Some notes on method

The questions
The aims of the thesis are

• To inquire into the translation methodology employed by the translators in question and their strategies to overcome the language barrier between Greek on the one hand and Syriac/Arabic on the other hand (I hold that the problems encountered in the previous Syriac 'translation phenomenon' are roughly comparable to those cropping up in the Arabic 'translation movement') and
• To trace two sets of influences exerted by the translation process: that of the source and target language structures on form and content of the target text and that of the translation phenomenon as a whole on certain areas of the Islamic 'scientific enterprise' (Gutas 1998), be it on the level of terminology, be it on the level of implicit or explicit adoption of 'thought patterns'.

Procedural overview
To answer the first question, a thorough analysis of both source and target text in terms of lexical, syntactic, textual and discoursal properties has to be undertaken. The methodological framework chosen is the text linguistic approach developed by Ian Mason and Basil Hatim (e.g. Discourse and the Translator, 1990 and Communication across Cultures, 1997). Supplementing the framework of text linguistics, Muhammad Farghal's extension of the concept of 'managing' in translation (Farghal 1993) could serve as a tool to analyse translators' ideologically motivated interventions in the course of the translation process.

   At the end of such an analysis, it should be possible to offer an assessment of the texts in terms of translational equivalence or adequacy in expressing the rhetorical purpose(s) of the author, i.e. Aristotle, and to provide an account of the modifications of the source text content necessitated by the specific problems of Greek-(Syriac)-Arabic translation of philosophical texts.

   While mainly based on the above linguistic model, I intend to make use of the wealth of methodological and terminological work by the likes of Gerhard Endress, Hans Daiber, Henri Hugonnard-Roche et al. and integrate, as far as possible, their textual findings into it.

Text linguistics, some more detail
By conceptualising translation as "a communicative process which takes place within a social context" (Hatim/Mason 1990, p.3), Hatim and Mason are able to overcome artificial boundaries between literary and non-literary translation and open up all sorts of texts and utterances to the same theoretical treatment.

   In general terms, text linguistics as understood by Hatim and Mason stresses the importance of contextual factors in establishing the meaning of a text. To relay meaning between text producer and user, a constant process of negotiation influenced by textual, pragmatic and semiotic dimensions of the message takes place. According to Hatim and Mason, the message of a text is constituted by the interplay of three contextual components:

• A communicative aspect, subsuming user-related variables such as dialects, temporal language variations, idiosyncrasies of personal language use etc. on the one hand and use-related variables such as field of discourse (the social function of utterances, e.g. 'religious sermon', 'political speech'), mode of discourse (medium of utterances, e.g. newspaper article, telephone conversation) and tenor of discourse (relationship between addresser and addressee, expressed by degrees of formality) on the other hand;
• A pragmatic aspect, covering elements of intentionality of a text act and dealing with factors such as the speech act sequence (cf. Austin/Searle's 'speech act theory'), inference of unstated content by the text receiver, implicature and the role of the cultural background for an understanding of texts;
• A semiotic aspect, approaching texts as signs in a semiotic system and classifying and structuring them along the lines of genre (textual conventions connected to specific social occasions, e.g. rules governing the structure of a newspaper editorial or an obituary), discourse (reflecting the text producer's attitude, e.g. 'committed', 'evaluative') and text type (divisions in discourse relaying a specific rhetorical purpose, e.g. exposition vs. argumentation); also covered by the semiotic component of context are phenomena of intertextuality, allusions and references to cultural connotations and knowledge structures, situating a specific text in a particular culture and textual tradition (cf. 'hypertext').

   These dimensions of context are expressed in a given text by means of texture, the combination of lexical and syntactic elements serving a particular rhetorical purpose. While contextual factors may be universal to different languages, the rules governing their linguistic expression are normally not. Thus, a translator has to exploit textual cues (texture) and his own background knowledge (intertextuality) to reconstruct contextual variables (communicative, pragmatic and semiotic values) and their rhetorical purpose. After having established the message of a text, he has to transfer it into the target language, reconstructing the message context and its linguistic expression according to its the textual conventions.

   Translation assessment operates on the same lines: identifying the rhetorical purpose of a source text and the contextual elements employed to convey that purpose; analysing the texture of the source text to reach a detailed understanding of the way rhetorical purposes are linguistically expressed; establishing the rhetorical purpose of the target text with recourse to the source text findings and prior knowledge of both the historical situation obtaining at the time of the translation and the structural constraints placed on the expression of source text rhetorical purposes in the target language. The results of such an analysis should offer ample material for an evaluation of the translator's work, his success and, if applicable, an explanation of shifts in rhetorical purpose and ideological meaning.

Text linguistics, some more problems
Before venturing into the details of linguistic analysis, a number of complications have to be addressed:

   While emphasizing the universal applicability of their theoretical framework in general, Mason and Hatim have based themselves on an exceedingly small sample of languages, mainly English and Modern Standard Arabic. According to Hatim, some of his findings relating to MSA are directly applicable to the classical literary language as well; nevertheless, he fails to substantiate his claim. Worse, no attempt whatsoever has been made to work with non-contemporary languages inaccessible to the linguistic intuition of literate and capable native speakers (such as Hatim in the case of MSA and English), especially ancient Greek. Therefore, the canon of lexical, syntactical and discoursal markers of rhetorical intentions established by Hatim and Mason for their languages of choice have to be assembled from a number of sources, esp. rhetorical literature on Greek, Syriac and Arabic. In the case of lexis, it should be fairly straightforward to identify at least the most obvious cases of lexical choice and variation motivated by particular rhetorical purposes and semiotic dimensions such as intertextuality. Much work in this field has already been done by Endress (including very detailed analyses of syntactic structures across the languages) and others for both translations and relevant philosophical texts; I hope to be able to transpose some of their findings onto the text samples I have chosen.

   The complex and at some points terminologically confusing framework developed by Hatim and Mason relies too often, as noted above, on the linguistic intuition of capable native speakers. Many factors adduced for the analysis of rhetorical purpose seem to be geared towards a situation in which it is possible to compare notes with both text producer and translator to validate one's findings. With the dearth of 'meta-translational' sources such as the notorious Risála by Hunayn ibn Isháq or contemporary accounts on Aristotelian writings that could shed some light on the intentions behind his lecture notes (the complicated textual history of the Metaphysics cannot but completely rule out any assessment of rhetorical purpose on the strength of textual clues alone), I cannot hope to replicate the level of detail apparent in the sample analyses put forward by Hatim and Mason. Both authors' insistence on the universal applicability of their theoretical model on all sorts of texts complicates the resulting model even further.

   Instead of ruling it out, the problems set out above actually simplify the application of Hatim's and Mason's brand of translation analysis. Taking into account the highly specific nature of the texts we intend to work with and the resulting restrictions on contextual factors and textual phenomena, it seems perfectly legitimate to modify their theoretical framework to suit the particular case I am dealing with, i.e. the Greek-Arabic translation phenomenon. With the obligation to fit all sorts of communicative occurrences out of the way, there seems to be ample scope for paring down Hatim's and Mason's somewhat unwieldy model and shed a number of analytical categories that are either inapplicable to our text or irrelevant to the task of translation assessment I intend to pursue.

   Since I will be dealing with written texts, semiotic categories appear to be especially significant in text analysis. Aspects of discourse, genre and text type, bound to specific instruments and mechanisms of linguistic realization (lexis, syntax, texture etc.), offer a core set of text features analysable and comparable across languages. Additional dimensions such as intertextuality, cultural relativity (e.g. inference, implicature) and communicative variables could provide a second set of findings to test assumptions reached by an analysis centred on semiotic properties of the texts in question. Anchored as they are in readily identifiable patterns of texture and terminology, these analytical instruments could serve as a basis for the translation assessment at the heart of the thesis and as pointers towards shifts in the ideological orientation of the texts bound up with the historical, cultural and intellectual circumstances of the 'translation phenomenon'.

Hoping for results
In keeping with Hatim's stress on contextual dimensions of meaning, a close reading of the texts in question and their terminology should provide ample pointers towards contemporary ideological conflicts. These should be fleshed out and substantiated by placing the texts in their proper literary and historical context, i.e. by relating them to the historical and literary circumstances characterizing the authors' lifetime.

   Supplementing the findings of translation analysis, early philosophical texts written on the basis of the translations in question (e.g. al-Kindí; the Theology of Aristotle) illustrate firstly the changing 'gestalt' of philosophical concepts thus transplanted into an Islamic intellectual environment and secondly their influence and instrumentalisation in contemporary ideological conflicts, illustrated e.g. by lexical choices in the text and other indicators for their historical relativity. Finally, they serve to illustrate the reciprocity of influence between areas of Islamic culture on the one hand and the sample of Greek philosophical thought transferred by the translators on the other hand.

   Obviously, the task of exhaustively examining source texts and translations on the lines proposed above cannot be tackled in the restricted framework of a Ph.D. thesis. To keep the amount of work and data in reasonable boundaries, one could e.g. concentrate on a number of shorter representative stretches of text; additionally, one could selectively choose from a rich supply of analytical instruments those that seem particularly relevant to chosen passages.

References

Beaugrande, Robert-Alain de / Dressler, Wolfgang: Introduction to Text Linguistics
   London (1981)

Farghal, Mohammed: Managing in translation: A theoretical model
   META (1993) 38 (2):257-267

Farghal, Mohammed: Ideational Equivalence in Translation
   Language, Discourse and Translation in the West and Middle East
   edited by Beaugrande, R.-A. de: Amsterdam (1994)

Gutas, Dimitri: Greek Thought, Arabic Culture
   London, New York (1998)

Hatim, Basil and Mason, Ian: Discourse and the Translator
   London (1990)

Hatim, Basil: The pragmatics of argumentation in Arabic: The rise and fall of a text type
   Text (1991) 11 (2):189-199

Hatim, Basil: Rescuing Arabic Rhetoric from Neglect: A Text-Linguistic Approach
   Tradition and Modernity in Arabic Language and Literature
   edited by Smart, J.R.: Richmond/Surrey (1996)

Hatim, Basil: Communication Across Cultures
   Exeter (1997)

Heinemann, W. / Viehweger, D.: Textlinguistik. Eine Einführung
   Tübingen (1991)

Koller, Werner: A linguistic approach to translation: its range and limitations
   Translation Theory in Scandinavia
   edited by Chaffey, P.N. et al: Oslo (1990)

Mason, Ian: Discourse, Ideology and Translation
   Language, Discourse and Translation in the West and Middle East
   edited by Beaugrande, R.-A. de: Amsterdam (1994)

Nord, Christiane: Text Analysis in Translation
   Amsterdam (1991)