ArticlePDF Available

Comparing original and translated Spanish: A corpus-based analysis of adjective position

Authors:

Abstract and Figures

It is a well-known fact that translated texts present a number of peculiarities which distinguish its language from the one found in texts produced originally. Many studies have tried to name some of these phenomena, which are usually grouped together under the umbrella term of ‘translation universals’. It has been demonstrated that translations do share a number of features irrespective of the source or target languages involved. Other divergences between original and translated texts are due to source language interference and are, therefore, language-dependent. This paper is a corpus-based study of several highly frequent Spanish adjectives in original texts and in texts translated from English. The unmarked position of attributive adjectives is the pre-modifying one in English and the post-modifying one in Spanish, though. Spanish also allows for the pre‑modifying position with certain connotations. The aim of this study is to identify differences in behavioral patterns with respect to adjective position in original and translated Spanish and explain these differences in terms of translation universals and/or source language interference. The results have revealed cases of simplification, unique item under-representation and untypical collocations in Spanish translations of English source texts.
Content may be subject to copyright.
is is a contribution from Babel 61: 4
© 2015. All rights reserved.
is electronic file may not be altered in any way.
e author(s) of this article is/are permitted to use this PDF file to generate printed copies to
be used by way of offprints, for their personal use only.
Permission is granted by the publishers to post this file on a closed server which is accessible
only to members (students and faculty) of the author’s/s institute. It is not permitted to post
this PDF on the internet, or to share it on sites such as Mendeley, ResearchGate, Academia.
edu. Please see our rights policy on https://benjamins.com/#authors/rightspolicy
For any other use of this material prior written permission should be obtained from the
publishers or through the Copyright Clearance Center (for USA: www.copyright.com).
Please contact rights@benjamins.nl or consult our website: www.benjamins.com
John Benjamins Publishing Company
Babel :  (), –. © Fédération des Traducteurs () Revue Babel
 ./babel...ram  – - –
Comparing original and translated Spanish
A corpus-based analysis of adjective position
Noelia Ramón
University of León
It is a well-known fact that translated texts present a number of peculiarities
which distinguish its language from the one found in texts produced originally.
Many studies have tried to name some of these phenomena, which are usually
grouped together under the umbrella term of ‘translation universals’. It has been
demonstrated that translations do share a number of features irrespective of
the source or target languages involved. Other divergences between original
and translated texts are due to source language interference and are, therefore,
language-dependent. is paper is a corpus-based study of several highly
frequent Spanish adjectives in original texts and in texts translated from English.
e unmarked position of attributive adjectives is the pre-modifying one in
English and the post-modifying one in Spanish, though. Spanish also allows for
the pre-modifying position with certain connotations. e aim of this study is
to identify dierences in behavioral patterns with respect to adjective position
in original and translated Spanish and explain these dierences in terms of
translation universals and/or source language interference. e results have
revealed cases of simplication, unique item under-representation and untypical
collocations in Spanish translations of English source texts.
Keywords: translation universals, corpora, adjective position, simplication,
untypical collocations
1.  Introduction
When reading a Spanish translation of a text written originally in English one of
the rst things that strikes the reader if he/she is a native speaker of Spanish is the
abundance of adjectives in the pre-modifying position. is abundance does not
make the target text illegible, but it certainly detracts from its idiomaticity and
results in an accumulation of untypical word combinations in the translation. is
fact has prompted the current study.
Adjective position is one of the typological features where English and Spanish
dier more clearly, since the unmarked position of adjectives is the
pre-modifying
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
one in English and the post-modifying one in Spanish. is dissimilarity is related
to the dierent origin of these two languages, English being a Germanic language
and Spanish a Romance language. However, the pre-modifying position of descrip-
tive adjectives is possible in Spanish too, with a number of semantic connotations,
involving mainly aectivity. Spanish translations of English texts may therefore
be expected to present a higher rate of pre-modifying adjectives than texts pro-
duced originally in Spanish due to the inuence of the source language English,
where the pre-modifying position is dominant. Previous studies have shown that
this is one important divergence between naturally occurring Spanish and Spanish
translations from English (Rabadán et alii 2009). Adjective position can therefore
be considered one of the main problems in translations between these two lan-
guages, and consequently a key point to take into account in translator training as
well as in translation quality assessment.
In this paper I will provide a detailed analysis of some of the most frequent
adjectives in English and Spanish, with reference in particular to their position in
the noun phrase. e working hypothesis is that there will be a strong trend to
locate adjectives in Spanish translations in pre-modifying positions more oen
than if these same adjectives occurred in original Spanish texts. e empirical
material used for this analysis will be extracted from the English-Spanish parallel
corpus known as P-ACTRES compiled at the University of León, Spain. is corpus
contains 2.5 million words of contemporary English texts and their corresponding
Spanish translations. A variety of registers are represented: ction, non-ction,
press, etc. e CREA corpus (Corpus de Referencia del Español Actual), a large refer-
ence corpus of Spanish, will be used for the comparison with original Spanish texts.
e aim of the paper is to illustrate how quantitative and qualitative factors
such as the overuse of the pre-modifying position of adjectives, among others
(Ramón 2009; Ramón and Labrador 2009), may be used for rating the quality of
Spanish translations and should therefore be considered major aspects in transla-
tor training and in translation quality assessment. e analysis of empirical data
extracted from the translation corpus may also provide a more general insight into
the linguistic features of Spanish translations and translation universals, from a
broader perspective.
.  eoretical background
.1  Trends in translated language
It is generally assumed that translations in any particular language read some-
what dierent from texts written originally in that same language. Many of
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 
these dierences are due to the interference of the source language in the target
language and are therefore, language-dependent; but many other of these pecu-
liarities have been found to occur in translated texts in general, irrespective
of the languages involved in the translation process. All translated texts seem
to share an aura of ‘secondaryness, of being somehow dierent from original
texts. Many scholars have given specic names to this phenomenon calling
translated language ‘the third code’ (Frawley 1984) or using the terms ‘transla-
tionese’ (Gellerstam 1986) or ‘hybrid language’ (Trosborg 1997; Schäner and
Adab 2001).
In the past 20 years or so, the availability of large corpora (monolingual, bilin-
gual, and multilingual) and the development of powerful tools for language analy-
sis have enabled scholars to investigate translated language in great detail and from
many dierent perspectives (Laviosa 2011). Numerous studies have since been
devoted to identifying these common linguistic characteristics shared by all trans-
lated texts simply because they are translations, features called by Mona Baker
‘universals of translation’ in her seminal paper published in 1993. Some of these
characteristics typically associated to translated texts include issues such as explic-
itation, simplication, and normalization. Translated texts have been found to
include additional information and linguistic material to the one present in source
texts, thus making translations more explicit (Olohan and Baker 2000). Anum-
ber of studies have also found that translated texts simplify the target language
to some extent, showing a lower lexical density and a higher frequency of occur-
rence of already common items in the target language (Laviosa 1996, 1998,2002).
Overlapping with this trend towards simplication, we nd a tendency of trans-
lations towards linguistic conventionality or normalization, trying to normalize
grammar, punctuation, dialect or other exaggerated target-language features, thus
avoiding peripheral issues and remaining within the mainstream conventions of
the target language (Toury 1995).
More recently, Mauranen (2008) has described further candidates for becom-
ing universals of translation, as the result of extended corpus-based research
using translation corpora from the early 2000s, and these include the under-
representation of unique-items in the target language (Tirkkonen-Condit 2004),
source language interference, and untypical collocations in the target language
(Mauranen2000, 2004).
In our study on the dierent usage patterns of adjectives in original and trans-
lated Spanish, we will contribute to this line of research by describing the collo-
cational behavior of these particular linguistic items in translations with respect
to original texts. Source language interference and untypical collocations are both
issues that will have to be considered here and could appear in an analysis of the
behavior of adjectives in original and translated Spanish.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
.  Adjective position in English and Spanish
Typological dierences between Germanic and Romance languages are many
and varied, and word order is one of the areas where these dierences are most
obvious. e dierences relating to adjective position in these two language types
are particularly important and have clear implications in the translation process
between English and Spanish, as we will see below. From a morphological per-
spective, adjectives are invariable lexemes in English, whereas in Spanish all adjec-
tives vary morphologically to express number, and many of them also to express
gender. But most importantly, as for adjective position, the unmarked position for
attributive adjectives is the pre-modifying one in English and the post-modifying
one in Spanish. is radical dierence is one important source of error in for-
eign language learning as well as in the translation process involving these two
languages.
Typologically, this dierence is clearly related to the relatively strict word
order we nd in English as the result of a poor inectional morphology. Except
for a few fossilized expressions, the pre-modifying position of adjectives in attrib-
utive function is the only option available in English. In contrast, a very rich mor-
phology in all content words in Spanish allows for a rather freeer word order,
including the possibility to locate descriptive adjectives in pre- or post-modifying
positions:
(1) un hombre pobre vs. un pobre hombre; a poor man’ vs. ‘a poor
(unhappy)man
Descriptive adjectives may occupy both positions within the noun phrase, though
generally with dierent meanings or connotations, mainly involving aectiv-
ity in the pre-modifying position and a neutral meaning in the unmarked post-
modifying position. A few very common adjectives even have morphologically
distinct forms for the pre-modifying position, such as gran for grande, or buen
for bueno. Classifying adjectives, on the other hand, are restricted to occur in the
unmarked post-modifying position in Spanish:
(2) un oso polar vs. *un polar oso; a polar bear’
When is it idiomatic to place a particular descriptive adjective in front of a noun
in Spanish and when is it not so idiomatic, but results in an untypical collocation?
ere is no clear answer to that question. No hard and fast rules exist in Spanish
with respect to the use of the pre-modifying positions of descriptive adjectives.
Phonemics seems to play a role, since mainly short adjectives may pre-modify lon-
ger nouns. It is also common to nd pre-modifying adjectives in original Spanish
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 1
when the head noun also has a post-modier, especially if it is a classifying adjec-
tive, which cannot occupy the pre-modifying position, as in (3):
(3) un gran oso polar; a big polar bear’
When multiple modication is involved, Spanish prefers locating one adjective
before and one aer the head noun, rather than coordinating both aer the head
noun. Despite these general trends, each particular adjective presents its own pat-
tern with respect to dominant pre- or post-modifying positions and this pattern
may vary depending on the head nouns. I will investigate in this paper the pat-
terns of several of the most common Spanish adjectives in texts written originally
in Spanish and, subsequently, compare the use of these same adjectives in texts
translated into Spanish from English source texts to try and shed some light on
this issue.
.  Methodology and data
e aim of this paper is to reveal the collocational patterns of some of the most
common Spanish adjectives with respect to their position in the NP, pre- or post-
modifying position, in original and in translated texts. An overuse of the pre-
modifying position in translated Spanish might be attributed to the inuence of
the source language English, where the pre-modifying position is the only one
available. e empirical data for this study have been extracted from two elec-
tronic corpora: a monolingual reference corpus of Spanish and a parallel corpus
of texts written originally in English and their corresponding translations into
Spanish.
CREA (Corpus de Referencia del Español Actual) is a large monolingual ref-
erence corpus of the Spanish language available online. It contains slightly over
154million words of texts in all registers, oral and written, and in all geographic
varieties of contemporary Spanish, from 1975 until our days. From this large ref-
erence corpus we have selected a subcorpus of texts from the year 2000 on, pub-
lished in Spain only (Peninsular variety), and from the written section only. All in
all there are 18,500,104 million words in our selection of original Spanish texts.
is previous selection is done to restrict the analysis to the contemporary stage
of the language and also to make the results comparable with the ones extracted
from our corpus of translated Spanish, which includes only translations published
in Spain from the year 2000 on.
P-ACTRES is an English-Spanish parallel corpus compiled at the University
of León, Spain, and which currently contains nearly 2.5 million words. e corpus
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
includes texts written in English and belonging to various registers with the corre-
sponding translations into the European variety of Spanish. All original texts were
produced in the year 2000 or later. For more details on the compilation process
of this corpus, see Izquierdo et alii (2008). Table 1 shows the number of words in
each subcorpus by register and language.
Table 1. Register distribution of P-ACTRES
English Spanish Tota l
Books – ction 396,462 421,065 817,527
Books – non-ction 494,358 553,067 1,047,425
Newspapers 115,502 137,202 252,704
Magazines 119,604 126,989 246,593
Miscellanea 40,178 49,026 89,204
Tota l 1,166,104 1,287,349 2,453,453
For this paper I will use the whole of the Spanish translation corpus, so all in all
1,287,349 words of translated Spanish. e analysis will consist of two stages:
a. First, I will carry out a quantitative study of some of the most frequent adjec-
tives in Spanish to determine whether there is a clear under-use or overuse of
any of these adjectives in the translations when compared to the use in origi-
nal texts. is is done to see whether our translations show untypical item
f requencies, as found in other studies (Gellerstam 1996; Laviosa1996).
b. Second, I will carry out a syntactic analysis of a representative number of cases
of some of these adjectives in both corpora to reveal the tendencies in use with
respect to the position of these adjectives within the NP. Dierences in the
distribution of the pre-modifying position in particular could be attributed to
the inuence of the source language English.
e starting point for our analysis was the frequency list in the CREA corpus for
Spanish, which was used to identify the most frequent adjectives in this language.
e list of the 25 most frequent adjectives runs as follows: gran, general, mayor,
nacional, mejor, nuevo, pasado, nueva, social, grandes, posible, importante, nal,
unidos, cierto, largo, claro, español, buena, internacional, igual, española, interior,
buen, especial. e 25 most commonly employed adjectives in English, accord-
ing to the Cobuild Wordbanks Online are: new, good, old, long, little, great, high,
best, big, national, small, full, young, free, public, important, white, local, black, able,
early, political, real, hard, available. We can see that many of them are semantic
equivalents, some even cognates.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 
As for the Spanish list, some of the adjectives included are the short forms of
adjectives for the pre-modifying position (gran, buen), so they are excluded from
our qualitative analysis, as they may never occupy the post-modifying position.
Other adjectives are of the classifying type (nacional, social, internacional), and
must always appear aer the noun they modify, so they are not useful in this study
either. I have selected 5 adjectives from this list – grande, bueno, nuevo, impor-
tante, and largo (‘big, good, new, important, and long’) – to determine the fol-
lowing issues in comparing original and translated Spanish: the overall frequency
of use of these particular lexical items in original texts and in translations, and
the statistical signicance of the dierences identied. e qualitative part of the
analysis consists in the extraction of concordances of 3 common adjectives to ana-
lyze their particular patterns with respect to position in the NP in original and
translated texts.
4. Results and discussion
4.1 Frequency analysis
e frequency analysis of the most common adjectives in Spanish has revealed
interesting dierences between original and translated texts. In this exclusively
quantitative part of the study, all the morphological forms (masculine singular
and plural, and feminine singular and plural, as well as short forms and irregular
comparatives) of the 5 Spanish adjectives selected (grande, bueno, nuevo, impor-
tante, largo) were searched for in CREA and P-ACTRES for raw gures. e aim
here was to identify divergences in the distribution of particular lexical items. It
was found that in all cases there were statistically signicant dierences in at least
one of the morphological forms.
Table 2 shows the number of cases in the two corpora used. Because the cor-
pora vary greatly in size (over 18 million words in original Spanish and 1.5 million
words in translated Spanish), it was necessary to apply a hypothesis test for two
independent proportions to determine the statistical signicance of the dier-
ences found. is test measures the likelihood that the dierences found between
the two corpora are due to chance. Here we have considered statistically signi-
cant only p-values under 0.01, thus allowing for a condence interval of 99%. e
dierences identied in this study can therefore be condently attributed to rea-
sons other than chance, in our case to the fact that one set of data was collected
from original texts and the other from translations. Table 2 also shows the p-value
for each lexical item and whether the dierence is an overuse or an under-use in
the translations.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
4 Noelia Ramón
Table 2. Number of cases and p-value of Spanish adjectives in CREA and P-ACTRES
Cases
CREA
Cases
P-ACTRES p-value Type of divergence
GRAN 11.697 1.047 0 Overuse
GRANDE 1.674 188 0 Overuse
GRANDES 6.135 477 0.019 Not signicant
MAYOR 10.941 712 0.08 Not signicant
MAYORES 2.508 162 0.35 Not signicant
BUEN 3.819 198 0.00005 Overuse
BUENO 1.701 264 0 Overuse
BUENA 4.058 274 0.62 Not signicant
BUENOS 1.117 106 0.002 Overuse
BUENAS 962 108 7e-8 Overuse
MEJOR 9.141 650 0.59 Not signicant
MEJORES 1.939 142 0.55 Not signicant
NUEVO 8.863 360 0 Under-use
NUEVA 6.037 402 0.39 Not signicant
NUEVOS 3.683 177 0.000001 Under-use
NUEVAS 3.602 186 0.00006 Under-use
IMPORTANTE 6.272 368 0.001 Under-use
IMPORTANTES 2.550 169 0.53 Not signicant
LARGO 5.194 321 0.03 Not signicant
LARGA 1.436 126 0.012 Not signicant
LARGOS 506 48 0.03 Not signicant
LARGAS 472 49 0.007 Overuse
We will now see each case in greater detail and try to explain the dierences
identied in terms of the possible translation universals underlying the patterns
found. As for the adjective grande and all its morphological variants, the short
form for the pre-modifying position and the singular form of the post-modifying
position are both overused in translations to a degree which is statistically sig-
nicant. e plural and the comparative forms show small dierences which are
not statistically signicant. e general high frequency of this adjective, which is
the most frequent one in Spanish, makes us think that the strong trend to over-
use it in translations is a feature of the lexical normalization of translated texts
in general. is phenomenon is known as the ‘simplication hypothesis (Baker
1993; Toury 1995; Laviosa 1996) and suggests that translations tend to boost
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 
the use of typical features of the target language, showing a proportional over-
representation of the most frequent lexical items in that language. However, this
adjective will not be a good candidate for the second part of the analysis, as there
are two separate morphological forms for the pre- and the post-modifying posi-
tions in Spanish.
A similar situation is the one observed in the case of the adjective bueno and
all its variants. e short form, the masculine singular, and both masculine and
feminine plural forms are all overused signicantly in Spanish translations when
compared with texts written originally in Spanish. Again this indicates a trend
towards the normalization or simplication of translated language in general,
i.e.,high- frequency lexical items or grammatical or syntactic resources tend to be
preferred as translation solutions at the expense of other target language options,
which would make the translations less at and more lexically varied. is adjective
presents two dierent morphological forms for the pre- and the post-modifying
positions, so it will not be considered in the qualitative part of the analysis, where
we will try to identify cases of overuse of the pre-modifying position in Spanish
translations through the inuence of the source language English.
In contrast to these two cases of overuse, the adjective nuevo and several of
its morphological variants appear to be underused in Spanish translations. e
adjective nuevo is very frequent in Spanish and has a very general meaning appli-
cable to virtually any noun, so the under-use in translations may be due to the
fact that other more precise and less general adjectives are being used instead by
professional translators to avoid homogeneity, repetition or too much simplic-
ity. e adjective new is the most frequent one in English texts, so the under-use
of the formal and functional equivalent in Spanish may not be attributed to the
inuence of the source language, but rather to an attempt to achieve lexical varia-
tion by using other more specic adjectives with a similar meaning to convey the
lexical content carried by the adjective new. is result seems to go against the
simplication hypothesis observed in the previous two cases, but it must be said
that many studies have also found tendencies that oppose simplication, particu-
larly when collocational patterns are considered (Mauranen 2000). Contrary to
other translation universals, in particular explicitation, simplication has not yet
been completely supported or refuted. e analyses of patterns of lexical combi-
nations most oen indicate that translations present untypical trends, suggesting
a wider rather than a narrower variety in the use of the resources available in the
target language. is seems to be the tendency observed in the case of the adjec-
tive nuevo.
On the other hand, because the same morphological forms are possible for
the pre- and the post-modifying positions, this adjective will be a good candi-
date for the qualitative part of the analysis in this paper. A detailed observation
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
of a suciently representative number of cases of this lexical item in original
and in translated texts will reveal whether the pre-modifying position is actually
overused in translations because of the inuence of the source language English
or not.
As for the adjective importante, only two forms have been considered, since
this item varies only in number in Spanish, but not in gender. e singular form
appears to be underused in translations when compared to original Spanish texts
and this under-use is statistically signicant. e formal equivalent in English,
important, is also one of the most frequent adjectives in that language, so the
under-use in the Spanish translations of this source language can only be attrib-
uted, as in the case of the adjective nuevo, to an attempt on the part of transla-
tors to avoid repetition and simplication by using other less general adjectives.
Again we nd evidence against the simplication hypothesis, and again this may
be related to the more varied combinatory patterns in which this particular adjec-
tive seems to be involved.
Finally, I have checked the raw frequencies of the adjective largo and all its
morphological variants in Spanish. e only form with a statistically signicant
overuse is the feminine plural form largas, but it is so infrequent in both cor-
pora that this nding cannot be considered relevant. e ndings indicate that
there is a slight trend towards simplication in the case of the feminine plural
form largas, but in the remaining forms, which are all far more frequent, the
gures are very similar in originals and translations. One possible reason for
this may be the fact that the adjective refers to a physical quality which can be
observed and is therefore not susceptible to interpretation in translations. In
any case, this adjective does not seem to be one that is very prone to cause prob-
lems in translations between English and Spanish with regard to its frequency
of occurrence. However, it may be used in both the pre- and the post-modifying
position, so it is a good candidate for the qualitative analysis in the next section
of this paper.
4.  Pre- and post-modication in translations
e next step in this analysis consists in looking at suciently representative
numbers of concordance lines of dierent adjectives to compare the gures of
pre- and post-modication in originals and in translations. e adjectives grande
and bueno cannot be used for this purpose, as mentioned above, because they
have specic morphological forms for the pre-modifying position. I will there-
fore choose the other three adjectives for this study: nuevo, importante, and largo,
in their masculine singular forms (unmarked gender in Spanish grammar),
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 
which are the most numerous ones in all cases. To be sure to study a suciently
representative number of cases of each adjective in original and translated
Spanish, Ihave used a statistical formula to determine how many occurrences
are required:
2
N
n
(N 1)E 1
=
−+
e element n is the nal sample we will analyze, N is the whole sample of occur-
rences, and E is the estimated error, in this case 0.05 for a 95% condence margin.
Table 3 below shows the number of instances analyzed of each of the adjectives
selected in both corpora, CREA for original Spanish and P-ACTRES for translated
Spanish.
Table 3. Number of cases selected for the study
CREA P-ACTRES
Tota l Selected Tot a l Selected
Nuevo 8,863 383 360 190
Importante 6,272 376 368 192
Largo 5,194 371 321 178
TO TAL 20,329 1,130 1,049 560
4..1  e case of nuevo
A careful analysis of the concordance of the adjective nuevo has revealed that this
lexical item clearly prefers the pre-modifying position in its attributive function,
both in original Spanish texts as well as in Spanish translations. Apart from the
use of nuevo as a single pre- or post-modier I have also distinguished other syn-
tactic combinations, in particular those that refer to multiple pre- and multiple
post-modication (always involving nuevo, of course), or a combination of both
pre- and post-modication. Table 4 shows the number of cases found in the vari-
ous positions with the corresponding percentages to make the data comparable in
both corpora.
It can be seen that nuevo is a predominantly attributive adjective, as
the predicative position is very infrequent (0.5% and 2.6%, in originals and
ranslations, respectively). e two most frequent positions in both corpora are
the pre-modifying positions, with or without additional post-modiers, account-
ing together for 75% of the cases in original Spanish and for 73% in translated
Spanish. Other uses are marginal in both languages, particularly multiple pre- or
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
post-modication. e chi-square test1 was applied to this set of data where pos-
sible2 and the p-value obtained was 0.0071, i.e., the dierences are statistically sig-
nicant. Figure 1 shows the 4 most common functions identied in the analysis,
those with 10% or more of the cases in at least one of the two subcorpora.
44.3
pre- &
post-modif.
single
pre-modif.
xed
expression
‘de nuevo
single
post-modif.
31.8
16.4
5.4
39.4
34.2
7.3 10
0
5
10
15
20
25
30
35
40
45
50
Original Spanish
Translated Spanish
Figure 1. Percentage of occurrence of the various functions of nuevo in original and translated
Spanish
1.  e chi-square test is a statistical tool to determine whether the differences between two
samples of different sizes are significant or not. A p-value lower than 0.01 indicates that the
difference is not due to chance, but must be due to another reason, in this case the influence
of the source language.
.  e test can only be applied when the expected frequency is over 5, so only the four most
common functions shown in Figure 6 were used. e same restriction was followed in the
remaining two cases.
Table 4. Syntactic functions of nuevo in original and translated Spanish
Function in context
Number of cases in
CREA
Number of cases in
P-ACTRES
Pre- + post-modication 170 – 44.3% 75 – 39.4%
Single pre-modier 122 – 31.8% 65 – 34.2%
Fixed expression ‘de nuevo 63 – 16.4% 14 – 7.3%
Single post-modier 21 – 5.4% 19 – 10%
Adjective used as noun 4 – 1.04%
Predicative position 2 – 0.5% 5 – 2.6%
Multiple pre-modication 1 – 0.2% 5 – 2.6%
Multiple post-modication 7 – 3.6%
Tota l 383 190
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 
e frequency distribution of the most common syntactic positions of this adjec-
tive seems to be quite similar in original and translated texts, with only small dif-
ferences in most cases. In original Spanish the most common use of nuevo – 44%
of the cases – is as the pre-modier of a noun which also carries post-modication
of various types: other adjectives, especially of the classifying type as in (4), parti-
ciple clauses as in (5), and relative clauses as in (6), among others.
(4) el nuevo planeamiento urbanístico consistiría en …; the new housing
plan would consist of …
(5) un nuevo caso revelado ayer; ‘a new case revealed yesterday’
(6) un nuevo movimiento que sorteaba la censura; a new movement which
escaped censorship.
e same pattern is also the most frequent one in translations, though slightly less
frequent than in original texts, with 39% of the cases:
(7) un nuevo plan nacional; a new national plan
(8) un nuevo café que acaba de abrir; ‘a new café which has just opened’.
e second most frequent option, single pre-modication, occurs in 31% of the
cases in original texts, as in (9), and in 34% of the cases in translations, as in (10):
(9) el nuevo sistema; the new system
(10) un nuevo conductor; a new driver’.
We notice here a slight overuse of this option, which may be inuenced by the
source language, but the dierence is too small to be considered signicant.
e use of nuevo in the xed expression de nuevo, meaning ‘again’, occurs
in 16% of the cases in original Spanish and appears to be much less frequent
in translations, with less than half the cases (7%). is is a clear case of under-
representation of a typical item in the target language. Because no similar expres-
sion exists in English involving the adjective new to be the source of this collocation
in the target texts, de nuevo is dramatically underused. e chi-square test has
shown that this function is the one that contributes most to the statistical dier-
ence between original and translated Spanish (52.5%).
Finally, single post-modication is an infrequent option in both corpora,
although slightly more frequent in translations (10%) than in original texts (5%).
Again here we may speak of a deviation from the typicality of source language
patterns, but not because of the source language, as the post-modifying position
is not possible in most constructions in English. is case is another example of
an untypical collocation or structural option in translations. Despite the small
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
4 Noelia Ramón
number of occurrences, this deviation is relevant here, contributing 39.2% in the
chi-square test to making original and translated Spanish dierent.
In general, the data reveal that there are not many dierences between the
use of nuevo in originals and translations when considering its position in the NP,
which is predominantly the pre-modifying one, with or without post-modiers.
What is very frequent in originals is slightly less frequent in translations, and fre-
quent combinations in original texts appear under-represented in the translations.
In particular, the low frequency of occurrence of the common expression de nuevo
in translations may be considered an important dierence between both corpora.
Our initial hypothesis expecting a clear overuse of the pre-modifying position is
not conrmed here, mainly because the pre-modifying position is also very fre-
quent in original Spanish for this particular adjective nuevo. ere is indeed a
small overuse of nuevo as a single pre-modier of 3%.
4..  e case of importante
As for the Spanish adjective importante, the analysis has revealed that this adjec-
tive is typically of the post-modifying type, with only rare occurrences in the pre-
modifying position in both original and translated texts. One of the reasons for
this clear trend may be its length, with 4 syllables, which makes pre-modication
by importante sound cumbersome in Spanish. Table 5 shows the number of cases
found in each corpus in each syntactic position. e chi-square test applied to this
set of data revealed a p-value of 0.0001, so the dierences found are statistically
signicant.
Table 5. Syntactic functions of importante in original and translated Spanish
Function in context
Number of cases
in CREA
Number of cases
in PACTRES
Single post-modication 112 – 29.7% 59 – 30.7%
Predicative position 95 – 25.2% 78 – 40.6%
Multiple postmodication 94 – 25% 27 – 14.06%
Pre- + post-modication 50 – 13.2% 24 – 12.5%
Single pre-modication 24 – 6.3% 3 – 1.5%
Adjective used as noun 1 – 0.2%
Multiple pre-modication 1 – 0.5%
Tota l 376 192
ere are important divergences in the use of this adjective in original and trans-
lated texts. e single post-modifying position is the most frequent one in original
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 41
texts, with nearly 30% of the cases, as in (11), and a very similar percentage was
found in translated texts, as in (12):
(11) una decisión importante; an important decision
(12) un riesgo importante; an important risk
However, the predicative position is much more frequent in translations (40% of
the cases, as in (13)), than in original texts (25% of the cases, as in (14)):
(13) es importante recordar aquí …; ‘it is important to remember here …
(14) le afectan numerosos factores y es importante analizarlos; many factors
aect him and it is important to analyze them.
is seems to point towards the inuence of the English source texts. e English
adjective important is also very frequent and may appear more oen in the pred-
icative position than its Spanish cognate, leading to this clear dierence in use. e
predicative position is the biggest contributor to the statistically signicant dier-
ence between original and translated Spanish, with 42.7% of chi-square.
e third most common option found shows that original Spanish texts pres-
ent another 25% of the cases of importante in a multiple post-modifying structure,
where our descriptive adjective occurs immediately aer a classifying adjective:
(15) información estadística importante; ‘important statistical information.
Translated texts present this combination in only 14% of the cases:
(16) un valor simbólico importante; ‘an important symbolic value
What we nd here is a common language pattern which is under-represented in
translations and contributes 30.3% of the chi-square.
e combination of pre- and post-modifying positions occurs in a very simi-
lar number of cases in both original (13%) and translated texts (12%), as shown
in (17):
(17) una importante presencia military; ‘an important military presence’.
As in the case of other adjectives analyzed here, the descriptive adjective occurs in
pre-modifying position mainly because there is a classifying adjective occupying
the position immediately aer the noun.
e remaining positions identied are rather infrequent in both corpora,
although the single pre-modifying position occurs more oen in original texts
(6%) than in translations (1%), in another case of under-representation of a pos-
sibility which is more common in original texts. is result actually contradicts
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
4 Noelia Ramón
our initial hypothesis which expected an overuse of the pre-modifying position in
Spanish translations. e reason in this case is obvious: the length of the adjective
importante, with 4 syllables, plays against its idiomaticity in the pre-modifying
position, up to the point of making it even less common in translations than in
original texts. e single pre-modifying option is rather marginal in both original
and translated texts, with less than 10% in both cases, but because the dierence is
so large, it contributes 26.5% of the chi-square.
Figure 2 illustrates the dierences found in the four most common syntactic
positions of the adjective importante in original and translated texts.
29.7
single
post-modif.
predicative multiple
post-modif.
pre- &
post-modif.
25.2 25
13.2
30.7
40.6
14.06 12.5
0
5
10
15
20
25
30
35
40
45
Original Spanish
Translated Spanish
Figure 2. Percentage of occurrence of the various functions of importante in original and
translated Spanish
e post-modifying position is dominant in original Spanish for the adjective
importante, be it as a single post-modier (nearly 30% of the cases) or in multiple
post-modifying combinations (25%). All in all, 55% of occurrences of importante
in original Spanish are post-modifying positions. In principle, this would make
it a good candidate for a deviation in translations showing an overuse in the pre-
modifying position. However, the data reveal that this is not so. e pre- modifying
position is very infrequent in original Spanish (only 6%), mainly because of the
length of the adjective importante, and this tendency is even more marked in
translations (1.5%), as discussed above.
So where is the dierence then? e dierence between original and trans-
lated texts here lies in the predicative position. In original texts the predicative
position accounts for 25% of the cases, but this position is dominant in transla-
tions with over 40% of the cases. is deviation may only be explained by the
inuence of the source language.
To sum up, in the case of the adjective importante we can see a trend towards
fewer pre-modifying positions in translations than in original texts. Here we nd
many more predicative uses of this adjective in translations than in originals. A
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 4
dierent distribution pattern arises as the consequence of the translation process,
resulting in untypical collocations and syntactic combinations.
4..  e case of largo
Finally, the Spanish adjective largo shows a clear preference for the pre-modifying
position in both originals and translations. Table 6 shows the number of cases in
each corpus with the corresponding percentage of occurrence. e chi-square test
applied to this set of data revealed a p-value of 6.55e-27, so, as in the previous two
cases, the dierences found are statistically signicant.
Table 6. Syntactic functions of largo in original and translated Spanish
Function in context
Number of cases
in CREA
Number of cases
in PACTRES
Fixed expression ‘a lo largo de … 195 – 52.5% 8 – 4.4%
Single pre-modication 85 – 22.9% 74 – 41.5%
Pre- & post-modication 35 – 9.4% 29 – 16.2%
Noun 26 – 7% 24 – 13.4%
Single post-modication 8 – 2.1% 21 – 11.7%
Predicative 8 – 2.1% 6 – 3.3%
Multiple post-modication 7 – 1.8% 14 – 7.8%
Multiple pre-modication 5 – 1.3% 2 – 1.1%
Numeral 2 – 0.5%
TO TAL 371 178
e analysis has revealed that in original texts the adjective largo appears mostly
(52% of the cases) as part of the adverbial expression a lo largo de …, meaning
‘along’, as in Example (18):
(18) a lo largo de la historia; ‘along history’.
is is interesting because only 4% of the cases in translations show this option.
is dierence contributes 56.3% to the chi-square value. Here we come across a
similar case to the one discussed above for the xed expression de nuevo: a clear
under-use of a particular expression typical of the target language. Of course trans-
lated texts are conditioned by their source texts, but translators seem to use other
expressions with similar temporal meanings instead of a lo largo de …, perhaps
because no exact cognate with the word long is available or equally frequent in
English. As in the case of de nuevo, I claim that this is the result of source language
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
44 Noelia Ramón
interference. In the case of largo the deviation is so important that the remaining
syntactic positions will always be far more frequent in translations.
In translations, the most frequent use of the adjective largo is as a single pre-
modier (41%):
(19) su largo cuerpo; ‘his/her long body’
And this is also the second most common use in originals, though with only half
of the cases (22%):
(20) el largo verano; ‘the long summer’.
ere is, consequently, an overuse in the pre-modifying position in translations,
an expected result if we consider the source language English. We may consider
this again a case of untypical collocation in the target language, contributing 11.1%
to the chi-square value between originals and translations.
e combination of pre- and post-modication occurs in nearly 10% of the
cases in original Spanish (21) and in 16% of the cases in translated texts (22):
(21) un largo proceso evolutivo; ‘a long evolutive process’
(22) un largo camino que desciende hacia el río; a long road which leads
down to the river’.
In this case again we can see a trend towards a more frequent use of the pre-
modifying position in translated texts, contributing 3.7% of the total chi-square
value.
e use of the word largo as a noun, mainly in expressions of measure indicat-
ing length occurs twice more oen in translations than in original texts, with 13%
and 7% of the cases, respectively, contributing 4.2% of the chi-square value:
(23) tiene más de 4.500 kilómetros de largo; ‘it is over 4,500 km long’
is phenomenon may be attributed to the inuence of the source language Eng-
lish, where the use of long is required in expressions where the adjective largo may
be idiomatically omitted in Spanish.
Finally, the use of largo as a single post-modier is also much more frequent
in translations (24) than in originals (25), (11% versus 2%), contributing in this
case 16.1% of the chi-square value:
(24) memorizar un poema largo; ‘learn a long poem by heart’
(25) tenían el morro largo; they had long snouts
Similarly, largo in multiple post-modifying structures is also more frequent in
translations (26) than in original texts (27), with 7.8% versus 1.8%, contributing
8.5% to the chi-square value:
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 4
(26) el pelo rubio y largo; ‘long and blond hair’
(27) un trazo largo y recto; a long and straight line’
Both cases are examples of untypical collocations in the target language, since the
data have revealed that native speakers of Spanish clearly prefer the adjective largo
in pre-modifying positions or in the xed expression a lo largo de …
e remaining cases are very infrequent in both corpora. Figure 3 shows in
percentages the most frequent positions in original and translated texts.
‘a lo
largo de’
single
pre-modif.
pre- &
post-modif.
noun single
post-modif.
52.5
4.4
22.9
41.5
9.4 16.2 13.4
72.1
11.7
0
10
20
30
40
50
60
Original Spanish
Translated Spanish
Figure 3. Percentage of occurrence of the various positions of largo in original and translated
Spanish
Figure 3 illustrates very clearly that the ndings are extremely biased by the
dierent frequency of use of the xed expression a lo largo de in original and
translated texts. From a lexical perspective, we may consider this dierence as an
under-use of a unique item in the target language (Tirkkonen-Condit 2004). e
target texts do not become ungrammatical or dicult to read by this dierent dis-
tribution, but it certainly detracts from idiomaticity. e explanation for this fact
seems to be the non-existence of an expression with a similar meaning including
the functional equivalent long in the source language English.
As for the remaining syntactic functions, we may say that the adjective largo
shows a strong trend towards the pre-modifying position, with or without addi-
tional post-modiers in the same NP, both in original and in translated texts. is
means that it is not a good candidate for being overused in the pre-modifying
position in translations, which was our initial hypothesis. e data show that,
apart from the great dierence with respect to the xed expression a lo largo de
, there are no major divergences in the positioning of this adjective in original
and translated texts. In fact, the second largest contributor to the chi-square sig-
nicance value (16.1%) is the single-post-modifying position, which occurs nearly
10 times more oen in translations than in originals, reecting a case of untypical
collocations in the target language Spanish which may not be attributed to the
inuence of the source language English.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
4 Noelia Ramón
.  Conclusions
From a typological perspective English and Spanish are very dierent as far as
adjective position is concerned, English having the pre-modifying position as the
unmarked one, and Spanish having the post-modifying position as the unmarked
one, but admitting the pre-modifying position in descriptive adjectives with cer-
tain semantic connotations. Previous studies have shown a generalized overuse of
the pre-modifying position in translated Spanish (Rabadán et alii2009), something
which may be attributed to the inuence of the source language English. is
paper has tested this initial hypothesis by looking at the patterns of the most fre-
quent Spanish adjectives, in originals and translations.
In this paper I have analyzed raw frequency data and representative numbers
of concordance lines of the most common Spanish adjectives in original texts and
in texts translated into Spanish from English originals. e adjectives analyzed
are grande, bueno, nuevo, important, and largo, with all their morphological vari-
ants in Spanish. e empirical data for the analysis have been extracted from two
large corpora: CREA for original Spanish and P-ACTRES for translated Spanish. A
corpus-based methodology is the most reliable way to conrm or reject our initial
hypothesis of a systematic overuse of the pre-modifying position of adjectives in
Spanish translations. e study has revealed the various usage patterns in original
and translated texts for the most frequent Spanish adjectives, and the dierences
found have been interpreted with the help of statistical tests. For each case, I have
tried to provide explanations for the dierences in terms of translation universals
and/or source-language interference.
e analysis contains two separate parts: (1) a quantitative study for deter-
mining overuse or under-use of the raw gures of the most frequent adjectives
in original and translations, and (2) a qualitative analysis of the actual use of the
most common adjectives in context, in particular with respect to their syntactic
position with respect to the noun they were modifying.
e quantitative analysis revealed interesting dierences between original
and translated language, with a statistically signicant overuse in translations of
the adjectival forms grande, bueno, and largas, but also a signicant under-use of
nuevo and importante in translations. e cases of overuse may all be explained by
the simplication hypothesis, as all these adjectives are highly frequent in Spanish.
e simplication hypothesis states that translations tend to present already com-
mon items in the target language in even higher frequencies of occurrence in
translations. In contrast, the cases of under-use point in the opposite direction,
namely towards an attempt to avoid the homogeneity conveyed by high-frequency
adjectives in the target language.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 4
e qualitative analysis has also yielded very interesting data, although not
in the sense intended initially with this study. e ndings of the empirical analy-
sis have actually rejected our hypothesis that the pre-modifying position would
be statistically more frequent in translations than in original Spanish texts. e
adjectives nuevo and largo show a clear preference for the pre-modifying position
in original Spanish texts already. ere is, indeed, a slight overuse of this position
in translations, but the dierence is either very small, or the number of cases is
too low to make the dierence relevant. In fact, the signicant dierences lie else-
where. In the case of the adjective importante, the pre-modifying position is even
signicantly less common in translations than in original texts!
So the data clearly contradict our initial hypothesis, but other very signicant
dierences have been unveiled contributing to our understanding of translation
universals and source language interference in translated texts. e main nding
refers to the fact that two of the adjectives analyzed (nuevo and largo) are used
less frequently in xed expressions in translations than in originals (de nuevo, a
lo largo de), and this under-use is statistically signicant and, therefore, not due
to chance. is clear trend illustrates a lower degree of typicality and idiomaticity
in translated language, and can be explained in terms of the unique item under-
representation in translated texts. Because these two adjectives appear in common
everyday expressions, it was possible to determine that translators under-use them,
possibly because no expression with a similar meaning exists or is equally frequent
in the source language English including the functional equivalent adjectives in
that language, new and long. Similar studies carried out with texts translated from
a dierent source language where these expressions do exist (for example French),
may yield dierent results, and would then conrm the source language interfer-
ence factor here.
Another relevant nding refers to the behavior of the adjective importante in
our two corpora. is adjective occurs much more oen in the predicative posi-
tion in translations than in original texts and this dierence is statistically sig-
nicant. Because the cognate English adjective important is equally frequent in
that language, we may argue that this dierence in the syntactic position is due to
source language interference. It may well be the case that the adjective important
occurs more oen in the predicative than in the attributive position in English,
and this trend is then replicated in the Spanish translations, as this option also
exists in the target language, though with a much lower frequency of occurrence.
At any rate, we may talk here of a case of untypical collocations in the target lan-
guage, most probably due to source language interference.
An additional dierence found in the data shows that the adjective importante
occurs less oen in multiple post-modifying structures in translations than in
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
4 Noelia Ramón
original texts. Translators show a clear trend to under-use this particular resource
available in the target language, and this fact may be related to the length of the
adjective in Spanish, with 4 syllables, which makes its use rather cumbersome in
multiple modifying constructions. Here we are confronted again with the under-
representation of a unique item in the target language, but also with what seems
to be a trend towards simplication in translations, with fewer complex structures
as modiers.
In conclusion, each adjective shows dierent usage patterns in original texts
and dierent divergences with respect to their translations. e initial hypoth-
esis of an overuse of the pre-modifying position in translations has not been
conrmed by the empirical data of the study, and what’s more, no general trend
can be applied to all the adjectives studied. Traces of translation universals have
been identied, as in the case of simplication or under-representation of unique
target language items, but other pieces of evidence argue against simplication.
Dierent patterns of use with respect to the occurrence of the predicative posi-
tion have been identied, and attributed to the inuence of the source language.
Most commonly, however, untypical collocations have been detected in all cases
as causing statistically signicant dierences between original and translated
Spanish.
is study has focused on a very small number of adjectives, and no general
conclusions can be drawn for the whole class of adjectives in Spanish as such. But
the data very clearly illustrate dierent types of divergences between original and
translated language, from simplication, to untypical collocations or unique item
under-representation. e ndings highlight here the importance of the lexical
factor, as each item behaves in its own particular way. It is the untypical com-
binations resulting from locating one particular adjective next to one particular
noun that oen gives translations their peculiar foreign avor. is fact makes it
extremely dicult to pinpoint any syntactic category or lexical item as the one
responsible for the foreign touch of a translation, as this eect is not due to a
single unit, but to a syntactic position of many dierent lexical items, which vary
greatly from one text to another. In Mauranen’s words, “detailed analyses tend to
show that the behavior of dierent linguistic items is not identical; therefore we
need to consider factors other than overall tendencies of a very general kind.
(Mauranen2008: 40).
e results of this study shed light on the various possible translation univer-
sals existing between Spanish originals and Spanish translations. Untypical collo-
cations were found to be the most common dierences identied in our analysis,
together with simplication or unique item under-representation. ese ndings
may be useful in elds such as translator training or translation quality assess-
ment, where the linguistic quality of the target language is essential.
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 4
Acknowledgements
is paper has been partly funded by the following grants: Ref. LE025A09 (Junta de Castilla y
León) and Ref. FFI2009-08548 (Spanish Ministry of Science and Technology).
References
Baker, Mona. 1993. “Corpus Linguistics and Translation Studies”. In Text and Technology. In
Honour of John Sinclair, ed. by Mona Baker, Gill Francis, and Elena Tognini-Bonelli,
233–250. Amsterdam: John Benjamins. doi: 10.1075/z.64.15bak
Frawley, William. 1984. “Prolegomenon to a eory of Translation. In Translation: Literary,
Linguistic and Philosophical Perspectives, ed. by William Frawley, 159–175. Newark:
University of Delaware Press.
Gellerstam, Martin. 1986. “Translationese in Swedish Novels Translated from English. In Tran s-
lation Studies in Scandinavia, ed. by Lars Wollin and Hans Lindquist, 88–95. Lund: CWK
Gleerup.
Gellerstam, Martin. 1996. “Translation as a Source of Cross-linguistic Studies”. In Languages in
Contrast, ed. by Karin Aijmer, Bengt Altenberg, and Mats Johansson, 53–62. Lund: Lund
University Press.
Izquierdo, Marlen, Knut Hoand, and Øystein Reigem. 2008. “e ACTRES Parallel Corpus:
An English-Spanish Translation Corpus. Corpora 3 (1): 31–41.
doi: 10.3366/E1749503208000051
Laviosa, Sara. 1996. “Comparable Corpora: Towards a Corpus Linguistic Methodology for the
Empirical Study of Translation”. In Translation and Meaning: Part 3, ed. by Marcel elen
and Barbara Lewandowska-Tomaszczyk, 153–163. Maastricht: Hogeschool Maastricht.
Laviosa, Sara. 1998. “Core Patterns of Lexical Use in a Comparable Corpus of English Narrative
Prose”. Meta 43 (4): 557–570. doi: 10.7202/003425ar
Laviosa, Sara. 2002. Corpus-based Translation Studies. eory, Findings, Applications.
Amsterdam: Rodopi.
Laviosa, Sara. 2011. “Corpus-based Translation Studies: Where does it Come from? Where
is it Going?”. In Corpus-based Translation Studies. Research and Applications, ed. by Alet
Kruger, Kim Wallmach, and Jeremy Munday, 13–32. London: Continuum.
Mauranen, Anna. 2000. “Strange Strings in Translated Language. A Study on Corpora. In Inter-
cultural Faultlines. Research Models in Translation Studies I: Textual and Cognitive Aspects,
ed. by Maeve Olohan, 119–141. Manchester (UK): St. Jerome Publishing.
Mauranen, Anna. 2004. “Corpora, Universals and Interference. In Translation Univer-
sals. Do ey Exist? ed. by Anna Mauranen and Pekka Kujamäki, 65–82. Amsterdam:
JohnBenjamins. doi: 10.1075/btl.48.07mau
Mauranen, Anna. 2008. “Universal Tendencies in Translation. In Incorporating Corpora.
e Linguist and the Translator, ed. by Gunilla Anderman and Margaret Rogers, 32–48.
Clevedon: Multilingual Matters.
Olohan, Maeve, and Mona Baker. 2000. “Reporting that in Translated English. Evidence
for Subconscious Processes of Explicitation?” Across Languages and Cultures 1 (2):
141–158. doi: 10.1556/Acr.1.2000.2.1
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
 Noelia Ramón
Rabadán, Rosa; Belén Labrador, and Noelia Ramón. 2009. “Corpus-based Contrastive Analysis
and Translation Universals: A Tool for Translation Quality Assessment English-Spanish?”
Babel 55 (4): 303–328. doi: 10.1075/babel.55.4.01rab
Ramón, Noelia. 2009. “Translating Epistemic Adverbs from English into Spanish: Evidence
from a Parallel Corpus”. Meta 54 (1): 73–96. doi: 10.7202/029794ar
Ramón, Noelia, and Belén Labrador. 2009. “Translations of -ly Adverbs of Degree in an English-
Spanish Parallel Corpus. Target 20 (2): 275–296. doi: 10.1075/target.20.2.05ram
Schäner, Christina, and Beverly Adab. 2001. “e Idea of the Hybrid Text in Translation: Con-
tact as Conict”. Across Languages and Cultures 1 (2): 167–180. doi: 10.1556/Acr.2.2001.2.1
Tirkonnen-Condit, Sonja. 2004. “Unique Items – Over or Under-represented in Translated
Language?” In Translation Universals. Do ey Exist? ed. by Anna Mauranen and Pekka
Kujamäki, 177–184. Amsterdam: John Benjamins. doi: 10.1075/btl.48.14tir
Toury, Gideon. 1995. Descriptive Translation Studies and Beyond. Amsterdam: John
Benjamins. doi: 10.1075/btl.4
Trosborg, Anna. 1997. “Translating Hybrid Political Texts”. In Analysing Professional Genres, ed.
by Anna Trosborg, 145–158. Amsterdam: John Benjamins. doi: 10.1075/btl.26.12tro
Résumé
Il est notoire que des textes traduits présentent un certain nombre de particularités qui dis-
tinguent leur langue de celle que lon trouve dans les textes originaux. De nombreuses études
ont tenté de nommer certains de ces phénomènes, que lon regroupe généralement sous le
terme général ‘universaux de traduction’. Il a été démontré que les traductions partagent eec-
tivement un certain nombre de caractéristiques, indépendamment des langues sources ou
cibles concernées. Dautres divergences entre les textes originaux et les textes traduits sont
dues à une interférence de la langue source et dépendent par conséquent de la langue. Cet
article est une étude basée sur un corpus de plusieurs adjectifs espagnols extrêmement fré-
quents dans des textes originaux et dans des traductions de langlais. Cependant, la position
non-marquée des adjectifs attributifs est une position de pré-modication en anglais et de
post-modication en espagnol. Lespagnol autorise également la position de pré-modication
avec certaines connotations. Lobjectif de cette étude est didentier les diérences dans les
modèles comportementaux en ce qui concerne la place de ladjectif dans les textes originaux
et les traductions espagnoles et d’expliquer ces diérences en termes d’universaux de traduc-
tion ou dinterférences de la langue source. Les résultats montrent des cas de simplication,
de sous-représentation déléments uniques et de collocations atypiques dans les traductions
espagnoles de textes sources anglais.
Mots-clés: universaux de traduction, corpus, place de ladjectif, simplication, collocations
atypiques
About the author
Noelia Ramón is Associate Professor at the Department of Modern Languages of the University
of León, Spain, where she teaches courses on English grammar and specialized translation. She
holds a Ph.D. in Translation (2002) and has published extensively in international journals on
© 2015. Fédération des Traducteurs () Revue Babel
All rights reserved
Comparing original and translated Spanish 1
corpus-based contrastive analysis and translation studies involving English and Spanish, espe-
cially with regard to Translation Quality Assessment (TQA). She has published articles in Meta,
Languages in Contrast, Linguistica Antverpiensia, International Journal of Corpus Linguistics,
Babel and Target. As a member of the ACTRES project, she has been actively involved in the
compilation of the English-Spanish parallel corpus P-ACTRES.
Address: Department of Modern Languages, University of León, Campus de Vegazana s/n,
24071 León, Spain
E-mail: noelia.ramon@unileon.es
La Collection Unesco a pour but de contribuer à l’appréciation mutuelle des
cultures par une aide à la traduction, à la publication et à la diusion d’œuvres
littéraires écrites dans des langues de diusion restreinte. Créée en 1948, elle
compte maintenant quelque 1000 titres représentant environ 80 littératures
diérentes.
Pour tout renseignement:
Collection Unesco d’œuvres représentatives
Division éditoriale et des droits
Éditions UNESCO 1, rue Miollis
75732 Paris Cedex 15
France
Télécopieur/Fax: +33 (0)1 45 68 57 2
Courriel/E-mail: publishing.promotion@unesco.org
http://www.unesco.org/publications
... Language-specific associations between grammatical function and formal resource are defined as "anchor phenomena": "those language-pair specific resources that can be empirically singled out as recurrent problem-triggers in cross-linguistic communication" (Rabadán 2010, 10). In the case of the language pair English-Spanish, the list of anchor phenomena includes the following grammatical items: personal pronouns, modal verbs (Rabadán 2006), adverbs (Ramón and Labrador 2008), verbal tenses (Rabadán 2009), demonstratives (Labrador 2011), negation (Rabadán and Izquierdo 2013), adjective-noun combinations (Ramón 2015), etc. ...
... However, the post-editing procedure benefits from a larger number of quantitative and qualitative anchor phenomena, which cover other grammatical divergences which may occur in this particular language pair. The ACTRES project has proposed a number of guidelines in various areas -all of them developed following a similar procedure to the one described here for pronouns -including modal verbs (Rabadán 2006), adverbs (Ramón and Labrador 2008), verbal tenses (Rabadán 2009), demonstratives (Labrador 2011), negation (Rabadán and Izquierdo 2013), adjective-noun combinations (Ramón 2015), etc. All these data have been used to build PETRA, a computerized tool for the assessment and postediting of Spanish translations of English originals. ...
Article
Full-text available
This paper presents a corpus-based descriptive research procedure for the identification of significant divergences between original Spanish and Spanish translated from English. When considering the language pair English-Spanish, personal pronouns seem to be good markers of significant differences (anchor phenomena), since they must obligatorily occur in English, but not in Spanish. To test this hypothesis, empirical data have been extracted from a large reference corpus in Spanish (CREA) and from an English-Spanish parallel corpus (P-ACTRES), in both cases from the fiction subcorpora. Statistically significant differences have been found in some of the uses of personal pronouns, having textual and pragmatic implications in the target texts. The aim is to use the results obtained in the case of personal pronouns, together with results from other linguistic areas, to build a semi-automated tool for the post-editing of Spanish translations of texts written originally in English.
... Estudios recientes, como los llevados a cabo a partir del corpus ACTRES inglés-español, muestran el interés de este tipo de recursos para la investigación contrastiva y la enseñanza del ELE (cf. Labrador 2015; Ramón 2015). Igualmente, algunos de los trabajos publicados en Doval y Sánchez Nieto (2019) apuntan nuevas posibilidades de uso de los corpus paralelos para el aprendizaje de idiomas. ...
Chapter
El capítulo ofrece un panorama actualizado de la lingüística de corpus aplicada a la enseñanza del español como primera y segunda lengua, atendiendo tanto a la vertiente investigadora como pedagógica. En uno y otro ámbito, se muestra la utilidad de los corpus para los distintos agentes implicados en el proceso de enseñanza-aprendizaje (investigadores, profesores, estudiantes, autores de materiales, evaluadores y formadores de profesores). La exposición pone de manifiesto que el tipo de corpus (oral, escrito, general, especializado, multilingüe, de aprendientes, entre otros) determina las opciones de uso y explotación no solo en lo que se refiere a los agentes involucrados sino también en cuanto al área específica de estudio (nivel fónico, morfosintáctico, léxico y pragmático-discursivo). En relación con este aspecto, se observa que el tratamiento y anotación que han recibido los materiales de los diferentes corpus es de crucial importancia para su aprovechamiento en la enseñanza del español, pues condiciona la información que puede ser objeto de búsqueda y recuperación, algo que resulta relevante desde la perspectiva del corpus como fuente de datos y también para su uso como herramienta de aprendizaje. A partir del examen de los diferentes recursos, investigaciones y materiales desarrollados para la enseñanza del español, el capítulo da cuenta del alcance actual y las direcciones futuras de la lingüística de corpus aplicada a un campo de gran interés y proyección en los próximos años.
... 4 Román (2015), en cambio, no logra corroborar la hipótesis de la sobrerrepresentación de la anteposición en su estudio de cinco adjetivos de alta frecuencia (grande, bueno, nuevo, importante, largo). No obstante, la limitación de dicho estudio a un número tan pequeño de adjetivos hace muy difícil la generalización de sus resultados. ...
Chapter
Esta aportación consiste en un estudio de la colocación del adjetivo en textos literarios en castellano original y traducido. Partimos de la constatación de estudios previos de que los textos traducidos recurren con mayor frecuencia de lo habitual a la anteposición del adjetivo. Algunos de estos trabajos presuponen, además, que la causa de ello se encuentra en la influencia de la lengua de partida, en especial del inglés. A través de un análisis cuantitativo de un corpus comparable de textos traducidos y originales, este trabajo se propone comprobar, por una parte, si se confirma esta tendencia para la combinación alemán-castellano. En un segundo paso cualitativo del análisis, se pone en relación la posición del adjetivo y su categoría semántica, con el fin de indagar en una posible explicación del fenómeno. Los datos obtenidos confirman la tendencia, pero únicamente para adjetivos calificativos de tipo valorativo. Puesto que una de las funciones principales de la anteposición del adjetivo es resaltar su carga valorativa o enfática, llegamos a la conclusión de que, al menos para la traducción de textos literarios, la sobrerrepresentación del adjetivo antepuesto se debe más a la influencia de los patrones de la lengua meta (convencionalización) que a una posible interferencia de la lengua origen. Palabras clave: anteposición del adjetivo, estudios de traducción basados en corpus, traducción literaria, rasgos del lenguaje traducido, interferencia, convencionalización. Abstract This contribution studies the placement of Spanish adjectives in original and translated texts. We start from the observation of previous studies that translators place the adjective before the noun more frequently than is usual in original Spanish texts. Some of these works also presuppose that the cause of this lies in the influence of the source language, especially English. Through a quantitative and qualitative analysis of a comparable corpus of translated and original literary texts, this study relates the position of the adjective to its semantic category. The data obtained confirms the over-representation of pre-noun adjectives, but only for those that qualify as evaluative. Since one of the main functions of the Spanish adjective's anteposition is to highlight its evaluative or emphatic load, the conclusion is reached that, at least for the translation of literary texts, the over-representation of the pre-noun adjective is due more to the influence of the patterns of the target language (conventionalization) than to a hypothetical interference of the source language.
Article
There has been a tremendous increase in the number of published pedagogy-oriented translation and interpreting studies. This increase is associated primarily with establishing a number of international specialized journals. Since developing a typology of pedagogy-oriented translation and interpreting research is an important step in identifying what has been done and what still needs to be done in its areas, this paper reports an attempt to categorize the distinct areas of such research. Reviewing 208 studies published in 11 international journals in a 10-year period (2006–2015), the author identified 6 main types or areas of pedagogy-oriented translation and interpreting research: proposed training, training programme evaluation, assessment, translation and interpreting processes, translated and interpreted products, and professionals’ experiences and perceptions. The author highlights the main issues addressed and the ones left unaddressed in each of these 6 pedagogy-oriented research areas. The paper shows distribution differences in the translation versus interpreting studies within some research areas, and it ends with providing suggestions for further reviews of relevant studies.
Article
Full-text available
The expression of modal meanings is an area of great complexity in the relationship between form and function in a single language and cross-linguistically. Modal adverbs in particular are considered to be problematic from a contrastive perspective due to their multifunctionality in English (Aijmer 2005). This paper is a corpus-based study of the expression of epistemic possibility by means of three common modal adverbs in English (certainly, probably, possibly) and the translational options chosen in Spanish for expressing those meanings. The aim is to identify trends in the translations of these epistemic adverbs that contribute to a better understanding of the various semantic functions that native speakers attribute to these units. The analysis shows that the omission of modal adverbs in the translations may be considered as an indicator of the degree of grammaticalization attained by these adverbs in English. The results provide useful information not only in the field of translator training and practice, but also in descriptive linguistics.
Article
Full-text available
Résumé Cet article étudie la nature linguistique des textes anglais traduits. L'auteur part d'une sous-section du English Comparable Corpus (ECC). La sous-section contient deux sélections de proses narratives en anglais : un premier jeu constitué de traductions à partir de diverses langues sources et un second contenant des textes originaux en anglais rédigés à la même épo-que. L'étude révèle quatre différents modèles d'utilisation lexicale dans les textes traduits par rapport aux originaux. Abstract This paper investigates the linguistic nature of English translated texts. The author' cor-pus consists of a sub-section of the English Comparable Corpus (ECC). It comprises two collec-tions of narrative prose in English: one is made up of translations from a variety of source languages, the other includes original English texts produced during a similar time span. The study reveals four patterns of lexical use in translated versus original texts.
Article
This paper describes the compilation of the ACTRES Parallel Corpus, an English–Spanish translation corpus built at the Department of Modern Languages at the University of Leon (Spain) by the ACTRES research group. The computerisation of the corpus was carried out in collaboration with Knut Hofland and Oystein Reigem, from the Department of Culture, Language and Information Technology, Aksis, at the UNIFOB/University of Bergen (Norway). The corpus is conceived as a powerful tool for cross-linguistic research in the fields of Contrastive Analysis and Descriptive Translation Studies. It was the need to bridge the gap between these disciplines and to extend applications that encouraged the building of a parallel corpus as a suitable tool to achieve these goals. This paper focusses on the practical aspects of building the corpus. A brief account of the research which prompted this endeavour precedes the description of this process.44This paper is an account of the building of the ACTRES Parallel Corpus, so no ...