Чмиль Л.В. Опыт применения статистических методов в изучении керамики Среднего Поднепровья XVI–XVIII вв / Chmil L.V. Statistical methods in the study of the ceramics of the Middle Dnieper region of the XVI–XVIII centuries (original) (raw)

The main task in the study of ceramics of the XVI–XVIII centuries has been to search for the signs that reflect chronological changes. Due to the large volume of material, statistical methods were used for working with the most widespread varieties of ceramic utensils, i.e. pots and plates. The majority of features which it was possible to formalize were morphological. They were the size, proportions, design features and schemes of ornaments. The dimensions, volume and proportions of the pots were calculated with the help of V.F. Gening’s methodology. Similar indicators were also developed for plates. Among the structural features of the pots are the design of the rim, the neck, the shoulder and the bottom, and of the plates the design of the rim, the bottom and the inflection. I. Sarachev’s methodology was applied, according to which the shape of the rim was determined by the method of its forming. Among the technological features taken into account were: the amount of iron in the clay, the type of glazing, the presence or lack of glaze, the place of its application and the colour, and the technique of decoration. Some of the technological features, such as the composition of the pottery pastes and the way of shaping of the vessel turned out to be common (with minor exceptions) for the entire array of ceramics, and hence of no use for micro-chronology. The next step was to study the correlation between the features through the use of the statistical methods developed by G.A. Fyodorov-Davydov. The correlation between the quantitative traits was determined with the help of correlation analysis, whereas for the quantitative and qualitative traits analysis of variance was used. The quantitative indicators which had the strongest correlation with the qualitative ones were divided into 2-3 intervals and considered as qualitative. The mutual occurrence of quality traits was determined using the criterion of χ2 and coefficients of correlation Q and association К. For fragmented material the method of comparison of frequencies was also used. On the basis of the obtained data we have identified the groups with strongly interconnected features, which are regarded as the most informative chronologically. On this basis the types of pots and plates have been identified. The features were also used for calculating the coefficient of absolute similarities for ceramics assemblages, which allowed to group them and to determine the relative chronology. The correlation of these data with numismatic material and other date sources allowed to obtain an absolute chronology for ceramics assemblages, types of pots and plates, and other varieties of ceramic tableware.