Abstract
Traditional approaches to human dental arch forms are mainly focused on categorical classifications and/or parametrizations by applying mathematical functions. In this study I examined intragroup variation of the arch forms, using a multivariate expression newly introduced for the present analysis. The following three components were extracted: 1) the ratio of the width at the canine part to the length at the incisal part, 2) the lateral expansion of the molar part, 3) squareness of the anterior part. On the basis of these results, the classification systems by HRDLICKA and THOMPSON were critically reviewed. My results revealed that these systems are different in character, in that they focused attention on different aspects of the arch form. The merits and demerits of the parametrization were also discussed. Parametrization's merit is the simplicity of its expression; however, when used for classification, the introduction of an artificial order into the arch form variability which is predetermined by an applied function is a demerit. In contrast, multivariate methods can provide natural order for arch variability.
