Introduction

Chapter 1 outlines the goals of a theory of verb meaning. It presents a justification for a constrained theory of verb meaning, and outlines the basic parameters of event structural approaches and surveys several types of such approaches to isolate their key ingredients. It also considers the predictions such theories make about possible and impossible verbs, and justifies the need for a further study of constraints of root meanings in particular to realize the predictive potential of event structural approaches. The chapter outlines the two hypotheses considered in depth in this study — The Bifurcation Thesis of Roots and Manner/Result Complementarity — and the methodology used to show they do not hold and the consequences for a theory of verb meaning if they do not. It concludes by outlining some preliminary assumptions on the semantics of causation and change.

Entailments of change in the roots of change-of-state verbs

Chapter 2 examines the semantic and morphological predictions of Bifurcation, focusing on the roots of change-of-state verbs and their stative correlates. Using evidence from entailment, morphology, and sublexical modification it shows that English verbal roots fall into two classes. The flat class describe simple states, while the crack class describe states but also entail that the state came about from change, a templatic notion. A formal analysis of the semantic and morphological properties of the two classes is proposed, albeit one that rejects Bifurcation. It then considers a variety of alternative analyses that preserve Bifurcation, which all come at some theoretical and empirical cost.

Conclusion

Chapter 5 summarizes and synthesizes the results of the overall study. The deeper motivation for why roots would not be subject to constraints such as those required by Bifurcation and Manner/Result Complementarity is explored, whereby the semantic specificity of idiosyncratic root meanings may in some cases require also entailing more basic templatic notions and other types of idiosyncratic meanings. This chapter also considers alternative conditions on roots and why these may not hold. It concludes with an outline of the larger typology of roots this study predicts, and in conjunction with a theory of templates how this typology still makes predictions about possible and impossible verbs despite the fact that root meanings can be as complex and unconstrained as the case studies explored here suggest.

Manner/Result Complementarity and causation in verbal roots

Chapter 4 explores the question of Manner/Result Complementarity. It proposes that there are systematic verb classes that entail both meanings at once, including verbs of manner of killing, cooking, and ballistic motion, demonstrated by applying various diagnostics for manner and result in a verb’s meaning and showing that certain verb classes pass both sets of tests. Data from sublexical modification further show that the manner and result meanings are both coming from the sole root, arguing against an alternative by which manner plus result verbs have two roots, each introducing one meaning. The roots of these verbs can either be syntactically like canonical manner roots or canonical stative roots, and a formal analysis for how roots of each type can introduce both meanings at once is developed. These roots also all entail causation as well, arguing once again against Bifurcation.

The roots of ditransitive verbs of caused possession

Chapter 3 examines English ditransitive verbs, which show the dative alternation between indirect object and to frames, each supposedly reflecting a different template for a single manner-describing root. It shows that these two templates are semantically highly underspecified, and it is the root that fleshes out many of the surface verb’s basic entailments. These entailments include change-of-state, possession, and co-location, all of which are independently known to be templatic meanings, arguing again against Bifurcation. The root also governs whether the verb even shows the dative alternation, a root-conditioned syntactic effect. A formal analysis of root/template composition is developed that relies on manner roots being able to impose conditions on the template’s result states in ways that predict the verb’s grammatical and semantic behavior. Counterproposals that might retain Bifurcation are also considered, though it is argued that they are dispreferred for various reasons.