Marius Sala, Ion Dãnãilã * Lexicography at the "Iorgu Iordan" Institute of Linguistics




For producing out the CDSAA, we have been using the Explanatory Dictionary of the Romanian Language (DEX) [6] because it was - and still is - the richest one-volume explanatory dictionary of Romanian2. Our aim has been to include in the new dictionary the synonyms for, the analogies to, and the associations with each of the approximately 180.000 meanings registered in DEX3, which will turn CDSAA into a computerised lexical thesaurus of the Romanian language. But the DEX was not compiled by using a computer, nor was it conceived such as to be processed by an electronic device that is "ignorant of word meaning unless supplied by man". If we add to this the fact that the editors of CDSAA had no direct access to a computer prior to 1995, it is understandable why the authors of the project had in mind the division of labour between man and computer, each of the parts having well-defined tasks in a strict order: first, the man prepares the input and feeds it into the computer4, than the latter processes it in order to sort synonyms, analogies and associations.

Making ready the input means to render uniform the various definition formulas used in DEX (i.e. simple analytical definitions, complex definitions, encyclopaedic definitions, complex definitions, synthetic definitions by means of a unique synonym, by a chain of synonyms etc.) and to reduce them to one type of definition to be processed ultimately by the computer. The software was based on the Socrate logical system - which means that it should be converted into a system compatible with the present-day PC's.

At this point, we have got the input ready up to the letter Î, but - for reasons independent of the authors and of the editors - the revision of the data on file proceeded only to the letter C, which obviously is a very slow rate. Initially, we were teamed up with 20 lexicographers (covering 12 full-time jobs), but, as time went by, their number decreased steadily by reason of personnel quitting the institute or getting retired. At the same time, the researchers in the department had to tackle also other directions of study and research projects. The lack of computers was a serious drawback drastically diminishing the working time scheduled for CDSAA. Consequently, since about 1985 only one person has done revision work of the input (but only 2-3 months a year) which indeed does not help the project to progress a great deal, but maintains it on the research program of the department and of the institute.

Let us mention the fact that the sorting into analogies and association is much more accurate if done by the computer, which, unlike man, does not operate impressionistically, but by strict rules. Further on, precise rules will place correctly each analogy and association under a given meaning according to a number of distinguishing features such as: nearness to the meaning under which it appears, spreading area, stylistic peculiarities, etc.

As far as the computer processing of the input is concerned, we have to limit ourselves to generalities put forward in the above paragraph, at least until the publication of the dictionary. After a twenty-year lapse of time, as of 1996, things are turning favourable to the achievement of the project, the authors having now at their disposal the necessary electronic devices and the skilled manpower to use them.

References

  1. R. MICHÃESCU, I. DÃNÃILÃ, Principii de analizã confruntativã a vocabularului. I, Studii ºi cercetãri lingvistice, Bucureºti, 28, 2, (1977), p. 111 - 124; II, 3, (1977), p. 247 - 260.

  2. R. MICHÃESCU, I. DÃNÃILÃ, Dicþionarul confruntativ al limbii române (DCLR) - Partea a II-a - Aspecte teoretice ºi practice ale confruntãrii distributive semantice (sinonimie) - Cu privire specialã la Dicþionarul confruntativ de sinonime, de analogii ºi de asociaþii al limbii române (DCSAAS), Bucureºti, preprint, Institutul de Lingvisticã din Bucureºti, 1977, 178 p. typewritten.

  3. *** Dicþionarul confruntativ al limbii române (DCLR) - Dicþionar de sinonime, de analogii ºi de asociaþii (DCSAAs) - Specimen - Editori: Ion Coteanu, Ion Dãnãilã, Radu Michãescu, Mircea Seche, Bucureºti, preprint, Institutul de Lingvisticã din Bucureºti, 1978, XL + 48 p. typewritten.

  4. I. DÃNÃILÃ, R. MICHÃESCU, Banca de date fono-morfo-semantice a limbii române (BANDASEM), Limba românã, Bucureºti, 29, 5, (1980), p. 471 - 475.

  5. I. DÃNÃILÃ, Pentru un Inventar general al lexicului limbii române (ILEX), Limba românã, Bucureºti, 42 (1993), p. 61 - 68.

  6. *** Dicþionarul explicativ al limbii române (DEX), Editori: Ion Coteanu, member of the Romanian Academy, Dr. Mircea Seche, Fr. Luiza Seche, Editura Academiei, Bucureºti, 1975.




2 The 56,588 entries of DEX fall into 52,876 headwords and 3,712 variants referees to the headwords.
3 According to our estimation there is an average of three meanings for the 52,786 semantically full words in DEX.
4 That is the input that was done in handwriting on paper slips. Now, after having been equipped with computers, the lexicographers will feed the input directly into the computer.

71

Previous Index Next