Contragram

Newsletter 8

No 8
December - 1996

CONTENTS

To be or not to be after believe/croire-like verbs

Dirk Noël & Bart Defrancq

In English as well as in French there is a category of verbs -- which we could, very roughly, call 'opinion verbs' -- whose members have in common that they can be complemented by either a finite or an infinitival clause. Members of the English category are assume, believe, consider, find, judge, know, suppose and think.¹ Members of the French category are considérer, croire, dire, espérer, juger, penser and savoir. Here are a few examples:²

(1) a. I believe that music-lovers are deluded when they claim to find artistic pleasure in any but a fraction of this music.
(1) b. The Ocean is calm, so calm you could believe it to be lacquered wood rather than water and that if you were in a hurry you could leave the boat and walk to shore.
(2) a. Je crois que l'histoire de la Révolution enseigne que la violence exercée par le parti de la liberté finit toujours par se retourner contre la liberté.
(2) b. Un chômeur aux abois dévalise une banque et prend en otage un truand juste libéré que la police croit être son complice.

In English the infinitival option is virtually restricted to cases in which the complement verb is either stative or perfective, the dominant verb being to be. The infinitival complement also does not allow the addition of modal auxiliaries and is less tolerant of modification by adverbials and additional subclauses than that-complements. On the other hand, the infinitive is reported to be the preferred choice in relative clauses like (3) because the catenation of finite verbs as in (4) would make such structures difficult to process (Mair 1993).

(3) "The picture shows a man who we believe to be the double burglar cuddling his girlfriend," said a spokesman.
(4) They had one daughter, who I believe is not right in the head; but that may not be the one you met.

Elsewhere (in Noël 1996) we have questioned this explanation for the choice of an infinitive over a finite complement in terms of processing difficulty for three reasons: (a) there is no convincing psycholinguistic evidence that sentences like (4) would be more difficult to process than those like (3); (b) a sequence of finite verbs is not all that uncommon in English; and (c) the explanation does not account for the choice between the two patterns in syntactic environments other than (4), where the use of an infinitive does not lead to the immediate juxtaposition of two finite verbs. Noël (1996) reports on a corpus investigation of about 100 contextualized instances each of believe that X is and believe X to be. This revealed two things: (a) everything else being equal (i.e. in cases when the above-mentioned restrictions do not apply) the infinitival option seems only to be available when the subject of the infinitive is textually 'given', and (b) the infinitival pattern occurs more often in subordinate clauses than the finite pattern. The latter fact might be indirect evidence that infinitives figure less prominently in texts than that-clauses, or, conversely, that the choice for a finite complement makes the proposition expressed by the complement more textually 'salient', if indeed it is true that propositions expressed by subclauses are less central to the text producer's goals (cf. Matthiessen and Thompson 1988). The choice between the finite and the non-finite pattern might therefore be textually motivated and this might ultimately explain the formal restrictions on the infinitival complement: if the propositions they express are less central to the text producer's goals (and -- therefore? -- do not introduce new referents), they need to be less specific, and the virtual ungrammaticality of, for instance, adverbial modifiers with infinitival complements may have resulted from this.

It would be interesting to explore whether the variation between the finite and the non-finite patterns after verbs like croire in French is susceptible of a similar explanation. The verb + NP + infinitive pattern in which the subject of the infinitive is not coreferential with the subject of the matrix verb is much less widespread in French than in English, however.³ Grevisse (1993: 1278) informs us that the French infinitival pattern is restricted to written language and mainly occurs "quand le sujet est le pronom relatif que" and a quick search in the le Monde corpus indeed revealed that combinations of croire-like verbs and the infinitive être only occur in cases like (5) and (6), i.e. in both adnominal and nominal relatives, though mainly the latter.

(5) Les militants de l'arrondissement auraient préféré comme chef de file l'un d'entre eux, le conseiller de Paris JeanLuc Gonneau ou le conseiller d'arrondissement Claude Pigement, plutôt que Mme Macle, qu'ils estiment être une candidate "parachutée".
(6) C'est la confiance en soi, non pas fondée sur l'orgueil, la vanité ou l'autosatisfaction, mais la confiance que l'on a dans sa détermination à accomplir ce que l'on considère être son devoir.

Such examples are perhaps not amenable to a textual explanation like the one offered above for English, but Grevisse also lists examples "où le sujet est un nom ou un pronom personnel":

(7) Il jugeait cette récréation lui devoir être profitable. (Flaubert)
(8) Elle [= une certaine "maison du Greco"] a été construite, il y a longtemps, par un admirateur du maître, sur l'emplacement de la vraie maison où on le sait avoir vécu. (Henriot, dans le Monde)
(9) Je les crois valoir d'être connues. (Yourcenar)
(10) Pour qu'on la [= la licorne] crût détenir les vertus surnaturelles qui font la matière de plusieurs récits. (Caillois, dans le Monde)

Interestingly, all the subjects of the infinitives in these (one-sentence) examples seem to be anaphoric, and this does warrant further investigation of contextualised instances of them... if we can find them. We'll report back later.

Notes

1. For a fuller list, see Francis et al. (1996).
2. Unless otherwise indicated the English examples are from the British National Corpus and the French examples from Le Monde sur CD-ROM.
3. An explanation that has been adduced for this is that in French the infinitive with non-coreferential subject is in competition with an infinitive whose (implied) subject is the same as that of the matrix verb, as in Après coup, j'ai cru comprendre que c'était la technique qui rapprochait les deux images, whereas in English the latter kind does not exist, excepting cases where the subject of the infinitive is a reflexive pronoun, as in Each village believes itself to be totally different from any other and often marks itself off in a variety of symbolic ways from those which surround it.

References

  • Francis, G., S. Hunston and E. Manning (1996) Collins COBUILD Grammar Patterns 1: Verbs.London: HarperCollins.
  • Grevisse, M. (1993) Le Bon Usage: Grammaire française. Refondue par A. Goosse. Treizième édition revue. Paris: Duculot.
  • Mair, C. (1993) A crosslinguistic functional constraint on believe-type raising in English and selected other European languages. Papers and Studies in Contrastive Linguistics 28: 5-19.
  • Matthiessen, C. and S. A. Thompson (1988) The structure of discourse and 'subordination'. In J. Haiman and S. A. Thompson (eds.) Clause Combining in Grammar and Discourse . Amsterdam: Benjamins. 275-329.
  • Noël, D. (1996) The "meaning" of believe X to be vs. believe that X is. Paper presented at the 6th Odense Valency Seminar, University of Odense, 26-27 November 1996.

[table of contents]

Current issues in contrastive linguistics Dutch-French

L. Beheydt & J.-P. Colson (Louvain, Unité d'Etudes néerlandaises )

Deceptive cognates Dutch-French

Deceptive cognates or "faux amis" have received much attention from comparative linguists. They can be defined in the strict sense as "pairs of words which are etymologically related, similar in form but semantically divergent." (Granger & Swallow 1988, 108), e.g. E. actually and F. actuellement. In the case of French and English, much attention has been paid to the existence of faux amis and their influence on foreign language learning (The 'Dictionary of false friends' by Van Roey, Granger, Swallow is a typical example here).

As far as Dutch and French are concerned, this problem has not yet been thoroughly investigated. Apart from a few interesting articles in the seventies and eighties, there has been no attempt, so far, to give an overall view of the problem, nor has there been made a list of frequent deceptive cognates Dutch-French.

The Unité d'Etudes néerlandaises has already carried out some preliminary research on this subject (Beheydt 1995; Colson 1992, 1996; Burnon 1990; Depouhon 1991, Wéry 1996). A first corpus of about 600 false friends has been delineated, and a few tentative results can already be mentioned.

Following the typology for deceptive cognates French-English (Van Roey and Granger), the different false friends can be classified into four main categories, which show approximately the following distribution:

INCLUSION D/F              30 %
INLCUSION F/D              24 %
OVERLAP                    30 %
TOTALLY DECEPTIVE          16 %

A striking result of the first analysis of the corpus is that the second category (inclusion of French into Dutch) is quite frequent, viz. 24 % of the cases. In comparison, the inclusion of French into English represents only about 10 % of the cases in Van Roey and Granger's Dictionary. In other words, it often occurs that a word borrowed from French into Dutch develops new meanings of its own in Dutch. If these results are confirmed, they seem to show the dynamic qualities of the Dutch language, borrowing a lot from neighbours but always adapting and adding new elements.

This kind of comparison between deceptive cognates French-Dutch and deceptive cognates French-English is a kind of contrastive analysis at a second level. Of course, the first impression yielded by the results needs further investigation.

From a socio-linguistic point of view, deceptive cognates French-Dutch illustrate the complex lin guistic situation of Belgium. For many pairs of words, it is indeed necessary to make a distinction between four language varieties: standard Dutch (Netherlandic), southern Netherlandic (Flemish), standard French and Belgian French. In many cases, the Dutch word borrowed from French is a deceptive cognate in standard Dutch but not in the Flemish usage: "conducteur, quasi, econoom". In other cases, something like a Belgian usage seems to exist. For the sea front, Flemish and Walloons use the same word, respectively "de dijk" and "la digue", whereas a Duchtman would use the deceptive cognate "de boulevard", and a Frenchman "le front de mer". More generally speaking, the study of deceptive cognates French-Dutch reveals that the two main languages of Belgium tend to influence each other and to suppress a few differences that are very subtle, such as those holding for deceptive cognates. An example taken from French is "les files", which is only used in Belgium for traffic jams, under the influence of the Dutch word "files". Without adopting the point of view of a purist, one might say that this kind of suppression of subtle differences between European languages, which is favoured by the European institutions, represents some threat to the integrity of French or Dutch.

Another type of deceptive words that has been analyzed in our research unit has to do with the so-called academic vocabulary (Beheydt and Charles 1995). By this, we mean the specific vocabulary thought to be required for fruitful studies in third cycle education (university and higher professional education). This vocabulary was selected using criteria which had been established by a.o. Nation (1990). It contains a lot of words and phrases that may lead to errors with non-natives, including cases of false friends, such as "quasi, catastrofaal, theoretisch", etc. The errors made by students as far as academic vocabulary is concerned fall into two main categories: errors due to diverging connotations between French and English, and errors due to incorrect morphological endings.

Another frequent type of errors made by the students has to do with the divergent structural relations holding for French and Dutch verbs, as in the case of "considérer" vs. "beschouwen":

(1a) Nous considérons que cette solution est erronée.
(1b)* Wij beschouwen dat deze oplossing fout is.

This example reveals an interesting common ground between contrastive valency and false friends. D. Noël (1996) gives an overview of the valency differences between "considérer", "beschouwen" and "consider". But errors such as (1b) might also be explained as a case of structural false friends.

Among such structural differences that often lead to errors, it is sometimes possible to find a certain regularity. Typical examples are verbs that are transitive in one language and intransitive in another:

(2a) Nous écoutons la radio.
(2b) * Wij luisteren de radio.

Comparable frequent differences concern the causative use of certain verbs:

(3a) The congress opens with.. .
(3b) Het congres opent met .. .
(4a) The president opens the congress.
(4b) De voorzitter opent het congres.
(5a) He grows flowers.
(5b) Hij kweekt bloemen.
(6a) The flowers grow.
(6b) De bloemen groeien.

In these examples, there is a one-to-one equivalence in Dutch and English causative use for senten ces (3) and (4) but not for (5) and (6).

Structural differences between French and Dutch are also illustrated by the reflexive use of some verbs:

(7a) Hij klaagt.
(7b) Il se plaint.
(8a) Hij zwijgt.
(8b) Il se tait.

Sometimes, there is a different structural use between the reflexive French verb, indicating a possibility, and its Dutch counterpart "zijn te" followed by an infinitive:

(9a) Cela s'explique facilement.
(9b) Dat is gemakkelijk te verklaren.
(10a) Une telle abondance ne se rencontre nulle part.
(10b) Zulke overvloed is nergens te vinden.

Apart from these cases, some structural differences between verbs are explainable in semantic terms. The French "faire" receives a lot of translations, but it would certainly be interesting to find a rule governed choice pattern for the corresponding "doen"/ "maken"/ "to do"/ "to make"/ "tun"/ "machen". Some work has already been done in this respect, e.g. by D. Nehls (1991).

A similar semantic choice-pattern could be attempted for the use of modal verbs. There is indeed a problem with the Dutch equivalent of such French verbs as "pouvoir" or "devoir". J. Weisshaupt (1980; 1993) has already tackled this problem.

A contrastive study of causation

Another domain of contrastive linguistics in which research is being done at the Unité d'Etudes néerlandaises concerns the equivalence of causation in various languages (mainly Dutch, French and English). This problem has been investigated for a couple of years by Liesbeth Degand and will be published as a Ph. D. in Dutch linguistics at the Université catholique de Louvain under the title "A situation-based approach of causation in Dutch, with some implications for text generation."

It is mainly an attempt to work out one aspect of Natural Language Generation (NLG), viz. causation, within Systemic Functional Linguistics (SFL), in such a way as to be able to correctly implement the choice of causative connectors (F. "puisque, parce que, car....; D. want, omdat, aangezien...") in different languages. This combined approach in NLG and SFL should eventually lead to an ade quate selection in automatic translation.

A contrastive approach to Dutch idioms

Idioms remain a major stumbling-block in the acquisition of Dutch by French-speaking students. J.-P. Colson has already published a few articles on fraseology and is directing student dissertations on Dutch idioms at the Institut Marie Haps and the U.C.L. (Jaspers 1996, Stéphany 1996). L. Beheydt and J.-P. Colson are now considering a cooperation project with prof. S. Theissen and Ph. Hiligsmann of the Université de Liège, that might lead to a dictionary of common Dutch idioms for French-speaking students.

One of the interesting results at the first stage of this research is that Dutch verbal idioms are often very difficult to translate into French. The translation dictionary gives a rough idea of the meaning of the idiom, but when you take a number of contexts into account, you realize that a different translation is required. A typical error made translation dictionaries is comparable to the "attraction of the cognate" mentioned by Granger & Swallow (1988) for deceptive cognates: they tend to translate an SL idiom by a TL idiom. This kind of reaction also characterizes beginning translators. Translation practice demonstrates that a very large number of Dutch verbal idioms actually correspond to specific verbs in French, as illustrated by the following examples:

DUTCH                           FRENCH
Aan de kaak stellen             fustiger, critiquer
Naar voren brengen (argumenten) avancer 
Als beschermheer optreden       parrainer
Door elkaar halen               confondre, mélanger
Uit elkaar houden               distinguer
In elkaar zetten                monter, assembler

This implies that a significant part of the French lexicon will actually correspond to a part of the IDIOMATIC component in Dutch. This is just the reason why French-speaking students are often unable to translate very specific French verbs into Dutch: they don't imagine for one moment that the solution to a lexical problem might be a Dutch idiom. This observation is confirmed by error analysis, but this would fall beyond the scope of this article. For this reason, also, it seems preferable to start from the learner's mother tongue to design a dictionary of idioms.

Another kind of information that students would need is actually related to the problem of valency. As it stands, valency theory and phraseology are not at all contradictory. It is true that valency lays stress on individual words and the so-called "slots" that they accept, whereas phraseology sees a lot of verbal constructions as inseparable units. It is also true that valency and collocation, one of the favourite phraseological themes, are not always easy to distinguish. On the one hand, valency can include lexical or semantic information (see F. Devos, 1996) and comes therefore very close to collocation. On the other hand, some phraseologists make a distinction between lexical collocations and grammatical collocations (Benson 1985, Klimaszewska 1990), and this grammatical collocation is actually comparable to valency. The precise label one uses is of relatively minor importance compared to the methodological implications of such linguistic information.

Idioms, by definition have to do with phraseology but valency certainly plays an important part in their use. For instance, "met een korreltje zout nemen" and "take with a pinch of salt" require a NP or Pro as a direct object. As to lexical restrictions, it is also of great importance to know if the subject of the verbal idiom can be human or non human. For instance, it is not obvious for French-speaking students learning Dutch that a verbal idiom like "roet in het eten gooien" can be used with persons but also with words like "cijfers, resultaten". Another type of lexical information about idioms that would be very useful for students is a list of the most frequent lexical items associated with them. "Op de tocht staan", for instance, is almost always used with a subject belonging to the semantic fields of "banen" (jobs).

References

  • BEHEYDT, L. (1984). Woordenschat in het vreemde-talenonderwijs. In: Neerlandica extra Muros, 42, 17-22.
  • BEHEYDT, L. (1986). Optimalisering van de woordenschatverwerving - met een praktische toe passing voor huiscomputer. In: Levende Talen , 416, 630-637.
  • BEHEYDT, L. (1987). The semantization of vocabulary in foreign language learning. In: System, 15, 55-67.
  • BEHEYDT, L. (1989). De vergeten leerprincipes. Cognitieve aspecten van de woordenschatverwerving. In: Toegepaste Taalwetenschap in Artikelen, 34, 42-50.
  • BEHEYDT, L. (1995). Leermiddelen voor woordenschatverwerving Nederlands als tweede en als vreemde taal. In: Handelingen van het internationale colloquium "Leermiddelen voor het Neder lands als vreemde taal", BEHEYDT, L. & L. CHARLES (1995). Academisch vocabularium. Een oefenboek voor anderstaligen. U.C.L., Louvain-la-Neuve.
  • BENSON, M. (1985). Collocations and Idioms. In: R. Ilson, Dictionaries, Lexicography and Language Learning, Oxford: Oxford University Press.
  • BURNON, C. (1990). Valse vrienden Frans-Nederlands, van abandon tot acteur. Onuitgegeven licentiaatsverhandeling. Louvain-la-Neuve, Université catholique de Louvain, Afdeling Germaanse filologie.
  • COLSON, J.-P. (1991). 'Phraséologie et enseignement des langues.' Communication lors du Colloque international "Phraséologie et terminologie en traduction et en interprétation" . Université de Genève, Ecole de Traduction et d'Interprétation, Genève, le 3 octobre 1991.
  • COLSON, J.-P. (1992a). Ebauche d'une didactique des expressions idiomatiques en langue étrangère. In: Terminologie et Traduction, 2/3, p. 165-18. Bruxelles, Luxembourg: Commission des Communautés Européennes.
  • COLSON, J.-P. (1992b). Valse vrienden in de Van Dale vertaalwoordenboeken Frans-Nederlands en Nederlands-Frans. In: Neerlandica extra Muros , 30, 35-39.
  • COLSON, J.-P. (1993a). De fraseologische exploitatie van nieuwsberichten. In: Bulletin de l'Association des Germanistes de l'Université Catholique de Louvain, 15, 3-6.
  • COLSON, J.-P. (1993b). L'usage moderne du passé composé en néerlandais et en français: une étude contrastive. In: L. Beheydt, red., Taal en leren. Een bundel artikelen aangeboden aan prof. dr. E. Nieuwborg. Bibliothèque des cahiers de l'Institut de linguistique de Louvain, 68. Louvain-la-Neuve, Peeters.
  • COLSON, J.-P. (1993c). Anglicismes et "néerlandismes" dans le français de Belgique : quelques perles récentes. In: Défense de la langue française , Paris, 169, 9-12.
  • COLSON, J.-P. (1993d). Het Nederlands als vreemde taal: fraseologie en luistervaardigheid. In: Le Langage et l'Homme, 27, 113-122.
  • COLSON, J.-P. (1995a). Quelques remarques sur l'enseignement de la phraséologie aux futurs traducteurs et interprètes. In : Le Langage et l'Homme 30, p. 147-156.
  • COLSON, J.-P. (1995b). L'enseignement de la phraséologie aux futurs traducteurs et interprètes. In : Le Journal du Traducteur, Paris (Le Chesnay), 12, p. 2-6.
  • COLSON, J.-P. (in press). Valse vrienden en dwaalduiders Nederlands Frans: de originaliteit van het Nederlands.
  • COLSON, J.-P. (in press). Nieuwe perspectieven voor contrastief onderzoek Nederlands-Frans.
  • DEPOUHON, A.-M. (1991). Morfologische valse vrienden. Een corpusanalyse en een poging tot classificatie. Onuitgegeven licentiaatsverhandeling. Louvain-la-Neuve, Université catholique de Louvain, Afdeling Germaanse filologie.
  • DEVOS, F. (1996). Contrastive verb valency: overview, criteria, methodology and applications. In: A-M. Simon-Vandenbergen, J. Taeldeman, D. Willems (eds.), Aspects of contrastive verb valency. Studia Germanica Gandensia , 40, 15-81.
  • GRANGER, S. & H. SWALLOW (1988). False friends: a kaleidoscope of translation difficulties. In: Le Langage et l'Homme, 23, 108-120.
  • JASPERS, S. (1996). Een contextuele studie van verbale vaste verbindingen in het Nederlands. Onuitgegeven licentiaatsverhandeling. Brussel, Institut Marie Haps.
  • KLIMASZEWKA, Z., Verbale Phraseologie des Niederländischen . Doctorale dissertatie, Warszawa, 1990.
  • NATION, P. (1990). Teaching and learning vocabulary. New York, Newbury House.
  • NEHLS, D. (1994). English DO/MAKE compared with German TUN/MACHEN and Dutch DOEN /MAKEN. In: IRAL 29, 303-316.
  • NOËL, D. (1996). Grammar in bilingual dictionaries: Contrasting English/French dictionaries and the CVVD. In: A-M. Simon-Vandenbergen, J. Taeldeman, D. Willems (eds.), Aspects of contrastive verb valency. Studia Germanica Gandensia, 40, 83-123.
  • STEPHANY, J. (1996). Vertaalproblemen Nederlands-Frans: 100 idiomatische uitdrukkingen on der de loep. Onuitgegeven licentiaatsverhandeling. Brussel, Institut Marie Haps.
  • VAN ROEY, J., GRANGER, S. & H. SWALLOW (1988). Dictionnaire des faux amis français -anglais. Gembloux, Duculot.
  • WEISSHAUPT, J. (1980). Moeten/devoir. Een contrastief-transformationele schets. In: S. Devriendt & E. Nieuwborg, eds. Leuven, Acco.
  • WEISSHAUPT, J. (1993). 'Pouvoir' kunnen interpreteren, een verkenning. In: L. Beheydt, ed., Taal en leren. Louvain-la-Neuve, Peeters.
  • WERY, M. (1996). Aanzet tot een woordenboek van faux amis Frans Nederlands, Nederlands-Frans. Onuitgegeven licentiaatsverhandeling. Louvain-la-Neuve, Université catholique de Louvain, Afdeling Germaanse filologie.

[table of contents]

Report on the First CoLLaTE Workshop

Bart Defrancq

The First Collate Workshop brought some 40 members of the Research Network together in the pleasant atmosphere of the newly restaured Reylof, Ghent for a fruitful workshop which concentrated on the activities of the different research units. The Collate pillars: valency, typology, multilingual corpora and applied linguistics were abundantly highlighted from different angles.

The first main issue of the Workshop concerned valency descriptions, with special attention to those descriptions which involve more than one language. Theoretical issues of valency description were discussed by Helmut Schumacher and Jacqueline Kubczak (Institut für Deutsche Sprache), Sabine Kirchmeier-Andersen (Odense University), and Susan Hunston (University of Birmingham), who introduced a more controversial note by questioning the usefulness of some of the basic notions in valency description. Lene Schøsler (Odense University) showed what advantages an extensive inventory of valency patterns can have for the description of syntactic phenomena like those linked to the notion of inalienable possession. Helmut Schumacher, on the other hand, stressed the usefulness of monolingual valency description for learners' purposes and focused on the series of valency dictionaries which are based on the valency dictionaries developed by the IDS but take into account the mother tongue of the learner of German. According to the prerequisites for contrastive lexicography enumerated by Filip Devos and Bart Defrancq (University of Gent) these dictionaries are not contrastive, but bilingual.

Ludo Melis (University of Leuven) shared some of his views on typological studies, explaining how difficult and risky it is, when comparing languages, to define the object of comparison. His colleagues drew a picture of the typological studies conducted at the University of Leuven, an effort which was repeated by Danielle Corbin (University of Lille) regarding the studies of SILEX. Ralph Salkie (University of Brighton) adressed another basic problem of contrastive research when he discussed the issue of equivalence and difference between languages. He and many other participants broke a lance for the use of multilingual corpora in contrastive and typological studies. Their research results on discourse connectors (Bengt Altenberg, University of Lund), epis-temic modality (Karin Aijmer, University of Gothenburg) and prepositions (Hans Paulussen, University of Namur) were more than convincing.

On the side of applied contrastive research, foreign language teaching was the main concern of the majority of the participants. Jean-Pierre Colson (Univerversity of Louvain) investigated the lexical sources of difficulties in foreign language learning, such as deceptive cognates and collocations, while Claire Blanche-Benveniste and André Valli (University of Provence) concentrated more on the difficulty of non-canonical word order in understanding other languages. Sylviane Granger (University of Louvain), on the other hand, linked the theoretical and applied aspects of contrastive research to the use of multilingual corpora which include corpora produced by non-native speakers. The comparison of these with native speaker corpora has important pedagogical advantages. However, not only language learners benefit from multilingual corpus reasearch. Emmanuel Herbigniaux (University of Namur) showed that multilingual corpora are also an important factor in the development of various machine translation applications.

It became all the more clear during and after the Collate Workshop that the different research units which take part in the Research Network all have common interests and a specific expertise, which, if it were shared, could enormously benefit the work of other members. Some steps in this direction were suggested at the round-table which concluded the Workshop: the compilation of a contrastive bibliography, a first version of which has already been made available, possible exploitation of the existing multilingual corpora and the organization of workshops on more specific topics. Taking into account the enthousiastic reactions afterwards, these first steps will not be the last ones.

[table of contents]

To the table of contents of other CONTRAGRAM issues