Recreating the Image of Chan Master Huineng: An Interpersonal Approach to Translations of the Platform Sutra

The study is a comparison of four English translations of a Chinese Chan Buddhist text, the Platform Sutra. It investigates translators’ choices on the level of lexico-grammar, the semantic consequences of such choices as the image of Huineng and the underlying contextual factors contributing to the variations among translations. Interpersonal system of SysFan, a computational tool specialized in systemic functional analysis, is used and choices in terms of SUBJECT PERSON, MODALITY and MOOD TYPE are analyzed and compared. It is demonstrated that in translation, lexico-grammatical choices will lead to variation in meanings, and there tend to be contextual motivations for the seemingly unconscious choices.


Introduction
Translating, according to Levý, is a process of making choices (2012: 72), and it is true that two translations of the same source text will not be identical in their choices of words and/or syntactic structures.Then the questions will be: in which way do translations of the same source text differ in their lexico-grammatical choices?What kind of semantic consequences, that is, meanings, can these linguistic choices lead to?And finally, why are different choices made and different meanings realized?
The three levels involved in the above questions: lexico-grammar, semantics and context, are closely related to the nature of language as a stratified semiotic system.According to systemic functional linguistics (SFL), there are different strata within language, which include phonetics, phonology, lexico-grammar and semantics in a bottom-up sequence, and there is also an extra-linguistic level of context which is closely related to the use of language.These different strata are penetrated by the three metafunctions of language: ideational (to construe human experience), interpersonal (to enact personal and social relations with others) and textual (to construct texts).
Focusing on the interpersonal metafunction, the present study is a comparison of four English translations (Wong 1930a;Heng 1977;Cleary 1998;Cheng 2011) of a 13th-century Chan Buddhist canon in China, the Platform Sutra.It will investigate different translators' choices of SUBJECT PERSON, MODALITY and MOOD TYPE on the level of lexico-grammar, the semantic consequences of these lexico-grammatical choices as the image of Huineng, and the influence of the contextual factor, Tenor, on the translators' choices.The scope of this study is illustrated in Figure 1.Huineng (638-713) is a great Chan master in the Tang dynasty of China.He is venerated as the founder of Chinese Chan, Japanese Zen, Korean Sŏn and Vietnamese Thiên (Jorgensen 2005: 1).The source text, the Platform Sutra, is a collection of the public sermons and personal conversations of Huineng and 'one of the best known, most beloved and most widely read of all Chan texts' (Schlutter 2007: 382).

Data
Due to its important status, fascinating ideas and plain language, the Platform Sutra is 'extremely popular in Chan and Zen communities around the world' (McRae 2003: 67) through translation.Up to now, it has more than 15 English translations, of which four are selected according to the criteria of heterogeneity in terms of translator's identity, publishing time and agency.Translations by two Chinese translators, Wong Mou-lam (1930) and Cheng Kuan (2011), and translations by two American translators, Heng Yin (1977) and Thomas Cleary (1998), are included.These translations are published at different times, spanning a time gap of nearly 80 years.Three out of the four translations are sponsored and published by Buddhist institutions, and the only exception, the translation by Thomas Cleary, is published by Shambhala Publications.It is hypothesized that this heterogeneity is likely to contribute to difference in translators' lexico-grammatical choices and semantic meanings of the texts.
Direct speeches of Huineng in five randomly chosen chapters, Chapter Two, Four, Five, Six and Seven are extracted from each translation as data of analysis.This is due to the fact that the analytical tool adopted in this study, SysFan, mainly depends on time-consuming manual analysis.However, representativeness of the data is guaranteed in that these five chapters comprise both Huineng's public sermons and private conversations.

Analytical framework
The translated texts are first divided into clauses and then imported into SysFan (Wu 2000), a computational tool specialized in systemic functional analysis.All the clauses (4,836 in total) are identified and only free clauses (3,348) are further analysed in terms of SUBJECT PERSON, MODALITY and MOOD TYPE in the interpersonal system of SysFan, which is presented in Figure 2.This is because only free (or major) clauses can serve as an exchange (Matthiessen 1995).After the analysis, result can be summarized automatically and specific choices of all clauses can be retrieved.SUBJECT PERSON, MODALITY and MOOD TYPE are selected as bases of comparison because they are the main lexico-grammatical strategies to realize interpersonal functions in English.It should also be noted that significant difference exists between the source and target languages in terms of SUBJECT PERSON, MODALITY and MOOD TYPE.Firstly, the source language has the tendency to leave clause subject implicit (Halliday and McDonald 2004;Wang 2002) and to omit personal pronouns (Lü 1999: 8).This may provide translators with relative freedom to make their own choices in terms of SUBJECT PERSON.Secondly, MODALITY in Chinese is lexicalized (Halliday and McDonald 2004: 339), while modal auxiliaries, which are part of the Finite of clauses (Halliday and Matthiessen 2004: 115-17), constitute a major realization of MODALITY in English.The fact that Chinese does not have corresponding finiteness (Matthiessen and Halliday 2009: 10) may lead to variation in terms of MODALITY in the four English translations.Thirdly, though Chinese and English both have three mood types, they differ from each other in that imperative clauses are more commonly used in Chinese as realization of commands and requests than in English without being considered impolite (Lee-Wong 1994; Gao 1999).

Choices of SUBJECT PERSON
In SFL, SUBJECT PERSON can be classified into two types: interactant, which refers to 'person within the dialogue', and non-interactant, which refers to 'person outside the dialogue' (Matthiessen 1995: 687).The interactant type includes speaker (I), addressee (you) and speaker-plus (we).The non-interactant type includes all the third person references.
Choices of SUBJECT PERSON in each translation are presented in Table 1.As can be seen from Table 1, the interactant type of SUBJECT PERSON has similar frequencies in all the four translations.Within the interactant type, however, a significant feature of using high frequency of speaker-plus (we) in relation to speaker (I) and addressee (you) is demonstrated in Wong's translation.An example of this is presented in the following.

ST:
善知识 既忏悔已 与善知识发四弘誓愿 (Chapter 6) Wong: Learned audience, having repented of our sins, we shall take the following four all-embracing vows.
Heng: Good Knowing Advisors, as you have repented and reformed, I will now teach you to make the four all-encompassing vows.
Cleary: Good friends, once you have repented, I will make the four universal vows for you.

Cheng:
Good Mentors, after having repented, you and I together will generate the Four Grand Vows.
In the Platform Sutra, Huineng is either giving public teachings to a large audience or answering questions from his students.There exists an inherent inequality in terms of social status between Huineng and his listeners.The use of 'you' to refer to the listener, in combination with the use of 'I' as self-reference, may create an 'Other' and a kind of distance between Huineng and his audience (Hyland 2001;Pennycook 1994).The use of inclusive 'we' in Wong's translation, on the contrary, can construct a 'chummy' and 'intimate' tone (Wales 1996) and help Huineng establish solidarity with his listeners.

Choices of MODALITY
Whenever one speaks, one is also expressing attitudes, such as certain or uncertain, to the information being conveyed (Coates 1990: 55).These attitudes are expressed in the form of MODALITY, which covers the semantic space between 'yes' and 'no' and is 'a form of participation by the speaker in the speech event' (Halliday 1970: 335).In English, MODALITY can be realized in the form of modal auxiliaries (such as can/could/may/might), modal adverbs (such as possibly/certainly), or separate clauses (such as I think/ it is possible that).
Percentages of clauses containing MODALITY out of total free clauses (indicative as well as imperative clauses) in each translation are presented in Table 2.It can be seen that Cheng's translation has the highest percentage of clauses containing modality (43.4%) among all the translations.This is followed by Wong (23.9%).In contrast, translations by Heng and Cleary have far fewer modalized clauses.This indicates that in comparison with Heng and Cleary, Cheng and Wong are more likely to qualify clauses through modality.Further investigation into variation in terms of types and values of MODALITY is presented in the following sections.

Types of MODALITY
MODALITY can be categorized according to the speech function of the clause as a proposition or proposal.MODALITY in propositions, termed modalization, is about how probable or frequent the information is valid.MODALITY in proposals, termed modulation, is about the obligation and inclination of the participants.Choices of the two types of MODALITY in each translation are presented in Table 3.  Wong: Mere reciting it without mental practice may be likened to a phantasm, a magical delusion, a flash of lightning or a dewdrop.On the other hand, if we do both, then our mind would be in accord with what we repeat orally. Cheng: If it is merely muttered in the mouth without Mental Implementations, it would be like Phantasm or Metamorphosis, or like dew drops and electricity.Chanting verbally and implementing mentally at the same time could make both the mouth and the Mind congruently corresponding. Heng: When the mouth recites and the mind does not practice, it is like an illusion, a transformation, dew drops, or lightning.However, when the mouth recites and the mind practices, then mind and mouth are in mutual accord.

Cleary:
Verbal repetition without mental application is illusory and evanescent.When it is both spoken of and mentally applied, then mind and speech correspond.
Although both modalization and modulation are expressions of the speaker's involvement in the exchange (Martin and White 2005), the two differ in their functions.Modalization is used by the speaker to provide information, while modulation is used by the speaker to realize proposals (such as to ask the listener to do something).Therefore, it can be said that Huineng presented in the translations of Wong and Cheng tends to inform the audience on the ideas of Buddhism, and Huineng presented by Heng and Cleary tends to make a stronger impact on the audience by indicating their obligation and inclination to do something.

Values of MODALITY
Apart from being classified into two types of modalization and modulation, MODALITY can also be graded into high, median and low values according to the strength of assessment.Choices of different values of MODALITY in each translation can be seen in Table 4. Table 4 shows that while median-valued modality is dominant in all translations, high-valued modality is used differently.Modal expressions with high value are used far less frequently by Wong (6.0%) and Cheng (7.2%) than by Heng (17.3%) and Cleary (15.1%).This can be seen from the following example, where median-valued modal auxiliary 'should' is adopted by Wong and Cheng, but highvalued modal auxiliaries 'must' and 'need to' are used by Heng and Cleary.

ST:
欲拟化他人 自须有方便 (Chapter 2) Wong: Those who intend to be the teachers of others should themselves be skilled in the various expedients which lead others to enlightenment. Cheng: With a view to edifying other people, you should be equipped with Expedient Dexterities. Heng: If you hope and intend to transform others, you must perfect expedient means.

Cleary:
If you want to try to teach other people, you need to have expedient methods yourself.
Interpersonally, the higher the value of obligation, the more likely the speaker is expecting the listener to respond (Croft 1994: 469).The choice of high-valued over median-valued modulation in Heng's and Cleary's translations helps to construct the image of Huineng as more authoritative and powerful.

Choices of MOOD TYPE
According to SFL, there are two mood types in English, indicative and imperative.While indicative clauses are typically used to exchange information, imperative clauses are the congruent way of realizing commands (Matthiessen 1995: 438-44).Choices of MOOD TYPE in each translation are presented in Table 5.The table shows that more imperative clauses are used by Heng (13.3%) and Cleary (10.1%) than by Wong and Cheng (7.6% and 4.0 respectively).Imperative clauses are considered the most direct and bald way to give commands and are potentially face-threatening (Brown and Levinson 1987: 69) in English.When used in conversation, imperative clauses may serve as an indication of the powerful status of the speaker over the listener, especially in comparison with another way of realizing commands---indicative clauses containing modulation, such as 'You should…' (Eggins and Slade 1997: 88).
It can also be seen from the table that Wong is the only one who uses a particular kind of imperative, suggestive (Let's/Let us).Different from a jussive imperative (Do it/Don't do it), where the listener 'you' is the only one responsible to carry out the action, a suggestive clause assigns the proposal to both the listener and the speaker and can be seen as intermediate between an offer and a command (Matthiessen 1995: 423, 25).In this way, the command is softened and the image of Huineng also becomes more polite.The following example can serve as an illustration.

ST:
善知识 于念念中自见本性清净 自修自行 自成佛道 (Chapter 6) Wong: Learned Audience, let us realise this for ourselves at all times (literally from one Ksana, momentary sensation, to another).Let us train ourselves, practice it by ourselves, and attain Buddhahood by our own effort.
Heng: Good Knowing Advisors, in every thought, see your own clear and pure original nature.Cultivate, practice, realize the Buddha Way! Cleary: Good friends, see for yourself the purity of original essential nature in every moment of thought, cultivating yourself, practicing yourself, attaining buddhahood yourself.

Cheng:
Good Mentors, you should endeavour to perceive your own Original Nature in its ever-pure state from instant to instant; in such a way of practice you would be carrying out the Self-cultivation, Selfimplementation, and Self-realization on Buddha Bodhi In this example, an indicative clause with the modulation 'should' is used by Cheng, a suggestive 'let us' is used by Wong, while imperative clauses are used by Heng and Cleary.This difference in the choice of mood type will surely have an impact on the presentation of Huineng as either a polite or an authoritative Chan master.

Summary
In the previous sections, choices in terms of SUBJECT PERSON, MODALITY and MOOD TYPE in each translation are investigated.Results are summarized in Table 6 with distinctive features of each translation highlighted.

Table 6. Summary of lexico-grammatical choices in each translation
In terms of SUBJECT PERSON, Wong uses inclusive 'we' as the dominant interactant type, which is in contrast to the juxtaposition of speaker 'I' and addressee 'you' in the other three translations.In terms of MODALITY, Cheng uses the highest percentage of MODALITY, particularly median-and low-valued modalization, in comparison with others.In terms of MOOD TYPE, Cheng uses the lowest percentage of imperative clauses.Interpersonally, inclusive 'we' can help the speaker (Huineng) achieve solidarity with his audience, and median-and low-valued modalization is to present the information as less assertive and open to discussion.Avoidance of using imperatives can also be seen as a sign of being polite on the part of the speaker.

Contextual interpretation
The inseparability of text from its context is emphasized in both SFL and translation studies.In SFL, it is recognized that there exists a dynamic relation between text and context, in that context is realized in text and text reveals context (Butt et al. 2006: 182).In translation studies, the importance of relating a translated text to its context is also acknowledged by many scholars.Lefevere and Bassnett, for example, point out that 'There is always a context in which the translation takes place, always a history from which a text emerges and into which a text is transposed ' (1990: 11).Influence of context becomes more evident in comparing different translations of the same source text, where the linguistic constraint remains the same.
There are two kinds of context identified in SFL, context of culture and context of situation (Halliday and Hasan 1989).The latter, context of situation, provides an immediate environment for the text and can be described in three dimensions: Field (what is to be talked or written about), Tenor (the relationship between the speaker/writer and listener/reader) and Mode (the kind of text that is being made).As tenor is within the interpersonal system of language and exerts influence on choices of SUBJECT PERSON, MODALITY and MOOD TYPE, which constitute the focus of the present study, this section will mainly discuss Tenor, that is, the relationship between translators and their target readers.
Both Wong Mou-lam and Cheng Kuan are Chinese translators, with difference in their religious identities: Wong is a layperson and Cheng is a Buddhist.Wong was the first person to translate the Platform Sutra into English in history.His translation was undertaken at the request of the founders of the Pure Karma Buddhist Association, a society organized by rich businessmen and reformers in Shanghai.Literature shows that Wong did not possess a high social status as the translator of the Platform Sutra.He was 'discovered' in a law firm by a founder of the Buddhist association, as he was both good at English and interested in Buddhism.Wong was asked to quit the job and serve as a translator for the Buddhist association (Welch 1968: 180).When translating the Platform Sutra, Wong actually stayed at the home of Dih Ping Tsze, one of the founders of the association, for one and half year (Ko 1996: 9-10), and might have been largely dependent on Dih financially.Moreover, the fact that Wong translated the book in the 1930s, a period when China was facing both the invasion of the foreigners and civil turbulence, could have further contributed to the inequality of status between the translator and his target readers.In his preface to the translation, Wong keeps apologizing to the readers for his incompetence in providing a good translation (Wong 1930b).This explains why Wong is the only translator who uses so many inclusive 'we' and suggestive 'let us' in his translation.Since the translator was unconfident in his own identity as an ordinary man, a layperson and a Chinese, in front of his target readers, he chose to present Huineng as considerate and polite.
Born in Taiwan, Cheng Kuan studied English in college and was ordained in 1988.Later Cheng became a Buddhist master himself and the abbot of two temples (one in Taiwan and the other in the U.S.).He is also the person in charge of a publishing institution which published his translation.All these, as one can see, give him a relatively high social status, which can explain why also a Chinese translator, Cheng does not use 'we' to refer to the audience in Huineng's speech.On the other hand, however, Cheng's identity as a Chinese Buddhist master trying to spread Buddhist ideas to English readers and the fact that he does not have many native western disciples (Low 2010) may explain why modalization is used to such a high frequency in his translation.Modalization, being able to present Buddhist ideas in an unassertive and less threatening way, is utilized by Cheng to introduce this traditional Eastern religion to his western readers and enhance the possibility of acceptance.
In contrast, both Heng Yin and Thomas Cleary are American translators with different religious identities: Heng Yin was a Bhikssuni when she translated the Platform Sutra and Thomas Cleary is a professional translator of Eastern religion and philosophy.The reason why their translations are so similar may lie in the fact that both of them translated the book for target readers of similar interests.Heng's translation is mainly targeted at Western Buddhist practitioners as the purpose of the translation is to help its readers to 'realize Bodhi and accomplish the Buddha way' (Hsuan 1977: xvi).Cleary's translation, on the other hand, is mainly targeted at ordinary readers.This can be seen from the description of Huineng as 'perhaps the most respected and beloved figure in Zen Buddhism' and the introduction of Cleary as 'holds a doctorate in Eastern Asian language and civilizations from Harvard University' on the blurb of the book.Therefore, translating for readers with similar cultural background, interest and religious identity, Heng and Cleary may not think it necessary to shorten the distance between Huineng and his audience by using inclusive 'we', qualify the propositions with modalization, or change the original imperative clauses into modulated indicative clauses in giving commands.

Conclusion
Applying SFL to the comparison of four English translations of a Chan Buddhist text, the study investigates translators' choices of SUBJECT PERSON, MODALITY and MOOD TYPE on the level of lexico-grammar, the semantic consequences of such choices as the image of Huineng and the underlying contextual factors contributing to the differences among the translations.Chocies of inclusive 'we' in terms of SUBJECT PERSON and suggestive impeartive 'let us' in terms of MOOD TYPE by Wong help recreat the image of Huineng as a friendly Chan master, which can be seen as the result of the inequality of status between the translator and his target readers in a specific historical context.High frequency of modalization by Cheng results in the presentation of Huineng as an unassertive Chan master and is itself the result of Cheng's effort to preach Buddhist ideas in the modern American society.In contrast, choice of speaker 'I' and 'you' as interactive subject person, use of many modulation as well as modalization, and imperative clauses to realize commands by Heng and Cleary serve to recreate the image of Huinen as authoritative and powerful.Contextully this is because the translations are targeted at readers with similar cultural and religious background and thus no attempt has been made to alter the original ways of expression.Therefore, it is shown that in translation, lexocigrammatical choices would lead to variation in meanings, as meaning is choice (Halliday 2003), and there will always be contextual motivation for the seemingly unconscious choices.

Figure 1 .
Scope of the present study

Table 1 .
Choices of SUBJECT PERSON in each translation

Table 2 .
Modalized clauses in each translation

Table 3 .
Types of MODALITY in each translation

Table 3
shows that modalization is dominant in translations produced by Cheng (74.6%) and Wong (62.8%), while nearly same numbers of modalization and modulation are used by Heng and Cleary.The following example illustrates that in many cases, modalization is used by Wong and Cheng, but not by Heng and Cleary.

Table 4 .
Values of MODALITY in each translation

Table 5 .
Choices of MOOD TYPE in each translation