Corpus-based Study of Differences in Explicitation

Between Literature Translations for Children and for Adults

n this study, I will examine the explication variation between translated novels for children and for adults using the corpus-based analysis method. Explicitation refers to an overall tendency to "spell things out rather than leave them implicit in translation" and has been identified as one of the universal translational features through corpus-based translation studies (CTS) (Baker 1996, p.180). CTS was initiated by Mona Baker in 1990s and has been undertaken to investigate the recurrent features of "explicitation, simplification, normalization/conventionality and leveling out" (Baker 1996, p. 84). This study simply focuses on the feature of explicitation by analyzing two English-to-Chinese parallel literary corpora¹. The purpose is to observe how literature translations for children and for adults show different degrees of noun density² added connectives and added notes. Since children and adults read for different purposes, translators will use different lexicons and different sentence structures, either simple or complicated, to produce varied degrees of explicitation and fulfill different translation purposes.

Rarely do we avoid the issue of readability when discussing literature translation for children. To enhance the level of readability for children, translated novels for children must have an overall higher level of explicitation than translated novels for adults because children require translations to be more easily comprehended than adults do. Many people will agree with this assumption, but assumptions still require an empirical test for genuine assurance. At the onset of this test, I raise several questions for an investigation.

1. Does literature translation for children show a lower percentage of noun density than that for adults and therefore have a higher degree of explicitation?
2. Does literature translation for children show a higher frequency of added connectives than that for adults and therefore have a higher degree of explicitation?
3. Does literature translation for children show a higher frequency of added notes than that for adults and therefore have a higher degree of explicitation?
4. How is the explicitiaton variation relevant to different translation purposes for different types of target readers?

Answers to these questions are expected to verify the purpose-oriented assumption of the explicitation differences between literature translations for children and for adults. Since this study uses Hans J. Vermeer's skopos theory and Zohar Shavit's poetics of children's literature as the theoretical framework for argument, I have to introduce their main concepts at some length in the following section. After that, I will discuss the corpus-based analysis method, including corpus design and search criteria. In addition, I will report some findings and discuss the implications of these findings to elicit significant insights. Finally, I will summarize this study, describe its limitations and suggest some possibilities for future research.

This research mainly explores the relevance of differences in explicitation to different translation purposes for children and for adults, so the theories of explicitation, translation purpose and the translation poetics of children's literature will be discussed as follows.

The term "explicitation" first appeared in Vinay and Darbelnet's article ([1958]1995; qtd. in Klaudy 1998, p. 80). Vinay and Darbelnet (1958) define explicitation as "the process of introducing information into the target language which is present only implicit in the source language, but which can be derived from the context or the situation" (qtd. in Klaudy 1998, p. 80). In other words, explicitation is "the process of introducing information into the target language which is present only implicitly in the source language, but which [could] be derived from the context or situation" (ibid.). E. A. Nida (1964) looks at the explicitation phenomenon in the translated text as a shift from the implicit to the explicit status that occurs when "important semantic elements carried implicitly in the source language may require explicit identification in the receptor language" (p. 228; qtd. in Klaudy 1998, p. 81).

There are abundant corpus-based studies of explicitation. S. Blum-Kulka conducted a systematic discourse analysis of English-French translations and English-Hebrew translations and found a noticeable rise in the level of cohesive explicitness from source texts to translated texts (cited here in a 2000 reprint, p. 300). To test the explicitation hypothesis, L. Øverås (1998) investigated the bi-directional English-Norwegian Parallel Corpus and discovered that the level of explicitness in both translated English and translated Norwegian was greater than their originals. M. Olohan and M. Baker (2000) conducted a comparable study on the addition of the word that after the verbs say and tell, with the aid of two corpora, Translational English Corpus (TEC) and British National Corpus (BNC). The analytical result showed that the use of that with say and tell in translational English corpus was more frequent than in non-translational English corpus. Moreover, M. Mutesayire (2005) used a comparable corpus to investigate reformulation markers such as in other words, namely, that is to say, and to be more precise/specific. These lexical items were found to occur more frequently in translated English than non-translated English (Olahan 2003).

All these studies confirmed the universality of explicitation that was produced independently of the impact of language pairs and text types in the translation process. These studies have found the global pattern that explicitation was produced not only without interference from language pairs but also without much relevance to text types. However, we cannot overlook potential local differences existing behind the generalization of explicitation when we consider the factors of different translation purposes. Thus, this study will examine how the explicitation feature shows different degrees between literature translations for children and for adults because the novels translated for these two types of audiences are expected to meet different reading purposes.

Vermeer's skopos theory argues that "every translation can and must be assigned a skopos" (1989, p. 230). The Greek word, "skopos" refers to the purpose of any piece of translation (Baker 2001; Munday 2001). In Vermeer's view, translating is not only a process of linguistic transfer from one language to another, but the cross-cultural transfer for the purpose of effective communication. The translation performance cannot be appropriate and satisfying if the translator is unaware of the purpose and the prospective readers of the translation.

Vermeer regards the addressee of the target language (TL) text as the main factor in determining the skopos (purpose) of translation (ibid.). "The skopos can be said to vary according to the recipient" (Reiss & Vermeer, 1984, p. 101). It follows the principle that "the translator should use the translation strategies which are most appropriate for achieving the purpose for which TT is intended" (Shuttleworth & Cowie, 1997, p.156). For example, to increase the child reader's knowledge in certain domains, the translator supplies background information of certain proper nouns with added notes for educational purpose. In addition, added cultural notes can be effective communication because cultural references in the source language (SL) text are situated in the source culture and henceforth require additional explanations to ensure the TL reader's comprehension. This example suggests that the intended purpose of translation affects the translator's decision-making and determines his/her translation strategies.

Due to different TL audiences and different translation purposes, a SL text can be translated in several ways. In her poetics of children's literature, Shavit (1986) proposes a set of rules for children's literature translation as opposed to those for adult literature translation. Shavit (1986) maintains that literature translation for children can follow two principles for two purposes. One is "an adjustment of the text to make it appropriate and useful to the child in accordance with what society regards (at a certain point in time) as educationally 'good for the child'" (Shavit 1986, p. 113). The other is "an adjustment of plot, characterization, and language to society's perceptions of the child's ability to read and comprehend" (ibid.). These two principles denote the importance of high readability, easy comprehensibility and didactic usefulness for child readers.

In contrast, to meet the need of adult readers, Shavit proposes the high literary style for the effect of "literariness" per se after comparing Hebrew literature written for children and for adults (1968, p. 128). She declares that the translator can use the special technique of linguistic presentation to create literary beauty and meet adult readers' need for aesthetic appreciation. Shavit's theory resembles Vermeer's skopos theory because both of them agree that the translator should use different methods and techniques to achieve different purposes.

This explicitation investigation expects to gain a more objective and more reliable conclusion through the corpus-based analysis of a huge volume of authentic data retrieved from the real market. Besides, this study includes a statistical survey with the aid of corpus-processing tools.³ However, it is crucial to have appropriate corpus design and set up search criteria prior to the investigation if it aims at obtaining complete and relevant results. These two components will be introduced as follows.

In this study, two parallel corpora of literature translations are used as analytical data. One is composed of the first chapters of five English novels with their Chinese translations for children and contains a total of 24,175 words (hereinafter Literature Parallel Corpus for Children/LPCC). The other is composed of the first chapters of the same five English novels with their Chinese translations for adults and contains a total of 108,386 words (hereinafter Literature Parallel Corpus for Adults/LPCA). A combination of these two corpora has a total of 132,561 words. The translated Chinese novels for children contain much fewer words than the translated novels for adults because literature translation for children, motivated by the need to make the text more communicative and more understandable, involves much adaptation but transmits similar messages. Tables 1 and 2 show the internal structures of LPCC and LPCA.

The Internal Structure of LPCC

The Internal Structure of LPCA

A comparison of Table 1 and Table 2 shows that the five translated novels for children were published in the late 1990s⁴ while the same novels translated for adults were published between 2001 and 2006.⁵ The reason for this gap is that Taiwan's translated novels for children with added notes were published in the late 1990s while Taiwan's translated novels for adults with added notes were not published until the first decade of the 2000s. Noticeably, the five novels are selected mainly because their translations for children and for adults are available on Taiwan's market.

Following a study of the noun-density-resulted explicitation, I select certain types of nouns as key words for search. In addition, the explicitation phenomena generated from added connectives at the structural level and added notes at the contextual level are probed, so that certain connectives and culture-specific notes are selected for search.

The three types of nouns to be selected are general nouns, proper nouns and pronouns. Some of them are listed in Table 3.

Connectives that refer to conjunctions and transitional words are normally used to signify logical relations between two clauses within one sentence or between two sentences. The five types of logical connectives are temporal, contrastive, causal, coordinate and conditional. The addition of these connectives in the translation helps the target reader to catch the message easily because they may explicate the implicit logical relationship at the sentence level. Table 4 shows some of the selected five connective types.

The names of some people or places in English novels have significant socio-cultural implications and provide non-English native children readers with useful knowledge and information. I randomly select some of them to check if they are translated with added notes. Table 5 shows these searched words that fall into six categories: place's names, people's names, foods, measurements, diseases, objects and religious terms.

I used the Concordance tool to calculate noun density, and then obtained the following. Table 6 shows the difference in noun density between the literature translations for children and for adults.

Table 6 shows that LPCC has lower noun density and therefore uses fewer varied types of nouns than LPCA. Thus, the translation for children has a 3% higher explicitation degree at the lexical level than does the translation for adults. This discrepancy in explicitation is not large compared to the explicitation gap by connectives (22%) and the explicitation gap by notes (28%).

Additionally, the author used the ParaConc tool to analyze and calculate the added connectives, and then obtained the explicitation percentages (Table 7).

Table 7 reveals that the degree of explicitation resulting from added connectives in translated literature for children is 22% higher than that for adults. However, this gap is still smaller than the explicitation by notes, as shown in Table 8.

Table 8 shows that literature translation for children shows a much higher percentage of explicitation resulting from added notes than the translation for adults. This finding suggests that the addition of cultural notes is a crucial strategy in the translation for the younger audience. The combination of lower noun density and higher percentages of added connectives and added cultural notes justifies a higher degree of explicitation in the translated novels for children than for adults.

The above-mentioned statistical figures reveal some significant insights and these points will be discussed in the areas of linguistic effect vs. aesthetic effect, educational function vs. literary function, and communicative acceptability vs. literal adequacy.

In response to the first research question, the result of the investigation indicates that noun density is slightly lower in the translated novels for children (24%) than for adults (27%). Since noun density is a method of measuring the lexical complexity of a text, the lower density in the translated novels for children reveals the use of fewer types of nouns. The repetitive use of certain lexical items might reduce the number of noun types in the translation for children. For example, the English phrases "any thing," "Toys and Trifles," and "my new gain'd wealth" are all translated with the same Chinese word, "貨/貨物" [goods] in a translated novel for children, while these words are respectively translated as "什麼貨物" [what things], "小掛件和小擺設" [small pendants and small decorations], and "新掙到手的錢"[money newly earned in the hands] in the translated novel for adults. More interestingly, "Ben Rogers" and "little Tommy Barnes" are both translated as "孩子" [child/ hei-ze] for children, but are literally translated as "班．羅傑斯" [ban luo jie si] and "小湯米．巴恩斯"[xiao tang mi ba en si] for adults. The use of rich vocabulary in the translation for adults does not create the comprehensibility burden but it would probably interfere with children's learning of the target language. Table 9 provides more examples of the repetitive use of some lexical items in the literature translation for children.

Table 9 shows that the English entries such as "His excellency," "he" and HURGO" are all translated as "欽差大臣"[the government-appointed official], but are respectively translated as "這位大官"[this high-ranking official], "他" and "赫哥" [HURGO] for adults. The repeated translation of "欽差大臣"[the government-appointed official] impresses young readers and enables them to memorize it easily. Repeating a translated term also helps children to acquire new words in a more impressive way. This lingustic impact meets Shavit's calling for the adjustment of children's literature for easy comprehension and easy learning.

In contrast, the linguistic impact resulting from the repetitive translation is useless for adult readers. Adults require literal translation with higher lexical variation because most of them are eager for aesthetic appreciation. Adult readers with higher TL proficiency accept lexical complexity in translation and tolerate any subtle presentation. The rich vocabulary for adults does not create a comprehensibility burden but it would probably impair children's language learning. Some adult readers favour high lexical variation and view this as derived from the translator's literary style. As stated earlier by Shavit, the translation for adults aims to produce literary beauty and boost aesthetic effect, so it expects to use a higher variety of lexical items.

In short, the higher lexical density creates aesthetic values for adults, whereas the lower lexical density brings the benefits to easy comprehension and easy memorization to children. The use of different linguistic techniques lead to the higher lexicon-specific explcitation for children than for adults. This explicitaiton variation concurs with Vermeer's skopos theory that every translation must be assigned a specific purpose and different translation strategies are used to achieve these intended purposes.

With regard to the second research question, the result of investigation informs us that translated novels for children have a higher frequency of added connectives than those for adults do. Thus, a higher degree of explicitation at the syntactic level can be found in the translation for children.

The addition of connectives explicates the implicit logical relations between two clauses and clarifies the core message of the sentence for child readers. Because of children's weak logical sense, added connectives assist in their understanding of the contextual relations. Let us compare the following pairs of translations in Table 10.

With the added conjunction because and the transitional words however and next, children will find it is easier to catch the cause-effect, comparison-contrast and chronological relations implicit in the context, and therefore are able to catch the message quickly. Added connectives help achieve the communication effect for children, but added connectives are of little help for adults who are more capable of seeing the implicit logical relations between sentences or between two clauses of one sentence. Added connectives sometimes destroy the literary sense and feeling of the original work. Thus, it is advisable to use literal translation method when translating novels for adults.

The difference between syntactic adaptation for children and literal translation for adults aims to fulfill different translation purposes: a contrast between communicative acceptability and information adequacy. This contrastive concept was raised by G. Toury when he discussed the issue of translational equivalence. In Toury's viewpoint, translation involves "an encounter, if not a confrontation, between two sets of norms" (1980, p. 55). When translation adheres to the norms of the SL system, the result is literal adequacy. Nevertheless, when translation follows the norms of the TL system, it brings about communicative acceptability. Thus, the adapted version fits young readers for the purpose of effective communication, while literal translation is suitable for adult readers for the purpose of informational adequacy. In the meantime, this discrimination conforms to Vermeer's skopos theory that every translation behavior is aimed at accomplishing a specific purpose.

Regarding the third research question, the answer is that translated novels for children have a higher frequency of added notes than translated novels for adults do. This means a higher degree of explicitation at the contextual level in the translation for children than for adults.

The purpose of adding notes is to transmit useful knowledge or background information of places, names, foods and others to the audience. Young readers, particularly elementary school students, are curious about and need to acquire cultural information or specialized knowledge about the outside world. At this point, young readers can access the deeper aspects of the foreign culture and acquire some specialized knowledge if they read the translation of foreign novels dotted with cultural or specialized notes. They can expand their cultural and informational horizons. The higher frequency of notes aims to achieve the purpose that the translation for children should be educational (Shavit 1986). Table 11 shows the different translations with and without cultural and specialized notes

To consider the added note about Paris, for example, it helps children readers to know the location and cultural importance of France's most famous city. The added notes of latitude and snuff provide children with specialized knowledge of geography and smokeless tobacco.

Adult readers, however, have already heard of Paris and have known well what latitude refers to or know how to find information about snuff, so additional notes about these terms are not necessary and are not added in the translation for adults. So far, we have known that added notes may provide children with cultural input through access to cultural information and specialized knowledge, but these notes might be redundant and ruin the original aesthetic form for adults. The differences in explicitation resulting from added notes between the translations for children and for adults meets Shavit's (1986) argument that children's literature is educational, but adult literature has a literary and aesthetic function.

This corpus-based study with statistical analysis has proved that the degree of explicitation varies between translations of children and adults' literature. Translated novels for children have a lower noun density (24%), a higher frequency of added connectives (43%) and a higher frequency of added notes (53%), so they show a higher degree of explicitaiton. In contrast, translated novels for adults have a higher noun density (27%), a lower frequency of added connectives (21%) and a lower frequency of added notes (25%), leading to a lower degree of explicitation. The hidden reasons for this discrepancy are that translated novels for children emphasize easy-to-understand lexical presentation with effective linguistic input while the translated novels for adults seek more complex lexical presentation for aesthetic appreciation. Furthermore, children require adapted translation for effective communication while adults require literal translation for informational adequacy. Finally, children's literature is an educational tool, requiring easier-to-learn lexicons and additional information input, but adult literature has a literary purpose, pursuing spiritual amusement and sublimation. In summary, the explication variation has close relevance to the target audience, the reading purpose and the translation function.

Although this study is limited in its corpora size (132,561 words in total), it serves as a model of purpose-oriented explicitation study using the corpus-based analysis method. It may be used as a point of departure for future researches on other literary genres, e.g. science fiction, in other language pairs to gain more significant insights.

I extend my enormous thanks to two graduate students, Vicky and Ruby, at the National Kaohsiung First University of Science and Technology, for helping me develop two parallel corpora. I also thank Taiwan's National Science Council for granting me funds to complete this project.