Elementary semantic properties of verbs

I assume that what we might call the basic sense of a word is typically expressible as a set of components, and that while some of these components may be idiosyncratic to particular words, others are common to possibly quite large classes of words. The components themselves may be complex, since they may be required to characterize events or situations that are themselves complex, but the ultimate terms of a semantic description I take to be such presumably biologically given notions as identity, time, space, body, movement, territory, life, fear, etc., as well as undefined terms that directly identify aspects of or objects in the cultural and physical universe in which human beings live.We shall sample certain elementary semantic properties of verbs, in particular those relating to time, space, movement, and ‘will’.

Some verbs refer to activities viewed as necessarily changing in time, others do not; this contrast is frequently referred to with the terms ‘ momentary ’ and ‘ continuative ’ respectively. Sleep is a continuative verb, wake up is a momentary verb. A continuing activity, or state, necessarily occupies a span of time, and thus it makes sense to qualify a continuative verb with a complement which identifies one or both of the end-points of such a span, or a distance-measure of the span. Thus while the sentences in (n) make sense, those in (12) do not:

On the other hand, the negation of a momentary verb can identify a continuing state; hence, the sentences in (13) make sense:

Momentary verbs that represent acts that are repeatable may be understood ‘iteratively’. Wake up is not iterative, as is shown by example (12). Kick, however, can be understood iteratively, as we see in (14):

(14) He kicked the dog until 5 o’clock.

Momentary verbs that are also ‘ change-of-state ’ verbs cannot be used iteratively when a specific object is involved, as we see in example (15):

(15) *He broke the vase until 5 o’clock.

But if the same activity can be directed to an unspecified number of objects, then change-of-state verbs can be understood iteratively, too, as we see in (16):

(16) He broke vases until 5 o’clock.

It appears, in short, that a lexico-semantic theory will have to deal with those aspects of the meanings of verbs which relate to the occurrence in time of the situations which they identify.

Turning to other types of semantic properties, we note that verbs like hit and touch, though they differ in that the former is momentary while the latter may be either momentary or continuative, have in common the notion of surface-contact, a property they share with knock, strike, contact, impinge, smite, and many others. They differ in that the impact of the described acts is apparently gentler for touch than for hit.

The verbs leap and jump agree in implying a momentary change in vertical position (one has to leave the ground in order to perform either of these actions), but they differ in that leap seems to imply a change in horizontal position, too. Slide, like leap, refers to position-changes along a surface, but differs from leap in not implying movement away from the surface. Scuttle, like slide, suggests movement across a surface, but with the assumption that contact with the surface is interrupted and with the further sense that the motion is rapid. Dart is like scuttle in referring to rapid sudden motion, but fails to share with it any reference to a surface. Verbs of motion, in short, may be described by associating them with properties relating to direction, speed, gravity, surface, etc.¹

Sometimes a verb has a built-in reference to the outcome of an activity. Conceptually it appears that the actor engages in some activity and though the activity may be directed toward some specific outcome it is the activity itself which (by chance) leads to that outcome. These have been called ‘achievement verbs’. One of the tests of an achievement verb is that the modal may is usable in construction with such a verb only in its epistemic or predictive sense, not in its pragmatic or permission-granting sense. This is apparently because of the ‘by-chance’ relationship between the activity and the outcome: one doesn’t grant someone permission to have good luck. Hence we find (17) and (19) understandable only in the epistemic sense of may, while (13) and (20) can be understood in either the epistemic or the pragmatic sense:

(17) He may achieve his goal.

(18) He may try to achieve his goal.

(19) He may find the eggs.

(20) He may look for the eggs.

A final general property of verbs that we may point out has to do with the intentional or non-intentional involvement of one of the participants in the events described by use of the verb. If we compare (21) and (22):

(21) John means x by y

(22) John understands x by y

we note that in (21), but not in (22), the association between x and y is intentional on John’s part. The word mean can be used in the sense which understand has in (22), but in that case the sentence is differently constructed. (23) is a paraphrase of (22):

(23) Y means x to John.

¹ Possibly the richest source of insights into verbs of motion is the recent output ofJeffrey S. Gruber. See his dissertation, Studies in lexical relations, M.LT. (1965, unpublished); ‘Look and see Language (1967), XLIII, no. 4, 937-47 ; and Functions of the lexicon in formal descriptive grammars (1967), Technical Memorandum TM-3770/060/00, System Development Corporation, Santa Monica, California.