This Paper was Presented at
The First International Conference on Language, Literature, and Translation in the Third Millennium
Bahrain University
March 16-18, 2002

Abstract

It is conventionally believed that familiarity with the source and target languages, as well as the subject matter on the part of the translator is enough for a good translation. However, due to the findings in the field of text analysis, the role of text structure in translation now seems crucial. Therefore, the present paper sets out with an introduction on different types of translation followed by some historical reviews on text analysis, and will then describe different approaches to text analysis. As a case in point, a text analysis of the rhetorical structure of newspaper editorials in English and Persian and its contribution to the translation of this specific genre will be discussed. It will be indicated that newspaper editorials in these two languages follow a tripartite structure including "Lead," "Follow," and "Valuate" making translation of this specific genre possible and more accurate between the two languages. The paper will be concluded with the idea that text analysis can contribute and lead to more accurate and communicative translations.

Introduction

Conventionally, it is suggested that translators should meet three requirements, namely: 1) Familiarity with the source language, 2) Familiarity with the target language, and 3) Familiarity with the subject matter to perform their job successfully. Based on this premise, the translator discovers the meaning behind the forms in the source language (SL) and does his best to produce the same meaning in the target language (TL) using the TL forms and structures. Naturally and supposedly what changes is the form and the code and what should remain unchanged is the meaning and the message (Larson, 1984).

Therefore, one may discern the most common definition of translation, i.e., the selection of the nearest equivalent for a language unit in the SL in a target language. Depending on whether we consider the language unit, to be translated, at the level of word, sentence, or a general concept, translation experts have recognized three approaches to translation:

- translation at the level of word (word for word translation)

- translation at the level of sentence, and

- conceptual translation

In the first approach, for each word in the SL an equivalent word is selected in the TL. This type of translation is effective, especially in translating phrases and proper names such as United Nations, Ministry of Education, Deep Structure, and so on. However, it is problematic at the level of sentence due to the differences in the syntax of source and target languages. Translated texts as a product of this approach are not usually lucid or communicative, and readers will get through the text slowly and uneasily.

The structure of the source text becomes an important guide to decisions regarding what should or should not appear in the derived text.

When translating at the sentence level, the problem of word for word translation and, therefore, lack of lucidity will be remedied by observing the grammatical rules and word order in the TL while preserving the meaning of individual words. So, sentences such as "I like to swim," "I think he is clever," and "We were all tired" can easily be translated into a target language according to the grammatical rules of that language. Translation at the sentence level may thus be considered the same as the translation at the word level except that the grammatical rules and word order in the TL are observed. Texts produced following this approach will communicate better compared to word for word translation.

In conceptual translation, the unit of translation is neither the word nor is it the sentence; rather it is the concept. The best example is the translation of idioms and proverbs such as the following.

"He gave me a nasty look"	"Carrying coal to Newcastle"
"Do as Romans do while in Rome"	"He kicked the bucket"

Such idioms and proverbs cannot be translated word for word; rather they should be translated into equivalent concepts in the TL to convey the same meaning and produce the same effect on the readers.

In addition to word-for-word, sentence-to-sentence, and conceptual translations, other scholars have suggested other approaches and methods of translation. Newmark (1988), for example, has suggested communicative and semantic approaches to translation. By definition, communicative translation attempts to produce on its readers an effect as close as possible to that obtained on the readers of the source language. Semantic translation, on the other hand, attempts to render, as closely as the semantic and syntactic structures of the TL allow, the exact contextual meaning of the original. Semantic translation is accurate, but may not communicate well; whereas communicative translation communicates well, but may not be very precise.

Another aspect of translation experts have attended to is the translation processes. For instance, Newmark (1988: 144) contends that there are three basic translation processes:

1. the interpretation and analysis of the SL text;
   
2. the translation procedure (choosing equivalents for words and sentences in the TL), and
   
3. the reformulation of the text according to the writer's intention, the reader's expectation, the appropriate norms of the TL, etc.

The processes, as Newmark states, are to a small degree paralleled by translation as a science, a skill, and an art.

This paper is concerned with some aspects of the first process. It will be suggested that a major procedure in the interpretation and analysis of the SL text should be text analysis at the macro-level with the goal of unfolding rhetorical macro-structures. By macro-structures we mean patterns of expression beyond sentence level. In the next parts of the paper, first a brief history of text analysis will be presented followed by approaches to text analysis. The paper will then continue by indicating how two specific genres; namely, newspaper editorials and poetry, lend themselves to macroanalysis of texts and how this analysis will help translators.

Historical Perspectives on Text Analysis

It is a major concern of linguists to find out and depict clearly how human beings use language to communicate, and, in particular, how addressers construct linguistic messages for addressees and how addressees work on linguistic messages in order to interpret and understand them.

Accordingly, two main approaches have been developed in linguistics to deal with the transmission and reception of the utterances and messages. The first is "discourse analysis," which mainly focuses on the structure of naturally occurring spoken language, as found in such "discourses" as conversations, commentaries, and speeches. The second approach is "text analysis," which focuses on the structure of written language, as found in such "texts" as essays and articles, notices, book chapters, and so on. It is worth mentioning, however, that the distinction between "discourse" and "text" is not clear-cut. Both "discourse" and "text" can be used in a much broader sense to include all language units with a communicative function, whether spoken or written. Some scholars (see, e.g., Van Dijk, 1983; Grabe and Kaplan, 1989; Freedman, 1989) talk about "spoken and written discourses"; others (see, e.g., Widdowson, 1977; Halliday, 1978; Kress, 1985; Leckie-Tarry, 1993) talk about "spoken and written text." In this paper, we stick to "text analysis" with a focus on the structure of written language at micro- and macro-levels.

According to Connor (1994), text analysis dates back to the Prague School of Linguistics, initiated by Vilem Mathesius in the 1920s. Later on it was elaborated by Jan Firbas and Frantisek Dane in the 1950s and 1960s. Connor (1994) believes that The Prague School's major contribution to text analysis was the notion of theme and rheme, which describes the pattern of information flow in sentences and its relation to text coherence.

On the other hand, Stubbs (1995) states that the notion of text analysis was developed in British linguistics from the 1930s to the 1990s. In this regard, the tradition, as Stubbs (1995) continues, is visible mainly in the work of Firth, Halliday, and Sinclair (See, e.g., Firth 1935, 1957a, 1957b; Halliday 1985, 1992; Sinclair 1987, 1990). The principles underlying these works, as stated by Stubbs, demand studying the use of real language in written and spoken discourse and performing textual analysis of naturally occurring language.

As (Connor 1994: 682) states, "systemic linguistics, a related approach to text analysis and semiotics, emerged in the 1960s with the work of linguists such as Halliday, whose theories emphasize the ideational or content-bearing functions of discourse as well as the choices people make when they use language to structure their interpersonal communications (see, e.g., Halliday, 1978)." Halliday's systemic linguistics has influenced text analysis tremendously as well as curriculum models for language education (see, e.g., Mohan 1986). Following Halliday and Hasan's (1976) taxonomy, the notion of cohesion has been one of the popular issues in text analysis.

According to Connor (1994), in the 1970s and 1980s, many linguists, psychologists, and composition specialists around the world embraced text and discourse analysis. Connor believes that this New School of Text Analysis is characterized by an eclectic, interdisciplinary emphasis, placing psychological and educational theories on an equal status with linguistic theories (whereas the Prague and systemic approaches primarily orient themselves to linguistics). Examples of text analysis from this new approach include studies of macro-level text structures such as Swales's (1990) studies of the organization of introductions in scientific research articles; and Biber's (1988) multidimensional computerized analysis of diverse features in spoken and written texts.

Bloor and Bloor (1995) contend that by the process of analysis, linguists build up descriptions of the language, and gradually discover more about how people use language in social communication. The same thing can be considered with the dynamic process of translation in that the discourse and rhetorical structures encoded in the source language can be reconstructed in the target language, and then the translator goes for the appropriate syntax and lexicon. One of the indexes of a "good" translation would, therefore, be to see to what extent a translator has been able to reconstruct the rhetorical structures of the source text in the target language through text analysis.

Approaches to Text Analysis

We may roughly divide the available literature on text analysis into two groups. First, those aiming at providing a detailed linguistic analysis of texts in terms of lexis and syntax. This approach has mostly referred to as analysis at micro-structure. Second, those related to the analysis and description of the rhetorical organization of various texts. This approach has been labeled as macro-structure analysis of texts. In this paper, we are concerned with macro-analysis and its implication in translation. First, the macro-structure of newspaper editorials in two languages, English, and Persian, will be presented. Then, the macro-structure of the poems of a famous Persian Poet, Hakim O'mar Khayam, and the English translation of these poems by a well-known English translator, Fitzgerald, will be presented as two cases in point. It would, of course, be naïve to generalize these cases to all languages and all types of genres without adequate research and empirical evidence. However, the point of discovering and unfolding macro-structures in a SL with the goal of reconstructing nearly the same patterns in the TL in the process of translation deserves theoretical and practical attention.

The Case of Newspaper Editorials

Bolivar (1994) studied editorials of The Guardian. She selected 23 editorials from The Guardian during the first three months of 1981. Based on the analysis of these editorials, she found out that a tripartite structure called "triad" organizes the macro structure of the editorials. Bolivar explains that the function of the triad is to negotiate the transmission and evaluation in written text and that it consists of three turns or elements, namely, Lead, Follow, and Valuate, serving distinctive functions of initiation, follow-up, and evaluation of the two. It shares similarities with the "exchange," as the minimal unit of spoken discourse. The following excerpt taken from The Gardian, "Behind closed Irish doors." March 3, 1981, cited in Bolivar (1994: 280-1) is an example of a triad.

L	Britain and Ireland are now trying, at long last, to work out a less artificial link between them than that which binds two foreign states.
F	This is the most hopeful departure of the past decade because it opens for inspection what had lain concealed for half a century and goes to the root of the anguish in Northern Ireland.
V	The two countries now recognize that though they are independent of one another they cannot be foreign.

According to Bolivar, not all triads have three turns. Triads can exhibit more than three turns provided that the sequence LF is repeated and V is the final turn. Thus, triads such as LFLFV or LFLFLFV can be found when the V turn is delayed by the writer.

The study of editorials from other British newspapers conducted by Bolivar confirmed the existence of three-part structures in those newspapers.

Parallel to Bolivar's study, Riazi and Assar (2001) conducted a similar study on Persian newspaper editorials to see if the same macro-structures are detectable in this particular genre. The editorials of six currently published Persian newspapers were examined. A sample of 60 editorials, 10 for each newspaper, was randomly selected to be analyzed.

The editorials were analyzed at two levels 1) at a rhetorical macro-structure level, and 2) at a micro syntactic level. Each text (editorial) was segmented by sentence units and was codified according to its function; lead, follow, or valuate. The inter-coder reliability indices of the segmentation and codification of the editorials were then determined. An inter-coder reliability index above .80 was obtained. The following excerpt from Iran (June 27, 1997), one of the newspapers, is an example of a triad in Persian newspaper editorials.

L	The motivating command of the Late Imam in May 1979 was the beginning of a revolutionary era for the popular movement to construct and develop the villages through the establishment of an organization called Jihad-e-Sazandegy.
F	It was a revolutionary institution whose fundamental duty was the improvement of economic and social conditions of villagers in Iran.
V	The marvelous achievements of Jihad-e-Sazandegy and the fruitful actions of this public institution proved the Imam's correctness of recognition and depth of revolutionary perception.

Results of the analysis performed on the editorials indicated that the most frequent pattern pertaining to all the studied newspapers was LFV. In other words, we can say that the general macro-structure of Persian newspaper editorials is LFV. This finding is in line with that of Bolivar's (1994) as related to The Guardian newspaper. This common pattern between the two languages enhances the translatability of the newspaper editorials. The task of translators would be to look for the triads and go for the appropriate syntax and lexicon. It is interesting to point out that in both Bolivar's and our study, it was found that each turn is characterized by specific sentence types. For example, it was found that "Leads" were mostly expressed in interrogatives; "Follows" mostly used passive structures; and "Valuates" used conditional and copulas. The usage of special syntactic structures for specific turns can be justified partly in light of the discoursal function, attributed to each structure and reported in previous studies. Interrogative sentences, for example, are used with the goal of eliciting information or presenting some new topic for discussion. Since the main function of L turn is to introduce the aboutness of the triad and a subject, therefore, it seems quite reasonable to have interrogatives mostly in L turns. On the other hand, the correspondence of passive structures and F turns might be due to the fact that passives provide development and elaboration of the events. Reid (1990: 201) points out that "the passive voice is indicative of the formal interactional character of ...[a] prose as opposed to the more personal, interactive prose of narrative." As for V turns, we can say that the function of conditionals is to produce or suggest some kind of solution or desirable action on some conditions (Bolivar 1994), thus, the association between V turns and conditionals. Becoming aware of these macro- and micro-features of texts, we can make our translations of particular texts and genres more accurate, meaningful, and communicative.

The Case of Khayam's Robaiyat (Quatrains)

Omar Khayam was one of the most famous and beloved Persian poets of middle ages. The Robaiyat of Omar Khayam is among the few Persion masterpieces that have been translated into most languages, including English, French, German, Italian, Russian, Chinese, Hindi, Arabic, and Urdu. The most famous translation of the Robaiyat from Persian into English was undertaken in 1859 by Edward J. Fitzgerald. He has tried his utmost to adhere to the spirit of the original poetry.

Yarmohammadi (1995) studied the rhetorical organization of Khayam's Robaiyat (quatrains) and compared it with its English translation by Fitzgerald. His study revealed that the macro-structure of all Khayam's Robaiyat included three components, namely, "description," "recommendation," and "reasoning" which can be used as a criterion to distinguish between the real Khayam's Robaiyat and those erroneously attributed to him. Based on his analysis, Yarmohammadi came to the conclusion that the reason for Fitzgerald's successful translation of Khayam's Robaiyat is that he was able to reconstruct the same macro-structures in English and then apply appropriate sentence structures and lexis. The following is an example of one of the Khayam's quatrains as translated by Fitzgerald.

Fitzgerald:

And this delightful Herb whose tender Green
Fledges the River's Lip on which we lean—
Ah, lean upon it lightly! for who knows
From what once lovely Lip it springs unseen!

Literal:

The grass that grows by every stream
Like angelic smiles faintly gleam
Step gently, cause it not to scream
For it has grown from a lover's dream.

Conclusion

As Hatim and Mason (1997) state, a translator typically operates on the verbal record of an act of communication between source language speaker/writer and hearers/readers and seeks to relay perceived meaning values to a group of target language receiver(s) as an separate act of communication. However, according to Hatim and Mason (1990), we know little about what patterns there are and how equivalence could be achieved between them. One thing of which we can be confident, nevertheless, is that the patterns are always employed in the service of an overriding rhetorical purpose. This is an aspect of texture which is of crucial importance to the translator. The structure of the source text becomes an important guide to decisions regarding what should or should not appear in the derived text. The point that the present paper tried to make is the benefit translators may derive from text analysis in translation by determining the micro- and macro-indices of the texts to support them in their difficult task.

Text analysis is, thus, becoming a promising tool in performing more reliable translations. There are numerous studies done on text analysis, which can have interesting messages for translators. For example, the kind of structure frequently reported for argumentative genres include "introduction, explanation of the case under discussion, outline of the argument, proof, refutation and conclusion" (Hatch 1992: 185). As a final word, we may say that in translation we should first try to reconstruct the macro-structure and rhetorical structure of the source text in the target language and then look for the appropriate words and structures; this is a procedure that skillful translators perform in the process of translation consciously or unconsciously.

References

Biber, D. (1988). Variation across speech and writing. New York: Cambridge University Press.

, Bloor, T. & Bloor, M. (1995). The functional analysis of English: A Hallidayan approach. London: Arnold.

Connor, U. (1994). Text analysis. TESOL Quarterly, 28, 682-685.

Firth, J.R. (1935). The technique of semantics. Transactions of the philological society, 36-72.

Firth, J.R. (1957a). Papers in linguistics. London: Oxford University Press.

Firth, J.R. (1957b). A synopsis of linguistic theory, 1930-1955. Studies in Linguistic Analysis, Special Vol., Philological Society, 1-32.

Halliday, M. (1978). Language as social semiotic. London: Edward Arnold.

Halliday, M. & Hasan, R. (1976). Cohesion in English. London: Longman.

Halliday, M. (1985). An introduction to functional grammar. London: Edward Arnold.

Halliday, M. (1992). Language as system and language as instance: The corpus as a theoretical construct. In J. Svartvik (Ed.), Directions in corpus linguistics (pp. 61-77). Berlin: Mouton.

Hatim, B. & Mason, I. (1990). Discourse and the translator. London: Longman.

Hatim, B. & Mason, I. (1997). The translator as communicator. New York: Routledge.

Hartmann, R. (1980). Contrastive textology. Heildberg: Julius Groos Verlag.

Hinds, J. (1980). Organizational patterns in discourse. In T. Givon (Ed.), Syntax and semantics: Vol. 12: Discourse and syntax. New York: Academic Press.

Reid, J.M. (1990). Responding to different topic types: A quantitative analysis from a contrastive rhetoric perspective. In B. Kroll (Ed.), Second language writing. Cambridge: Cambridge University Press.

Riazi, A. M., & Assar, F. (2001). A Text Analysis of Persian Newspapers Editorials. Journal of Social Sciences and Humanities of Shiraz University, Vols. 31 & 32.

Sinclair, J. (1987). Collins Cobuild English Language Dictionary. London: Happer Collins.

Sinclair, J. (1990). Collins Cobuild English Grammar. London: Happer Collins.

Stubbs, M. (1995). Text and corpus analysis. Cambridge, MA: Blackwell Publishers Inc.

Swales, J. (1990). Genre Analysis: English in Academic and Research Settings. Cambridge: Cambridge University Press.

Yarmohammadi, L. (1995).The discoursal and textual structure of Khayam's poetry in Fitzgerald's English versification. In L.Yarmohammadi (Ed.), Fifteen Articles in Contrastive Linguistics and the Structure of Persian: Grammar, Text and Discourse. Tehran: Rahnama Publications.