1

Preliminaries to translation as a process

This chapter examines translation as a processÐwhat it is the translator actually does. But first, we must note a few basic terms that will be used throughout the course:

Text Any given stretch of speech or writing assumed to make a coherent whole. A minimal text may consist of a single wordÐfor example `Stupendo!'Ðpreceded and followed by a silence (however short). A maximal text may run into thousands of pages.

Source text (ST) The text requiring translation.

Target text (TT) The text which is a translation of the ST.

Source language (SL) The language in which the ST is spoken or written.

Target language (TL) The language into which the ST is to be translated.

Strategy The translator's overall `game-plan', consisting of a set of strategic decisions taken after an initial reading of the ST, but before starting detailed translation of it.

Strategic decisions The first set of reasoned decisions taken by the translator. These are taken before starting the translation in detail, in response to the following questions: `What is the message content of this particular ST? What are its salient linguistic features? What are its principal effects? What genre does it belong to and what audience is it aimed at? What are the functions and intended audience of my translation? What are the implications of these factors? If a choice has to be made among them, which ones should be given priority?'

Decisions of detail Reasoned decisions concerning the specific problems of syntax, vocabulary and so on, encountered in translating particular expressions in their particular context. Decisions of detail are made in the light of the strategy. However, problems of detail may well arise during translating which raise unforeseen strategic issues and oblige the translator to refine the original strategy somewhat.

8 PRELIMINARIES TO TRANSLATION AS A PROCESS

With these terms in mind, the translation process can be broken down into two types of activity: understanding an ST and formulating a TT. These do not occur successively, but simultaneously; indeed, it is often only when coming up against a problem in formulating the TT that translators realize they have not fully understood something in the ST. When this happens, the ST may need to be reinterpreted in the light of the translator's new understanding of it. This reinterpretation sometimes entails revising the original strategy, the revision in turn necessitating changes to some of the decisions of detail already taken. Nevertheless, it is useful to discuss ST interpretation and TT formulation as different, separable processes.

The processes of translation are not different from familiar things that everyone does every day. Comprehension and interpretation are processes that we all perform whenever we listen to or read a piece of linguistically imparted information. Understanding even the simplest message potentially involves all our experiential baggageÐthe knowledge, beliefs, suppositions, inferences and expectations that are the stuff of personal, social and cultural life. Understanding everyday messages is therefore not all that different from what a translator does when first confronting an STÐand it is certainly no less complicated.

In everyday communication, evidence that a message has been understood may come from appropriate practical responseÐfor example, if your mother asks you for a spoon, and you give her a spoon and not a fork. Or it may come from appropriate linguistic responseÐsuch things as returning a greeting correctly, answering a question satisfactorily, or filling in a form. None of these are translation-like processes, but they do show that the comprehension and interpretation stage of translation involves an ordinary, everyday activity that simply requires an average command of the language used.

However, one everyday activity that does resemble translation proper is what Roman Jakobson actually calls inter-semiotic translation (Jakobson 1971:260± 6), that is, translation between two semiotic systems (systems for communication). `The green light means go' is an act of inter-semiotic translation, as is `The big hand's pointing to twelve and the little hand's pointing to four, so it's four o'clock'. In each case, there is translation from a nonlinguistic communication system to a linguistic one. To this extent, everyone is a translator of a sort.

Still more common are various sorts of linguistic response to linguistic stimuli which are also very like translation proper, even though they actually take place within a single language. These sorts of process are what Jakobson (ibid.) calls intralingual translation. A brief look at the two extremes of intralingual translation will show what its major implications are. Take the following scenario. Jill is driving Jack through the narrow streets of a small town. A policeman steps out and stops them. As he leans in to speak to Jill, she can see over his shoulder that, further on, a lorry has jackknifed and blocked the street. At one extreme of intralingual translation lies the kind of response typified in this exchange:

The policeman's essential message is `Turn left'. But he does not want to sound brusque. So he mollifies the driver with a partial explanation, `There's been an accident', and then modalizes his instruction with `I'm afraid you'll have to'. `Down St Mary's Lane' gives a hint of local colour and fellowcitizenship; but he does add `here', just in case the driver is a stranger. Finally, he completes his explanation.

When Jack asks what he said, however, Jill separates the gist of the policeman's message from the circumstantial details and tonal subtleties, and reports it in her own words. This type of intralingual translation we shall call gist translation. The example also shows two other features which intralingual translation shares with translation proper. First, Jill's is not the only gist translation possible. For instance, she might have said `We've got to go down here'. Among other things, this implies that at least one of them does not know the town: the street name has no significance. A third possibility is `We've got to go down St Mary's Lane': if Jack and Jill do know the town, the policeman's gist is accurately conveyed.

The other feature shared by intralingual translation and translation proper is that the situation in which a message is expressed and received affects how it is expressed and how it is received. By `situation' here we mean a combination of three elements: linguistic context (for example, the policeman's words and Jack's question), the non-linguistic circumstances in which speaker and addressee find themselves (such as being stopped in a car and having to take a diversion), and all the experiential baggage they carry with them, all the time (knowing or not knowing the town; familiarity or unfamiliarity with conventions for giving and receiving instructions; liking or disliking the police, etc.). There are always so many variables in the message situation that it is impossible to predict what the gist translation will be or how the addressee will take it. For example, Jill might simply have said `Turn left', a highly economical way of reporting the gistÐno bad thing when she has to concentrate on driving. However, depending on how she says it and how Jack receives it, it could give the impression that the policeman was brusque.

Another reason why `Turn left' could sound brusque is that, grammatically, it looks like direct speech, an imperative, whereas all the other gist translations we have given are clearly indirect speech (or `reported speech'). Now all translation may be said to be indirect speech, inasmuch as it does not repeat the ST, but reformulates it. Yet most TTs, like `Turn left', mask the fact that they are

10 PRELIMINARIES TO TRANSLATION AS A PROCESS

indirect speech by omitting such markers as The author says that¼', or modulation of point of view (as in substituting `we' for `you', or `he' for I). As a result, it is easy for reformulation consciously or unconsciously to become distortion, either because the translator misrepresents the ST or because the reader misreads the TT, or both.

In other words, gist translation, like any translation, is a process of interpretation. This is seen still more clearly if we take an example at the other extreme of intralingual translation. Jill might just as easily have interpreted the policeman's words by expanding them. For example, she could build on an initial gist translation as follows:

We've got to go down St Mary's LaneÐsome fool's jackknifed and blocked the High Street.

This puts two sorts of gloss on the policeman's message: she adds details that he did not give (the jackknifing, the name of the street ahead) and her own judgement of the lorry driver. We shall use the term exegetic translation to denote a translation that explains and elaborates on the ST in this way. The inevitable part played by the translator's experiential baggage becomes obvious in exegetic translation, for any exegesis by definition involves explicitly invoking considerations from outside the text in one's reading of itÐhere, the jackknifed lorry, Jill's knowledge of the town, and her attitude towards other road-users.

An exegetic translation can be shorter than the ST, as in this example, but exegesis is usually longer, and can easily shade into general observations triggered by the ST but not really explaining it. Knowing the town as she does, Jill might easily have gone on like this:

That's the second time in a month. The street's just too narrow for a thing that size.

The explanation added in the second sentence may still just about be admissible as exegetic translation, but it does go much further than the policeman's. If she got a bit more carried away, however, her comment might still count as exegesis, but surely not as translation:

That was another Tory bright idea, letting juggernauts loose on British roads.

As the above examples suggest, it is sometimes hard to keep gist translation and exegetic translation apart, or to see where translation shades into comment pure and simple. It certainly seems very hard to achieve an ideal rephrasing, a halfway point between gist and exegesis that would use terms radically different from those of the ST, but add nothing to, and omit nothing from, its message

THINKING ITALIAN TRANSLATION 11

content. Nevertheless, with its constant movement between gist and exegesis, intralingual translation happens all the time in speech. It is also common in written texts. Students regularly encounter it in annotated editions. A good example is G.K. Hunter's edition of Macbeth, in which the text of the play is followed by about fifty pages of notes. Here are the opening lines of the Captain's report on how the battle stood when he left it, followed by Hunter's notes and rephrasings and, in square brackets, our comments on them:

Doubtful it stood,

As two spent swimmers that do cling together

And choke their art. The merciless Macdonwald

Worthy to be a rebel, for to that

The multiplying villainies of nature

Do swarm upon himÐfrom the Western Isles

Of kerns and galloglasses is supplied [¼].

choke their art make impossible the art of swimming. [An exegetic rephrasing, in so far as it makes explicit what is only implicit in the metaphor: neither army holds the advantage. At the same time, it only conveys the gist, losing the crucial implications of the image of `choking'.] to that as if to that end. [Exegetic rephrasing, explaining Shakespeare's

elliptical formulation.]

multiplying villainies of nature / Do swarm upon him hosts of rebels join him like noxious insects swarming. [Exegetic rephrasing which simplifies, but distorts, Shakespeare's image, turning the metaphor into an explanatory simile: in the ST, the rebels are not like insects, they are villainous manifestations of nature.]

Western Isles Hebrides. [Synonymous rephrasing, for readers unfamiliar with Scottish geography: a good example of how any rewording involves presuppositions regarding the target audience's experiential baggage.]

kerns and galloglasses light and heavy-armed Celtic levies. [Virtually synonymous rephrasing, this time for readers unfamiliar with medieval Irish and Scottish armies. However, `levies' may not be accurate in respect of galloglasses, who were mercenariesÐeven the simplest rephrasing may be misleading, intentionally or not]

(Shakespeare 1967:54, 140)

In all the examples we have been discussing, the dividing-lines between gist, exegesis, translation and comment are blurred. Things could not be otherwise. If one thing has become clear in this chapter, it is the difficulty of controlling (and even of seeing) how far an intralingual TT omits from, adds to or faithfully reproduces the ST message content. And, as we shall see in the next chapter and throughout the course, what applies to intralingual translation applies a fortiori to