrator. Obviously, the non‐diglossic nature of the modern day English language situation makes it difficult for translators to effectively deal with the diglossic situation presented in Ah Q Zhengzhuan. It may be possible, however, to make use of archaic English grammar and lexicon in order to represent the non‐vernacular characteristic of wenyan in the Chinese diglossic scenario. In fact, Lyell’s translation demonstrates the use of both archaic grammar and archaic lexicon.

Thirdly, only Lovell’s translation was able to recreate the parodic narrative of the original, since only hers had a clear, and more or less distinct, class‐related variance in register visible between the language of the narrator and that of the subject of the narration, the character of Ah Q. As I mentioned earlier, this difference could have been made more obvious still so that the parodic effect produced might be even greater. In addition, Lovell is the only one of the four translators to provide a preface outlining some of the socio‐historical background necessary for an understanding of the context in which Lu Xun’s parody functions. She also provides the reader with a list of “further readings” related to the historical moment in which Lu Xun lived and produced Ah Q Zhengzhuan. Lyell (1990) does provide extensive footnotes to accompany his translation. However, socio‐historical details are not discussed as thoroughly as they are in Lovell’s work. As has been mentioned already, within the field of Translation Studies, issues relating to the visible involvement of a translator within a text only began to come under scrutiny in the 1990s (Bassnett 2002:9). As a result, the historical context of the translator must also be considered when discussing the presence or absence of any overt translator involvement. Both Wang (1941) and the Yangs (1956) published their translations well before any discussion of translator visibility had taken place.

Fourthly, in terms of creating a distinction between the language of Ah Q and that of his narrator, both Lovell and Lyell were successful; however, because Lyell’s

74 narrator had such a propensity for the use of vernacular, his translation was, at the same time, unsuccessful in recreating any parodic narrative function.

In closing, I would like to suggest some possible areas for investigation. First, with respect to Translation Studies and the concept of diglossia, it may be fruitful to develop further the idea that translation into a diglossic language contact situation may have some effect upon the H‐L relationship present in the domestic language situation. One example, discussed briefly in the present paper, is the proportionately greater amount of L‐variant grammar and lexicon in translated Chinese Buddhist texts in comparison to that found in indigenously composed Chinese Buddhist texts (Mair 1994:712).

Second, with respect to Lu Xun and his works in translation, it may be interesting to compare his rhetoric of parody with that used in classic English literature. It has been documented, for example, that Lu Xun was greatly influenced by the works of great parody writers such as Nikolai Gogol and Jonathan Swift.19 A comparison of the parodic styles of authors who worked and lived in linguistic and socio‐historical contexts quite different to Lu Xun may produce some insight into possible options for the representation of Lu Xun’s parody in translation. Finally, further exploration of Pym’s (2000) notion of “syntagmatic alteration of distance” as basis for discussion and translation of linguistic variation is needed. In particular, it would be useful to investigate viability of the notion in terms of translation of texts that employ extensive linguistic variation such as George Bernard Shaw’s (1912) Pygmalion and Mark Twain’s (1884) Adventures of Huckleberry Finn.

19 There is a discussion in Patrick Hanan’s “The Technique of Lu Hsün's Fiction,” regarding the influence of Western authors on the literary consciousness of Lu Xun.

75 Bibliography

Altano, W. Brian (1988) Translating dialect literature: the paradigm of Carlo Emilio Gadda, Babel 34(3): 152–156.

Anderman, Gunilla (2007) Voices in translation: bridging cultural divides, Clevedon: Multilingual Matters.

Baker, Mona (2011) Routledge encyclopedia of translation studies, London; New York: Routledge.

Bassnett, Susan (2002) Translation studies, London; New York: Routledge.

Brisset, Annie (1996) A Sociocritique of Translation: Theatre and Alterity in Quebec, 1968‐ 1988, trans. Rosalind Gill and Roger Gannon, Toronto: U of Toronto Press.

Britto, F. (1991) ‘Tamil diglossia: An interpretation’, Southwest Journal of Linguistics 10(1): 60–84.

Catford, JC (1965) A linguistic theory of translation: an essay in applied linguistics, Oxford University Press.

Chen, Ping (1993) ‘Modern Written Chinese in development’, Language in Society, 22(4): 505‐537.

Chen, Ping (1999) Modern Chinese: history and sociolinguistics, Cambridge, U.K.; New York, NY: Cambridge University Press.

Corbett, J (1997) ‘Writtin in the langage of Scottis natioun: literary translation into Scots’, in Susan Bassnett (ed.) Translating Literature, Cambridge: Brewer, 95‐118.

Corder, S. Pit (1973) Introducing Applied Linguistics, Harmondsworth: Penguin.

Coulmas, Florian (1987) ‘What writing can do to language: some preliminary remarks’, in S. P. X. (ed.) Georgetown University Roundtable on Languages and Linguistics 1986, Washington, D.C.: Georgetown University Press, 107‐129.

Coulmas, Florian (1989) The writing systems of the world, Oxford, UK; New York, NY: B. Blackwell.

Coulmas, Florian (2005) Sociolinguistics: The Study of Speakers’ Choices, Cambridge: Cambridge University Press.

Cronin, Michael (1995) ‘Altered States: Translation and Minority Languages’, Traduction, Terminologie, Rédaction (TTR), 8(1): 85‐103.

Denison, Norman (1971) ‘Some Observations on Language Variety and Plurilingualism’, in John B. Pride and Janet Holmes (eds.) Sociolinguistics, Harmondsworth: Penguin, 65‐77.

76

Deshpande, Madhav (1991) ‘Conceptions of diglossia in the writings of Sanskrit grammarians’, Southwest Journal of Linguistics 10(1): 23‐40.

Davies, Gloria (1991) ‘The Problematic Modernity of Ah Q’, Chinese Literature: Essays, Articles, Reviews (CLEAR) 13: 57‐76.

Eber, Irene (1985) ‘The Reception of Lu Xun in Europe and America: The Politics of Popularization and Scholarship’, In Leo Ou‐fan Lee (ed.) Lu Xun and His Legacy, Berkeley: Univ. of California Press, 242‐273.

El‐Badarien, Mohammed and Muhammad Raji Zughoul (2004) ‘Diglossia in Literary Translation: Accommodation into Translation Theory’, Meta 49(2): 447‐456.

Fan, Shouyi (1999) ‘Highlights of Translation Studies in China Since the Mid‐Nineteenth Century’, Meta, 44:1, 27‐43.

Feng, Zongxin (2008) ‘Fictional narrative as history: Reflection and deflection’, Semiotica, 170(1‐4): 187‐200.

Ferguson, Charles (1959) ‘Diglossia’, in Thom Huebner (ed.) Sociolinguistic Perspectives: Papers on Language in Society, 1959‐1994, Oxford University Press, 25‐39.

Ferguson, Charles (1991) ‘Diglossia revisited’, Southwest Journal of Linguistics 10(1): 214 – 234.

Ferguson, Charles (1996) ‘Epilogue: Diglossia Revisited’, in El‐Said M Badawi and Alaa Elgibali (eds.) Understanding Arabic: essays in contemporary Arabic linguistics in honor of El‐Said Badawi, Cairo, Egypt: American University in Cairo Press, 49‐67.

Fernandez, Mauro (1993) Diglossia: a comprehensive bibliography, 1960‐1990, Amsterdam: John Benjamins.

Fishman, Joshua (1967) ‘Bilingualism with and without diglossia; diglossia with and without bilingualism’, Journal of Social Issues 23(2): 29‐38.

Fishman, Joshua (1971) Advances in the Sociology of Language, The Hague: Mouton Row.

Franke, Wolfgang (1972) The reform and abolition of the traditional Chinese examination system, Cambridge, Mass: Harvard Univ. Press.

Gafaranga, Joseph (2009) ‘The Conversation Analytic model of code‐switching’, In Barbara E Bullock and Almeida Jacqueline Toribio (eds.) The Cambridge handbook of linguistic code‐switching, Cambridge; New York: Cambridge University Press, 114‐126.

Gardner, Charles S (1961) Chinese Traditional Historiography, Cambridge, Mass: Harvard University Press.

Gibbs, Donald A and Yun‐chen Li (1975) A bibliography of studies and translations of modern Chinese literature, 1918‐1942, Cambridge, Mass: Harvard University Press.

77

Gilbert, Glenn (1975) ‘Review of Fishman's The Sociology of Language’, Language, 51(1): 234‐236.

Goldman, Merle (1982) ‘The Political Use of Lu Xun’, The China Quarterly 91: 446‐461.

Gumperz, John (1958) ‘Dialect Differences and Social Stratification in a North Indian Village’, American Anthropologist, 60: 668‐681.

Gumperz, John (1961) ‘Speech variation and the study of Indian civilization’, American Anthropologist 63: 976‐988.

Gumperz, John (1962) ‘Types of linguistic communities’, Anthropological Linguistics 4(1): 28‐40.

Gumperz, John (1971) Language in social groups: essays, Stanford, Calif.: Stanford Univ. Press.

Halliday, M.A.K., McIntosh, A. and Strevens, P. (1964) The Linguistic Sciences and Language Teaching, London: Longman.

Hanan, Patrick (1974) ‘The Technique of Lu Hsün's Fiction’, Harvard Journal of Asiatic Studies, 34: 53‐96.

Hass, Walter (2002) Comment, International Journal of the Sociology of Language 157: 109‐ 115.

Hatim, Basil and Ian Mason (1990) Discourse and the Translator, UK: Longman.

Hawkins, Peter (1983) ‘Diglossia revisited’, Language Sciences, 5(1): 1–20.

Hock, Hans Henrich, and Rajeshwari Pandharipande (1978) ‘Sanskrit in the Pre‐Islamic Sociolinguistic Context of South Asia’, International Journal of the Sociology of Language 16: 11‐25.

Huang, Martin W (1990) ‘The Inescapable Predicament: The Narrator and His Discourse in The True Story of Ah Q’, Modern China 16(4): 430‐449.

Hudson, Alan (1992) ‘Diglossia: a bibliographic review’, Language in Society, 21: 611‐674.

Hudson, Alan (2002) ‘Outline of a theory of diglossia’, International Journal of the Sociology of Language, 157: 1‐48.

Huters, Theodore (1984) ‘Blossoms in the Snow: Lu Xun and the Dilemma of Modern Chinese Literature’, Modern China, 10(1): 49‐77.

Ives, Sumner (1950) ‘A theory of literary dialect’, Tulane Studies in English, 2: 137‐182.

Kahane, Henry and Renee (1979) ‘Decline and Survival of Western Prestige Languages’, Language, 55: 183‐196.

78

Kao, George (1942) ‘Review: Ah Q and others: Selected stories of Lusin’, The Journal of Asian Studies, 1(3): 280‐281.

Kaye, Alan S (1991) ‘Is English diglossic?’, English Today 7: 8‐14.

Kaye, Alan S (2001) ‘Diglossia: the state of the art’, International Journal of the Sociology of Language, 2001:152, 117–129.

Keller, R. (1982) ‘Diglossia in German‐speaking Switzerland’, In Walter Haas (ed.), Standard languages: Spoken and written, 70 –93. Manchester: Manchester University Press.

Kowallis, Jon (1994) Review: ‘Diary of a Madman and Other Stories. By Lu Xun, translated by William A. Lyell. [Honolulu: University of Hawaii Press, 1990]’, The China Quarterly, 137: 283‐284.

Lane‐Mercier, Gillian. (1997) ‘Translating the Untranslatable: The translator’s Aesthetic, Ideological and Political Responsibility’, Target 9(1): 43‐68.

Lee, C. (2010) Cosmopolitan Publics: Anglophone Print Culture in Semi‐Colonial Shanghai (review). Comparative Literature Studies 47(3), 379‐381. Penn State University Press.

Lee, Gina M. (1986) ‘Diglossia in Ancient India’, Working Papers in Linguistics 34: 151‐164.

Lee, Leo Ou‐fan (1985a) Lu Xun and His Legacy, Berkeley: Univ. of California Press.

Lee, Leo Ou‐fan (1985b) ‘Contemporary Chinese Literature in Translation’, The Journal of Asian Studies 44(3): 561‐567.

Lefevere, André (1992a) Translating Literature: Practice and Theory in a Comparative Literature Context, Modern Language Association of America.

Lefevere, André (1992b) Translation, Rewriting and the Manipulation of Literary Fame, London: Routledge.

Lin Yü‐sheng (1979) ‘Review of William A. Lyell Lu Hsün's Vision of Reality’, The Journal of Asian Studies, 38: 365‐368.

Liu, Lydia (1995) Translingual Practice: Literature, National Culture, and Translated Modernity — China, 1900‐1937, Stanford, California: Stanford University Press.

Lovell, Julia (2006) The Great Wall: China Against the World, 1000 BC ‐ 2000 AD, New York: Grove Press.

Lovell, Julia, trans. (2009) The Real Story of Ah Q and Other Tales of China. By Lu Xun. London: Penguin.

Lovell, Julia (2010) Julia Lovell on translating Lu Xun's complete fiction: ‘His is an angry, searing vision of China’, Interview. Available online at

79 [http://www.danwei.org/translation/julia_lovell_complete_lu_xun_f.php] (accessed on June 2011).

Luo, Jing (2004) Over a Cup of Tea: An Introduction to Chinese Life and Culture, Maryland: University Press of America.

Lyell, William A (1976) Lu Hsün's vision of reality, California: Univ. California Press.

Lyell, William A., trans. (1990) Diary of a Madman and Other Stories. By Lu Xun. Hawaii: Univ. Hawaii Press.

Lu Xun (魯迅 著) (2002)《阿Q正傳》 , 臺北市: 臺灣麥克

Mair, Victor H. (1994) ‘Buddhism and the Rise of the Written Vernacular in East Asia: The Making of National Languages’, The Journal of Asian Studies 53 (3): 707‐751.

Marfany, Joan‐Lluís (2010) ‘Sociolinguistics and some of its concepts: a historian's view’, International Journal of the Sociology of Language, 206: 1–20.

McDougall, Bonnie (2009) ‘Authorising and Authorised Translation’, Conference on Translation and Authority, University of British Columbia, Vancouver, Canada, 6‐8 March. 1‐39. Internet. 3 Jan. 2012.

Meylaerts, Reine (2006) ‘Literary heteroglossia in translation’, in João Ferreira Duarte, Alexandra Assis Rosa, Teresa Seruya (ed) Translation studies at the interface of disciplines, Amsterdam; Philadelphia: J. Benjamins, 85‐98.

Mkilifi, M. H. Abdulaziz (1972) ‘Triglossia and Swahali‐English Bilingualism in Tanzania’, Language in Society, 1: 197‐213.

Müller, N. and Ball, M.J. (2005) ‘Code‐switching and diglossia’, In Ball, M.J. (ed.) Clinical Sociolinguistics, Oxford: Blackwell, 49‐62.

Munday, Jeremy (2009) The Routledge companion to translation studies, London; New York: Routledge.

Newmark, Peter (1988) A Textbook of Translation, UK: Prentice Hall.

Nida, Eugene (1976) ‘A Framework for the Analysis and Evaluation of Theories of Translation', in Richard W. Brislin (ed) Translation: Applications and Research, New York: Gardner Press, 48‐91.

Nilep, Chad (2006) 'Code Switching in Sociocultural Linguistics’, Colorado Research in Linguistics 19: 1‐22.

Pritchard, Earl (1942) ‘Outstanding Books on the Far East Published in 1941’, The Journal of Asian Studies, 1(3): 247‐252.

80 Pym, Anthony (2000) ‘Translating Linguistic Variation: Parody and the Creation of Authenticity’, in Miguel A. Vega & Rafael Martín‐Gaitero (ed) Traducción, metrópoli y diáspora, Madrid: Universidad Complutense de Madrid, 69‐75.

Rubin, Joan (1968) ‘Bilingual Usage in Paraguay’, in Joshua Fishman (ed.) Readings in the Sociology of Language, The Hague: Mouton, 512‐530.

Sanchez, Maria T (1999) ‘Translation as a(n) (Im)possible Task: Dialect in Literature’, Babel 45 (4): 301‐310.

Sanchez, Maria T (2009) The problems of literary translation: a study of the theory and practice of translation from English into Spanish, Bern, Switzerland: Peter Lang AG, International Academic Publishers.

Schiffman, Harold (1998) ‘Diglossia as a Sociolinguistic Situation’, The Handbook of Sociolinguistics, Ed. Florian Coulmas, Oxford, UK: Blackwell Publishers. 205‐216.

Schogt, Henry G (1988) Linguistics, literary analysis, and literary translation, Toronto: University of Toronto Press.

Shek, Ben‐Z (1988) ‘Diglossia and Ideology: Socio‐Cultural Aspects of «Translation» in Québec’, Traduction, Terminologie, Rédaction TTR 1(1): 85‐91.

Shek, Ben‐Z (1977) ‘Quelques réflections sur la traduction dans le contexte socio‐culturel canado‐québécois’, Ellipse 21: 111–117.

Shen, Shuang (2009) Cosmopolitan publics: Anglophone print culture in semi‐colonial Shanghai, New Brunswick, N.J.: Rutgers University Press.

Snow, Edgar (1937) Living China; modern Chinese short stories, New York, USA: Reynal & Hitchcock.

Snow, Don (2010a) ‘Diglossia in East Asia’, Journal of Asian Pacific Communication, 20(1): 124‐151.

Snow, Don (2010b) ‘Hong Kong and Modern Diglossia’, International Journal of the Sociology of Language 206: 155–179.

Solomon, Jon (1993) ‘Taking Tiger Mountain: Can Xue’s Resistance and Cultural Critique’, in Tani E Barlow (ed) Gender politics in modern China: writing and feminism, Durham, US: Duke Univ. Press, 238‐265.

Tambling, Jeremy (2007) Madmen and Other Survivors: Reading Lu Xun's Fiction, Hong Kong: Hong Kong Univ. Press.

Timm, Lenora A. (1981) ‘Diglossia Old and New: A Critique’, Anthropological Linguistics (23)8: 356‐367.

Venuti, Lawrence (1996) ‘Translation, Heterogeneity, Linguistics’, Traduction, Terminologie, Rédaction (TTR), 9(1): 91‐115.

81

Vogt, Hans (1954) ‘Language Contacts’, Word 10(2‐3): 365‐374.

Wallaert, Ineke (2001) ‘Translating Sociolects: A Paradigm of Ideological Issues in Translation’, in Janet Cotterill and Anne Ife (ed) Language Across Boundaries, London: Continuum, 171‐184.

Walters, K. (1996) ‘Diglossia, linguistic variation, and language change in Arabic’, In M. Eid (ed.) Perspectives on Arabic Linguistics 8: Papers from the Eighth Annual Symposium on Arabic Linguistics, Amsterdam: John Benjamins, 157 – 197.

Wang, Chi‐chen (1941) Ah Q: and others, selected stories of Lusin, New York: Columbia Univ. Press.

Wang, Edward Q (2001) Inventing China through history: the May Fourth approach to historiography, Albany, NY: State University of New York Press.

Weinreich, Uriel (1953) Languages in Contact, The Hague: Mouton.

Wilkinson, Jane (2005) ‘Staging ‘Swissness’: Inter‐ and Intracultural Theatre Translation’, Language and Intercultural Communication, 5(1): 72‐85.

Xu, Xiaomin (2011) ‘A comparative study of English translations of Lu Xun's works’, Babel, 57(3): 324‐341.

Yang, Gladys (1983) A Small Town Called Hibiscus, Beijing, China: Chinese Literature: Distributed by China Publications Centre (Guoji Shudian).

Yang Xianyi and Gladys Yang (1956) Selected Works of Lu Hsun Vol. 1, Beijing, PRC: Foreign Languages Press.

Yang, Xianyi (2002) White Tiger: An Autobiography of Yang Xianyi, Chinese University Press.

Zhelokhovtsev, A. N. (1982) ‘Lu Xin in American Sinology’, Far Eastern Affairs in English [Moscow ], 4: 67‐76.

董大中 [Dong Dazhong] (2007) 魯迅日記箋釋(一九二五年), 台北市, 台灣: 秀威資訊.

张杰 [Zhang Jie] (2002) 鲁迅: 域外的接近与接受, 北京, 中國: 社会科学文献出版社.

82