Lexical categories : legacy , lacuna , and opportunity for functionalists and formalists

root INDICATE is taken to be present in the syntax of all the examples in (3):in (3b), it is the complement of an n head; in (3c), it is the complement of an a head; in (3a) it is the complement of a v head. The driving force behind positing these extra heads does not come so much from the theory of lexical categories per se, but from the morphological goal of providing a syntactic position for derivational morphemes like -ion in (3b), -ive in (3c), or -ize in legalize or fossilize, as required by DM’s view that all complex words are formed in the syntax. Although not universally adopted, this view has become mainstream. One can question its “depth” as a theory of the lexical categories. To a large extent, it simply moves the interesting theoretical question around: now one needs a theory of what ns, vs, and as are, rather than a theory of what nouns, verbs, and adjectives are. No obvious progress is made on the substantive question by this reframing, and many sources use the uncategorized-root-plus-categorizing-head system in noncrucial ways, simply as a matter