Polarity and constraints on paradigmatic distinctness

Polarity is a type of syncretism where the syncretic items are distributed along the diagonals rather than the rows or columns of a paradigm. Polarity is only rarely found (and only rarely studied as a topic), but is not restricted to a particular linguistic area or language family.1 It is debated whether it occurs accidentally or is a systematic phenomenon sui generis. Correspondingly, linguists either ignore the phenomenon or seek examples that can challenge common theoretical assumptions. One reason to consider polarity is that it seems difficult to learn. In this overview, only instances of morphological polarity are considered, brought about by affixes or ablaut. Subject– object reversals are excluded because they are regulated by independent means for mapping argument encodings to argument roles. The notion ‘polarity’ was first proposed by Carl Meinhof (1912) to characterize the number–gender distribution found in some Afroasiatic languages (see section 5.5 below). A more formal definition was given by Hetzron (1967: 184):