"How is it possible to defeat not the authors but the
functions of the author, the idea that behind each
book there is someone who guarantees a truth in
that world of ghosts and inventions by the mere fact
of having invested in it his own truth, of having
identified himself with that construction of
words?"¹.

The lexical aspect of translation-oriented analysis is not to be neglected, even if it is the most self-evident and, therefore, easy to take for granted. The text's markedness is a culture-specific phenomenon, and the translator must consider marked traits of the source culture in order to decide how to translate them in the metatext or, whenever a textual translation is impossible, how to manage the loss.
  Lexical markedness is expressed, for example, by the use of out-of-context lexicon: technicalities outside the relevant technical field, or lexicon belonging to a different register that that adopted in the utterance's co-text, for example, figures of speech rare or unusual in that context, expressions belonging to specific regions within a standard lexicon, sociolects, idiolects etc. There is, moreover, syntactic markedness in which the sentence construction is purposely unusual, peculiar, for the pursuit of given expressive targets.
  Let us take the example of the syntactically standard sentence:

Red wine is good for the prevention of thrombosis.

  And let us see the same sentence in a marked form, with a syntactic dislocation:

Red is the wine that is good for thrombosis prevention.

  In this last sentence, the utterance markedness is used to force the reader's attention toward the red color of the wine, while in the previous standard sentence this detail was left in the background for the most part. It is essential for the translator to catch these aspects; her knowledge of the standard constructions in the source language/culture must be very exhaustive.
  Lexical collocation can also be either standard or marked, and the identification of this aspect requires great sensitivity, because even dictionaries, perhaps due to space constrictions, are rather sparing in words, and only with abundant luck can one find the exact collocation she's searching for among the reported examples. Even when an example is found in the dictionary, there can be no certainty about the spontaneity of the utterance (it could be produced by the dictionary author in function of the example) or on the frequency of use of such collocation. Let us make an example in this case, too.

  He wore a spacious cape.

Confronted with such sentence, an English speaker realizes that it is a peculiar collocation of the adjective "spacious". But the translator most often translates into her own language, and therefore reads the prototext in a foreign language. It is not always possible for her to have the linguistic sensitivity and/or competence necessary to decide beyond any doubt if a given collocation is marked.
  A peculiar collocation like "spacious cape" raises the question of its authority, of authoritativeness of the text. If the case of the writer's linguistic incompetence is to be excluded, it is advisable to think that the unusual collocation is made purposely. Normative considerations are not pertinent here, because, but for some rare examples, the translator is not required to do any corrective work (for which she is not even obliged to have a competence, since that is the editor's job). The standard attitude of a translator tends to be a descriptive one, and any corrective trespassing is to be attributed more to unconscious activity than to a deliberate appropriation of the other's jurisdiction.
  What a translator understands from a collocation like "spacious cape" is that the author wanted to express a differing notion from that conveyed by "wide cape": it therefore is up to the translator to decide how to render such markedness, if by a similarly improbable collocation or by another expressive device.
  Confronted with such doubts, the translator must ascertain whether her suspicions are founded or not in order to decide how to have the metatext reader understand that a peculiarity in syntactic association is present. In most cases, when the translator is not certain of the utterance's markedness, the translation ends up having a standardizing impact, by the suppression of its marked traits: but this is certainly not work well done with regard to either the prototext or the translation reader. The tools useful for the search and checking of such expressive details will be covered in the fourth part of this course.
  Since lexical markedness, like syntactic markedness, is culture-specific, the choices linked to the translation of such elements are not elementary: as Nord holds, a metaphor is not to be necessarily translated into a metaphor and a simile into a simile. This principle would hold true if all languages and cultures were isomorphic, but we know this is not the case.
  Another element that can be marked is the length and the type (affirmation, interrogation, exclamation, ellipsis) of the sentence. This is a culture-specific trait too; for example, it is well known that in contemporary English there is a coherent trend to use short sentences, while other languages - like French, Italian, etc. - tend to more complex and, sometimes, less decipherable structures.
  Within the pertinent cultures, therefore, the translator checks the sentence on its peculiarity - conformity axis, and, in case of peculiarity, checks if it is a regional, social, technical, individual (of the author or of the character) peculiarity etc.
  These are considerations that depend on the translation strategy adopted and on the type of text under examination, but interlingual translation does not necessarily imply an adaptation of sentence type and/or length. Such adaptation is often held as an axiom in practical translation courses, but it must be cautiously considered, and to be applied only whenever the translation has strongly informative and mildly expressive finalities.
  As to suprasegmental features (see about that units 17 and 28 of this second part), it is important to see what the impact on written texts is of this particularly evident feature in oral texts. Possible examples are found in the use of onomatopoeias, peculiar spellings (for instance the use of double letters, both vocal and consonants, to indicate a peculiar pronunciation), use of punctuation signs. Punctuation, too, has culture-specific norms, but transgression is more or less admitted even in cultures where rules are rigidly followed.
  Nord uses as examples of written suprasegmental features, comparing two sentences:

  John, Peter, Mary, Paul were there.

  and

  John and Peter and Mary and Paul were there².

  The simple choice preferring an asyndetic enumeration (with commas in this case) gives the sentence a stronger rhythm than polysyndetic construction (in this case with the "and" conjunctions) in which the reader dwells longer, in mental reading as well, on single items. Among other things please note that probably both quoted examples are marked in most cultures. because the standard form is often promiscuous, like this:

John, Peter, Mary and Paul were there.

Bibliographical references

CALVINO I. If on a Winter's Night a Traveller, London, Random House, 1998, ISBN 0-749-39923-6.

NORD C. Text Analysis in Translation. Theory, Methodology, and Didactic Application of a Model for Translation-Oriented Text Analysis, translated from the German by C. Nord e P. Sparrow, Amsterdam, Rodopi, 1991, ISBN 90-5183-311-3.