One of the characteristics of text was that individual texts resemble other texts and it is this resemblance which is drawn upon by the text-processor in 'making sense' of the text. This knowledge is, clearly, of crucial importance to the language user and any attempt to explain how texts are created and used must include an answer to the question 'How is it, given that each text is unique, that some texts are treated as the same?'
The question – posed in different forms – has, we quickly realize, arisen on no less than three previous occasions in our discussion: (1) in exploring the relationship between utterance, sentence and proposition; (2) in demonstrating the manipulation of syntactic structures to create a range of thematic variations; and, most recently, (3) in defining the notion 'text' itself and will re-appear in relation to the creation of conceptual categories.
The answers we gave earlier are germane to the one we seek now; the key concept is that of a type-token relationship; each individual text is a token – a realization – of some ideal type which underlies it just as the individual proposition underlies a set of clauses which, in their turn, underlie the infinite realizations of the utterance.
Unfortunately, the situation is not so simple. The infinity of utterances derives from a limited number of clause-types which, in their turn, derive from an even smaller number of propositions but, in a very real sense, the individual text is an utterance; a realization of something else. What, though, is this 'something else'? It is an interlinked series of clauses - the forms and order of which are only partially predictable - representing an interlocking series of speech acts (propositional content + illocutionary force) which are also predictable only to a limited extent.
The difficulty derives from the fact that a text-typology has to deal not with 'VIRTUAL SYSTEMS...the abstract potential of languages [but] ...with ACTUAL SYSTEMS in which selections and decisions have already been made' and, further, such a typology 'must be correlated with typologies of discourse actions and situations'.
This immediately calls to mind the form-function dichotomy which has been running through our discussions. Perhaps we could try a formal approach which focused on the topic (the cognitive content; the semantic sense) of the text and, as an alternative, a functional one focused on intention.
1. Formal typologies
Texts have traditionally been organized into informal typologies on the basis of topic – the propositional content of texts – making use of quantitative measures (frequency of occurrence of particular lexical items or syntactic structures) which were thought able to typify 'the language of science' and the like. Such work in register developed into the kind of discourse analysis and ran side-by-side with attempts at rather more ad hoc and intuitive groupings such as 'institutional', 'technical', 'literary' and so forth. In addition, where the typologies were set up as part of a programme of translator-training, they were used as a means of grading texts by ranking them along a scale of 'difficulty' and 'loss' from the extreme of poetry, through other literature, other texts and scientific and technical to mathematical texts which appear to be the least 'difficult' and in which there is virtually no 'loss'.
There are a substantial number of difficulties in working with such a typology but one is immediately obvious and significant. There is a fundamental problem of definition. What is meant by 'poetry', or 'literature' and how are 'scientific', 'technical' and 'mathematical' distinguished? There is, clearly, a substantial degree of overlap which suggests that content, per se, is inadequate as a discriminator. 'Poetry', for example, can presumably be about anything. It is how the poet treats the topic which marks it as 'poetic'. Perhaps, then, it is the formal characteristics (the linguistic structures) which are the defining characteristic. Such an approach will work with some highly ritualized genre (some types of poetry, for example) but not in the case of the majority of texts where again, and now at the formal level, there is overlap. Many of the linguistic characteristics of poetry, for example, recur in non-poetry, e.g. advertising copy. This suggests that a much more sophisticated view of 'topic' is required and this we can find in the notion of domain; the function of the text.
上一篇：New Types of Translation Work Generated by Technology
下一篇：Benefits and Drawbacks of Working with a TMS
微信公众号搜索“译员”关注我们，每天为您推送翻译理论和技巧，外语学习及翻译招聘信息。