The pursuit of beauty by an aesthete : A study of Harold Acton ’ s manuscripts of Popular Chinese Plays

This article examines Harold Acton‟s translation of Popular Chinese Plays. It examines the modifications Acton made in the final polishing stage of his translation and explores possible linguistic, aesthetic and social considerations for such alterations. Four plays—Ch‟un-hsiang Nao Hsueh 春香闹学, Shih Hou Chi 狮吼记, Lin Ch‟ung Yeh Pên 林冲 夜奔 , and Yü Chou Fêng 宇宙锋—were selected for this research. Computerized data were analysed using Wordsmith 4.0. This article demonstrates that at this later stage of the translation‟s genesis, Acton gives priority to the overall effectiveness and coherence of the translation while paying less attention to lexical and syntactical details. His corrections to the manuscript indicate that the translator departs from the source text, displaying more independence as a writer. The purpose of this analysis is to show how translators‟ strategies can evolve during the genesis of a translation, thus complicating the common assumption that translators adopt a consistent “stance” when translating.


Genetic criticism and translation
Genetic criticism defines itself in opposition to -textualist‖ approaches to the study of literature that consider the published text to the exclusion of its avant-textes (notes, earlier versions, drafts, its manuscripts) and with little concern for the contingencies of the text's production or the phases of its coming into being (de Biasi, 1996).It thus offers a paradigm shift for studying the translated text as a dynamic process involving plural, meaningful variations rather than as a static product (Deane-Cox, 2014, pp. 192-193; see also Gilberthorpe, 2015, pp. 141-147;Passos, 2008, pp. 150-151).Such a perspective brings the temporality of translation into view and presents translation researchers with -not the text, but texts‖ (Hay, 1977, p. 73).Genetic translation studies have advanced the notion that the translator -unfinishes‖ the source text, rendering it an avant-texte (draft) of the translation (Scott, 2006, p. 107), she or he -transforms the text back into a process of writing‖ and -multiplies its possibilities of being‖ (Scott, 2006, p. 107).This continuum of writing extends through the deletions, additions, substitutions, restructuring and other metamorphoses brought about by the translator.Using textual evidence of such processes, genetic scholars attempt to reconstruct the writing process, even if their resources sometimes appear like -the ashes remaining when a fire is consumed or […] the footprints on the ground after a dance is over‖ (Ferrer, & Groden, 2004, p. 11).Genetic studies have, nevertheless, been effective in developing hypotheses to account for -the birth of motivations, strategies and metamorphoses of writing‖ (de Biasi, 1996, p. 29) and in answering the difficult question of how to describe how a writer writes.Similarly, comparing and analysing the avant-textes of a translation can help to describe how a translator translates.Furthermore, such research can locate differences between the writing of the writer and that of the translator, thus revealing the different kinds of creativity used in these different activities.
The translator's creativity has been widely acknowledged in recent translation studies.However, outside of controlled experiments with working translators, descriptions of the actual processes of translation have remained largely theoretical.The purpose of the present study is to provide a descriptive linguistic analysis of Harold Acton's editing of his translations of Chinese plays.It attempts to account for the way in which the translator rendered the fictive universes of his Chinese plays differently in English by retracing a moment in their respective geneses.
To do so, this study will address the following questions: (1) How did Acton translate Popular Chinese Plays?(2) What are the general patterns of his creative interventions at the near-completion stage?(3) What considerations motivate these modifications?This analysis will attempt not only to categorize -the outcome of the translator's choices‖, but also to infer and thus step closer to -the motives, the pattern of instructions which informed the choices‖ (Hermans, 1999, p. 23).

Sir Harold Acton and his translation of Popular Chinese Plays
In China, Sir Harold Acton (1904Acton ( -1994) ) is remembered more as a translator of traditional Chinese plays, novels and poetry than as the writer and poet he is known as in the West.Acton described himself as both a -fervent votary‖ and -an irrepressible addict‖ of Chinese theatre (Acton, 1948(Acton, , p. 354, 1970, p. 5), p. 5), his -aesthetic nourishment‖ (Acton, 1970, p. 5).This passion led to his co-translation of Famous Chinese Plays (戏剧之精华) with L. C. Arlington (Arlington & Acton, 1937) and Peach Blossom Fan (桃花扇) in collaboration with Ch'ê n Shih-hsiang ( 陈世骧) (Kung, 1976).However, his project of translating Chinese theatre was much more ambitious than his publications would suggest.In his memoirs, Acton (1948) mentions that in 1936 he embarked on an ambitious programme of translation to -introduce English readers to a whole library of popular Chinese literature‖ (p.365), which included a new volume of his favourite plays.In his later memoir, he reflects upon these translations: The Peking theatre had become my chief pastime and hobby, and I spent half the day translating the plays I preferred with my teacher Mr. Chou who knew no English.[…] A stout volume of these translations was to have been published by Henri Vetch of the French Bookstore, but beyond correcting some proofs just before the outbreak of war, I have not seen or heard of them since.The manuscript must count as a war casualty.This was a grief to me at the time as the plays I had selected were in many respects superior to those I had rendered in condensed form with L. C. Arlington.With their lilt in my ears and the lithe actors leaping before my eyes I fancied I could convey some of their charm to Western playgoers, though the texts were the mere framework of many an exquisite performance.(Acton, 1970, p. 2) While Acton laments the loss of his manuscript, the Harold Acton Papers (GEN MSS 663, 1932-1941) at Yale University's Beinecke Rare Book and Manuscript Library contain a corrected typescript identified as -the unpublished manuscript for Popular Chinese Plays, which was once thought lost‖ (-Guide to the Harold Acton Papers‖, n.d.).Yet this late corrected typescript is not -the‖ manuscript of the translation, but the one surviving manuscript of a series of formulations and revisions.It is unlikely to be the corrected -proofs‖ to which Acton refers above, because there is variation between it and three plays published in 1939 in the T"ien Hsia Monthly (Acton, 1939a-c).According to the Beinecke Library, its manuscript -was contained in the original leather trunk in which it was packed for dispatch from Beijing, China, after the invasion of the Japanese in 1937‖ (-Guide to the Harold Acton Papers‖, n.d.).Furthermore, -[a]ccording to the dealer, the integrity of the archive was strictly maintained and nothing was removed‖ (-Guide to the Harold Acton Papers‖, n.d.). 2 One may conclude, however, noting the finished quality of the translation and the fact that Acton went to the effort to ship this corrected typescript from China to London, that if it is not the corrected proofs, then it is likely to be the most advanced state of the translation to have been in his possession.This corrected typescript is thus a late avant-texte and possibly Acton's final revision, from which the version he sent to publisher Henri Vetch derives.Acton's comments above suggest that he made a subsequent revision to his translation, correcting the proofs that Vetch returned to him, and that this version has been lost.If not the final revision, the manuscript in the Beinecke Library is a rich source of material for assessing Acton's strategies as a translator during at least three phases of his work: at the composition of the typescript (which is already an advanced state of the translation), and two subsequent revisions witnessed by this typescript as Acton corrected in pen and in pencil.Indeed, this rare avant-texte permits here the first genetic translation study of an English translation of traditional Chinese literature.

Research methodology
This study focuses on four works included in Acton's manuscripts.The first three are Kunqu Opera 昆曲 and have been published in the magazine T"ien Hsia Monthly (Acton, 1939a-c); the fourth is a Peking Opera 京剧.
1. Ch"un-hsiang Nao Hsueh 春香闹学 (CHNH), translated into English as Ch"un-hsiang Turns the Schoolroom Topsy-Turvy.This is an excerpt from the Mu-tan T'ing 牡丹亭 or The Peony Pavilion, a masterpiece of the Southern School of Ming Dynasty theatre.
2. Shih Hou Chi 狮吼记 (SHC), translated into English as The Roaring Lioness.This is a Chinese version of Shakespeare's The Taming of the Shrew.In this apocryphal, humorous Kunqu comedy, the hen-pecked husband is, unlike Petruchio, helped by his friends to rectify his wife's shrewish behaviour.
3. Lin Ch"ung Yeh Pê n 林冲夜奔 (LCYP), translated into English as Lin Ch"ung Flees by Night.This work portrays the desperate flight of Lin Ch'ung, a military officer.
4. Yü Chou Fê ng 宇宙锋 (YCF), translated into English as The Sword of the Universe.Yü Chou Fê ng owes its popularity to the masterly portrayal of feigned insanity by the virtuous heroine as she declines the marriage proposal of the wicked Emperor.
The three different phases of translation to which the avant-texte bears witness are described in this way: 1.The typescript of the English translation prior to any revision is referred to as draft 1(D 1 ); 2. The revised typescript with substantial handwritten modifications of D 1 in pen is referred to as draft 2 (D 2 ); 3. Minor, occasional revisions in pencil in D 1 , possibly modified after D 2 , referred to as draft 3 (D 3 ). 3  The genetic dossier for the present study also includes the published versions of CHNH, SHC and LCYP, which appeared in the T"ien Hsia Monthly (Acton, 1939a-c).
This article frames translation for the theatre within Poyatos' semiotic metamorphosis theory, where -all the acoustic, visual, tactile, kinaesthetic, olfactory and gustatory signs [of the theatre] are reduced to visually perceived ones by the writer‖ (2008, pp. 91-94), which -in turn, and depending on the writer's skills, elici[t] all manner of sensory and intelligible experiences in our [the readers'] imagination‖.The role of the translator is to generate within readers this -inverse process of signchannel amplification‖ (p.93).Views of gestic subtext outlined by Pavis (1989, pp. 25-45) and performability by Bassnett (1998) also inform the present study, which nevertheless proceeds through a linguistic analysis of the phases of translation by combining corpus analysis with process analysis.
Translation modifications are classified into paradigmatic and syntagmatic changes at the lexical level and the syntactic level.A paradigmatic modification refers to the relationship of substitution between words of the same parts of speech, which can be replaced for each other in the same position within a given sentence.Syntagmatic relationships occur along the horizontal axis, where items co-occur sequentially and interact within the same construction.In the corpus analysis we manually annotate the translator's revisions to the translation at the sentence level, using categories derived from Toury (1995/2012, pp. 82-85).Informed by his polysystem network of textual-linguistic, aesthetic and social norms that regulate translation, we analyse the text for colloquial language, coherence, conciseness, poetic effect and interpersonal motives.Computerized data are transformed into a text file and analysed with Wordsmith (4.0) corpus software for quantitative statistical analysis.Using corpus tools to analyse the linguistic features of genetic variants arising from the revisions of the translation manuscript helps to identify Acton's paradigmatic and syntagmatic modification patterns, which might otherwise have gone unnoticed.This approach facilitates the formulation of hypotheses concerning the nature and orientation of the translator's strategies.And, for a genetic translation study, it allows one to identify the prominent characteristics of the translator's writing and revision styles.

Results and discussion
The experience of translation has offered some translators -the sensation of being the writer at work‖ (Valéry, 1992, p. 119).Yet Valéry's model of the singular translator stepping into the shoes of the author cannot be transposed onto Acton's collaborative translation practice.In his preface to Popular Chinese Plays, the translator describes his method.
Though he has no English, my teacher Mr. Chou I-min 周逸民 has been my constant collaborator in that he patiently read all these plays aloud to me, sentence by sentence, explaining difficulties as they cropped up.By following his reading, which was so expressive that I often felt he had missed his vocation in not becoming an actor, the rhythm of the original, crescendo and diminuendo, was in my ears as I committed the words to paper.To insure greater accuracy I went over all the translations in typescript with my two friends Mr. Yen Yü-heng 颜毓蘅 and Mr. Yang Shan-ch'uan 杨善荃, and I owe more than I can easily define to their suggestions, encouragement, advice and patient scrutiny.To Mr. Yang I am particularly indebted for the elucidation of thorny problems which arose in the K'un-Ch'ü plays: his intuition for the mot juste has saved me from many a pitfall.To my learned friends Dr. H. H. Hu 胡先骕 and Mr. Chang En-yü 张恩裕 I am also indebted for valuable information incorporated in the notes and introductions.(Acton, 1933(Acton, -1941, pp. 2-3) , pp. 2-3) While Mr. Chou I-min does not speak English, one cannot be sure whether the revisions to the D 1 typescript were made with the assistance of Acton's other Chinese collaborators or not.But the fact that, as we shall see, most of those adjustments are stylistic rather than corrections to inaccuracies or specific translation problems suggests that this late revision of the English text was probably performed by Acton himself with the aim of improving the text's coherence and cohesion.
The present study prioritizes those plays that were published so as to reveal the proximity between the D 2 and D 3 revisions and the published versions of those texts.With respect to Shih Hou Chi, only modifications in Scenes I and II are calculated -its lengthy unpublished Scenes III and IV are thus omitted from the corpus analysis so that the sample passages from the four plays are of roughly the same length.This allows for a more balanced, consistent comparison.Acton's revisions at D 2 , D 3 and before the published versions of the four works do not show the crossing out of large passages or other examples of substantial reorganizing of the text.
In total, 372 modifications were observed in 356 sentences from the four sample texts.Among them, 141 are cases of paradigmatic modification at the lexical level.In descending order, there are 52 cases of the replacement of verbs, 41 of nouns, 33 of adjectives and 15 of adverbs.There are 49 cases of deletion and 16 cases of insertion.Paradigmatic replacement of words and deletion are the most frequent translation techniques observed.In contrast, miscellaneous insertions are negligible, and they merely include a couple of notes to clarify embedded cultural allusions and revisions of occasional articles, conjunctions and pronouns, which will not be analysed in detail here.
On the other hand, syntactic adjustments or reshuffling at the sentence level and above are clearly identified as the predominant features of the revision.A total of 231 such modifications are identified, which are subdivided in the following manner: 77 introduce more colloquial or idiomatic expression, 64 cases are revisions for textual coherence, 54 revise for more concise expression, 20 involve aesthetic considerations and 16 cases relate to the interpersonal relationships between participants.It can be seen from Table 1 that the type/token ratio (TTR) of text 1 (Acton's D 1 typescript) is 34, higher than the 32 of text 2 (which includes his final revisions at the D 2 and D 3 phase).The standardized TTR of Text 1 is 45.33, whereas that of the later revision is 47.87.This suggests that during his revising Acton enriched the vocabulary and linguistic diversity of his text. 4There are fewer sentences in later versions and the mean sentence length of later versions is 10 English words, shorter than 11 words of D 1 . 5Such a phenomenon is in accordance with the linguistic norm of using concise prose for oral delivery in the theatre.
To glimpse the creativity in Acton's translation decisions, we first examine cases of paradigmatic modifications below sentence level in section 4.1 before studying those at the sentence level or above in section 4.2.This detailed analysis will show a sample of his revisions from across the four texts.Observing features of his revision allows one to discern the consistency of Acton's translation strategy across a wide range of linguistic features of his text; it also helps with formulating hypotheses concerning such strategies.A sample page of the English revision of the translation draft is presented in section 4.3 so that one may witness all such features of his revision at play in a single passage.This study concludes with a summary of the major findings and their significance.

Modifications at the lexical level
Lexical modifications along the paradigmatic axis occur as replacements of verbs, nouns, adjectives, and adverbs.The following analysis shows how Acton's fine-tuning and creative attention to his lexis render the textual universe of his Chinese texts differently.

Replacing verbs
Most of the verbal modifications arise from considerations of the text's performability and the attempt to increase the impact of the language on the audience.This feature of theatrical discourse has been termed -gestic subtext‖, -the union of spoken text and the gestures accompanying its [rhythmic] enunciation‖ (Pavis, 1989, p. 36).Acton's revisions show how verbs can be employed to recreate a new -language-body‖ (Pavis, 1989, p. 36) in the mise en scè ne of the translated text.
(1) 柳氏 动也不许动！ Liu Shih.You are not to move<stir an inch> from where you are.(SHC, D 2 , p. 47) 6   [Moving from your position is not allowed.] 7 Here, generic -move‖ is transformed into a dynamic image that allows the audience to imagine the character's stirring extended into the theatrical space, albeit only -an inch‖.
Past participle verb forms can also serve an adjectival function to evoke feelings and emotions that invite audiences into the internal world of characters and share their feelings.Acton's attention to such details in his revisions can be seen, for instance, in the following example: (2) ……红尘中，误了俺五陵年少。 A valiant hero in the pride of youth My life is wrecked, and all my hopes are blasted <is> <destroyed> / <defeated>.(LCYP, D 2 , p. 7) [My promising youth is ruined in the vicissitudes of life.]The translator vacillates between -blasted‖, -destroyed‖, and -defeated‖, which have nuanced differences of meaning.Lin Chong's rebellion is not a sudden event or -blast‖, but rather a gradual process.As a respected instructor of 800,000 imperial soldiers, he bears the humiliation of his wife being molested by the son of his commander.He holds no grudge for it because he still has great dreams that otherwise may be ruined.Yet, he is unaware that he has been led into a trap set up to eradicate him by a villain who desires his beautiful wife.Eventually Lin realizes that he has been snared in an arson attack that may lead to his death and he becomes more and more resolute towards his final rebellion.Establishing a link with the sudden arson attack and its aftermath probably tempted Acton into translating Lin's hopes as -blasted‖ and then -destroyed‖, yet the translator's final choice, -defeated‖, is not only closer to the original Chinese word -误‖ (-to ruin‖), it also connotes how Lin has to abandon hope eventually after many attempts to compromise.

Replacing nouns
Acton's revisions across his manuscripts led to 41 cases of noun replacement.One of the puzzles Acton had to solve involves culturally specific terms dashed with local colour.Owing to his unfamiliarity with the source culture or his failure to find suitable English expressions, the translator sometimes struggles to find English terms for specific Chinese concepts or objects.Yet a genetic approach to translation can show the translator testing different solutions to the common translation problem of the target language lacking a lexical term of appropriate specificity.Example 3 serves as a good illustration: [Indeed there will be a 妓 (Ji).]These insulting remarks are made against a girl named Chin Ts'ao 琴操 by the cantankerous wife, Liu Shih.Acton's repeated corrections to his translation of -妓‖ (Ji) indicate how this term is the nexus of a cultural mismatch.The profession of the Chinese "妓" is different from that of a prostitute 娼.Like the Japanese geisha, these girls are artisans trained to provide an array of services from singing, dancing, chanting of poems and painting to entertaining their guests, which may not involve sex.The closest term -courtesan‖ (SHC, p. 25) also proves to be a red herring in that these young girls have no associations with a court of monarchs or powerful benefactors as in the West.Corrections on the page suggest that Acton hesitates between -loose woman‖, -prostitute‖, -trollop‖, and -harlot‖, all variations of derogatory terms that show Liu's contempt and disdain.The evidence of the translator's revisions demonstrates his attempt to find a way of expressing a concept for which no English term exists.This evidence not only defends him against the charge that the term he finally chooses is naive or lacking comprehension; it also illustrates the specificity of translation for the theatre.Were the scene not evident to the audience, with characters located in the scenes described, the translator may have needed to provide further explication, for instance, by calling the girl a -poor-man's courtesan and strumpet‖.The genetic evidence shows, on the other hand, a translator striving for theatrical economy and impact by using the more concise and harsher terms, -trollop‖ and -harlot‖.

Replacing adjectives and adverbs
Using the genetic materials to observe Acton revising, one becomes aware of his attention to the dynamism of his adjectives and their potential to awaken the senses of the audience.For instance, (4) [… 原来那边还有一座大花园，]桃红柳绿，好耍子呢。 Truly it"s a gorgeous sight with all the green <shooting>willows and peaches in blossom.(CHNH, D 2 , p. 17) [(Outside there is a big garden.)Peach blossoms are red and willows green.It is a lot of fun to play there.] Here, Acton's revising lets the audience feel more palpably the thriving spring garden, the insipid -green willows‖ in D 1 giving way to a garden brought to life with the energetic image of -shooting willows‖.
The translator's revisions to his adverbs show him enhancing the kinesic potential of his language, its ability to communicate performative gestures.
( In this frolicsome, hilarious scene, Liu Shih is determined to let her husband have a taste of the rod to remind him of the possible dire consequences of consorting with -loose women‖.Not having the courage to defy his wife yet unwilling to give in before such humiliating discipline, Ch'ê n attempts to delay the spanking to the last moment by walking -gingerly‖ to his venue of destination.The stage direction -slowly‖ is thus revised for the more specific -gingerly‖, which invites the actor to mimic the hen-pecked Ch'ên's timidity as he approaches his termagant wife.

Syntactic adjustments
The examples above are cases of paradigmatic replacement in the translator's revisions.Yet he also modifies syntax and replaces entire sentences to render his scene more effectively, adapting his language to the linguistic, social, and aesthetic norms of English and translation for the theatre.With her teasing remarks, Liu Shih is urging her husband to borrow the bamboo stick from the neighbour so that she can warn him with a vigorous beating not to associate with any -loose women‖.Ch'ê n, however, not surprisingly, finds complying with such an instruction embarrassing.In stark contrast with the rather flat language of D 1 -Be more brisk about it‖ and -How slow you are‖, the pointed tone of the idioms in the translation revision, -Look sharp!‖ and -slowcoach‖, adds an aggressive dimension to Liu Shih's language, animating the action on the stage.

Being concise
The intrinsic time constraints in theatre performance and a preference in Chinese theatre for sharp dialogues require concise and economical sentences.In Example 7 the revising translator is shown to chop away superfluous and inconsequential words to this end.The intended other party of the conservation -with him‖ is quite obvious from the context, thus further specification is unnecessary.In the process of translation such redundancy can easily creep into the text.While theatrical imperatives may have compelled Acton to tighten up language, it is interesting to see that only at this relatively late stage in the translation's genesis did the translator realize the redundancy and take steps to eliminate it.

Improving textual coherence
The textual-linguistic norm of coherence 8 is also prioritized at the latter stage of this translation, as is revealed in Example 8: The revision of the expression in D 1 , -Very well, since you propose it yourself‖, to -Since the proposal comes from you, that settles it‖ sustains the idiomatic quality one would expect from theatrical dialogue while enhancing the organisation of the scene's coherence.The addition of -that settles it‖ at the end of the sentence offers a counterpoint to the coming sentence --But we have no rod of correction in the house‖within this internal form of argumentation.The conjunction is thus moved to the front of the next sentence, -However, that's easily solved‖, to facilitate the logical transition in the character's speech.

Aesthetic considerations
Acton states he has found more -genuine poetry‖ in Lin Chong Ye Pen than in many of -the hallowed products of the laureates‖ (Acton, 1939b, p. 182).Aesthetic beauty and experience in Chinese theatre are codified through culturally specific expressions of natural or emotional states or the -soul‖ of a character, and are generated through recognized artistic images and poetic rhythms (Hu, 1997, p. 25).In his revision, Acton uses the resources of English to heighten such aesthetic experience.Consider Example 9: The Chinese source contains the onomatopoeia -喇喇‖ (lǎ-la), -声声‖ (shēng-sheng), -哀哀‖ (āi-ai) to connect the disheartening sound of wind and the wild animals.Responding to Acton's use of poetic, especially Romantic conventions in English, his audience will feel the environment to be charged with Lin's fear and panic, for the translator has mingled a personified, menacing nature with the behaviour of the scene's animalsthe total effect generated through such intensified verbal states as -whistling‖, -echoes‖, and -howl‖.The rhythm of impending doom is beaten with the alliterated -d‖ in words of negativity: -assailed‖, dismally‖, -doubts‖, -dire‖.The tiger's -roar upon roar‖ is connected to nature when it is revised with the simile, -like thunder-peals‖.The concert of nature, animals, and humans is reinforced by the heavy assonance connecting their vowel sounds, such as those that connect the -apes‖ with the -assailed‖ subject whose courage -fails‖, and off-rhymes which suggest that distinct elements are actually variations of each other, such as when the -thunder-peals‖ echo in the -stream‖ and the -I‖ who will -follow‖ echoes in the apes' -howl‖.
In the following Example 10, Acton's revisions show him gesturing towards the parallel rhyming pattern of his Chinese source.The revised version allows -fall‖ and -wall‖ to imitate the Chinese rhyming pattern, 窗 (chuāng) and 墙 (qiá ng).Here Acton has privileged the sound of his English over his more imagistic earlier version, sacrificing its suggestive atmosphere of a Chinese painting in which -the shadow of the moon pierces the gauze window‖ and the -plum-blossoms girdle the white washed wall‖ in D 1 .

Revising interpersonal dimensions
Speech acts are performative if their articulation doesn't just communicate a message but changes some aspect of the (social) world in which they are uttered, for example by asking a question, issuing a command, apologizing, promising (Austin, 1962, pp. 98-100).Stylistic variations in English signal different levels of interpersonal communication within such speech acts (Takahashi, 2012, p. 88), which is observable in Acton's revisions in the following Example 11.The casual -Greetings, Teacher‖ of D 1 gives the impression that Miss Tu is overly familiar with her teacher.Articulated within a staged traditional Chinese context, the expression implies too much inappropriate intimacy between student and teacher.In contrast, the more elaborate wording and the subject's deference in the revision communicates greater reverence towards the teacher.
From data in sections 4.1 and 4.2 we can extrapolate that when he revises his D 1 typescript, Acton is on the whole not correcting inaccuracies or errors that a close comparison with the source text might reveal.Rather, he tends to revise his text to make it more reader-oriented through introducing more idiomatic and creative expressions.This supports the hypothesis that at the stage of this late revision in the work's genesis the translator is behaving more like a reviser of translations, a task that is evidently both generic to the process of translation and often outsourced to another (usually more senior) translator.As a reviser of his own translation, Acton is fine-tuning his text, and quite possibly with little reference to the source text.If this is indeed the case, unlike a professional translation reviser who must keep a vigilant eye on the source text when checking the translation, Acton is revising more like a creative writer-sharpening his language, cutting extraneous or redundant material-except that as a translator he knows that while he is reshaping his own expression he is ethically bound not to modify significantly the content of his text.

The case study of a sample page
In this section, in order to offer a more holistic view of how Acton revised his avant-texte, we look at a sample page of the English manuscript from Shih Hou Chi.It is a passage sung by Fo Yin, a monk and friend of Ch'ên, which expresses Fo Yin's view on life and describes the journey of his visit to Su Tung-p'o's mansion.
Figure 1 Sample manuscript from Shih Hou Chi (Acton, 1933(Acton, -1941, p. 26) , p. 26) Fo Yin (enters and sings in man chiang hung or ‗the crimson stained river').○ 1 The world of desire is like a mirage.○ 2 Those who are immersed in it are always entertaining <harbouring> lofty hopes.○ 3 But when shall these hopes be fulfilled?○ 4 Men's eyes are always <ever> straining towards official promotion; they are always <continually> building castles high up in the clouds.○ 5 But alas, the affairs of this world are <as> evanescent, like <as> a dewdrop on a flower-petal, and the life of man i<a>s a candle before the wind.
○ 6 Consider <How often>, in the course of a year, how often can the moon be seen straight <directly> over head?○ 7 How often are the wine cups overturned?* ○ 8 (Speaks) I am Fo Yin,<.> and h<H>aving just received an invitation from Su Tung-p'o, I am going there directly<on my way to his house>.○ 9 (Walking around the stage): I walk beside <Walking by> the green waters and <pool's margin and> cross<ing> the red <painted> bridge.<,>○ 10 I have reached <come to> the door of the Su<his> mansion.○ 11 (Pauses) Here I am at his abode.○ 12 I'll knock at the door.* That is, to show that they have been drained.
(SHC, D 2 , p. 26, numbers introduced by the author) This corrected typescript may be compared with its published version.
Fo Yin (enters and sings in Man Chiang Hung or "The crimson stained river").The world of desire is like a mirage.Those immersed in it are always harbouring lofty hopes.But when shall these hopes be fulfilled?Men are ever straining towards official promotions; they are always building castles in the clouds.But the affairs of this world are as evanescent as a dewdrop on a flowerpetal, and the life of man as a candle before the wind.How often, in the course of a year, can the moon be seen directly overhead?How often are the wine-cups overturned?(SHC, published version, Acton, 1939c, p. 102, my emphasis) Here one witnesses Acton fine-tuning his text for fluidity and cohesion.For instance, at ○ 1 -who are‖ is cut, its syntax remaining implicit in later revisions.Those -entertaining lofty hopes‖ at ○ 2 are subsequently -harbouring‖ them, which not only renders them more intimate and concealed but also more coherent with Fo Yin's remark, -The world of desire is like a mirage.‖In sentence ○ 4 , -Men's eyes are always straining …‖ was a literal translation of -人眼总盯着官位升迁‖.This perfectly natural Chinese expression is perhaps a little awkward in English, though its revision introduces a lofty, poetic register: -Men are ever straining […].‖Also in ○ 4 of D 1 , the adverb in -always building castles high up in the air‖ is replaced with -continually‖ in D 2 , though in the final published version -always‖ is restored, probably to avoid the connotation of a never-ceasing, continuous effort, which would have been misleading.Sentence ○ 5 is a good case of rewriting for the purpose of coherence.It uses the -as … as‖ structure to join two discrete sentences into one comparative sentence, rendering the logic apparent.Sentence ○ 6 rejects the redundant -consider‖, and the moon is more appropriately -directly‖ rather than -straight‖ overhead.In sentence ○ 8 of D 1 , Fo Yin first introduces himself and then recounts his invitation to the house of Su Tung-p'o, to which he is setting off.Acton's revision separates these two dramatic elements into separate sentences, thus aiding the audience's understanding and heightening their dramatic effect, for Fo Yin may punctuate his introduction on stage before beginning his story.In addition, -On my way to his house‖ of D 2 is further shortened to -on my way‖ in the published version, the context allowing for this more economical expression.Yet the location is reinstated later when -going there directly‖ is revised to the more colloquial -on my way to his house‖ .Sentences ○ 9 ○ 10 ○ 11 in D 1 are succinctly reduced into one sentence in later revisions through the use of the gerund, which allows the action to unfold before the reader and audience.In ○ 9 , -red bridge‖ is changed into -painted bridge‖ because the Chinese 雕桥 (painted as decoration) is not necessarily red.In ○ 10 , the present perfect in -I have reached the door‖ is replaced with a simple present, -I come to the door‖, consistent with the revision for present-tense action.-Su [Tung-p'o's] mansion‖ is changed into -his mansion‖ in D 2 with a pronominal anaphora function to bridge with the previous message and thus aid with fluidity and coherence.The redundant message from sentence ○ 11 -Here I am at his abode‖ in D 1 is expediently suppressed in D 2 , for it can be deduced from the context.Thus, in the late phases of his translation, the translator is observed striving above all to increase the overall effectiveness of the passage.Except for the verb modification in the shift from -entertaining‖ to -harbouring‖ and the noun modification from -red bridge‖ to -painted bridge‖ below sentence level, most revisions occur at the syntactic level and are motivated by either (1) linguistic norms of textual coherence or (2) theatrical concerns for conciseness and comprehension.Redundancies are The pursuit of beauty by an aesthete 123 trimmed and cohesive devices are added.Inconsistencies are corrected and the whole passage reads more smoothly.

Conclusion
The above analysis demonstrates that while revising his translation Acton was more concerned with -how to say‖ on stage rather than -what to write‖ on the page.From his published comments and the revised typescript, one may surmise that his translation process involved becoming familiar with the original text by reading and visiting the theatre.He then translated the text with his teacher, Mr Chou I-min 周逸 民, and revised the translation with his two friends, Mr Yen Yü-heng 颜 毓蘅 and Mr Yang Shan-ch'uan 杨善荃.Acton (1933Acton ( -1941) also acknowledged his learned friends, Dr H. H. Hu 胡先骕 and Mr Chang En-yü 张恩裕, -for valuable information incorporated in the notes and introductions‖ (pp.2-3).Their contributions may have influenced the translations.The revised typescript attests to a phase of revision conducted at a developed stage of the translation's genesis.Here, Acton enters a phase of rewriting that involves editing for textual coherence.Although he does make paradigmatic replacements and revises for exactitude and aesthetic subtlety, more emphasis is placed on trimming and restructuring sentences to make them more harmonious and, most of all, appropriate for delivery on stage.At this late stage in the genesis, the translator appears, therefore, less focused on the source text than on the reader of the target text as well as the actors who will deliver the lines and the audience for whom they are intended.Like a playwright, he is considering the textual-linguistic, social and aesthetic norms of theatre, and as a translator he is testing ways to present this very foreign text to a contemporary audience.
The present genetic study has shown how the translator inhabits different roles at different points in the process of producing a translated text.It elucidates the fact that the contested status of the translator as handmaid to the source text, as its rewriter with his or her own -agency‖, and even as the -author‖ of the translation, each indicate positions that the same translator may occupy during different stages of the translation's genesis.Therefore, if studies of translations examine only the published version of a text, they must be wary of making any general claims with respect to a translator's imputed position or strategy (e.g.foreignizing/domesticating, source-oriented/target-oriented), for not only do such claims have ideological implications that may be contentious or misleading, they also tend to ignore the need for translators to use other strategies at different moments of their composition.TTR refers to type/token ratio, the number of types (the number of unique word forms instead of the total number of words in a text) divided by the number of tokens (the number of individual words in the text).It tells how rich or -lexically varied‖ the vocabulary in the text is.The higher the resulting value, the greater the number of different words used in a text.However, such a ratio is not totally reliable.The problem of TTR is that as the text gets bigger, the number of new word types being counted falls.To remedy this shortcoming, standardized TTR calculates TTR based on every 1,000 words and produces an average TTR, which is a more representative figure.
5 This is the threshold level for prose that is -easy to read‖, as defined by advocates of -plain English‖ (Sanyal, Cutts, & Sen, 2006, p. 49).Prose sentences of this length are less taxing for readers and their brevity helps -make every word tell‖ (Strunk & White, 1918/2007, p. 39).The fact that Acton revises his theatrical discourse for brevity suggests he is increasing the impact of its delivery.The English translations in square brackets are my own literal renditions, though I sometimes adjusted them to allow for a clearer understanding of Acton's translation.To avoid further complication, I did not use the Pinyin system to transcribe the Chinese characters.
8 Blum-Kulka (1986) defines coherence as -a covert potential meaning relationship among parts of a text, made overt by the reader or listener through processes of interpretation‖, and cohesion as -an overt relationship holding between parts of the text, expressed by language specific markers‖ (p.17).Cohesive devices deployed in the target language may create shifts in levels of explicitness.

(
Speaks) I am Fo Yin.Having just received an invitation from Su Tung-p'o, I am on my way.(Walking around the stage.)Walking beside the green waters and crossing the painted bridge to his house I reach the door of his mansion.(Pauses) I'll knock at the door.
Poyatos, F. (2008).Textual translation and live translation: The total experience of nonverbal communication in literature, theatre and cinema.Amsterdam: John Benjamins.Sanyal, J.,Cutts, M., & Sarbjit, S. (2006).Indlish: The book for everyEnglishspeaking Indian.New Delhi: Viva.Scott, C. (2006).Translating the literary: Genetic criticism text theory and poetry.In S.Bassnett & P. Bush (Eds.),The translator as a writer (pp.106-117).London: Continuum.Strunk, W., & White, E. B. (2007).The elements of style.New York: Penguin.(Original work published 1918) Takahashi, H. (2012).A cognitive linguistic analysis of the English imperative: With special reference to Japanese imperatives.Amsterdam: John Benjamins.Toury, G. (1995/2012).Descriptive translation studies and beyond.Amsterdam: John Benjamins.Valé ry, P. (1992).Variations on the eclogues.In J. Biguenet & R. Schulte (Eds.),Theories of translation: An anthology of essays from Dryden to Derrida (pp.113-126).Chicago, IL: University of Chicago Press._____________________________ 1 I wish to thank Anthony Cordingley for his substantial revisions to this article and for the many ideas which he contributed to it.This paper is part of a research programme funded by: the Science Foundation of Beijing Language and Culture University (-Fundamental Research Funds for the Central Universities‖) [Nos.15YBB20 & 15YJ020022]; the Research Fund for Beijing Social Science [No.15WYB029]; and the Research Fund for Young Scholars of Social Science, Ministry of Education, China [No.11YJC 740030]. 2 The collection includes the mostly unpublished manuscript and notes and correspondence concerning its projected publication.Beinecke's -Guide to the Harold Acton Papers‖ states the provenance of the archive: Purchased from Simon Finch on the Edwin J. Beinecke Book Fund, 2009.This archive was given by Harold Acton to his only godson, the Hon.Martin Parsons (1938-), younger son of Michael Parsons, sixth Earl of Rosse (1906-1979).The gift was made in memory of the Hon.Desmond Parsons, younger brother of Michael, who died prematurely in 1937 at the early age of 26.[...] The bulk of the material-translations, writings, correspondence-was created and accumulated by Acton during his residence in China and in the course of his literary work there as a translator and writer.Upon Acton's departure from China, the material was packed into a trunk and shipped to London.(n.d.) 3 See Figure 1 for illustration of D1, D2 and D3.In this case, the D3 in pencil in the left margin is indiscernible. 4

6
Unless otherwise indicated, revision examples are from D2. 7

Table 1 .
Statistics of wordlist