THE DURATIVE TEMPORAL SUBORDINATING CONJUNCTIONS: SINCE AND UNTIL

D. S. Brée · Journal of Semantics · 1985

The temporal subordinating conjunctions fall into two categories, durative and non-durative, depending on the length of time for which the main proposition is predicated to hold. Formally the two durative subordinating conjunctions in English, temporal since and until, are usually treated as though they were symmetric about the time of reference (henceforth the TOR). I examine this assumption from three points of view: the time relationships between the main and subordinate propositions, the truth conditions of the two propositions, and the causal relationship that may be inferred between the states or events referred to by the main and sub propositions. The analysis is based on samples from the Brown University corpus In order to distinguish between the temporal and inferential meanings of since four criteria are needed, corresponding to the four different uses to which temporal since may be put. The principal criterion is that the main verb must be in the past aspect, corresponding to the use of since to indicate that the main proposition is true at some moment prior to the TOR. There are also three special uses of since: in a comparative construction, as part of an explicit reference to a time period and as a qualifier in a noun phrase. In all uses the sub-proposition indicates the onset of a time period which ends at the TOR. The main proposition must be true either once, several times or for the whole of this period, depending on the durative aspect of the main verb. The main proposition is also conversationally implied to be false before the onset of the time indicated by the sub-proposition. The sub-proposition is presupposed to be true. Until marks the end of a time period that, by default, begins at the TOR. The main proposition is asserted to hold during this period, except when until is being used in one of the same three special uses as we found for since. The sub-proposition is presupposed to be true when the TOR is in the past and conversationally implied to be true when it is in the present or the future, except in special circumstances. Usually one of four causal connections may be inferred between the sub and main propositions; the sub-proposition is either: 1) the cause of the main proposition ceasing to hold, 2) the cause of the starting of the unnegated form of the main proposition (only with negative main propositions), 3) the goal of the actor in the main proposition or 4) an unintended result of the main proposition. In some respects the properties of temporal since and until are the same, e.g. in marking the boundary of a time period, in having the same three special uses and in hardly ever having a negated sub-clause. They are symmetric with respect to the TOR, as evidenced by the use of the perfect with since and its absence (with one exception) with until. But there are also other nonsymmetric differences: the time of utterance may be in the middle of the time period specified for since; the tense of the subclause is always in the past for since but depends on the time of the TOR with until; until is always durative; for until a negated non-durative main verb is interpreted as durative; since may indicate only the cause of the main proposition, whereas until is not confined to indicating the result of the main proposition, but any one of four different possible causal relations may be inferred.

Read the paper · More papers on PaperTik