INSTITUTE OF LINGUISTICS

Faculty of Philosophy, University of Zagreb

Ivana Lučića 3, 10000 Zagreb, CROATIA                                                                   
tel. (+385 1) 6120-011, 6120-142; fax. (+385 1) 6156-879
e-mail: zzl@ffzg.hr                                                                                         CROATIAN NATIONAL CORPUS

The Institute of linguistics is one of the organizational units of the Faculty of Philosophy at the University of Zagreb in Zagreb, Croatia.

The Institute is the center of most of the linguistic-oriented projects at the Facultyof Philosophy.

This document covers several topics on the Institute of linguistics:

1. History
2. Aims
3. Activities (projects, publications)
4. Staff

HISTORY

The Institute of Linguistics was founded in 1960 upon the suggestion of five distinguished professors of the Faculty of Philosophy (Mirko Deanović, Rudolf Filipović, Vladimir Gortan, Josip Hamm and Vojmir Vinja). The initial purpose of such an institution was to obtain more efficient organization of linguistic research at the Faculty. The first director of the Institute was professor Ljudevit Jonke (1960-1963), followed by professor Rudolf Filipović (1963-1983), professor Milan Moguš (1983-1992), Dr. Maja Bratanić (1992-1994) and Dr. Vesna Muhvić-Dimanovski (1994.-).

AIMS

The general aims of the Institute defined from its beginning have been:

• the development of linguistics and linguistic research methodology in the field of Croatian and other languages;
• research of the phenomena in different fields of general and comparative linguistics of Indo-European and non-Indo-European languages;
• the investigation of correspondences between Croatian and other languages.

ACTIVITIES

Projects

In the 1960-s the main activity of the Institute was predominantly oriented towards Croatian in contact with other languages. In the 1970-s the project of compiling the Croatian language corpora started. An overview of projects (concluded and current) can be divided in three areas:

1. Computer processing of the Croatian language
2. Contrastive projects (Croatian vs. other, primarily European, languages)
3. Current projects (Funded by Ministry of Science and Technology of the Republic of Croatia)

1. Computer processing of the Croatian language

Two areas of computational linguistics have been covered by projects:

• Corpus compiling and processing:

  1 million corpus of running text in Croatian (M-corpus) covering the 1937-1978 period and compiled by Prof. Milan Moguš (project MZT RH 6-03-048 which ended in 1996).
  The corpus is divided in 5 sub-corpora (prose, poetry, drama, secondary school textbooks, newspapers; 200.000 tokens each)
  The processing includes:

  • alphabetic dictionary of types and lemmas
  • frequency dictionary of types and lemmas
  • concordances (KWIC) lemmatized as well as non-lemmatized

  Croatian national corpus composed of two components: 30 million corpus of contemporary Croatian language (30m) and Croatian Electronic Text Archive (HETA) (see current project 130718 below)

  Croatian-Slovene parallel corpus (see current project 130821 below)

  Croatian participation in ELAN project (see current project 130729 below)

  For survey of corpus projects in Croatia and Institute see:

  • Marko Tadić, »Računalna obradba hrvatskih korpusa: povijest, stanje i perspektive«, Suvremena lingvistika 43-44, (1997), str. 387-394. (ISSN 0586-0296) rtf  pdf  ps
  • Marko Tadić, »Corpus-building projects in Institute of linguistics, Philosophical Faculty, University of Zagreb«, invited lecture, University of Tübingen, 1999-12-01, slides)

• Computational Morphology of Croatian (see current projet 130718 below)

  Computational model of the Croatian flective system has been designed and prototyped. As the result of that research the

  Morphological generator of Croatian word-forms has been developed. At this moment the filling of the lexicon is on its way. The result of generation will be Croatian morphological dictionary composed as the list of word-forms accompanied with MSDs (morphosyntactic descriptions) that can be used for semi-automatic lemmatization and corpus tagging. The dictionary will be coded according to the MulTextEast v2 specification.

2. Contrastive projects

The areas which are covered by Institute contrastive projects are:

• Phraseological Dictionaries
  of Croatian and 11 different languages (French, Slovene, Czech, Polish, German, Italian, Russian, Ukrainian, Latin, Greek, Modern Greek);
• Languages in contact
  the investigation of Anglicisms in several European languages;
• Contrastive analysis
  of Croatian and English, German, Russian, French, Spanish, Italian...

3. Current projects

• COMPUTER PROCESSING OF THE CROATIAN LANGUAGE (MZT RH 130718)
  The starting point for every linguistic research is the corpus. Today the investigation of the corpus of a language without the help of a computer seems to be hardly imaginable. Almost all important world languages - but also the Slavic ones - base their linguistic and lexicographic studies on representative multimillion corpora of the respective languages. As for Croatian there has not been any corpus compiled according to the present experiences provided by corpus linguistics (except the corpus of one million words of Croatian (M-corpus) compiled between 1976 and 1996 in this institution). The aim of the project is the compilation and processing of representative Croatian texts - both older and contemporary - in the form of systematically designed corpus. Particular goals are as follows:
  1. to compile and process a multimillion (30 M) corpus of contemporary Croatian;
  2. to bring selected texts and dictionaries of older Croatian authors as well as translations of the capital works of the civilization (Bible, Talmud, Kur’an) into the form of a corpus;
  3. to make a supplement of the orthography dictionary on the basis of corpus processing;
  4. to investigate on the basis of relevant corpora in diachronic and synchronic perspective the procedures in the formation of neologisms in the creation of Croatian terminologies;
  5. to publish the corpora as well as the results of their processing on opto-magnetic (CD-ROM) and electronic medium (CARNet).

  The method of compiling the material and processing the corpus has been tested on the pilot-research of the one-million corpus of Croatian worked on in this institution within the frame of a previous project financed by MZT (6-03-048). The influence of the project can be expected in the following areas:

  1. Croatistics (history of the Croatian language, orthography problems, the development of Croatian terminology, the investigation of Croatian flexion and derivation)
  2. lexicography and lexicology (Croatian-Croatian and Croatian-foreign language dictionaries, text concordances, orthography dictionaries, thesauruses, terminology dictionaries, dictionaries of neologisms etc.);
  3. information science (indexing and searching of textual data bases, processing of natural language, generating natural language messages, creation of computational tools for the natural language processing).

• CROATIAN PARTICIPATION IN ELAN PROJECT (MZT RH 130729)
  Today computationally processed corpora are unavoidable source of linguistic material for the linguistic description coping with all levels of natural languages. Until recently the main obstacle to data access in corpora was unstandardized format of texts included. Introduction of SGML (ISO 8879; 1986) and with TEI (Text Encoding Initiative: Sperberg-McQueen & Burnard 1994) the standardized format for corpus storing and annotating was established. Meanwhile, the set of procedures for access to standardized corpus data remained unstandardized. That's why EU, via two non-profit organizations (PAROLE and TELRI) which were constituted after the closing of two large European project bearing the same names, started the ELAN project (European Language Activity Network). ELAN project aims at delivering the standard tool for mantaining and querying corpora and/or lexica for more than 20 west-, central- and east-european languages. Final result of ELAN will be a network of corpora and/or electronically stored lexicons installed on servers of the institutions involved in project. That network of linguistic resources will be accessible by WWW service and standardized interface.
  Institute of linguistics on the Faculty of Philosophy at the University of Zagreb was invited to ELAN project because of its participation in TELRI I project (regrettably not as a full member). Since the Institute is in fact a referent institution for corpus linguistics in Republic of Croatia, it can take a participation in ELAN project producing a corpus of Croatian language of approximately 2.000.000 words. The Institute will get access to a newest generation of corpus tools which will became a standardized tool in the field of corpus linguistics.
  The new corpus methodology achieving will have direct impact to the Croatian national corpus which is also being collected at the Institute in the scope of MZT project 130718.
  ELAN project is also important because it is the first international EU-funded project where Croatian language is institutionally, organizationally and terminologically completely separated from the Serbian and that is the opportunity which should not be neglected.

• CROATIAN-SLOVENE PARALLEL CORPUS (MZT RH 130821)
  This parallel corpus should be the starting point for applied linguistic research of two neighbouring languages which are genetically close also. Croatian and Slovene have undergone significant changes after both stated gained independence in 1991. The corpus will be retrievable by means of WWW and is planned to represent important linguistic material for students of Croatian and Slovene on Faculties in Zagreb and Ljubljana, for translators from and to both languages. Because of old Croatian-Slovene dictionaries (usually Serbo-Croatian - Slovene), the compiling of such corpus would be a unavoidable resource for new dictionaries.

• SEMANTIC FIELDS AND SYNTAX
  The main focus of this project would be to analyse the principles of semantic field organization and their prototypical syntactic determiners expressed primarily on the syntagmatic or sentence level. Thus the main aim will be to show how meaning and vocabulary organization connectand intermesh with syntactic phenomena. An analysis of this kind that brings together semantic and syntactic characteristics of parts of vocabulary of different languages, sheds light on the basiv principles of how natural languages function. The analysis would primarily be centered on semantic fileds in Croatian and English, but would also include data from non-Indoeuropean languages. Croatian and English represented languages belonging to different families of the Indoeuropean group, and by contrasting them differences can be determined, but at the same time principles that go beyond language specific analysis appear. Inclusion of data from non-Indoeuropean languages represents a way of confirming the basic principles, of semantic field organization, and it is expected that they will provide additional proof for universal principals, of how linguistic phenomena are manifested on the semantic and syntactic level. Previous work of the principal investigator on the prototypical organization of semantic fields, and the researchers theory of "Role and Reference Grammar" are starting points that in combination should provide the basis for a more comprehensive understanding of the complexity of semantic an syntactic phenomena.

• CROATIAN-GERMAN LINGUISTIC RELATIONS
  The project will include both diachronic and synchronic levels of Croatian-German linguistic relations. On the diachronic level grammars, dictionaries and teaching books of German and written in German will be analyzed. German-language reviews, books and newspapers published in Croatia will be also dealt with. On the synchronic level the influence of the German language on 20th-century Croatian dialects will be investigated, decribed and analyzed. The social and institutional aspects of teaching German in Croatia between the 18th and 20th century will be investigated.

• CONTRASTIVE GRAMMARS: CROATIAN AND FRENCH
  The basic aim of the research is the elaboration of contrastive grammars of Croatian and French. The basic thesis of the project starts from the fact that there is no modern Croatian grammar with special reference to French grammar, nor any French grammar with special reference to Croatian. As far as the general theory of grammar is concerned there are different approaches which in principal do not enable the elaboration of a gramamr of a language. These grammars should contribute to the creation of a thorough grammatic theory and develop a general grammatic theory taking into account different models of grammatical description as well as modern linguistic methods. The result of the research would be the elaboration and publication of two grammars: the Croatian grammar with special reference to French and the French grammar with special reference to Croatian. This project would contribute to the introduction of Croatian into the world science, especially to its presence in the French language area. As the project will be using the results of computational linguistics and formal grammars, especially so called regular grammars, the project will contribute also to the development of general knowledge in this area, both in Croatia and in the world. The project plans the collaboration with Maurice Gross (Laboratoire d’automatique documentaire et linguistique - Université Paris VII et Paris VIII), with the Centre Lucien Tesniere (Université de Franche Compté - Besançon), with the Laboratoire de linguistique informatique (Université Paris XIII) and with Olivier Soutet (Sorbonne, Paris IV).

Publications

The Institute has developed a rather strong publishing activity which encompasses several series of publications such as:

• Computer concordances of the texts of the (older) Croatian literature;
• Contrastive studies of Croatian and other languages;
• Phraseological dictionaries;
• The English Element in European Languages series
• Linguistic textbooks

More than 60 books and studies have been published as the result of the work on the Institute projects.
The Institute is also co-publisher of the journal Suvremena lingvistika (Contemporary linguistics) which is quoted in the MLA and BL. It also publishes the Bulletin of the Institute of Linguistics where thorough bibliographical data as well as articles on ongoing projects are published.

STAFF

1. Permanent

Vesna Muhvić-Dimanovski, Ph.D..
Ph.D. in contrastive linguistics; her field of interest: languages in contact, anglicismsin Croatian and German, research on neologisms in Croatian and other European languages

Ida Raffaelli, Ph.D.
Ph.D. in linguistics; field of interest: semantics in historical perspective, mediaevalFrench, chronicles

Boško Bekavac Web-page
B.A. in linguistics and informational sciences; field of interest: computational linguistics, linguistic tools, corpus linguistics, SGML, XML

Sanja Fulgosi
B.A. in croatistics; field of interest: croatian corpora, morphology, POS tagging

Krešimir Šojat
B.A. in anglistics and germanistics; field of interest: croatian corpora, collocations, semantic tagging

Ivana Simeon
B.A. in general linguistics and Russian language and literature; field of interest: parallel corpora processing and annotation, M(A)T, translation theory

2. Principal researchers of the Institute's current projects

Professor Zrinjka Glovacki-Bernardi
Ph.D. in linguistics, field of interest: discourse analysis, languages in contact: Germanand Croatian; principal researcher of the project: Croatian-German linguistic relations.

Professor Dubravka Sesar
Ph.D. in linguistics, field of interest: Slavic languages, language standardization processes, Czech and Slovak; principal researcher of the project: The analysis of West-Slavic languages.

Marko Tadić, assistant professor Web-page
Ph.D. in computational linguistics; field of interest: corpus linguistics, corpora compiling and processing, computational morphology; principal researcher of the projects: Computational processing of the Croatian language, Croatian participation in ELAN project, Croatian-Slovene parallel corpus

Professor Milena Žic-Fuchs
Ph.D. in semantics, field of interest: cognitive linguistics, semantics, syntax; principal researcher of the project: Semantic fields and syntax.

Back to Faculty of Philosophy Homepage

Last change 2001-04-25 by Marko Tadić