Chapter 3: A typology of principal-part systems

As a preliminary to investigating the relative complexity of IC systems, we employ the notion of principal parts as a basis for typological comparison. We examine three significant dimensions of variation among principal-part systems: (i) the number of (static, adaptive, or dynamic) principal parts needed to determine a lexeme's realized paradigm; (ii) the number of dynamic principal parts needed to determine a given cell in a lexeme's realized paradigm; and (iii) the degree to which certain realized cells are favored in an IC system's optimal static principal-part sets. Logically, dimensions (i) and (ii) are correlates of an IC system's complexity. Variation in dimension (iii) is, as we shall see, logically orthogonal to IC systems' variation in complexity; nevertheless, (iii) exhibits an intriguing correlation with another correlate of complexity.