This paper examines the notion of “root” within the framework of canonical approach. Although this notion has been widely discussed, many controversies exist, especially regarding its linguistic properties. These controversies are motivated by a fact that in natural languages, the notion of “root” is instantiated differently across languages, or even within the same language. To analyze the diversity of roots systematically and uniformly, a theoretical space of roots is constructed under the canonical approach. The specific position of each instance within the theoretical space against the canonical roots can be determined by its behavior on four criteria, namely, boundness, positional flexibility, phonological richness and meaning lexicality. In practice, I make a thorough examination of 122 Chinese morphemes whose morphological status has remained ambiguous. These morphemes are ultimately divided into five groups, gradually deviating from the canonical roots, showing that even in a typical isolating language Mandarin, roots are not uniform – neither in their structures nor in their meaning. More importantly, the distinctions among individual roots in terms of their linguistic properties are clearly explained. Also, the theoretical space of roots enables cross-linguistic comparisons of roots, making it highly significant for both intra-linguistic and cross-linguistic studies of root variation.
© 2001-2026 Fundación Dialnet · Todos los derechos reservados