Abstract

 

Over the past few decades, lexicography has undergone dramatic development, which completely changed the original concept of dictionaries fixed on paper. Two illustrative examples of lexicographic structures hardly conceivable 15-20 years ago are: wiki dictionary (whose tribal representative is the Wictionary, or Wikislovník in the Czech version) and a linguistic thesaurus. A key feature of the first is the use of the creative potential of a large number of volunteers and/or volunteering (semi)professionals to produce an online dictionary of a size which would be extremely expensive and time consuming to do in a professional way. The second term is now defined as a structured system of semantic relations between lexical units, i.e. a database of terms having as a core part not the definitions of meanings, but the set of synonymic-antonymic and hyponymic-hyperonymic links; it was only by means of the web hyperlinking that this second concept became fully functional.

Both concepts actually stand in opposition to each other, even though they repeatedly try to integrate some features of their opponent. The first system out of necessity gives up some advantages of professionalism: the sophistically built-up glossary and the uniform processing of the entries is substituted by a large number of entries more or less standardized for a very reasonable price; as an added value there is the motivation of a large number of contributors to achieve the result. The second system is logically contrary, requiring high professionalism and processing in most cases to date, not used as a practical tool for a large number of users, but as a specialized tool in teaching, research or specific professional situations.

The advantages of both systems are of the type which is difficult to implement in signed language lexicography, due to several factors inherent in a signed language and to the culture of its users:

• an interest in using sign language for the communication between the hearing and the deaf, dates not so far back in history, and if we take into account the small number of native signers, the consequence is that there still is not much demand for standard lexicographical work
• technical requirements for building up a sign dictionary are extraordinary (high-quality video capturing, editing, data basing, retrieving), compared to a written dictionary 
• obstacles during the education of sign language users lead to the fact that those who manage technology for lexicographical work and who have the intellectual interest in it are not identical with the group of sign language users

Existing dictionaries of sign languages ​​therefore have several features in common:

• relatively small number of lexical units registered (due to the structure of the signed language and to its high iconicity which helps making concepts comprehensible without previously defining them)
• limited grammatical information (mainly confined to the determination of the part of speech where it is unambiguous, and to elementary stylistic assessment)
• poor structure of the glossary (phrases and compound lexical units are often ranged at the same hierarchical level as the units from which they are composed)
• absence of exemplification (samples used in the context)
• bilingual character, but with a very small number of semantic equivalents in the target language (often just one) and missing comments on the semantic conditions for differentiating if more of them are listed
• processing the glossaries of partial semantic fields (thematic areas) separately in various dictionaries
• no or minimal information about synonyms/antonyms and hyponyms/hyperonyms

The database and interface of the Czech sign language dictionary online are an attempt to overcome these traditional limitations, without making the structure too complex and impossible to implement some wiki features. This paper comments on the structure of the developed vocabulary and its anticipated system management.