KRYLOV LAW IN POLISH LINGUISTIC RESEARCH Cover Image

PRAWO KRYŁOWA W POLSKICH BADANIACH JĘZYKOZNAWCZYCH
KRYLOV LAW IN POLISH LINGUISTIC RESEARCH

Author(s): Marek Ruszkowski
Subject(s): Language and Literature Studies
Published by: Vilniaus Universiteto Leidykla
Keywords: language law; Krylov Law; statistical linguistics;

Summary/Abstract: In the early 1980’s, the Russian linguist Yuri K. Krylov carried out a number of tests taking advantage of the Russian language material and proposed a description of the relationship between a number of meanings of lexemes in a dictionary and their fre¬quency. The relationship is now termed the Krylov Law. It says that a share of entries having a specific number of meanings in a monolingual dictionary decreases in a geometrical progression with a common ratio of 1/2. It means that a share of lexemes with one meaning is app. 50%, 25% for lexemes with two meanings, 12.5% for lexemes with three meanings, 6.25% for lexemes with four meanings and 3.125% for lexemes with five meanings. The Russian linguist argued that this is a general principle applicable to any natural language. Values provided by Krylov are average, which means that their distribution in em¬pirical material shows some deviations. In respect of the Polish language, Wierzbicka randomly sampled 500 entries from The Polish Language Dictionary, ed. Doroszewski. Sambor analysed material taken from The Small Polish Language Dictionary, ed. Skorupka, Auderska, Łempicka. Three samples of 1,000 lexemes were taken into account to include a) whole dictionary b) nouns c) verbs. Hammerl and Sambor, in turn, analysed 3,000 entries from The Modern Polish Language Frequen¬cy Dictionary, authored by Kurcz, Lewicki, Sambor, Szafran, Woronczak. Kałucka, taking advantage of an online version of The Polish Language Dictionary issued by PWN, randomly sampled 20,000 entries. All the above studies generally prove that the Krylov Law holds valid for the Polish lexical mate¬rial. The Law shows a greater regularity in respect of the total number of entries in a dictionary than in respect of specific parts of speech. It is evident from Kałucka’s research that the number of lexemes fewer than 16,000 might be insufficient to obtain satisfying results, and any sample of 17,000 entries is sufficient for validation of the Krylov Law. However, the Law can be valid when a number of entries subject to anal¬ysis is several times smaller, which was proved by Wierzbicka’ (500 entries) and Sambor’s (1,000 entries) findings. Krylov himself analysed 9,500 and 2,700 entries in two dictionaries of the Russian language.

  • Issue Year: 2010
  • Issue No: 18 (23)
  • Page Range: 58-63
  • Page Count: 6
  • Language: Polish