"Punctuality" and Verb Semantics

Stefan Engelberg
University of Wuppertal

It has been a longstanding issue since the beginning of research on aktionsarten in the last century whether verbs have to be marked as punctual vs. non-punctual (durative). The question has been revived by recent work on aspect and aspect composition, and especially Verkuyl has argued against such a distinction. The question of the need for the punctual-durative distinction for linguistic reasons has always been accompanied by the question of what it means for a verb to be temporally punctual. The talk will present several linguistic arguments in favor of a punctuality-feature, and will provide an answer to the question of the meaning of punctuality by consulting psychological research on cognitive time concepts. Among the linguistic constructions whose grammaticality is based on the punctuality restriction is the German an-construction which alternates with an accusative object. The construction is restricted to verbs that express a change of state. Among these, only durative verbs like "schreiben" 'write', "bauen" 'build', "nähen" 'sew', but not punctual verbs like "sprengen" 'blast, blow up', "brechen" 'break', "knicken" 'fold' are admissible:

     
(1)  "sie baute ein Ferienhaus / an einem Ferienhaus"
      she built a cottage / at a cottage
    (aprox.: 'was still in the process of building a cottage')
     
(2)  "sie sprengte das Ferienhaus / *an einem Ferienhaus"
      she blasted the cottage / at the cottage

Aspectual adverbials of point of time ("at five o'clock"), span of time ("in five minutes"), and of duration ("for five minutes") show a strong affinity to the punctual-durative distinction: i) Adverbials of point of time describe the time of the event when combined with punctual verbs, whereas they trigger tense-specific interpretations when combined with durative verbs (which will be shown with German examples); ii) durative but usually not punctual verbs combine with time-span adverbials (though punctual verbs may be found with time-span adverbials if they carry a presupposition about the occurrence of a preceding event); iii) in the context of durative adverbials, punctual but not durative verbs receive an iterative interpretation.

Furthermore, punctuality is one of the factors that explains the incompatibility of some verbs with the progressive, and aspectual verbs like "stop", "finish", "start" either exclude punctual verbs as complements or provide them with a particular (e.g. iterative) interpretation. Obviously, there is much linguistic evidence for distinguishing punctual from durative verbs. But what is the meaning of the predicate "punctual" and what is it a predicate about? It has long been noticed that events described by punctual verbs like breaking, blasting, jumping are not completely without duration. The way out chosen by many researchers is to assume that punctuality is not a property of the event described by the punctual verb, but rather the property of the verbal predicate to present events as having no duration. But that doesn't make much sense either. If the event has duration why should we pretend it hasn't? Why - in other words - should we constantly lie about the true duration of events?
A closer look at the psychological research on time concepts points to a solution. Among the time concepts shown to have cognitive relevance is an interval of two to three seconds. The relevance of this interval shows up in the follwing experiments and observations (among others): i) the two visual "readings" of ambivalent pictures like the Necker-cube alternate involuntarily every two to three seconds; ii) the duration of visual or acoustic stimuli is overestimated with short stimuli and underestimated with long stimuli; the threshold between over- and underestimation lies around two or three seconds; iii) when listening to metronome beats, two adjacent beats are perceived as a unit (a "tick-tock"-effect), but only if the pause between the two beats does not exceed about two and a half seconds; iv) intercultural investigations of rhythmic actions like knocking, hammering, handshaking, scratching, or waving show that these activies come in iteration bundles which are separated by short pauses every three seconds; v) spoken lyrics tend - independent of language or literary tradition - to include short breaks every two or three seconds; a fact that cannot be reduced to breathing requirements.
This short interval of time has been interpreted as a unit of consciousness which constitutes our impression of the present. This "moment of cognition" is longer or about as long as the events we refer to with "punctual" verbs. That is, a punctual verb is a verb that refers to events that do not exceed the length of a moment of cognition (2 - 3 seconds) while a durative verb refers to events that do exceed that interval.
Formal semantics as it is employed in theories of aspect composition tends to avoid concepts that cannot be defined in logical theories of time, thereby ignoring the relevance of these conepts for an adequate explanation of linguistic phenomena. While there is no need for (lexical) semanticists to give up the theoretic rigor of formal semantics, the basic predicates and ontological sorts we employ in formal semantic theories do have to be given a cognitive interpretation.


< Back to List of Abstracts | Back to PLC23 Home Page >

About the PLC23 Committee
Previously held Penn Linguistics Colloquium: PLC22 (1998), PLC21 (1997)

Penn Department of Linguistics
University of Pennsylvania