Section 1: Multilingual databases

International Journal of Lexicography

Volumn 9, Issue 3, 1996, Pages 179-196

  Sinclair, John ^a  

^a UNIVERSITY OF BIRMINGHAM   (United Kingdom)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 29144508938     PISSN: 09503846     EISSN: 14774577     Source Type: Journal    
DOI: 10.1093/ijl/9.3.179     Document Type: Article

References (35)

1. Participants in the project were: M. Gellerstam (University of Gothenburg, Sweden), F. Kiefer (Hungarian Academy of Sciences, Budapest), W. Teubert (IdS, Mannheim, Germany), N. Calzolari and M. Monachini (ILC, Pisa, Italy)
2. M. Bratanic (University of Zagreb, Croatia), A. Ezquerra and G. Corpas Pastor (University of Malaga, Spain) and Helen Liebeck (COBUILD and OUP).
3. The analysis led to an improved understanding of features in the grammatical environment of the English node that can determine the selection from the range of German equivalents.
4. Many of these features have not been taken into account by existing bilingual dictionaries.
5. Features were chosen that are formalisable for possible machine translation.
6. The patterns of usage thus established ensure correct translations of the English words in most cases.
7. The project work contributed to a better comprehension of certain theoretical problems inherent in translation, such as the number of instances in which only an idiomatic expression involving a rephrasing of the whole sentence will render an adequate translation.
8. To filter those instances, a wider section of the environment (in many cases a complete sentence) has to be analysed in regard to its features.
9. Frequency was found to be an important parameter.
10. In cases where a literal translation leads to an acceptable expression that is of a considerably lower frequency than its English counterpart, preference should be given to an idiomatic expression of a comparable frequency.
11. In cases where there is the option of either a literal or an idiomatic translation, text type or genre, envisaged quality levels and parsing effort should be used as guiding parameters.
12. The English concordances were analysed in Pisa by Italian native speakers.
13. It was noticed that sometimes, even though an English word is very polysemous, it is almost always translated by only one Italian word, because it covers the same range of meanings, including the metaphorical ones.
14. Therefore it should be borne in mind that a single translation is not an indication of a single meaning at the monolingual level.
15. There were indicators of the effects of collocation in widening the number of adjectives in Italian which translate English small and are translated by English small.
16. Etymologically, the German equivalent of lend is leihen, of borrow is borgen.
17. Whereas English borrow is the semantic converse of lend and vice versa, German borgen and leihen are often used as synonyms.
18. Both can occur in literary usage, but in spoken language borgen is restricted to Northern German, while leihen can be found everywhere.
19. When used in the sense of borrow, borgen (very frequently) and leihen (frequently) are used as reflexive verbs.
20. When used in the sense of lend, the recipient is expressed as a dative complement.
21. There are subtle differences in the meaning of these verbs.
22. Where the act implies a commercial transaction, leihen and its compounds seem to be preferred to borgen and its compounds, which is also true for many nominal derivatives: Leihhaus (pawnbroker's shop), Leihbucherei (circulating or rental library).
23. Other than that, a systematization of the different shades of meaning connected with the basic verbs and their compounds does not seem possible.
24. Hungarian has five verbs of borrowing/lending which are all derived from the stem kolcson {loan): a) kolcsonoz (borrow/lend) :t{n) =ldna/ldna ut b) kolcsonad (lend) :t{n}=land ut c) kolcsonkap lit (receive as a loan) :t{n) =fa till Idns d) kolcsonker lit (ask for a loan) :t{n) =be att fa Idna e) kolcsoncesz lit (take as a loan) :t{n} =ldna It is only verb (a) which has more than one translation equivalent.
25. It should be noted that the pattern of cases for the lend meaning of (a) is the same as that for (b), and the case pattern for the borrow meaning of (a) is the same as that for (c), (d) and (e).
26. Other dative pronouns, such as ci, vi and low, do not appear in the concordances. This is particularly interesting in the case of loro, which according to the grammar books follows the verb.
27. There are three facts worth noting in the Swedish concordance.
28. First of all, the pattern Idna + OBJECT + OBJECT does not appear at all in the corpus although it is looked upon as a perfectly normal construction in Swedish.
29. Lend+ OBJECT + OBJECT is a very frequent pattern in the English corpus.
30. Second, the meaning 'lend' is invariably signalled by the particle ut (out) in Swedish.
31. It is difficult to say if this is due to sampling differences between Swedish and English.
32. Third, is interesting to notice the general frequency relation between 'give a loan' and 'ask for a loan', a piece of information that is available in Swedish where there is one word for both meanings.
33. Out of 151 concordance lines of Idna there are 32 instances of'giving a loan'.
34. This should be compared with the frequency relation between borrow and lend in English which-according to the COBUILD corpus-is 644 to 538.
35. The higher frequency of 'giving a loan' in English could be explained by the frequent metaphoric use of lend.