1“False cognate” or “false friend” are metalinguistic terms used to name ‘‘a word which has the same or very similar form in two languages, but which has a different meaning in each’’. (Platt, Richards & Weber 1985: 103). In a written text, it is the orthographical identity or similarity that can lead a second language learner to use the word wrongly, but when it comes to an oral text, a foreign word may sound familiar to us because of its phonological resemblance to a word in our mother tongue.
2The expression faux amis was used for the first time by Maxime Koessler and Jules Derocquigny in their work Les Faux Amis ou les trahisons du vocabulaire anglais in 1928 (Mounin et al. 1974: 139). Robert and Collins (1990: 296a) refer to these pitfalls as “deceptive cognates”. Other coinages for this lexical phenomenon are faux frères, mots sosies and falschen Freunde, among others. Anglophone linguists frequently use the term “false friend”, a calque from French, and there are purists that refer to them as “false cognates”. In Spanish the expressions falsos amigos and términos equívocos (Cuenca 1987) or palabras de traducción engañosa (Navarro & Hernández 1992, 1994) are generally used indistinctly.
3Several scholars have classified these word pairs differently (Larson 1989, Laufer 1990, Moss 1992, Gläser 1995, Maillot 1997), although a distinction between “total” false friends and “partial” ones has been made. In this regard, we adopted here the definitions and classification on semantic criteria offered by R. Gläser in 1995 that establishes three categories of deceptive cognates:
Total false friends are interlingual word pairs with morphological similarity or even identity, but with a completely different meaning […].
Partial false friends usually cover two subgroups:
1. interlingual word pairs with identical [or similar, we believe] morphological structure, but only a partial correspondence of their meaning.
2. intralingual word pairs which share the same [or similar] morphological structure, but have different terminological meanings in different subject areas or even within the same subject area. The polysemy […] is discarded in the textual environment […]. (Gläser 1995: 17).
4Examples of these different false friend types (with the misleading cognate capitalized) are given in Table A – Total false friends (hereafter T or TFF), Table B – Partial 1 false friends (hereafter P1 or P1FF) and Table C – Partial 2 false friends (hereafter P2 or P2FF).
Table A. Total false friends
|
English
|
Spanish
|
while Spanish
|
English
|
|
carbon
|
carbono
|
CARBÓN
|
coal
|
|
embarrassed
|
turbado, avergonzado
|
EMBARAZADA
|
pregnant
|
|
carbuncle
|
ántrax
|
CARBUNCO
|
anthrax
|
Table B. Partial 1 false friends
|
English
|
Spanish
|
|
drug
|
DROGA / fármaco
|
|
to retract
|
RETRACTARSE / retraer, retirar, replegar
|
|
to divert
|
DIVERTIR / desviar
|
Table C. Partial 2 false friends
|
English
|
Spanish
|
|
marked difference
|
NOTABLE …
|
|
marked improvement
|
NOTORIA …
|
|
marked accent
|
…FUERTE
|
|
marked contrast
|
… MAYOR, or … MÁS ACUSADO
|
|
a marked man
|
… FICHADO
|
5In Table C “marked” is likely to be translated as marcado/a by the inexperienced reader who does not know of the polysemy of the cognate; thus, the wide variety of lexical equivalences/ alternatives (capitalized in the examples) provided by the dictionary.
6The same holds true for the term control (see Table 8). As so aptly stated by Navarro,
en el inglés médico se abusa hasta la saciedad del galicismo ‘control’ […]. Si recurrimos a la traducción control cada vez que en inglés se utiliza esta palabra estaremos empobreciendo nuestro idioma hasta límites insospechados’’. (2000: 110)
7Most FF are of Greek or Latin origin and although these word pairs may share identical or very similar spelling in English and Spanish, for some reason, they have acquired different meanings in the two languages:
“ability” means talent, faculty, power in English while
“habilidad” means skill or cleverness in Spanish
“sensible” means wise, aware, appreciative in English while
“sensible” means sensitive in Spanish
8On the other hand, certain words from extractions other than Greek or Latin have evolved to orthographical forms similar or very close to Spanish lexemes. For instance, the term gripe (Cuenca 1987) derived from Old English gripan (to hold) is likely to be translated into Spanish as gripe (influenza).
9A number of researchers have approached the discussion of these word pairs in English and Spanish general language vocabulary from different aspects: Hill (1982) and Cuenca (1987) with reference to lexicography; Moss (1992) and Lerchundi & Moreno (1999) focused on the importance of cognate recognition in teaching ESP reading courses to Spanish speakers. Several authors have also dealt with this lexical phenomenon with regard to English for Medical Purposes (EMP): Mackin & Weinberger (1970), Smith Kline & French Sae (1979), Navarro & Hernández (1992, 1993), Congost (1994), Divasson (1997) and Navarro (2000), among others. The Diccionario enciclopédico University de términos médicos (1981) also offers a list of FF in medical contexts.
10Nevertheless, research in contrastive studies so far has focused mainly on General English, one of the exceptions being Gläser’s in 1995 on false friends in English and German in LSP vocabulary with special reference to foreign language teaching. She stated that deceptive cognates are also important, ‘‘often more intriguing in LSP lexicology’’ (Gläser 1995: x-xi). But we do not know of any contrastive English-Spanish study dealing comprehensively and systematically with this lexical phenomenon in the field of ME. This paper is thus devoted to the study of deceptive cognates in today’s ME written discourse.
11We gathered a sample that can be considered representative of current ME prose, consisting of twenty research papers, published between 1994 and 1996 in various leading medical journals (The New England Journal of Medicine, The Lancet, Journal of the American Medical Association, etc.), and pertaining to twenty major specialties (for the list of references, see Appendix 1). The corpus is made up of 89,688 running words. This research is part of a wider project carried out at the University of la Laguna, Tenerife, which studies FF in the four ME genres (Editorials, Review Papers, Articles, Research Papers and Case Reports).
The objectives of this paper are threefold:
-
To identify the false friends present in the corpus.
-
To classify them as TFF or partial, and within the latter category as P1FF and P2FF.
-
To determine the word classes they belong to as well as their ranking.
12Our final aim is to elaborate a list of deceptive cognates which could serve in EMP courses and in the training of translators.
13Our analysis is limited to the orthographical properties of the interlingual word pairs that cause interference. Over the years we have gathered an ever increasing list which we use as a Reference List of FF. Most of them became evident in the classroom situation with medical students and postgraduates in Medicine, or when correcting exams. Others have been collected from dictionaries of false cognates in General English (Hill 1982, Cuenca 1987), Medical English dictionaries (Stedman Bilingüe 1999), the above mentioned works of Mackin & Weinberger (1970), Navarro & Hernández (1992, 1994) and Navarro (2000), among others.
14In this article every item of our compiled reference list of FF is considered a lexeme, that is, a word which has an entry in the dictionary. Their multiple inflected or case forms in the corpus are referred to as occurrences.
15Every FF from our reference list has been identified by means of a contextual analysis. Their frequency was recorded and their percentages with regard to the total number of running words were computed. The deceptive cognates found were classified according to two groups of variables: the three FF types (T, P1 and P2) and the four main word classes (the so-called full, lexical or content words) – nouns, verbs, adjectives and adverbs. TFF were easily identified. Conversely, P1/P2FF were difficult to differentiate and we have encountered many borderline cases which have not been included in our analysis.
Table 1. Results: Final number of FF (lexemes and occurrences)
|
TOTAL NUMBER OF FALSE FRIENDS (FF)
|
TOTAL NUMBER OF THEIR OCCURRENCES (OCC.)
|
|
TOTAL (T)
|
73 (18.86%)
|
982 (20.50%)
|
|
PARTIAL 1 (P1)
|
62 (16.02%)
|
730 (15.24%)
|
|
PARTIAL 2 (P2)
|
252 (65.11%)
|
3,077 (64.25%)
|
|
FINAL NUMBER
|
387 (100%)
|
4,789 (100%)
|
Figure 1. Final number of FF (lexemes and occurrences) distributed in types
Table 2. Results: Final number of FF (lexemes and occurrences) distributed in word classes
|
TOTAL NUMBER OF FALSE FRIENDS (FF)
|
TOTAL NUMBER OF THEIR OCCURRENCES (OCC.)
|
|
NOUNS
|
92 (23.77%)
|
814 (16.99%)
|
|
ADJECTIVES
|
182 (47.02%)
|
2,873 (59.99%)
|
|
VERBS
|
96 (24.80%)
|
1,053 (21.98%)
|
|
ADVERBS
|
17 (4.39%)
|
49 (1.02%)
|
|
FINAL NUMBER
|
387 (100%)
|
4,789 (100%)
|
Figure 2. Final number of FF (lexemes and occurrences) distributed in word classes
16There were three times as many P2FF as TFF, and four times as many P2FF as P1FF (see Table 1 and Figure 1). Besides, the total number of noun occurrences was three times more than that of adjectives and verbs and almost 60 times more than that of adverbs (see Table 2 and Figure 2).
Table 3. Overall percentage of T, P1 and P2 FF occurrences with respect to the total number of running words in the corpus
|
CORPUS SIZE (running words)
|
FALSE FRIENDS (occurrences)
|
|
89,688
|
4,789
|
|
100%
|
5.339%
|
Figure 3. Overall percentage of T, P1 and P2 FF occurrences with respect to the total number of running words in the corpus
17Table 3 and Figure 3 illustrate that the 387 FF lexemes gave 4,789 occurrences in the 89,688 running words. The proportion of those occurrences in the biomedical corpus then reached 5.3%. One can thus infer that in a 6,000-word-research paper (the length of about half of those analysed in our corpus), a reader may encounter about 318 FF forms, or an average of 159 if the article were 3,000 words long, which means more than five deceptive cognates in every 100 running words of text). This may imply a considerable risk of misinterpretation of the original article.
Table 4. Distribution of FF types in word classes (number of lexemes)
|
FF
|
Nouns
|
Adjectives
|
Verbs
|
Adverbs
|
|
T
|
44 (24.17%)
|
17 (18.47%)
|
8 (8.33%)
|
4 (23.52%)
|
|
P1
|
33 (18.13%)
|
14 (15.21%)
|
13 (13.54%)
|
2 (11.76%)
|
|
P2
|
105 (57.69%)
|
61 (66.30%)
|
75 (78.12%)
|
11 (6470 %)
|
|
FINAL NUMBER
|
182 (100%)
|
92 (100%)
|
96 (100%)
|
17 (100%)
|
Figure 4. Distribution of FF types in word classes (number of lexemes)
18As we can see in Table 4 and Figure 4, among the 387 FF lexemes identified in the corpus, the number of adjectives and verbs was similar, while there were about twice as many nouns, and only a few adverbs. With regard to the three types of FF, nouns accounted for roughly a 60% of the total number of FF lexemes.
Table 5. Distribution of FF types in word classes (number of occurrences)
|
FF
|
Nouns occ.
|
Adjectives occ.
|
Verbs occ.
|
Adverbs occ.
|
|
T
|
764 (26.59%)
|
149 (18.30%)
|
48 (4.55%)
|
21 (42.85%)
|
|
P1
|
322 (11.20%)
|
214 (26.28%)
|
190 (18.04%)
|
4 (8.16%)
|
|
P2
|
1787 (62.19%)
|
451 (55.40%)
|
815 (77.39%)
|
24 (48.97%)
|
|
FINAL NUMBER
|
2,873 (100%)
|
814 (100%)
|
1,053 (100%)
|
49 (100%)
|
Figure 5. Distribution of FF types in word classes (number of occurrences)
19As can be seen in Table 5 and Graph 5, among the 4,789 FF occurrences found in the corpus, the number of adjectives and verbs was similar, while there were almost three times as many nouns, and again only a few adverbs. Considering the different types of FF, nouns accounted for almost a 60% of the total number of FF occurrences. A list of some examples of T, P1, and P2FF, including the four word classes analysed in this research, is provided below. Appendix 2 presents all the data in detail (Distribution of FF types in word classes in relation to the number of lexemes and occurrences).
Table 6. Example of list of false friends
|
English
|
Spanish
|
while Spanish
|
English
|
|
Nouns
|
|
|
|
|
influenza
|
gripe
|
INFLUENCIA
|
influence
|
|
injury
|
lesión herida, traumatismo, daño
|
INJURIA
|
insult,offence, slanderous allegation
|
|
parent
|
progenitor
|
PARIENTE
|
relative
|
|
physician
|
médico
|
FÍSICO
|
physicist
|
|
Adjectives
|
|
|
|
|
actual
|
real, verdadero, efectivo, concreto, propiamente dicho
|
ACTUAL
|
present, current
|
|
major
|
importante, principal, grave, serio
|
MAYOR
|
greater, larger, bigger, older
|
|
consistent
|
constante, coherente, indicativo
|
CONSISTENTE
|
solid, sound, firm
|
|
eventual
|
final, definitivo, permanente
|
EVENTUAL
|
possible, casual, temporary
|
|
Verbs
|
|
|
|
|
to remove
|
extirpar, quitar, sacar, extraer
|
REMOVER
|
to stir, bring up again
|
|
to stretch
|
extender, estirar, forzar
|
ESTRECHAR
|
to make (...) narrower
|
|
to rest
|
descansar
|
RESTAR
|
to subtract, deduct
|
|
to record
|
anotar, registrar
|
RECORDAR
|
to remember, remind
|
|
Adverbs
|
|
|
|
|
ultimately
|
en última instancia, finalmente, llegado el momento, en el fondo
|
ULTIMAMENTE
|
recently, lately
|
|
currently
|
actualmente, hoy en día
|
CORRIENTEMENTE
|
ordinarily
|
Table 7. Examples of P1FF collected in the sample
20We present in capital letters the Spanish cognate that the inexperienced translator is most likely to use instead of the correct word in a particular context.
|
English
|
Spanish
|
|
Nouns
|
|
|
canal
|
CANAL/ conducto (tubular structure)
|
|
catheter
|
CATÉTER / sonda / drena
|
|
condition
|
CONDICIÓN / afección, enfermedad, proceso
|
|
disorder
|
DESORDEN / enfermedad, trastorno
|
|
lens
|
LENTE / cristalino
|
|
relative
|
RELATIVO / pariente, familiar
|
|
Adjectives
|
|
|
adequate
|
ADECUADO / suficiente
|
|
clinical
|
CLÍNICO / frío or aséptico (unemotional)
|
|
domestic
|
DOMÉSTICO / local or nacional
|
|
dramatic
|
DRAMÁTICO / espectacular / drástico
|
|
aggressive
|
AGRESIVO / dinámico, audaz / traumático
|
|
apparent
|
APARENTE / claro, notorio, evidente
|
|
Verbs
|
|
|
to alter
|
ALTERAR, cambiar / castrar
|
|
to compose
|
COMPONER / calmarse, sosegarse, tranquilizarse
|
|
to discharge
|
DESCARGAR / dar de alta (from hospital) / supurar /despedir
|
|
to prevent
|
PREVENIR / impedir, evitar
|
|
Adverbs
|
|
|
adequately
|
ADECUADAMENTE / suficientemente
|
|
evidently
|
EVIDENTEMENTE /al parecer, por lo visto
|
|
|
21We present in capital letters the word that the inexperienced translator is most likely to use to the detriment of other terms (also capitalized in our examples and taken from Navarro 2000) which would be more appropriate.
Table 8. Examples of P2FF identified in the corpus
|
English
|
Spanish
|
|
Nouns
|
|
|
test
|
TEST
|
|
Apgar test
|
ÍNDICE de...
|
|
blood test
|
ANÁLISIS de...
|
|
diagnostic test
|
PRUEBA de...
|
|
vaginal smear test
|
CITOLOGÍA...
|
|
control
|
CONTROL
|
|
birth control
|
ANTICONCEPCIÓN
|
|
malaria control
|
LUCHA...
|
|
self-control
|
AUTODOMINIO
|
|
control of symptoms
|
SUPRESIÓN ...
|
|
control visit
|
DE REVISIÓN
|
|
control of an epidemic
|
CONTENCIÓN ...
|
|
control group
|
... DE REFERENCIA
|
|
tract
|
TRACTO
|
|
digestive tract
|
TUBO...
|
|
auditory tract
|
VÍAS...
|
|
genital male tract
|
APARATO...
|
|
olfatory tract
|
CINTILLA or PÉNDULO...
|
|
outflow tract
|
...INFUNDÍBULO (of left ventricle)
|
|
posterior tracts
|
CORDONES...
|
|
rubrospinal tract
|
HAZ...
|
|
Adjectives
|
|
|
established
|
ESTABLECIDO
|
|
established custom
|
...ARRAIGADA
|
|
established diagnosis
|
... DE CERTEZA
|
|
established fact
|
...ADMITIDO or DEMOSTRADO
|
|
established staff
|
...FIJO or DE PLANTILLA
|
|
established use
|
...CONSAGRADO
|
|
established name
|
...OFICIAL
|
|
mental
|
MENTAL
|
|
mental ability / work
|
...INTELECTUAL
|
|
mental clarity
|
GRADO DE DE LUCIDEZ or CONSCIENCIA
|
|
mental stress
|
...PSÍQUICA, NERVIOSA
|
|
mental hospital
|
...PSIQUIÁTRICO, MANICOMIO
|
|
mental medicine
|
PSIQUIATRÍA
|
|
systemic
|
SISTÉMICO
|
|
systemic circulation
|
...GENERAL or MAYOR
|
|
systemic disease
|
...GENERALIZADA
|
|
systemic infection
|
...DISEMINADA
|
|
systemic vascular resistance
|
...PERIFÉRICA
|
|
Verbs
|
|
|
to check
|
CHEQUEAR
|
|
to check a hemorrage
|
DETENER...
|
|
to check an epidemic
|
CONTENER...
|
|
to check blood
|
ANALIZAR...
|
|
to check the results
|
VERIFICAR, COMPROBAR...
|
|
to control
|
CONTROLAR
|
|
to control a hemorrage
|
DETENER, RESTAÑAR...
|
|
to control the blood pressure
|
MEDIR, VIGILAR, ESTABILIZAR...
|
|
to control the effects of an overdose
|
NEUTRALIZAR...
|
|
to control the temperature
|
REGULAR...
|
|
to control the anemia
|
CORREGIR...
|
|
to control a crisis
|
YUGULAR...
|
|
Adverbs
|
|
|
entirely
|
ENTERAMENTE, totalmente, exclusivamente, unicamente
|
|
precisely
|
PRECISAMENTE, con precisión, exactamente, minuciosamente
|
22As was expected, some FF are also found in General English. Differentiation of deceptive cognates into P1 and P2 categories has not been an easy task, and we found many borderline instances, a conclusion also reached by other researchers (Gläser 1995). From the previously detailed data one can draw general conclusions about the importance of the different FF types and their relevance according to the word classes. Firstly, FF of the T and P1 types can be expected to be the least common in the biomedical research written discourse (in our corpus, 982 and 730 occurrences respectively of the 4,789 instances found). However, students’ or readers’ overlooking them may have serious consequences for the correct interpretation of the source language text (they usually consist in meaning transgressions). Secondly, although the kind of mistake caused by the misunderstanding of P2FF instances may be much more a kind of inaccuracy (generally misunderstanding of very subtle nuances), its effect may cause a progressive erosion and impoverishment of the target language, and much more so if we consider their much greater frequency (3,077 occurrences in the sample). Therefore, we need to be on the alert for misleading collocations. Thirdly, nouns ranked first in all the frequency lists (close to 60% of the total number of FF occurrences were nouns), followed by adjectives and verbs with a similar distribution, except for the cases of P2 (in which verbs considerably outnumbered adjectives) and T lexemes (in which the opposite holds true). Therefore, regarding word classes, a representative FF reference list for pedagogical purposes should only include a very limited number of adverbs (just 49 occurrences in 89,688 running words), its major goal being the noun class.
23Our findings differ slightly from those of the technical texts analyzed by Lerchundi & Moreno (1999), for whom the most frequent kind of FF was the P1 type (in their work, falsos amigos 2), closely followed by TFF (falsos amigos 1). On the other hand, in their study, P2FF (falsos amigos 3) were the least frequent ones, while in ours the opposite held true. The results of Gläser’s analysis (1995), based on a corpus of ESP text-books from a variety of fields, show a word class distribution quite similar to ours: nouns ranked first, with little difference in the number of verbs and adjectives.
24The results of our analysis lead us to believe that deceptive cognates deserve careful consideration in the teaching of EMP to postgraduates and medical students as well as in the training of technical translators. This is of paramount importance in the Spanish scientist’s case if we take into account that for the last thirty years a great proportion of biomedical Spanish literature has been the product of translations from English, frequently considered quite inadequate (Herranz 1984, Tapia 1991, Navarro & Hernández 1992), false cognates being one of the most common kinds of mistake. Although it is practically impossible to indicate every single source of error, we suggest that more attention should be paid to them in academic syllabi.
25It is thus our contention that some reference tables with the most frequent word pairs and their equivalent, together with well-chosen exercises to practise them, could be of great help if offered at early stages of reading comprehension and translation courses. They may increase the trainees’ – and their teachers’ – awareness of this complex and pervasive problem. The results obtained with some exercises we have designed consisting of identification and translation of contextualized FF are encouraging. In conclusion, similar studies based on other Romance languages could be of interest for the subsequent comparison of results.
Flux RSS
