|Abstract||This study deals with the categorial status of a set of words known as 'determiners' and the features encoded in them.
We argue, following a suggestion in Chomsky, 1975, that 'Optimality' conditions on grammar require grammatical categories to be 'primitive' in the sense that they must be unanalysable into further entities. The status of many words classified as determiners does not conform with this suggestion. The category 'D' is therefore a non-standard grammatical category. We have provided mathematical, morphological and syntactic arguments and facts drawn from a number of languages including Arabic, English, French, German, Hebrew, Italian, ...etc. showing that the category 'D' is not a 'primitive' category and therefore should be suspended and replaced by its 'primitive5 components.
We have shown that 'genuine' determiners are morphologically complex where each element encodes a functional feature, such as 'person9, 'number9, 'gender9, 'proximate', 'definite', ...etc. These features, we claim, are functional categories. Many words classified as belonging to the category 'd' are in fact nouns or adjectives.
Adopting an articulated theory of 'D' in which functional features are taken as functional categories is not only theoretically motivated but also has implications for language learn ability.