English affixal negation translated into Spanish 71
rendering negative meanings. The question arises, however, whether this central-
ity also applies to non-translated Spanish; this will require verification in CREA,
along with the other translational options whose frequency of use is significant
(Section 4.4). The variability of lexical negation has prompted a verification stage
which is lexis-based, so the results obtained are mainly quantitative. As for para-
phrasing, the absence of any regularity whatsoever makes it impossible to identify
any grammatical pattern that can serve as input query in CREA for verification of
target language fit. The information derived from the descriptive stage is qualitative:
there are two main (though not unique) paraphrasing strategies: (i) interpretation,
i.e. an explanation of the translator’s reading that may include negative elements,
and (ii) some type of free transposition, i.e. total or partial transfer of the nega-
tive meaning to other parts of speech (Vinay & Darbelnet 1977) (see Example 24).
Certain semantic changes have been observed when this strategy is applied (see
Example 15).
(24) Not sexless, because he was leanly handsome enough, H-D, and sure of himself
(FBW1E.s310)
No le faltaba atractivo sexual, porque era guapo, dentro de su delgadez, y
además seguro de sí mismo (FBW1S.s308)
Negative –less corresponds to faltaba (lacking in) in the Spanish translation.
Atractivo (attractive) adds further positive characterization which cannot be
derived from the original English text.
Of the two remaining solutions, borrowing, in this analytical context, belongs
rather in the realm of lexical negation since the borrowed items have been used as
units, and there is no evidence that these non-translated items (e.g. topless) have
combined with the affixal capabilities of Spanish.
Omission (Ø) resists any interpretation other than functionally redundant and,
for obvious reasons, will not be considered for target language verification.
At this point, we are interested in whether the translations reflect Spanish
grammatical usage or rather correspond to translation-induced grammatical fea-
tures of the ‘third code’ (Frawley 1984), i.e. a separate variety of the target language
showing characteristics of the original language that do not usually appear in non-
translated language (Øverås 1998).
To make such an assessment, we need to contrast the diagnostic P-ACTRES
findings with data from the CREA control corpus so as to obtain empirical evi-
dence that corroborates or falsifies our working hypothesis, namely that the use of
affixal negation and other negative resources is distributed differently in translated
and non-translated Spanish. To do so, we will start with a quantitative comparison
and then proceed to qualitative interpretations.