Chomskyan generative linguistics treats inflectional verbal endings as heads of the node IP, meaning that tense and mood morphemes must in some sense count as parallels to lexical heads. Finding substantial arguments for this alleged similarity is not an easy task. Taking critical stances by Haspelmath as a starting point, I agree that some tests for head status (hyperonymy, distributional equivalence, status as a valency bearer) will not apply, but contrary to Haspelmath, I hold that dependency does in fact apply, provided that a version of the classical distinction between endocentric and exocentric constructions is reintroduced. Tense and mood morphemes are obligatory and therefore heads, in the sense that they contract exocentric constructions. At the content side, however, these verbal inflections are similar to free adjuncts: sentence adverbials and dialogic particles, in that both modify a part of the clausal hierarchy: a predication or a proposition. Tense modifies predications, mood modifies propositional structure, and free adjuncts – depending on the subclass – modify predications or propositions. This is not captured in generative frameworks since free adjuncts are modifiers and inflectional endings are heads. As parts of morphology, however, verbal inflections are turned into exocentric co-heads of the predicate structure of the clause. Leading up to the analysis of tense and mood, I present an argument that in English and the Modern Scandinavian languages, exocentric constructions are a necessary part of the descriptive and theoretical inventory of grammar. This view is at odds, not only with minimalism, but also with central works in construction grammar and cognitive grammar, in which exocentric constructions are at best marginal. My point is not a formal alternative, but rather some principles for the construction of syntax as sets of empirically based rules. It is hoped that these will be found relevant by both formalists and functionalists.
Lars Heltoft (Mon,) studied this question.