Measuring Events

The telic-atelic distinction has been argued to hinge on the presence of a (bounded) internal argument measuring out the event, or, alternatively, a resultative small clause providing an end point for the event. Both perspectives are partially correct and partially incorrect. On the one hand, the resultative is more adequately seen as a measure than an end point; on the other, it is the resultative predicate rather than the internal argument that performs this measuring function. Empirical evidence is adduced in support of this point of view: resultative predicates are subject to the requirement that they denote a bounded scale. Only bounded predicates can delimit an event by providing it with minimal parts. As a matter of conceptual necessity, unbounded predicates, though potentially denoting end points, cannot function as event measures.*

Introduction

The distinction between telic and atelic (bounded and unbounded, or delimited and nondelimited) events has a long and venerable tradition, both in the philosophical and the linguistic literature (Kenny 1963, Ryle 1949, Garey 1957, Potts 1965, Vendler 1967, Bauer 1970, Verkuyl 1972, Comrie 1976, Nordenfelt 1977, Declerck 1979, Dowty 1979, Bach 1981, Dahl 1981, Bennett 1981, Mourelatos 1981, Brinton 1988, Hoekstra 1988, Talmy 1988, Binnick 1991, Smith 1991, Tenny 1994, Depraetere 1995, Jackendoff 1996, to name just a few). For example, in the aspectual classification developed by Vendler (1967) and Dowty (1979), given in 1, the states and activities may be taken as unbounded, the accomplishments and achievements as bounded.

1. 1. states    know, believe, have, desire, love

   activities    run, walk, swim, push a cart, drive a car, tease Mary

   accomplishments    paint a picture, make a chair, deliver a sermon, draw a circle, run across the street

   achievements    recognize, spot, find, lose, reach, die

As the examples make clear, and as Verkuyl has pointed out (1972), the aspectual classification requires that one look not just at verbs, but at least at VPs.

1. 2.

   1. a. Mark knitted a sweater.

   2. b. Mark knitted sweaters.

Thus 2a is telic or bounded, that is it has an inherent end point, whereas 2b, with a bare plural object, is unbounded. In other words, depending on what the VP contains, the interpretation of any particular verb may shift from unbounded to bounded, or vice versa. Resultatives also illustrate this phenomenon. Thus Hoekstra (1988, 1992) has argued that any activity verb may be turned into an accomplishment by adding a resultative small clause to it, as in 3.

1. 3.

   1. a. Freddy cried.

   2. b. Freddy cried the handkerchief wet.

Used intransitively (3a), the activity of crying has no inherent end point. But in 3b, the interpretation is different in that the activity is inherently bounded. According to Hoekstra, the combination of the postverbal NP and secondary predicate expresses a state that constitutes the end point of the crying activity. Syntactically, the combination the handkerchief wet is argued to form a constituent, viz. a (resultative) small clause.

A different perspective on accomplishments is presented by Tenny (1994). Consider 4.

1. 4. Sam mowed the lawn.

The direct object is said to ‘measure out’ the activity of the verb: as the activity progresses, more parts of the lawn are mowed, until there is no lawn left to be mowed, at which point the activity must terminate. The lawn may consequently be seen as a scale or measurer of the amount of activity that has taken place. Thus, whereas Hoekstra considers the postverbal constituent in accomplishments to represent an end point, in Tenny’s view it is an event measurer. While there is no inherent incompatibility between these points of view, I shall present evidence suggesting that it is possible to construe an empirical difference between them. The evidence will show that in terms of measuring out Tenny’s view is empirically more accurate with respect to the role played by the postverbal constituent in accomplishments. This implies that Hoekstra’s view on telic sentences as involving a small clause that denotes an end point is incorrect. But I shall argue that Hoekstra is correct with respect to the essential role played by the secondary predicate itself in the aspectual interpretation of the event. Tenny’s analysis is therefore wrong in...