"[...] words [...] are in and of themselves
metaphorical and therefore imprecise, and cannot be
imagined without ornament"¹.

In this course, the translation process is dealt with in stages, those in which it is usually divided. After a general review of problems and terminology in the first part, in the second we examined perception, assimilation, and interpretation of the prototext from various aspects. In this third part, we deal with the next stage, in which the mental construction deriving from interpretation seeks an outer realization.
  In this actualization stage, it is possible to outline two sub-stages from close-up range. One outline is aimed at expression, the other at cohesion. The translator, having finished her interpretive work, has two needs: first, to externalize the set of impressions caused by the text perception, translate into speech acts the mental material produced by contact with the prototext; second, to make this product coherent within itself, i.e. transform a set of speech acts into a text (the metatext).
  First in a somewhat unaware way a realization, as such, is formed: a sort of unorganized voice, like a chimney flue hosting the smoke rising from a fire out of strictly physical reasons, by draft, by being pulled upwards by heat etc.
  Secondly, this smoke signals - not necessarily belonging to a semiotic system, to a smoke signal code that renders them intelligible - are transformed, from voices isolated from context, into a coherent discourse.
  The reason why the two stages differ from one another is very simple. We know that interpretation concerns a whole text, not single words. The interpretation of a speech act varies according to the context in which it is inserted. But the notion "context" implies a degree of elasticity in the delimitation of concrete contexts considered in single speech acts. Depending on the time available for carrying out a specific translation work, and on the need for interpretive exploration felt by the client or the translator or both, the context considered can vary from a few words to the whole universe of meaning.
  As an example, let us consider the case of a simultaneous interpreter, maybe the one in which the time factor most cruelly demanding. In the best hypothesis, before starting the mediation work, the interpreter is informed generally about the subject of the speeches she will have to orally translate. Consequently the interpreter can tune in on the activation of the senses relevant to that subject and tune out the senses presumably non relevant.
  Apart from this general briefing, the interpreter does not have tools for deep analysis, because the expression of her translation effort often must start even before being able to examine (perhaps "catch a glimpse of" would be better) the prototext as a whole. The context the simultaneous interpreter can consider for the formulation of the metatext is minimal, and the possibility for a re-creation of the metatextual cohesion is minimal too. If we follow a conference or convention through earphones and less than perfectly grammatical, lexical or syntactically cohesive sentences reach our ears among the various elements, we are willing to tolerate these little inconveniences from the start in view of the speed with which the core of the speaker's message is conveyed.
  An opposite example, from some points of view, is that of the philologist preparing the critical edition of a classical text. In this case, time at her disposal abounds. The contextual elements considered span texts that the author might have consulted, or to whose influence he might have undergone, historical facts the author might have known, philosophical doctrines that might have influenced him and so on.
  Not only: the interpretation of the single elements of the whole text can be later reviewed in the light of new conjectures activated by connections between texts made after the initial drafting of the critical edition.
  What is the difference between these two cases, let alone the pragmatic or contingent differences in terms of translation process?
  The difference lies in the dimension of the units the prototext is divided into.
  We stated at the beginning, "interpretation concerns a whole text, not single words". That is to say, interpretation implies the delimitation, although unwitting, of a text, a context, referring to which interpretation is made.
  In physics, there is the notion of "quantum". The main feature of this element is that it is a minimal, but variable entity. Something similar happens with the broadness of the perspective used when dealing with a text. Depending on the possible or desired exploration, the minimal entity of considered text differs but only when a "quantum" of text is reached the possibility of interpreting it becomes real. While reading, the process might be imperceptible but, while listening a simultaneous interpreter, it is easier to recognize it. An interpreter's speech is neither homogeneous nor parallel to the speaker's; it alternatively races and stops short.
  Following this stop-and-go, and its intervening pauses, whether phonically amortized by vocal streams/free-flow and slowdown/accelerations of the discourse, one can realize the limits (beginning and end) of the units the interpreter considers for disassembly in the prototext, one by one, in recoding.
  The segment following the procedure - nearly absent in the simultaneous interpreter for contingent causes, but very significant in the case of written translation - transforms the product of the primary elaboration into a text.   Every culture expresses itself in the form of the texts that are produced in that culture. In a sense, every text is a translation of the culture it is born from, it is an actualization of its culture, as incomplete as any other translation. Any text enriches the context in which it is published because it contains new elements, i.e. alien elements. The innovative contribution of a text consists of the ingredients that differ from what is implicit in the receiving culture.
  If a text comes from an interlingual translation, it often also originates from a geographically or politically different culture. A translator, mediator between two cultures, when choosing the translation strategy decides the degree a text must adapt to the culture and how much a culture must adapt to the text it receives.
  The only fixed point is that, from the linguistic point of view, the text must be adapted to the receiving culture: for this reason the translator is consulted. (But this point is not rigid because an undetermined portion, large or small, of the metatext can be preserved in the original language.)
  As in dreams, there is first a personal recollection of the memory and the impressions (primary elaboration), a personal record not shareable because no other is able to understand the syntax of what is told by the dreamer, unless it is followed by a secondary elaboration transforming such unshaped material in a cohesive text, recountable outside, so it is in translation, after the primary mental elaboration the transformation into text occurs.
  Besides the natural code of the metatext, the degree of adaptability of the text is variable, depending on whether an adequacy or acceptability approach is preferred. The difference lies in the way such secondary elaboration occurs.
  Having to manipulate mental material produced by elaboration of the prototext to make a cohesive text of it, in the adequacy approach the text produced satisfies criteria of maximum prototext description. The metatext, although readable, announces explicitly its identity as metatext, and has a subordinate role regarding the prototext, of which it is a sort of satellite.
  In the acceptability approach, the text produced satisfies the criteria of maximum readability. Tending to hide its identity as a translated text, acceptable translation tries to disguise itself as original, re-creation occurs according to the canons of the receiving culture.
  

  

Bibliographical references

MARÍAS J. Negra espalda del tiempo, Punto de lectura, 2000 (original edition 1998), ISBN 84-663-0007-7.

---

¹ Marías 2001, p. 8. "[...] la palabra [...] es en sí misma metafórica y por ello imprecisa, y además no se concibe sin ornamento [...]". Marías 2000, p. 10.