Towards a Functional Model for Translation Quality Assessment
2. Theoretical principles
2.2 Textual analysis
For the analysis of both target and source text, and their ensuing comparison, one needs a set of instruments for textual analysis. The framework I adopted derives from an action theoretical approach: texts are considered communicative acts in which a sender attempts to acquire a certain effect on the intended addressee(s). Therefore, the guiding principle in developing the analytic apparatus is the relationship between the intended communicative aim - which is determined by the situational context of this kind of text - and the textual means used to achieve this aim. For this purpose I have tried to establish criteria by which the textual realization of the communicative function can be analyzed on different levels. In this way, phenomena that are often studied in isolation, such as cohesion, thematic structure and coherence, can be inter
related, and aspects on the textual microlevel can be related to macrostructural and communicative notions like ‘function’, ‘text type’ and ‘overall text structure’.
Section 1. Users’ Expectations
2.2.1 The concept of connectivity
After an extensive study of the existing approaches in the field of the applied branch of translation studies, discourse analysis and text linguistics, I have opted for an approach in which the central notion is ‘connectivity’: the assumption is that the text’s organization with respect to structure and content can be looked upon from a functional perspective. Although this concept is close to that of ‘coherence’, I prefer the label ‘connectivity’ because the former notion is often associated with the ‘cohe
sion-coherence’ discussion in text linguistics (cf. De Beaugrande & Dressier 1981), and because, more recently, the term ‘coherence relations’ is increasingly used to refer to relations between sentences and utterances (cf. Hobbs 1983, Pander Maat 1994, Sanders 1992). The multi-layered concept of ‘connectivity’ is made into a workable apparatus, which provides in a way of analysis that proceeds from the macrolevel to the microlevel (the ‘top-down’ approach) and is able to account for the relationship between communicative purpose and textual realization on various levels. This appa
ratus is partly based on the work of the Dutch text linguist Pander Maat (cf. Pander Maat 1994).
To begin with, I have examined the ‘macrostructural’ concept of ‘text type’ with respect to its usability. As it turns out, this concept - that is to say in the way it is currently applied - is not fit to serve as point of departure for the analysis of actual texts, because it is not able to satisfactorily account for the many ways in which the text function can be realized by linguistic means. To solve this problem the notion of ‘text act’ is introduced, which may be considered analogous to ‘speech act’. This notion enables one to bridge the gap between both aspects of the text as a commu
nicative act (i.e. the functional and text internal aspects), because it is possible to establish a distinct relationship between text acts and text function on the one hand, and text acts and the global organization of the text on the other. The introduction of the concept of ‘text act’ enables one to make a clear-cut distinction between func
tion and means, and, at the same time, to make the relationship between these enti
ties visible. For example, a persuasive goal may be achieved by supplying neutral information, leaving the conclusions to the reader, but also - in a more direct persua
sive way - by offering statements which are supported by arguments. In both cases, the intended function remains the same, whereas the textual strategies differ.
2.2.2 Macrostructural connectivity
Linking the concepts of ‘text act’ and ‘global text structure’ - the second step - makes it possible to distinguish between two types of macrostructure, according to the nature of the text act: the ‘topic centred’ structure, in which all utterances have the same value and provide information on the central topic, and the ‘hierarchical’
structure, in which a main act is supported by functionally subordinate utterances.
Both have to be considered ideal types: in actual practice hybrid structures may occur.
Because text acts can be realized in different ways (ranging from implicit to explicit) one has to be careful in automatically linking specific text acts to one of the types of global text structure. Bearing this in mind, however, it is still possible to make cau
tious generalizations. The text act ‘description’, for instance, tends to be linked with a topic centred structure, whereas in the case of a persuasive text act, depending on
Jacqueline Hulst
the degree of directness in presentation, both structures may occur. A directive text act, except in the case of highly implicit realizations, usually is realized by means of a main act and supporting utterances.
Macrostructural connectivity is the product of the relationships between sentences or utterances on the microlevel, the so-called ‘coherence relations’, which are also called ‘rhetoric relations’ (Mann & Thompson 1988). In my approach, there are five types of coherence relations, distinguishing between relations based on elements of content (in the broadest sense of the word) and relations pertaining to the commu
nicative situation. Coherence relations based on elements of content are: descriptive relations, explicatory relations, epistemic relations and adversative or concessive rela
tions. In the first category, i.e. the descriptive relations, both connected utterances are of equal value; in all the other categories a relation of subordination exists between both. The only group of relations pertaining to the communicative situation are those in which utterances providing interactional support are involved, the so-called ‘meta- communicative elements’ such as titles or introductory remarks.
Textual connectivity is not only produced by coherence relations, but also by the links between referents; this is particularly true in the case of topic centred texts.
I distinguish two ways to establish ‘referential connectivity’: the analysis of cohesive links between referents (or situations and referents) on the one hand (roughly based on Halliday & Hasan 1976), and the analysis of the theme-rheme structure on the other (cf. Danes 1974, Gerzymisch-Arbogast 1987). The microlevel links with respect to structure and content between referents and the thematic structure yield a referen
tial macrostructure. Together with the coherence macrostructure, which is the prod
uct of the accumulated coherence relations, the referential macrostructure gives a comprehensive picture of textual connectivity.
Although the set of instruments is mainly devised for the analysis of the structural organization of the text, the central point is the functionality of the means employed.
In establishing the relationships between textual elements (utterances as well as ref
erents) with respect to hierarchical structure and content, it is always examined to what degree they contribute to the realization of the text act in question, and, conse
quentially, of the intended communicative purpose. The choice of textual means, after all, is always based on a certain intention of the sender. Whether or not this inten
tion is realized depends on the adequacy of the means opted for in relation to the intended purpose and the target group.
In this way it is possible, on the basis of the various aspects of analysis as indi
cated in scheme (A) (see appendix), and proceeding as schematically represented in (B), to arrive at a quality assessment of the text’s adequacy. As already stated, the main emphasis is on structure and content, and the way in which the text is phrased is a only secondary factor in the analysis: the phrasing is always examined in relation to the recognizability and acceptability of the relationships concerning structure and content.
2.2.3 Relationship between function and means
Within my approach for translation quality assessment, this analytic apparatus serves as an instrument for comparison: the comparison between target and source texts takes place on all levels of connectivity that have been distinguished earlier.
Section 1. Users’ Expectations
Because the focus is on the relationship between functions and means, it is always examined to what extent differences in textual realization of source and target texts can be related to the intended communicative effect.
The relationship between function and means constitutes the tertium comparationis, the invariant of the comparison, and this tertium is made operational by way of the set of instruments for analysis enabling one to relate the various levels of connectiv
ity to the intended purpose. This basic principle fits in with the nature of the selected corpus, which contains translations aiming at ‘functional equivalence’.
Together the description and the ensuing assessment of both source and target text enable one to relate the possible differences between both texts with respect to the organization of structure and content to the intended communicative purpose.
Some differences may be called ‘neutral’, that is, they do not affect the realization of the text act and the communicative purpose. When differences do affect the realiza
tion of the intended function, they may do so either positively or negatively. Differ
ences with a positive effect might be called ‘functionally adequate’, and they may occur for several reasons. They may be caused, for instance, by the fact that certain obscure source text passages have been improved during translation, or by the fact that the nature of the intended audience necessitated a shift in emphasis with respect to the textual realization.
This goes to show that the occurrence of differences in itself is no indication for the quality of the target text as a translation. In this respect, the clear-cut distinction between function and means employed in the analysis turns out to be very useful:
the fact that a translation aimed at functional equivalence - i.e. a similar communica
tive effect on an intended audience from a different linguistic community - does not automatically mean that the source text serves as the sole standard for the assess
ment. The very point of functional equivalence is that the intention on which the textual realization of the source text is based is conveyed in the target text as well.
With respect to the assessment this means one has to account both for the demands of the target culture and the possibility that the intention has not been realized opti
mally in the source text, in which case the differences may constitute an improve
ment. By taking this into account both perspectives on the target text (as a derivative or an independent product) may be brought into agreement with each other: from either point of view functional adequacy is the standard.
Jacqueline Hulst
References
Beaugrande, R. de, & Dressier, W. 1981. Introduction to Text Linguistics. London: Longman.
Bühler, K. 1934. Sprachtheorie. Jena: Fischer.
Danes, F. 1974. Papers on Functional Sentence Perspective. The Hague: Monton
Gerzymisch-Arbogast, H. 1987. Zur Thema-Rhema-Gliederung in amerikanischen Wirtschafts- fachtexten. Tübingen: Gunter Narr.
Halliday, M.A.K. 1973. Explorations in the functions of language. London: Edward Arnold Publishers.
Halliday, M.A.K. 1978. Language as Social Semiotic: The Social Interpretation of Language and Meaning. Londen: Edward Arnold.
Halliday, M.A.K. 1985. An Introduction to Functional Grammar. Londen: Edward Arnold.
Halliday, M.A.K. & R. Hasan 1976. Cohesion in English. London/New York: Longman.
Hobbs, J.R. 1983. Why is discourse coherent? In: F. Neubauer (ed). Coherence in natural lan
guage texts. Hamburg: Buske Verlag. 29-70.
House, J. 1977. A Model for Translation Quality Assessment.Tüb'mgzrv. Narr.
Hulst, J. 1995. De doeltekst centraal. Naar een functioneel model voor vertaalkritiek. Amsterdam:
Thesis.
Mann, W.C. & S.A. Thompson 1988. Rhetorical Structure Theory. Towards a functional the
ory of text organization. In: Text: an Interdisciplinary Journal for the Study of Discourse 8.
243-281.
Nord, C. 1988. Textanalyse und Übersetzen. Theoretische Grundlagen, Methode und didaktische Anwendung einer Übersetzungsrelevanten Textanalyse. Heidelberg: Julius Groos.
Pander Maat, H. 1994. Tekstanalyse, een pragmatische benadering. Groningen: Martinus Nijhoff.
Reifi, K. 1971. Möglichkeiten und Grenzen dér Übersetzungskritik. Kategorien und Kriterien fiir eine sachgerechte Beurteilung von Übersetzungen. München: Hueber.
Sanders, T. 1992. Discourse structure and coherence. Aspects of a cognitive theory of discourse repre
sentation. Dissertation, Katholieke Universiteit Brabant.
Toury, G. 1980. In search of a theory of translation. Tel Aviv: The Porter Institute for Poetics and Semiotics.
Section 1. Users’ Expectations Appendix:
A. LEVELS OF TEXT ANALYSIS