Who needs the “Lexicograph” and what for?

The system “Lexicograph” is a semantic dictionary in the form of a data base.

Modern linguistic semantics proceeds from the assumption that the lexicon, as well as other levels of language, i.e. its grammar and phonetics, is a system. The basis of the lexical system is constituted by recurrent meaning components, such as ‘know’, ‘see’, ‘move’, ‘cause’, ‘thing’, and parameters of word meaning, such as ontological category, thematic class, roles of participants of the denoted situation, and thematic classes of participants.

Not a single word in a language is identical to any other word, either with respect to meaning, or combinability, or any other aspect. However, most individual differences are determined by the components of meaning and parameters recurring in dozens or, sometimes, in hundreds or thousands words.

A word’s linguistic behavior, i.e. its combinability with other words, the set of possible grammatical forms and their meanings, and the ability to be under the sentence stress, is to a large extent predetermined by its meaning. Thus, the verb teret’ ‘to rub’ has the meaning of the Progressive in the Imperfective, whereas teriat’ ‘to lose’ does not; the grammatical form of the Imperative trite! ‘rub!’ (from teret’) has a standard meaning of an urge, whereas teriajte! ‘lose!’ (from teriat’) has none. A possible explanation is the difference in category; in fact, teret’ is an action, while poterial ‘has lost smth’ is a happening, i.e. something that has happened to one. Parameters and meaning components generate classes of words with similar behavior.

Having realized that a word’s behavior depends on its meaning, a linguist, however, has to face the fact that words have usually more than one meaning, as polysemy is an inherent property of any language. Interestingly, it is parameters and recurrent semantic components that are the aspects of word meaning most prone to change. Therefore two meanings of the same word often differ by a parameter. Thus, in zastilat’ stol skatertju ‘to cover the table with a tablecloth’, zastilat’ is an action, whereas slezy zastilaiut glaza ‘tears fill [lit. cover] the eyes’ it is a state; the verb trebovat’ ‘to require’, which belongs to the class of speech act verbs, becomes a modal verb in the context Mashina trebuet remonta ‘The car wants[lit. requires] to be fixed’, i.e. changes its lexical class; stuchat’ ’to knock’ in Veter stuchit stavnej ‘The wind knocks with the shutter’ is a verb of sound, whereas in Otkroi, kto-to stuchit ‘Open the door, somebody is knocking’ it is an information verb.

This double role of parameters is quite natural. The fact is, multiplicity of meanings (namely, regular polysemy as it was defined by Ju.D. Apresjan) can be presented as a set of meaning shifts. Shifts of meaning are systematic, which means that the words characterized by the same semantic feature (i.e. a meaning component or a parameter) tend to undergo semantic shifts of the same kind. Therefore, parameters distinguish not only different words but also different meanings of one and the same word.

In traditional dictionaries, each word is supplied with a list of separate meanings. In the same way, modern semantics works with separate word meanings, ignoring the word’s semantic unity. The problem has been to recreate this unity. Having this in mind, we have set for “Lexicograph” two interrelated tasks, namely:

• to explicate each word meaning in such a way that the peculiarities of a word’s behavior in this very meaning could be explained;
• to show how word meanings are related, i.e. to construct a hierarchy of word meanings or even a paradigm of meanings shared by words of one and the same taxonomic class.

“Lexicograph” is not just a dictionary; it is a data base. This corresponds to the assumption that meaning explications should be formalized. Parameters and meaning components are basic instruments of formalization.

The project is experimental. The dictionary is by no means complete – rather, it is a collection of representative examples. We suggest formats of explication for basic types of word meanings, so that the users having the base at their disposal can add new words on the basis of the pattern formats. Thus, we supply not only the ready entries, but the instruments to process new lexical units. The present version of the system includes verbs only.