Uncovering the Scale: On the Interaction Between the Semantic Content of Roots and Functional Structure

The aspectual flexibility of verbs is an established fact, but the mechanisms for accounting for this flexibility remain controversial.  One approach, articulated most succinctly in Borer’s work (2005), suggests that functional structure completely determines aspectual composition. Non-functional listemes, or roots, are said to be devoid of any grammatically relevant content – they are grammatically inert.  The semantics of roots always yield to the dictates of functional structure.  Any perceived constraints on the distribution of verbs in particular aspectual frames comes from a ‘mismatch’ between the non-grammatical conceptual content of the non-functional listemes (roots) and the semantics of the surrounding functional structure, but the constraints are not, strictly speaking, grammatical.

In this talk I put forth a different view in which roots have grammatically relevant semantic components and these components constrain the distribution of roots in functional structure.  I support this view with a detailed case study which contrasts the aspectual potential of degree achievement verbs with a class of locative verbs represented by the verb cover.  The roots from both classes minimally contrast in a strikingly instructive way: the two classes of roots are found productively in stative, non-causative eventive, and causative eventive structures with fully compositional derived syntax and semantics.  Yet, I will show that the verbs from the two classes differ in their aspectual potential: the verbs in each class have aspectually defined readings that the verbs in the other class lacks.  For example, the inchoative uses of degree achievement verbs have a range of aspectually defined readings which the inchoative use of cover verbs lack.  I will argue that the contrast in aspectual flexibility follows from constraints on the interaction between the semantics of roots and functional structure.  In particular, I will argue that the fact that degree achievement verbs are built on roots which lexically encode scale structure, whereas the cover verbs are built on roots that lack scale structure dictates that the roots from the two classes get integrated in different surrounding functional structure.  Finally, I resolve the claim that the roots of cover verbs do not lexically encode scale structure with the fact that they appear with modifiers which are classic hallmarks of scalar expressions.

The constraints illustrated are grammatical in nature – they show no flexibility.  If the content of lexical roots is grammatically inert as argued by Borer and always accommodates the interpretation imposed by functional structure, we would expect the roots from the two classes to be able to ‘accommodate’ to functional structures and show the full range of aspectual flexibility.  The fact that they do not, and the fact that it is possible to pinpoint the exact elements of semantics which determine the different surrounding functional structure, argues strongly that there are elements of meaning encoded in roots which serve as the interface with grammar.