An Appraisal Analysis of the English Versions of “Qiang

Appraisal meanings, as a crucial feature of classical Chinese poetry, are often conveyed implicitly. Such resources can be misinterpreted due to translators’ subjectivity, and thus mistranslation occurred, which deserves greater attention. “Qiang Jin Jiu”, a classical Chinese poem written by Li Bai, covers rich appraisal resources reflected in multiple images and themes. This study attempts to apply the attitude and engagement systems of the Appraisal Framework developed by Martin and White to analyze the classical Chinese poem “Qiang Jin Jiu” and its three English versions. It aims to investigate the reasons for their appraisal differences and explore their translation quality, hoping to further enrich studies on the translation of classical Chinese poetry.


Introduction
Due to cultural and linguistic differences, translating classical Chinese poetry into English has always been a difficult task, particularly the translation of implicit meanings (Wang, 2008). A main feature of classical Chinese poetry is that appraisal meanings are often expressed indirectly via rhetorical devices such as personification, allusion and synesthesia. Such implicit appraisal meanings can be distorted in translation due to translators' subjectivity, thus causing mistranslation (Zhang, 2002). Greater attention should be attached to this phenomenon.
As a classical Chinese poem written by Li Bai, "Qiang Jin Jiu" represents "one of the best examples for Chinese lyric poetry" (Xiao et al., 1999) and has been appreciated by readers and scholars both at home and abroad for many years. This poem contains rich appraisal resources reflected in multiple images and themes to convey poet's praise for self-confidence and discontent over the ruling class. Although "Qiang Jin Jiu" has been translated into English by various famous scholars, few studies have analyzed the poem's English versions, especially from the perspective of appraisal meanings. In this regard, the present study, by adopting the attitude and engagement systems of the Appraisal Framework, attempts to analyze and compare the appraisal resources of "Qiang Jin Jiu" and three of its English versions. It aims to investigate the differences of appraisal meanings conveyed in three English versions respectively and possible reasons behind and to explore their translation quality.

The Application of Appraisal Framework in Translation Studies
Proposed by Martin and White (2005), the Appraisal Framework contains three interacting subsystems: attitude, engagement, and graduation. It has been developed for nearly 20 years and applied in different research areas, offering a new angle to study evaluative resources in discourse. Most studies centered on discourse analysis and foreign language teaching, demonstrating that the Appraisal Framework is feasible to analyze attitudes and voices in different discourse of various contents and genres.
It can also be applied to translation study, an important type of intercultural discourse. Studying translation with the help of Appraisal Framework is a relatively new research area, but its achievements are significant. Zhang (2002) found out that a translator would consciously or unconsciously adjust the appraisal resources in the source text based on his or her understandings and attitudes, leading to "unfaithfulness" in the translated text. Liu (2012) attempted to optimize translation processes so that appraisal resources in translation can be better analyzed. Three sub-systems of the Appraisal Framework should be taken into consideration to correctly translate appraisal meanings of the source text whether they are at the lexical, syntactic, textual, or semantic level. The Appraisal Framework can also help recognize translators' subjectivity and inter-subjectivity to better assess translation quality (Si, 2018). All studies mentioned above are general and macroscopic, which lack in-depth analysis of specific text. Research data are also insufficient.
With regards to further development, Munday (2018) compared the source text of the U.S. President Trump's inaugural address with its simultaneous interpretations and a written translation in Spanish based on Appraisal Framework, showing that translator's interpretation or sometimes intervention is of vital importance. The study proved that Appraisal Framework is useful to identify the power behind evaluative languages and how they can influence communication. Scholars have also adopted the Appraisal Framework to analyze both literary and non-literary texts and their translated versions. As to non-literary texts, researchers compared the attitudinal resources in news (Chen, 2007) and speeches (Munday, 2012, pp. 42−83;Chen, 2019) with their translations to explore both similarities and differences. Studies on literary texts such as novels (Xia & Li, 2009), autobiographical proses (Hao, 2013), and short stories (Yu, 2019) are increasing, especially comparative studies on poems and their English versions. In poem translation, the equivalence of source text and target text in appraisal meaning is crucial to evaluating translation quality (Yu & Qi, 2007). In addition, Chen and Wen (2016) conducted a comparative and statistical analysis of Bai Juyi's poems and its English versions translated by Levy, arguing that the translated versions contain different attitudinal resources with regards to the reader's readability. Zhang and Li (2016) stated that Appraisal Framework, relevant to stylistic and aesthetic features, can offer important criteria in evaluating the aesthetic values in both sources and translated poems.
Researchers have proved the possibility and significance of Appraisal Framework to study translation from a macroscopic level. More importantly, this framework can be employed to analyze source texts of various genres and their translated versions, thus further developing the criteria for evaluating translation quality and providing guidance for translation practices.

Studies on the English Versions of "Qiang Jin Jiu"
There are various English versions of "Qiang Jin Jiu" and the translated ones from Xu Yuanchong, Sun Dayu, Arthur Waley and Stephen Owen are widely spread and appreciated. However, studies on this poem and its translated versions are not sufficient. Wang (2005) analyzed how translators transferred the beauty in images, emotions and rhythms of the poem into its English versions. Liao (2013) discussed the influence of translator's subjectivity on retranslation process, showcasing that retranslation is useful to convey meanings in "Qiang Jin Jiu" because different social and cultural backgrounds, as well as ethical values of a translator, will lead to different interpretations of the poem. Through investigating the "cultural untranslatability" in the poem "Qiang Jin Jiu", Wang (2013) argued that the principle of "cultural intimacy" should be applied.
Studies largely focus on the transfer of cultural meanings or aesthetic values and the influence of translator's subjectivity. Different research perspectives have been employed to study the English versions of the poetry, such as descriptive translation, reception and systemic functional linguistics, but few analyses have been conducted mainly with the Appraisal Framework. Therefore, this study can be a new attempt to explore the appraisal resources in a classical Chinese poem and it hopes to give reference to further translation studies on classical Chinese poetry. Martin and Rose (2007, p. 17) noted, "Appraisal is concerned with the attitudes negotiated in a text, the strength of the feelings involved and the ways in which values are sourced and readers aligned." Accordingly, Appraisal Framework can be categorized into three major domains: attitude, engagement and graduation. As the primary system in Appraisal Framework, attitude is classified into three sub-systems: affect, judgement and appreciation. Affect deals with resources for construing positive or negative emotional reactions, involving un/happiness (e.g., happiness or hate), in/security (e.g., trust or anxiety), dis/satisfaction (e.g., respect or displeasure) and dis/inclination (e.g., desire or fear) (Martin & White, 2005, p. 45). Judgement is the assessment of people's behavior according to ethics, morality or other principles. It can be further divided into social esteem and social sanction (Martin & White, 2005, p. 52). Appreciation, evaluating the aesthetic quality of certain semiotic and natural phenomena, has three sub-categories: reaction (notability and likeability), composition (balance and complexity) and valuation (social significance) (Martin & White, 2005, p. 56).

The Appraisal Framework: A New Approach for Evaluating Meanings
The engagement system, dealing with sources of attitudes, includes monoglossia and heteroglossia. Martin and White (2005, p. 99) stated that monoglossia, with no reference to other voices and viewpoints, is generally regarded as a fact or bare assertion; heteroglossia provides dialogical alternatives either contracting or expanding the dialogic space.  (2019) and Stephen Owen (1981) are selected for further analysis.

Research Method
This study adopts both quantitative and qualitative methods. Quantitative method is used to present frequencies and distributions of appraisal resources in both source and target texts. Specifically, all figures are put in tables manually with the help of Microsoft Excel. Qualitative method is applied to further identify and investigate appraisal resources in the classical Chinese poem "Qiang Jin Jiu" and its English versions based on Appraisal Framework. The source text of the poem and the three English versions are analyzed successively according to attitude and engagement systems since the number of graduation resources in the data is negligible. For the attitude system, the study also explores whether the appraisal resources are expressed directly or indirectly and whether they are positive or negative.

Appraisal Resources in "Qiang Jin Jiu"
"Qiang Jin Jiu" was written by Li Bai, an influential poet from Tang Dynasty. The poet not only drank down his sorrow and expressed depression and contempt over the ruling class, but he also carried forward the spirit of seizing the day and the sense of self-confidence. In addition, this poem also tried to explore the secret of nature and life. Compared with the majestic Yellow River that runs permanently, an individual's life is short and insignificant. The appraisal meanings in this poem are mainly expressed via lexical choices and grammatical structure.
After calculation, the poem contains 32 appraisal resources, which are tagged and analyzed manually according to the Appraisal Framework. The following is an appraisal analysis of the source text from attitude and engagement systems.

Attitude
No judgement resource is found in "Qiang Jin Jiu" because it mainly contains the poet's feelings and description of things rather than the evaluation of people's characters. As to effect, tokens of positive attitude outnumber those of the negative one, expressing the willingness to indulge in wine, forget all misfortunes and enjoy the pleasant time and the poet's high spirit, self-awareness and vigor. Table 1 presents all affect types and their percentages in the source text. It is worth noticing that some of the attitudinal resources are expressed or invoked indirectly. For example:

1) 君不见，高堂明镜悲白发，朝如青丝暮成雪。
In line 1 which describes that the once silk-black hair has turned snow-white overnight, the poet adopted hyperbole to grieve over the transient life of an individual. Although this description does not directly showcase any emotion, readers can interpret the poet's deep sympathy for this unusual phenomenon after understanding the context: the whole life of an individual is nothing but a process from sunrise to sunset. Line 1 thus conveys the feeling of unhappiness.

2) 陈王昔时宴平乐，斗酒十千恣欢谑。
The verb "恣" (be free from constraint) in line 2 signifies the poet's desire for joy and laughter without directly employing adjunct to modify the action. Although the poet felt angry about the ruling class when he was exiled from the then capital Chang'an, he was inclined to indulge himself in wine to get rid of the sorrow. Therefore, line 2 implicitly expresses people's inclination.

3) 钟鼓馔玉不足贵，但愿长醉不复醒。
"钟鼓馔玉" (bell, drum and costly dishes) in line 3 implies the lavish lifestyle of the rich, who would prepare a sumptuous meal accompanied by a grand Chinese orchestra. Instead of rare dishes, getting drunk together with friends can make the poet more satisfied. The image here is an irony conveying the poet's dissatisfaction or even disdain about luxurious life in an implicit manner.
In terms of appreciation, most of the appraisal resources are implicitly expressed. Some of them are related to the perception or aesthetics of items, describing the Yellow River and a person's appearance; Others show the poet's evaluation of the value of things and actions. Table 2 shows all appreciation resources and their percentage in the source text. Both positive and negative appreciation resources are presented in this poem since the poet experienced a shift of feelings. The use of exaggeration can magnify a person's appreciation of natural scenery.

4) 君不见，黄河之水天上来，奔流到海不复回。
In line 4, "天上来" (come from the sky) is an imagination of the poet, who was awed by the great momentum of the rushing Yellow River. The poet's appreciation of the grand river is conveyed through vivid images instead of a plain description of running waters. Faced with the magnificent scenery, the poet realized the tininess of humans. Exaggeration is often used in classical Chinese poetry to express emotions and it is necessary to be reproduced in translation.

Engagement
Both monogloss and heterogloss are found in the source text of "Qiang Jin Jiu". The poet explicitly stated his assertions of natural or social phenomena in most cases while utilizing dialogic contraction and expansion to convey emotions. The following analysis focuses on heteroglossic resources since monogloss is easy to be identified and interpreted by readers. Table 3 shows all the heteroglossic types in the source texts, their percentages and examples. As to dialogic contraction, all are disclaim resources and most are deny resources, as the poet uses "不" (no or not) for five times and "莫" (another word to express not) for twice in the source text, which can be seen as a prominent feature of the source text in terms of engagement. As to proclaim, entertain resources enjoy higher frequency, which are evidenced by the use of "须" (should), "请" (please) and "与" (together), because one of the main purposes of this poem is to invite friends to wine and hope they can bend ears to listen to the poet's opinions. Some words in this poem deserve special attention.

5) 钟鼓馔玉不足贵，但愿长醉不复醒。
The Chinese word "但" in line 5 often means "but" in English in modern times, serving as a conjunction for countering expectation. But in most ancient Chinese poems ("Qiang Jin Jiu" included), "但" means "only", a continuative employed to adjust expectation. Therefore, although "但" belongs to "counter" resources, its function is not like the concessive conjunction "but".
Appraisal meanings in allusion should also be emphasized because they are easy to be neglected. For example:

6) 陈王昔时宴平乐，斗酒十千恣欢谑。
In line 6, "陈王" (Prince Chen), a cultural-loaded word, literally means Cao Zhi, a prince of the state of Cao Wei in the Three Kingdoms period. He once held a feast in Pingle Palace for noblemen to drink wine and enjoy themselves. The allusion here not only serves as an analogy with the poet's status but also reflects his disdain over hedonism and extravagance of the noblemen. Allusion is frequently used by Chinese poets to implicitly express emotional reactions or opinions towards certain behaviors or personal characters.
Above all, more than half of the attitudinal resources belong to affect, among which positive feelings such as satisfaction and confidence made up for nearly 60%. About 40% of appreciation resources describe aesthetic or perceptual aspects of items. As to the heteroglossic resources of engagement system, most of them belong to contract and they are all disclaim resources to assert the poet's stances. Implicit meanings are reflected via culture-loaded words and rhetorical devices such as allusion and hyperbole.

A Contrastive Analysis of Three English Versions from an Appraisal Perspective
After appraisal resources are collected and classified, the data reflects that Xu Yuanchong's version contains 28 appraisal resources, Stephen Owen's 26 and Arthur Waley's 24. The following part will be a contrastive analysis of "Qiang Jin Jiu" and the three English versions from attitude and engagement systems to investigate their appraisal meanings and the translation quality.

Attitude
Attitudinal resources are abundant in both the source text and its English versions. The similarities and differences are presented according to three subtypes of attitude system.

Affect and Judgement
It is worth noting that all three English versions contain judgement resources, which are not in line with the source text of "Qiang Jin Jiu" due to translators' understandings or intentions to explain or to rhyme. Table 4 shows the distribution differences of affect and judgement resources in the source text and its three English versions. It can be inferred from Table 4 that three English versions are similar to the source text in terms of dis/inclination while the un/happiness and dis/satisfaction resources vary obviously. For more details, Table 5 shows the different correlations of affect resources in three English versions respectively. The translation will be labeled as "=" if the affect resource in the translated text belongs to the same type as the source one; if not, it will be labelled as "≠". If there is no corresponding translation of the affect resource in the English version, it will be labelled as "0" and if there is additional resource of the same type in the translated text, it will be labelled as "+". Table 5. The correlation of affect resources between the source text and three English versions Table 5 demonstrates that Owen's version, where there is only one omission and one different type of affect, is mostly equivalent to the source text while Xu's use of affect resources is different from the source text to a large extent. It is possible that he intended to clarify or maximize the feeling of joy implicitly conveyed by the poet. For example: 7) Kill a cow, cook a sheep and let us merry be;

Source text Xu's version Owen's version Waley's version
And drink three hundred cupfuls of wine in high glee.
(Source text: 烹羊宰牛且为乐，会须一饮三百杯。) The affect resource "乐" (be happy) is illustrated twice in line 7. "Glee" can achieve end rhyme with "be". In ells.ccsenet.org English Language and Literature Studies Vol. 12, No. 1;2022 addition, the word "merry" can convey the pleasant feeling alone, but the additional resource "glee" can make readers get a vivid impression of the poet's happiness of drinking wine with his friends so as to strengthen the poet's high spirit and pave the way for the exciting verses in the latter part of the poem.
In contrast, Waley omitted several resources for the sake of rhyme or simple and unified sentence structure. As a result, he integrated similar emotions into one meaning instead of translating every character directly. The following example can help illustrate Waley's translation strategies.
8) Snatch the joys of life as they come and use them to the full; Do not leave the silver cup idly glinting at the moon.
(Source text: 人生得意须尽欢，莫使金樽空对月。) In line 8, Waley assumed that "得意" (obtain satisfaction) and "欢" (joy) convey the same meaning, so he combined the two resources as one to avoid redundancy and to make rooms for rhyme ("full" and "moon"). However, such kind of omission has abandoned the sense of satisfaction and thus affected the equivalence of affect between Waley's version and the source text.
There are similarities among the three English versions in terms of affect resources. First, all three English versions have translated the negative feelings such as "愁" (sorrow) and "悲" (to grieve) appropriately despite different wordings. It is partly because the poet expressed his sadness explicitly, which can be easily identified, so the translators do not have to dig into the deep meanings. Second, every line of three translated texts contains a subject. Due to the differences in sentence structure between Chinese and English, the subjects of clauses in classical Chinese poetry are often omitted while in English, a subject should be clarified in every sentence. This translation strategy can help readers, who are unfamiliar with classical Chinese poetry, better understand the poem.
It should be noticed that judgment resources appear in three English versions, reflecting translator's subjectivity in interpreting the poem. Here "aged", a judgement resource, shows people's sympathy towards the passing time, but there is no explicit description in the source text. The addition may be used to highlight the emotion of sorrow so readers can resonate with the fact that time is eternity while people are getting old with each passing day. The following is another example: 10) But real drinkers are more famous than sober sages. (Translated by Xu Yuanchong) And only the greatest drinkers have a fame that lingers on! (Translated by Stephen Owen) (Source text: 惟有饮者留其名。) Both Xu and Owen added judgement resources to evaluate the fame of drinkers to clarify the poet's positive attitude towards drinkers. Since there are no corresponding judgement resources in the source text, the appraisal meanings are expressed differently in the above-translated versions.

Appreciation
For better analysis, Table 6 presents the distribution differences of appreciation resources in the source text and its three English versions. The table above indicates that Waley's version is closest to the source text in terms of appreciation resources. As for more details, Table 7 shows the different correlations of appreciation resources in three English versions respectively. The labelling standard is consistent with that of Table 5. Table 7. The correlation of appreciation resources between the source text and three English versions

Source text Xu's version Owen's version Waley's version
The table above implies that Owen's version contains all appreciation resources in the source text and at the same time, the nonequivalence of appreciation resources is less apparent than that of affect resources. Certain translation strategies deserve greater attention. First, omission or alternation to different types of appraisal resources are more frequent in Xu's version, for example: 11) What difference will rare and costly dishes make?
(Source text: 钟鼓馔玉不足贵。) Instead of directly translating "不足贵" into "be not precious", Xu posted a rhetorical question first and later provided an answer to indirectly convey the contempt of extravagant lifestyle. Therefore, the appraisal resource has transferred from attitude to engagement to invite readers to probe into the poet's deep feelings.
Second, in Waley's version, an appreciation resource was altered to an attitude one, as can be seen in the following example: 12) Snatch the joys of life as they come and use them to the full; Do not leave the silver cup idly glinting at the moon.
(Source text: 人生得意须尽欢，莫使金樽空对月。) "空" (empty) in the source text literally describes the status of the cup. In line 12, the translator changed the static status into action with an adjunct "idly", giving people a sense of urgency to seize the pleasant time indirectly.
Third, additional appreciation resources can be seen in the versions of Waley and Owen. Waley added new appreciation resources such as "deep sea" to describe the destination of the Yellow River, while Owen utilized "lovely" to reflect the poet's fondness of wine.
Although appreciation resources are not as abundant as affect ones, it is still important to translate them into English to create an appropriate atmosphere to express self-confidence and discontent to the ruling class.

Explicitness and Polarity of Attitudes
To investigate whether the attitude resources are conveyed directly or indirectly and whether they are positive or negative, Tables 8 and 9 show the distributions of the attitudinal resources of the source text and the three English versions in terms of explicitness and polarity.  According to the data presented above, Xu's version is closest to the source text from these two aspects while Waley's version is the most different one since three translators employed translation strategies like omission, explanation, or amplification based on their subjectivity and interpretation of the poem. It should also be noted that the source text tends to express attitudes more implicitly than the three English versions, partly because Chinese people tend to express emotions in a more implicit way and, in most cases, they would leverage cultural-loaded words or rhetorical devices to convey meanings.

Engagement
Regarding engagement resources, Table 10 presents all the types of heteroglossia and their percentages in the source text of "Qiang Jin Jiu" and its three English versions. As is mentioned above, the frequent use of disclaim resources is a prominent feature of the source text. According to the data, Xu's version uses a similar proportion of negatives and concessive words, whereas Owen's version contains the smallest proportion in this regard. It is partly because Owen saw into the meanings of the wordings and transferred the negative clauses into assertive ones, losing the effect of the dialogic contract.
In terms of proclaim resources, Waley's version is less consistent with the source text compared with the other two. The reason may be that he did not fully understand the intention of the poet. This poem can be regarded as an invitation to friends, so that the poet used a large proportion of entertain resources such as imperative sentences ("请君为我倾耳听", Please hear), modal auxiliaries ("须", should, may) and rhetorical devices (hyperbole, allusion). However, certain entertain resources are omitted in Waley's translation. In comparison, Xu's version contains a larger proportion of entertain resources than the source text. The reasons for it are multiple. Xu preferred to use rhetorical questions to invoke resonance among readers by changing several appreciation resources into proclaim ones, as can be seen in line 11. Xu also tried to add imperative clauses to make the poet's intention of inviting friends more explicit. In addition, interjections such as "O", "Oh" were utilized to attract readers' attention and encourage them to dig into the deep meaning.
Three English versions have delivered the attribute resources, but their strategies are slightly different. Owen and Waley translated "陈王" literally as "the Prince of Ch'en", a proper name not familiar to target readers. Instead of adopting literal translation, Xu paraphrased it as "the Prince of Poet" to reflect the implication in this allusion. The poet employed the culture-loaded word "陈王" to make an analogy, so there must be similarities between the two. To specify the prince character as a poetry-lover can illustrate the same hobby of the poet and the prince, strengthen the ironic effect, help readers better align with the poet's disappointment to the ruling class, and encourage them to know more about the connotation of allusion.

Findings and Discussion
After analyzing and discussing the appraisal resources in "Qiang Jin Jiu" and its English versions, certain points should be highlighted.
First, the attitudinal resources are abundant in the poem and its three English versions, particularly the affect resources. Owen's version is closest to the source text concerning the type of affect while Xu's version is the most different one. As to appreciation, all the three translators, due to subjectivity, have used various translation strategies like omission, amplification, or alteration to different appraisals. In terms of the explicitness and polarity of attitudinal resources, Xu's version is almost equivalent to the source text, which tends to express attitudes more indirectly than all three English versions. It is worth noticing that Owen and Waley's versions present judgement resources. For engagement system, there is no English version that is particularly close to the source text, because translators tend to alter certain resources according to their understandings.
Second, reasons for the differences in attitude and heteroglossia include the translators' intention to rhyme, to use rhetorical devices, to specify the poet's intention, to create an emotional atmosphere, to uniformize the sentence structure of two lines, and to encourage readers to probe into deep meanings of the poem. The translators' subjectivity and lack of knowledge on Chinese culture also matter. All differences can be explained by the linguistic and cultural differences between China and English-speaking countries.

Conclusion
Based on the Appraisal Framework, this study investigates the attitudinal and engagement resources of the source text and three English versions of "Qiang Jin Jiu". The reasons for their appraisal differences are also analyzed. Appraisal meanings, which are significant to illustrate themes and underlying meanings of the source poem, deserve the attention and research of translators. In this regard, Appraisal Framework provides an effective method to identify appraisal resources so as to understand emotions in the source text and minimize the influence of translator's subjectivity. However, it should be admitted that as the language in classical Chinese poetry is different from modern Chinese, intralingual translation plays an important role in getting the complete meaning of the poem and deviations are unavoidable. Different translators may also have diverse understandings of the attitudes or themes in "Qiang Jin Jiu", so the classification of appraisal resources is somehow subjective. Therefore, further research of the Appraisal Framework and detailed studies of the intralingual translation of Chinese classical poetry are necessary.