Morphological asymmetry was first noted by Karcevskij (1929), who found that any given morphological ending may realize several morphological functions and any function may be realized by several endings. The Russian declensional ending -a, for example, expresses NomSg among Declension II nouns like sestr-a 'sister' and knig-a 'book'. On Declension I stems, however, it realizes GenSg, e.g. stol 'table' : stol-a 'of (the) table', and if the stem is neuter like sel-o 'village', it may mark NomPl, sel-a 'villages'. The Genitive case, on the other hand, is realized by -a on Declension I nouns, -i on Declension II nouns and -ej, -ov, and null on plural nouns. The separation of derivation and affixation accounts for this asymmetry quite neatly. Asymmetry is not to be confused with cumulative and extended exponence, which are similar. Asymmetry refers to the a mismatch in the number of phonological realizations and functions in different contexts; cumulative and extended exponence refers to such a mismatch in one context. Bazell (1949, 1952) also discusses this problem.