A Corpus-Based Cognitive Study of the “Rustic Literariness” of Translated Chinese Fiction

Abstract

This paper is a corpus-based cognitive study of Anglo-American sinologists’ English translation of Chinese fiction during the last four decades. With a domain analysis of the semantically tagged corpus data, it shows that animals and plants are key concepts in the corpus of Chinese Fiction Translation in the four decades (CCFT), in sharp contrast with the reference corpora, that is, OTC, the Other-source-language Translational Corpus of Fiction, and EFC, the English Fiction Corpus of Balanced Selection. Data shows that the “rusticness” embodied in the construction of animal and plant images not only exists in the four periods, but also has no diminishing tendency in the translated Chinese Fiction. An analysis of the concordance-line corpus of animal and plant words reveals a variety of key cognitive domains like Emotional, Sensory, People, and Relationship, in the proximate surrounding contexts, clearly indicating the existence of metaphors, and further analysis is conducted on metaphors embedded in “as…as” construction. The diversified metaphors with animal and plant images are constructed along the Great Chain of Being, and fully reflect the rich imagination of the authors and translators. It is argued that the animal and plant concepts are not simply indexes of the rustic environment, but are essential in the construction of the literariness of translated Chinese fiction. Therefore, any criticism that neglects or denies the “rustic literariness” of Chinese Fiction and its translation is partial. The study has implications for introducing new theoretical models and empirical methods into literary translation studies.

The general impression of modern and contemporary Chinese fiction is that artistic imagination is insufficient and “rusticness”(Xīangtǔ Qì 乡土气)is too heavy, which leads to poor literary performance. This is basically an impression based on the examination of translated Chinese novels before 1990s, especially made by Duke (1990). Duke (1990, 198) echoed Ou-fan Lee’s chastisement of contemporary Chinese writers for their excessively pedestrian social realism and lack of literary imagination, and he agreed with C. T. Hsia’s conclusion that “the work of these writers does not evince great imaginative power or technical brilliance; the intrusive presence of utilitarian ideals precludes the disinterested search for excellence”(Duke 1990, 201).

Duke’s primary assumption is that the international reputation of modern and contemporary fiction written in Chinese anywhere in the world is chiefly dependent upon English-language translations of that fiction (Duke 1990, 200), and he was concerned only with the fate of Chinese fiction when translated into English and read by people who do not understand the Chinese language, but who are avid and knowledgeable readers of serious English-language fiction. He presented the problems of modern and Contemporary Chinese fiction in relation to the “implied readers” as follows: the fact remains that the overwhelming bulk of their works, even when read in Chinese, appear to the contemporary reader as overly parochial, excessively topical, and lacking in artistic distinction (Duke 1990, 201).

In Duke’s critique, “overly parochial” (Duke 1990, 201) is in more specific terms “overly rustic.” Duke (1990) especially relates rural themes or setting to the lack of artistic distinction. As he later pointed out, “the modern Chinese novel became an increasingly rural-based panorama of social realism in which humanistic and ideologically engaged writers attempted to depict the desperate problems of poverty, ignorance, cruelty, and backwardness that cried out for solution at the time. This shift from individual to social concern and from urban to rural setting was one symptom of a striking urban-rural polarity that emerged for the first time in Chinese social history in this century…. modern Chinese fiction became predominantly concerned with rural life, or with the contrast between rural and urban China, and it has remained so on the mainland into the 1990s.” (Duke 1991, xiii).

There are universal values in literary expression, and there is also social-culturally distinctive literariness. Defamiliarization effect, emotional and aesthetic function, and the experience of characters that can resonate with or even entertaining ordinary readers (McDougall 2007) may be universal literary values. But there are obvious differences in the concept of literariness between China and the West. And we would argue that rusticness may be considered a special feature of the literariness of Chinese fiction, rather than a belittling feature as viewed from a eurocentric perspective, or from the perspective of Duke’s implied readers. In Xīangtǔ Zhōngguó (《乡土中国》) (Fei Xiaotong 费孝通 1984), a set of essays written shortly after World War II, Fei Xiaotong, the world-renowned sociologist and anthropologist, clarifies the meaning of the title in the book’s first sentence, “Chinese society is fundamentally rural.” He is using xīangtǔ 乡土 to modify zhōngguó 中国 (China), and conveying a subtle meaning to his readers that Chinese society has grown out of its ties to the land (Fei 1992, vii). The translators Hamilton & Wang Zheng (Fei 1992, vii) explained the translation of the Chinese title in their foreword: the book’s title is usually rendered in English as “rural China,” but this rendering is inexact. xīang 乡 means “countryside” and tǔ 土 means “earth”; but the combination, xīangtǔ, is a set phrase meaning “one’s native soil or home village.” Rusticness is closely related to the literariness of Chinese fiction, given the social-historical significance it carries for China. Therefore, we coin the term “rustic literariness” to discuss how Chinese fiction is different from western ones in this respect.

Given the impression or critique above-mentioned is mainly about Chinese fiction produced and translated before 1990s, it is inviting to wonder whether the rusticness remains in the English-translated Chinese fiction after the reform and opening-up till the present, and it is necessary to further explore whether the rustic feature of English-translated Chinese novels hinders its emotional and aesthetic functions. If we examine the literariness of Chinese fiction, especially the literariness of translated Chinese fiction in the English world, a more holistic study is needed and a diachronic perspective is essential.

Therefore, the questions of this research are: in the 40 years since the reform and opening up, has Chinese fiction as presented in English translation gradually changed, and is it still rustic and insufficient in literary imagination in front of foreign readers?

To answer these questions a holistic and diachronic study of the rusticness of translated Chinese fiction is to be conducted, and we resort to corpus-based methods for the purpose. A corpus is constructed of Anglo-American sinologists’ English translation of Chinese fiction during the last four decades, and for a corpus-based study, linguistic indexes of rusticness should be identified for automatic processing. Rusticness is related to the countryside or land, and the first things or images that come to mind when thinking of the countryside are plants or animals, hence the corpus-based analysis will focus on animals and plants, which are basic in the rural environment, and in close interaction of the characters. As the interaction between the people and the environment is, in light of the theory of embodied cognition and cognitive linguistics, essential to the social-cultural cognition and as the close link between land and Chinese spirit is embraced by previous sociologists and anthropologists, it is meaningful for us to examine how the prominent environmental marks of rusticness are related to literary expression. The structure of the paper will be arranged as follows:

Section 1 will be a description of the target corpus, reference corpora, and tools for analysis. Section 2 will make a statistical analysis of the corpus data of animal and plant concepts. Section 3 will analyze the animal and plant metaphors in light of cognitive linguistics and discuss the rustic literariness of Chinese fiction as related to animal and plant metaphors, and Sect. 4 is the conclusion with implications for literary translation studies.

1 Target Corpus, Reference Corpora, and Tools of the Research

The target corpus of the research is the Corpus of Chinese Fiction translated into English by Anglo-American Sinologists in the four decades (1979–2018) (CCFT). CCFT is a L2-L1 translational corpus, rendered by over 30 influential sinologist translators and consisting of over 130 works¹—some are translated volumes composed of more than two novels or short stories, and some are co-translated works. Table 1 is the statistics of CCFT, showing the number of word tokens, TTR and STTR.

Table 1 Statistics of the target corpus CCFT

Three reference corpora are used in the research, namely, BNC-Sampler Written Corpus, an equipment of Wmatrix, 968,267 words (Abbreviated as BNC-SW; OTC (the Other-source-language Translational Corpus of Fiction), and EFC(English Fiction Corpus of Balanced Selection). According to the help files of Wmatrix, written texts were selected for inclusion in the BNC-SW corpus according to three independent selection criteria: domain, time, and medium. Target proportions were defined for each of these criteria. As to domain,75% of the written texts were to be chosen from informative writings: of which roughly equal quantities should be chosen from the fields of applied sciences, arts, belief and thought, commerce and finance, leisure, natural and pure science, social science, world affairs. 25% of the written texts were to be imaginative, that is, literary and creative works.

Written texts for EFC were selected based on the ranking list of four institutions or websites, namely, (1) The Top 10 Novels by the Top 10 For Everything (top10for.com); (2) Best Novels of All Time by The Top Ten (https://www.thetoptens.com/best-novels); (3) The Books of the Century 1900–1999 by UC Berkeley; and (4) The bestselling novels by Ranker. The data were selected and balanced based on the following criteria:

• a. Balanced across four periods: 1980–1989, 1990–1999, 2000–2009, and 2010–2018, each period account for 1,200,000–1,500,000 words.

• b. Balanced across six themes for each period, which are romance, science fiction, fantasy, west adventures, detective, and general (Table 2).

  Table 2 Statistics of EFC

OTC is a reference corpus of translated fiction from source languages other than Chinese. Written texts were selected based on the ranking list of 2 institutions or websites, namely, (1) Best Novels of All Time, selected by the Top Ten (https://www.thetoptens.com/best-novels), and (2)The Books of the Century: 1900–1999 by UC Berkeley (Table 3).

Table 3 Statistics of OTC

The corpus analysis is mainly conducted with Wmatrix (https://ucrel-wmatrix4.lancaster.ac.uk/), a research platform developed by Paul Rayson, Lancaster University. The knowledge-based Wmatrix 4.0 include POS tagger CLAWS and Semantic tagger USAS (UCREL Semantic Annotation System). USAS can automatically tag semantic domains/fields) (Rayson 2008, 519–549). Its semantic tagset is based on Longman Lexicon of Contemporary English, including 21 general domains and 212 secondary domains, 94 tertiary domains, for example: emotion, food, law, time, etc. With a reference corpus, Wmatrix 4.0 can produce key words, key pos, and key concept/domains of a target text.

An example of Wmatrix 4.0 output data line is as follows:

Item  o1   %1  o2      %2      LL   LogRatio

Z8 29552   12   72023 7.44 +  4.45    0.69      Pronouns

Item is the grammatical or semantic item, Z8 is the Wmatrix marker of Pronouns

O1 is observed frequency in target corpus

O2 is observed frequency in reference corpus

%1 and %2 values show relative frequencies in the texts (frequency per 100 words)

+ indicates overuse in O1 relative to O2,

− indicates underuse in O1 relative to O2

The table of output data is sorted on log-likelihood (LL) value, an indicator of keyness. The cut-off value is 6.63, i.e., the item is statistically significant if the LL is above 6.63. With cut-off value set within Excel, key terms or concepts can be extracted and insignificant terms or concepts can be filtered out.

It is assumed that reference corpus is essential for revealing or filtering the key features of the target corpus. For our study, three types of reference corpora have been constructed and used. To reveal or filter characteristics of the CCFT or the English translation of Chinese novels, our basic assumptions as to the use of different reference corpora are:

• (1) BNC-SW is representative of general English written texts, thus with BNC-SW as a reference, the key literary and translational features of the target corpus text, can be revealed.

• (2) with EFC (original fiction corpus) as a reference, some if not all features of the fictional style of CCFT can be filtered out, but its features as translated fiction will be highlighted.

• (3) with OTC (translated fiction corpus from other source languages) as a reference, some common features of fictional translation can be filtered out, thus presenting the unique characteristics of translated Chinese fiction.

Some preliminary tests are conducted to verify the assumptions. Below is a brief test with CCFT as the target corpus (only the first four data lines shown).

Referencing BNC-SW

Item  o1        %1      o2     % 2     LL         LogRatio

Z8 1741538 12.74 72023 7.44 + 23712.43  0.78     Pronouns

Z99 730219 5.34   22165 2.29 + 20408.86  1.22     Unmatched

B1 224688   1.64   5489   0.57 + 8703.71    1.54     Anatomy and physiology

S4 137344    1.00  4002    0.41 + 4115.21    1.28     Kin

Referencing EFC

Item o1        %1    o2         % 2           LL       LogRatio

S4 137344   1.00 14503      0.38 +    16268.82 1.41 Kin

Z99 730219 5.34 162228    4.23 +    7582.84   0.34 Unmatched

F1 64365      0.47 10432     0.27 +     3074.17  0.79 Food

L3 38728     0.28 5437      0.14 +       2710.18  1.00 Plants

Referencing OTC

Item   o1     %1      o2        % 2           LL        LogRatio

N1 206915 1.51   32067     0.63 +     26324.74 1.27 Numbers

S4 137344 1.00    24611     0.48 +     13206.43 1.06 Kin

Z1 213446 1.56    45537     0.89 +     13153.39 0.81 Personal names

Z99 730219 5.34  216737    4.25 +    9088.70   0.33   Unmatched

With BNC-SW as the reference corpus, the statistics show that Pronoun is the highest in keyness among all concepts of CCFT. The value of RF is above 12, which is very remarkable, for it indicates that there are 12 pronouns on average in 100 words. The data lines verify the assumption (1), as the use of Pronoun is a key feature of literary narratives. Out of the need of narrative plot, especially plot coherence, literary works tend to use pronouns to refer to the environmental elements, characters, and events mentioned above, thus playing a role of textual cohesion. Francis & Kucera (1982), Biber (1985), and Tuldava (2005) share the view that one of the stylistic features of typical narrative articles is the “greater use of personal pronouns.” Fang Mengzhi (1988) compares the PB report of the US Business Publishing Bureau with the corpus of English short stories and finds that the number of pronouns in the novel is more than ten times that in the report.

However, when referencing OTC and EFC reference corpus, Pronouns didn’t occur in the key concept list of the target corpus or has not reached the critical cut-off value. And that verifies the assumptions (2) and (3), that is, some if not all features of the fictional style of CCFT can be filtered out. With BNC-SW as the reference corpus, the frequency of pronouns in the CCFT (Chinese fiction translation corpus) reaches the critical value with a high LL. However, relative to OTC and EFC, the critical value is not reached. OTC and EFC help us filter out this feature of literary narrative.

As statistics show, the keyness of Personal Names are critically high when referencing OTC, but the concept or domain of Personal Names doesn’t occur in the key concept list when referencing EFC, which reveals or highlights a remarkable feature of translated Chinese fiction translation texts. And this finding verifies the assumption (3).

Our purpose of using multiple reference corpora is to reveal more. OTC can help filter out some common features of literary translation, like the use of Pronouns, but it will certainly obscure some features, such as the Anatomy and Physiology domain revealed by BNC-SW. We use the three reference corpora together to reveal comprehensively and objectively the distribution characteristics of the key semantic domains of the target translated text, and to distinguish the key features of each 10-year period.

2 The Rusticness of Translated Chinese Fiction

As mentioned above, plants and animals are the prominent rustic images. Therefore, in this section we will focus on the distribution of animal and plant concepts or domains in CCFT (1979–2018), with BNC-SW, EFC, and OTC as reference corpus, respectively. The data lines will be listed and analyzed below.

Referencing EFC

Item o1     % 1   o2       %2        LL        LogRatio

L2 70221 0.51 13500   0.35 +   1757.8   0.55    Living creatures: animals, birds, etc.

L3 38728 0.28 5437     0.14 +   2710.18   1       Plants

F4 12681 0.09 1847     0.05 +   812.66   0.95    Farming & Horticulture

With EFC as reference corpus the concepts of plant and animal are obviously highlighted in CCFT, and the LL (log-likelihood) values, which reflects the keyness, is in the 812-1758 interval, which rank the fourth and ninth in the list of key concepts (to save space, the irrelevant concepts are not listed). Farming and Horticulture concepts related to plant domain ranks 17.

Referencing OTC

Item o1      %1   o2        %2       LL          LogRatio

L2 70221 0.51    14773 0.29 +   4510.77 0.83 Living creatures: animals, birds, etc.

L3 38728 0.28     6771  0.13 +   3923.55 1.09 Plants

Referencing BNC-SW

L2 70221 0.51 3225      0.33 +   661.58   0.62 Living creatures: animals, birds, etc.

L3 38728 0.28 2439      0.25 +    32.7     0.17  Plants

With OTC as reference corpus, the concepts of plant and animal are even more prominent. The LL values, which are in the 3923-4511 interval, and are far higher than that referencing EFC, rank 6 and 7, respectively, in the list of key concepts. When referencing BNC-SW, the concepts of animals and plants have also reached a critical level, much higher above the cut-off value, 6.63.

In sum, when referencing the three corpora BNC-SW, OTC, and EFC, animal and plant domains of CCFT reached a highly critical prominence, and most prominent when referencing OTC, and then EFC. As statistics show, the two domains represent the specialty of CCFT.

To further explore the historical change of this specialty, we will examine the situation of each period separately.

Statistics of animal domain in the four periods of English translations of Chinese fiction, referencing EFC, BNC-SW, and OTC, respectively, will be listed below.

Animal domain in 1979–1989

Animal domain is not a key domain referencing EFC and BNC, hence no data output here, and a little above the critical value when referencing OTC. Its relative frequency is 0.31, LL is 17.89.

Referencing OTC

Item o1     %1     o2     %2      LL    LogRatio

L2   5136 0.31 14773 0.29 + 17.89    0.1    Living creatures: animals, birds, etc.

Animal domain in 1990–1999

Referencing EFC

Item o1      %1 o2       %2       LL       LogRatio

L2 13728 0.44 13500 0.35 +  350.27  0.33     Living creatures: animals, birds, etc.

Referencing BNC-SW

L2 13728   0.44 3225 0.33 +  220.18  0.41     Living creatures: animals, birds, etc.

Referencing OTC

L2 13728 0.44 14773 0.29 +  1250.85 0.61    Living creatures: animals, birds, etc.

Animal domain in 2000–2009

Referencing EFC

Item o1     %1   o2        %2      LL       LogRatio

L2 31031 0.7    13500 0.35 +  4885.94   1        Living creatures: animals, birds, etc.

Referencing BNC-SW

L2 31031 0.7     3225  0.33 +  2010.55    1.08  Living creatures: animals, birds, etc.

Referencing OTC

L2 31031 0.7     14773 0.29 +  8537.13  1.28   Living creatures: animals, birds, etc.

Animal domain in 2010–2018

Referencing EFC

Item o1      %1 o2        %2       LL     LogRatio

L2 19925 0.45 13500 0.35 + 451.81     0.34         Living creatures: animals, birds, etc.

Referencing BNC-SW

L2 19925 0.45 3225    0.33 + 251.25     0.42       Living creatures: animals, birds, etc.

Referencing OTC

L2 19925 0.45 14773 0.29 +  1592.36    0.62      Living creatures: animals, birds, etc.

Through data comparison, we can find that the animal domain in the first period is relatively less prominent than the other three periods. The RF of the third period is the highest, with LL values in the 2010-8538 interval, then the fourth period with LL values in the 251-1593 interval, and then the second period with LL values in the 220-1251 interval.

Plant domain in the four periods

Below is the Plant domain data of CCFT with BNC-SW, EFC, OTC as reference corpora, respectively:

Plant domain in 1979–1989

Referencing BNC-SW

Referencing EFC

L3 4717 0.28 5437    0.14 + 1188.62    1.01        Plants

Referencing OTC

L3 4717 0.28 6771    0.13 + 1518.04    1.1        Plants

Plant domain in 1990–1999

Referencing EFC

L3 7920 0.25 5437  0.14 +  1133.75     0.85      Plants

Referencing OTC

L3 7920 0.25 6771 0.13 +   1550.77     0.94      Plants

Referencing BNC

not statistically significant

Plant domain in 2000–2009

Referencing EFC

L3 12861 0.29 5437 0.14 + 2163.88      1.04       Plants

Referencing OTC

L3 12861 0.29 6771 0.13 + 2922.47      1.14       Plants

Referencing BNC-SW

L3 12861 0.29 2439 0.25 + 46.53         0.21       Plants

Plant domain in 2010–2018

Referencing EFC

L3 12788 0.29 5437 0.14 + 2027.86     1.01      Plants

Referencing OTC

L3 12788 0.29 6771 0.13 + 2754.33   1.11      Plants

Referencing BNC-SW

L3 12788 0.29 2439 0.25 + 33.72      0.18      Plants

Through the statistical data, we can find that plant domain is a key domain in each of the four periods. Even in the second period with the lowest saliency, the LL value reached 293.05. The plant concept is the most prominent in the fourth period (2010–2018), LL value reaching 2470.19.

With a domain analysis of the semantically tagged corpus data, it is found that animals and plants are key concepts in the corpus of Chinese Fiction Translation in the four decades (CCFT), in sharp contrast with OTC, the translational fiction corpus of other source languages, and EFC, the Balanced Corpus of English Original Fiction. In terms of relative frequency and log-likelihood, the animal domain is the most prominent in the third period (2000–2009), and the plant domain is the most prominent in the fourth period (2010–2018), i.e., the “rustic atmosphere” embodied in the construction of animal and plant images not only exists in the four periods, but also has no diminishing tendency in the translated Chinese Fiction.

In order to clarify the framing effect of animal and plant concepts on the “rusticness” of translated Chinese fiction, we will examine the distribution of specific animal and plant words, respectively. EXCEL is used to obtain the target words from the concordance lines of the CCFT, and then Antconc is used to retrieve the word forms, and the distribution data of top target words are shown in Table 4.

Table 4 The top 20 animal words in CCFT

As the table shows, 10 of the 20 most frequent and prominent animal words in CCFT are HORSE, DOG, GOAT, CHICKEN, PIG, EGG, BARK, DONKEY, OX, and MULE, accounting for 35.4% of all animal words. They are all obvious signs of family and agricultural production in traditional Chinese rural society. Among them, pigs, cattle, sheep, horses, chicken, and dogs constitute the group of “Six Domestic Animals” (六畜) in rural China, accounting for 30.21% of all animal words. Cattle, horses, donkeys, and mules are closely related to the production and transportation of the Chinese agricultural society, accounting for 16.17% of all animal words. Sheep, chicken, pig, and egg are the main sources of protein food in rural areas, accounting for 13.84% of all animal words. Statistics show that the animal words that account for the largest proportion of all animal words in CCFT are clear markers of the Chinese rural society.

Next, we will examine the specific distribution of plant words (Table 5).

Table 5 The top 20 Plant words in CCFT

By observing the keyword data in the above table, we can find that tree, flower, grass, leaf, bamboo, garden/vegetable garden, (generally) plant, seed, willow, and root are the top ten plant words that mark rural scenes. As mentioned above, the Farming & Horticulture domain of rural agriculture has also reached a high degree of keyness, with a LL value of 812.66 when referencing EFC.

Based on the above data analysis, it can be concluded that the English translation of Chinese fiction from 1979 to 2018 is characterized by “rusticness” embodied in the images or concepts of animals and plants, which is in sharp contrast with English original fiction  (EFC) and translated English fiction from other source languages (OTC). Moreover, according to the relevant data of the four periods, this remarkable feature does not show a gradual weakening trend with the changes of history and with the influence of different incoming western literary movements—sometimes dominant.

3 The Rustic Literariness of Translated Chinese Fiction

In Chinese fiction translated into English as presented by sinologist translators, the images of animals and plants provide the “rustic” scene of literary narrative, or engage it with rustic atmosphere.

Then, what is the specific role of these animal and plant words in literary narrative? Are they just offering a “rustic” scene where a story takes place? What role do they play in the construction of literariness? To answer these questions, we need to further explore the context of these animal and plant words in relationship with other domains or domain matrices.

3.1 Rusticness and the Great Chain of Being

In light of cognitive linguistics, the matrix of life domain constituted by the animal and plant concepts is, more importantly, a link in the Great Chain of being, which ranges over the full gamut of forms of being in the universe, and which provides a cognitive framework for literary narrative (Lakoff and Turner 1989).

The Great Chain is a scale of forms of being—human, animal, plant, inanimate object—and consequently a scale of the properties that characterize forms of being—reason, instinctual behavior, biological function, physical attributes, and so on. Commonly, the Great Chain of Being is taught as background to literature and the history of ideas, as essential to an understanding of the worldviews of classical authors. In a larger sense, the Great Chain of Being is a cognitive model that we use to make sense of, and impose order on, the universe. It is a cognitive framework for literary construction and has a powerful influence on the construction of literary images and literary expressions (Lakoff and Turner 1989, 170–176).

e.g.,

He is a fox. (human-animal)

He is iron-hearted. (human-inanimate object)

In the two examples, metaphors are constructed along the scale of forms of being, and with the mappings between human and animal and between human and inanimate object constructed, respectively, we are able to understand the human characters in terms of the features of animals and inanimate objects.

Based on the Great Chain of Being, sometimes we also understand and recognize those low-level forms of existence from the perspective of human beings. For example, food can be understood as having a certain character, such as “unpretentious” wine, “tempting” dessert, “bold” perfume, etc. Animal life domain is one of the most complex cognitive domains for us to understand the non-human existence form from the human point of view. We usually use human characteristics to describe and understand animal characteristics metaphorically. For example, the animal schema about pigs contains dirty, chaotic, and rude features, which are metaphorical elements.

Examples above are conventional examples of metaphors involving the Great Chain of Being, in the expressions of which human attributes can be used to understand the attributes of lower-order things in the chain.

With the Great Chain of Being, we can place Chinese fiction and its translation works in the common field of world literature, observe the image construction along the hierarchy of the chain, and conduct comparative studies across languages and cultures based on a common cognitive model or framework. Next, we will explore the characteristics of the cognitive construction of animal and plant images in the CCFT (Chinese fiction translation corpus) within the framework.

3.2 Metaphors in the Surrounding Context Corpus of Animal Words

For examining the metaphorical construction of animal and plant images, we need to look at the surrounding context. The surrounding context corpora of animal and plant words, which consist of concordance lines of animal and plant words (80 characters on the left and right), respectively, were uploaded to Wmatrix 4.0 for further analysis. The Wmatrix platform enables us to look into the variety of key cognitive domains in the surrounding context.

With BNC-SW as reference corpus, the key surrounding domains of animal concepts in the CCFT were revealed and listed as follows:

• (1) Emotional domains in sequence of saliency

Item o1  %1    o2    %2     LL      LogRatio

E3- 778 0.31 1647 0.17 + 167.63 0.85 Violent/Angry

E5- 402 0.16 757   0.08 + 119.04 1.01 Fear/shock

E5 + 74 0.03 134   0.01 +  24.05   1.07 Bravery

E5 ++ 20   0   0          0 +   6.28      3.93 Bravery

E4.1- 330 0.13 979  0.1 +  14.64   0.36 Sad

• (2) Physiological domains in sequence of saliency

Item o1   %1   o2     %2         LL    LogRatio

B1 4461 1.75 5489 0.57 + 2878.38 1.63 Anatomy and physiology

B2- 476 0.19 1275 0.13 + 40.34      0.51 Disease

B4- 8         0   2          0 + 16.04      3.93 Dirty

• (3) Social domains in sequence of saliency

Item o1   %1 o2      %2    LL     LogRatio

S2 1093 0.43 2896 0.3 + 98.04    0.52 People

S2.1 494 0.19 1364 0.14 + 35.36 0.46 People: Female

S3- 2          0        0      0 + 6.28  3.93 Relationship

S3.2 221 0.09 508 0.05 + 36.35 0.73 Relationship: Intimacy and sex

• (4) Other domains

Item o1      %1    o2     %2     LL      LogRatio

X3.2 999   0.39 1271 0.13 + 615      1.58 Sensory: Sound

X3.4 1280 0.5 2795 0.29 + 250.46  0.8 Sensory: Sight

A1.1.2 324 0.13 815 0.08 + 37.23   0.6 Damaging and destroying

L1- 769      0.3 1585 0.16 + 179.25  0.88 Dead

An analysis of the surrounding context corpus of animal words reveals a variety of key cognitive domains like Emotional, Sensory, People, and Relationship, which are mostly abstract domains associated with human beings, and lead us to speculate that the target corpus may contain mappings along the Great Chain, especially mappings from animals to human beings. To further verify this speculation, specific metaphorical expressions need to be extracted from the corpus, and for this purpose metaphorical triggers need to be identified with corpus tools.

It is found that in the key concept list of the surrounding context corpus referencing BNC, Degree ranks the 7th, which is very prominent. The data line is as follows:

Item o1    %1 o2   %2      LL     LogRatio

A13 270 0.11 0    0.00 + 847.57 11.00 Degree

In the Degree domain, “as…as” is the most frequent item as compared to “more…than” etc., which is a conceptual metaphor trigger or signal (Tan 2016, 99). With an analysis of the “as…as” construction in the Degree concordance lines based on the metaphor identification procedure developed by Pragglejaz Group (2007, 3), a variety of animal metaphors in CCFT can be identified.

First, there are image metaphors that refer to the color, appearance, speech, hardness, weight, etc., with animal features mapping onto the inanimate features of things or human beings.

• (1) Color metaphors

E.g., as pale as a fish’s belly/as red as a monkey’s ass/His hair had been as red as pig’s liver

• (2) Layout or appearance metaphors

E.g., as closely arrayed as fish scales or swarming bees/as coarse as pig bristles/as crooked as the tail of the Mongol ox/as jagged as scattered dog’s teeth/as filthy as a pig/and I never cleaned/as lousy/as filthy as a pig/as matted as a magpie’s nest/as bare as a newly plucked chicken/as gaudy as a butterfly/as weather-beaten as bark/as hoarse as tree bark/as messy as a birds nest/as neat and tidy as a groom/The yellow plateau resembles a wind-dried skeleton/and its insides are as pitted as an ants’ nest/as swollen as spring silkworms about to spin their cocoons/they looked like they would burst if you touched them/as tiny as a fly’s legs/as minute as fine animal hair

• (3) Speed metaphors

E.g., as fast as a trotting horse/Time limped by as slowly as a decrepit ox pulling a broken cart

• (4) Hardness and weight metaphors

E.g., as hard as hog bristles/as hard as a mule’s hooves/as light as a feather/as light as a sparrow/as light as the wings of a cicada/as lightly as a little bird

Second, along the Great Chain of Being, there are emotional metaphors, which endow animals with human emotions.

• (1) Anxiety metaphors, using the images of ants, dog, pig, rabbit,fish, etc.

E.g., as anxious as an ant in a hot pot/as anxious as an ant on a hot frying pan/as anxious as ants in a hot wok/As restless as a sow in heat/Wufu was as restless as a sow in heat and kept muttering to himself/as restless as an ant on a hot frying pan/As flustered as a dog who has lost his way home/As flurried as a fish who has escaped the net/as nervous as a rabbit/He felt as desperate as an ant trapped in a hot wok

• (2) Fear metaphors, using the images of rat/mouse, sheep, chicken, etc.

E.g., as frightened as a little mouse/as frightened as the rats/as terrified as a mouse/as timid as sheep/as timid as sheep or chickens they find/But the real proof was in the eyes/which darted around as if wanting to take in everything while daring to look at nothing/like those of a cornered rat/the girls and their clients must have been as startled as a flock of sheep suddenly encountering a pack of wolves./

• (3) Happiness metaphors, using the images of fish and clam, etc.

E.g., as happy as a clam/as happy as fish in water/Yang gave her his word/and they were as happy together as two fish sporting in the water

• (4) Loneliness metaphors

E.g., as lonely as a caged dog

Third, there are a variety of physiological metaphors.

E.g., Strangely enough, however, those concubines, who in other homes were as fertile as hens and ducks/as barren as spayed sows or bitches/as strong as a calf/as mighty as leopards/as powerful as a galloping horse/as strong as a calf/as strong as a leopard/as strong as a young ox/as limp as a dying rabbit/Whenever he returned home/he felt as frail and vulnerable as a silkworm that had just shed its cocoon

Fourth, there are personality metaphors.

E.g., as fierce as tigers or eagles/as docile as a lamb/as stubborn as a donkey/as stubborn as a mule/as stupid as pigs/as sturdy as those of a workhorse

Fifth, there are social relationship metaphors.

E.g., as coquettish as monkeys/sweet dates as inseparable as fish from water

Sixth, there are generic-specific metaphors.

E.g., those who become men among men are as rare as phoenix feathers and unicorn horns/they would catch him as easily as a turtle in a jar/as easily as an eagle grabbing a chick/when I was in the army I could kill a man as easily as an ant/as busy as a mule pulling a millstone

Lastly, there are a variety of perception and mental metaphors.

• (1) mapping visual and tactile features of animals

E.g., he was as shiny as the horns of an old bull/as slippery as live fish/as skinny as a monkey/as skinny as a praying mantis/as slender as a water snake swayed past/as slippery as an eel/as slippery as live fish/a young man with a pimply face and a nose as flat as a calf’s/covered with sweat.

• (2) mapping auditory features of poultry, kitten, fish, cat, ox, mosquito, etc.

E.g., as noisy as a poultry farm/as peaceful as a kitten/as quiet as a cat/as silent as a school of fish, as silently as a fish he leaves me/Her voice became as soft as a little chick’s/in a voice as soft as a mosquito/Five said in a voice as faint as a mosquito/in a voice as steady as an ox/as cacophonous as a chicken farm/

• (3) mapping olfactory features onto abstract human characters

E.g., not as chaffly as common goose, more sedate than cat, more nutritious than rat, less demonic than weasel, and more common than lynx. Our meat tops the charts

• (4) mental metaphors

E.g., as carefully as a circus goat on a tight wire/as meticulously as a bird building its nest/It seemed like only yesterday, but now, suddenly, Huzhu had become as enigmatic as a river turtle./You’re accountable for everything you say, so don’t start going after people as indiscriminately as a mad dog!

3.3 Metaphors in the Surrounding Context Corpus of Plant Words

The surrounding context corpus of plants does not form valuable conceptual connections like that of animals. However, when referencing BNC-SW, we still find that Degree is a prominent concept/domain, ranking sixth in the key concept list of the surrounding context corpus of plant domain:

Item O1   %1 O2     %2     LL     LogRatio

A13 219 0.12  0    0.00 + 802.12 11.16 Degree

The same exploration procedure is repeated with the plant domain, and no lesser variety of metaphors is found.

Based on the Great Chain of Being, metaphors triggered by “as…as” map plant images to various aspects of human characteristics, including physical, physiological, emotional characteristics, appearance and temperament, life and perception features, etc. Examples are as follows:

• (1) Using plants to metaphorically describe human body features

E.g., his head was as big as a willow basket/His withered body was as thin as a straw swaying in a whistling wind/as thin as winter twigs/he was as skinny as a bamboo shoot/She extended her slender fingers, as fine as mist-wreathed bamboo shoots pointing to the heavens/The city mayor’s face appeared as small as an apple,/he becomes as fat as a pumpkin/Her form as lithe as a willow and in a breeze

• (2) Using plants to metaphorically describe facial features

E.g.,his face turned as red as the pumpkins/as red as a blossom in spring/his face was now as green as a spring bud.

• (3) Using plants to metaphorically describe the human perceptual features

E.g.,as smooth as sesame-seed glutinous rice dumplings/As stiff as dried mushrooms/Chen Di’s face was as rough and pale as hemp paper./

• (4) Using plants to metaphorically describe the beauty, alluring and pure temperament of a woman

E.g., The girls are as pretty as flowers/a daughter who is as fair as jade and as pretty as a flower./Without leggings, they would be as unsightly as a flower with no leaves around it/As captivating as a flower/As lovely as a flower or a piece of jade/Her face as lovely as a petal on a stream/As graceful as a tall narcissus/saw the face of Fenghuang, as pure as an apricot blossom/What would be the fate of a girl—as pure and delicate as a flower pistil/the girls… had to be as clean as a flower/as clean as a magnolia blossom, not a hint of anything unclean./they want to help them make their characters as beautiful as gardens/closer scrutiny revealed that she was as faded as her flowers.

• (5) Using plants to metaphorically describe human vitality and motion features

E.g., as ephemeral as a delicate flower/Tiankuan turned fifty, though he couldn’t figure out how he’d managed to live this long/what kind of guts filled his belly. He was as lively as a tree stump/In the glow of dawn her smile was as fresh as a new leaf./My legs felt as light as grass, and every step bounced up as if the ground were a spring/As feeble as a piece of straw/You were as free as a dandelion in a field/He could be as still as a tree or float motionless like a cloud in his room./

• (6) Using plants to metaphorically describe people’s perception, emotion, and ethical characteristics

E.g., Genzhu’s smile grew as thick as tree bark./That backward glance, as she was walking away—a glance as rich as a garden overflowing with sensory delights—made their hearts flutter and kept them up at night./he looked as pitiful as a leaf caught in the wind/You shore people are as rotten as your potatoes.

In the above four cases, smile is compared to thick bark, people’s glance to gardens, person to a leaf and people to rotten potatoes, forming complex metaphors, which map human perception, ethics, and emotional experience onto plants.

Examples of auditory perception metaphors are: in a voice as rough as tree bark/in a voice as hoarse as tree bark/You listen to the voices floating around you, as enviously as a tree trunk staring at falling leaves.

Here, the tangible plant features are used to describe the invisible voice, which can be said to be an  ingenious application of the traditional rhetoric–synesthesia.

Finally, there are examples of specific-generic metaphors, that is, using specific instances to metaphorically explain general or natural laws, such as:

E.g., As for the chances of my family ever returning to Ding Village, it was about as likely as a pile of leaves hopping back on to the tree they had fallen from./Exactly! We’ll soon have him in our power! And as another old saying goes, he’ll be as helpless as falling flowers and running streams./Because when a person from the lowliest class of all performs the noblest deed of all, it is fully as remarkable as a magic mushroom growing out of a dunghill, and it deserves to be publicized./The women gathered together as easily as pine nuts.

All in all, we can find diversified cross-domain mappings of animal and plant images with corpus data extracted from the CCFT. The corpus-based analysis is far from being exhaustive, yet we can see colorful metaphors based on the Great Chain of being, which fully reflect the rich imagination of the Chinese fiction as presented by sinologist translators. Therefore, it is argued that the animal and plant images or concepts are not simply indexes of the rustic environment, but are essential in the construction of the literariness of translated Chinese fiction. Based on the cognitive framework of Great Chain of Being, the metaphorical construction of the animal and plant images in the English translation of Chinese fiction can be interpreted in a broader sense of literariness. When the Chinese folk culture endows animals and plants with such human characteristics as intelligence, loyalty, courage, rudeness, beauty and changeability, and when the authors put the endowment into practice in their literary creation, the readers are led to metaphorically understand the attributes and behaviors of animals and plants in terms of human attributes, and also conversely. In this sense, any criticism that neglects or denies the literariness of Chinese Fiction and its translation is partial. If the rustic literariness presented in translation is not appreciated by the English readership, then we could only say that there is no universally appreciated literary values, or universally appreciated literary values need to be found elsewhere.

4 Conclusion

McDougall (2003, 30–31) pointed out that few literary specialists seem to have acknowledged the huge conceptual gap between Western critical methodologies and non-Western literary texts. Despite the indisputable influence of Western literature on twentieth-century Chinese literature, the position of literary texts in the previous two thousand years of Chinese history is unique to China, and even today it continues to enjoy a significance which is almost incomprehensible in the West. McDougall (2003, 38) urged western scholars to be alert to the way in which modern Chinese literature and its criticism might inform world literary theory and criticism. In this paper, we have used empirical methods to prove that the old impression of rusticness of the translated Chinese fiction is still true to the fact, but the associations of rusticness to bad literariness are problematic or eurocentric. Borrowing from eco-translatology, we may consider to change the term from “rustic literariness” to a more positive one, that is, “eco-literariness”, which proves to be a special feature of Chinese fiction as presented by sinologist translators.

Finally, there are some implications of the present research for future Cognitive Translation Studies and corpus-based translation studies. Cognitive Translation Studies can and need to be related to other approaches or fields of translation studies, which may lead to insightful findings as to important notions or core models in translation studies broadly defined. Literary translation is surely an important field for cognitive exploration, and corpus-based methods can be more cognitively revealing or no less revealing than other empirical methods. In this research the corpus-based methods enable a diachronic perspective and a holistic study of the rustic literariness of the Chinese fiction translation. Via an adequate variety of reference corpora and the semantic tagging tool of Wmatrix, the holistic and salient cognitive features of the target corpus are revealed. The advantage of corpus-based methods lies in that by using big data and text mining methods we can focus on the prominent or even unique features of a translation text, avoiding repetitive research around non-prominent and minor features in case studies, and thus avoiding over-generalization and too subjective deductions.