Quantitative measurement of affinity between originals and translations is one of the most difficult tasks of quantitative linguistics. Statistical methods are efficient in the description of measurable, formal features of text, e.g. lengths or frequencies of linguistic units, but fail to grasp their semantic content. Consequently, they allow for reliable comparisons of parallel test samples based on well-defined criteria only. The subject of this contribution is the comparison of parallel Polish-Russian samples with regard to the distribution of stressed and unstressed syllables in the line of text. The time-series generated by the sequences of syllables are modelled by means of the ARIMA method (Box, Jenkins 1970). Numerical parameters of the estimated models reveal the most important sequential text characteristics, such as depth and strength of the contextual relationship, and make it possible to determine the general level of the rhythmical orderedness in text.
Financed by the National Centre for Research and Development under grant No. SP/I/1/77065/10 by the strategic scientific research and experimental development program:
SYNAT - “Interdisciplinary System for Interactive Scientific and Scientific-Technical Information”.