Word-based Morphology from Aristotle to Modern Wp

The dominant morphological models developed in the modern period can both trace their origins to ancient Indo-European grammatical traditions. Bloomfieldian models have their ultimate roots in the Sanskrit grammatical tradition, particularly as represented in the work of Pān ̇ ini. These models adopt a formative-based perspective in which the central analytical task of morphology involves the disassembly of words into arrangements of sub-word units.1Modern word-based models likewise lie at the end of a continuous tradition that arose from attempts first to describe and subsequently to teach the grammar of ancient Greek and Latin. These models project morphological analysis primarily upwards from the word, and treat the association of words with paradigms or other sets of forms as the most fundamental morphological task. Subword sequences figure in the patterns of modification that related basic forms of an item to their ‘inflected’ variants. However, the classical model has no counterpart of morphemes—not even roots (Law 1998: 112)—and its proponents perceived no need for any unit intervening between sounds and words. The primacy of words in the