Reflection: Presuppositional resources in other languages and change-of-state verbs

As briefly noted in Table 3.1, English has at its disposal a group of full lexical verbs that presuppose a state or situation that contrasts with the one they express. Thus, for example, if one stops something, it presupposes that something had started. There are two main groups of such verbs, one relating to the beginnings of such states (e.g. begin, start, commence, get underway) and one relating to the endings of such states (e.g. stop, finish, terminate, break off). However, other languages can achieve similar presuppositions through grammatical aspect.

Aspect refers to the completeness or wholeness of an action, event or state – hereafter in this box, the situation. It is not to be confused with grammatical tense, which refers to the location of the situation in time. English uses certain combinations of auxiliaries and participles to express a contrast between the perfect and imperfect aspect. For instance, he had washed the dishes means that the situation of washing the dishes is complete, whereas he was washing the dishes means that the situation of washing the dishes is ongoing. Aspect presents the speaker’s perspective on the situation and not some kind of absolute reality. So it is perfectly possible to talk about situations that are not complete at a particular point in the past (e.g. he was washing the dishes), even though they are now complete at the point when the speaker talks about them.

The point of relevance for us here is that other languages can express other aspectual meanings through the grammar of verbs. “Ingressive”, “inceptive” or “inchoative” aspect, broadly understood, involves an inflection, affix, auxiliary verb or grammatical particle placing a focus on the beginning of a situation. Let’s imagine English had an inflection “begin”, then the sentence he was wash-begin the dishes would express the meaning “he was just beginning to wash the dishes”. “Egressive”, “cessive” or “terminative” aspect, broadly understood, involves an inflection, affix, auxiliary verb or grammatical particle placing a focus on the ending of a situation. Let’s imagine English had an inflection “end”, then the sentence he was wash-end the dishes would express the meaning “he was just ceasing to wash the dishes”

A genuine ingressive example from another language would be the case of the prefix za- and the role it plays in certain Slavic languages, notably Russian, Ukrainian, Belarusian and Bulgarian (cf. Dickey 2000). Thus, in Russian zapetj means “to begin singing”, zaplakat “to begin crying” and so on. Another example would be the inflection –sco in Latin, thus giving, for example, pallesco “to begin to turn pale” or nigresco “to begin to turn black” (cf. Kennedy 1985). The key point is that such aspectual resources presuppose a state or situation that contrasts with the one they express. If one expresses the sense “to begin singing” or “to begin to turn pale” employing aspectual resources, one presupposes that the target was not already singing or already pale.

Presuppositional triggers do not determine presuppositions, but rather have the potential to trigger an inference about what is presupposed. What is presupposed is defeasible, which means that what is inferred can be cancelled or suspended in particular contexts. But what are those contexts? Consider a negative version of our earlier sentence: we are not glad to have written this topic. Here, the presupposition that we have truly written is not cancelled. The ability of presuppositions to survive the negative counterpart of a positive sentence is seen as a key test of presuppositions, the constancy under negation test. One reason why what is presupposed generally survives such negation is that a presupposition is not, generally, what is asserted by the sentence (a point made as early as Frege 1892: 69, quoted in Atlas 2004: 31). What is asserted is often new information (the fi gure); presuppositions are often old information (the background) (we will return to this point in section 3.4). Thus, in we are/are not glad to have written this topic, we assert our gladness or lack of it in having written this topic, but we always presuppose that we truly have written it.

However, there are also more indirect kinds of negation that can cancel presuppositions. Consider example [3.8]:

The because is a definite NP, triggering the presupposition that a “because” exists. Note that the fact it appears in a negative sentence (Do not seek ...) does not cancel this presupposition (cf. constancy under negation). However, in love there is no because does cancel it. This particular kind of indirect negation has been termed metalinguistic negation (e.g. Horn 1985). Metalinguistic negation is a somewhat controversial notion, but generally such constructions invite the reader/listener to reflect back on a previous utterance and correct their default understanding of it in the light of some objection. Note that it is not simply a matter of contradicting what was previously said. This is why metalinguistic negation would not work with a positive version of the first clause. Seek the because – in love there is no because results in contradiction,– it seems to be incoherent nonsense (if the two parts were said by different speakers in a dialogue, it is more plausible). With the negative version, as in [3.8], the second part clarifies the negation of the first: don’t seek the because, because in love there is no because.

The survival or otherwise of presuppositions in the context of complex sentences has been the subject of much debate and is referred to as the projection problem (Karttunen 1973). This refers to the question of whether the presuppositions of an embedded clause “project” up, or survive, at higher levels of a sentence. In some instances this question arises due to the way in which presuppositional triggers can interact. Example [3.9], reporting what the captain of the ship Costa Concordia said about its sinking, is a case in point.

Consider the embedded clause in isolation: “The rocky reef that the ship hit on the night of Jan 13 2012 was not marked on his nautical charts.” Here, the definite NPs, the rocky reef, etc. and his nautical charts, trigger existential presuppositions – assumptions that these things exist. But the presence of the main verb, claims, a non-factive verb, introduces a rather non-committal stance; we are not sure whether the rocky reef, etc. and nautical charts exist. Of course, generally it is the case that if presuppositions in such complex sentences are considered in the context of the broader discourse and background knowledge, it is quite clear how one should take them (there was ample evidence, not least from TV footage, that the ship was on a rocky reef, and we can readily assume that a ship’s captain would have nautical charts). Cases where a presuppositional clash remains are thus likely to be taken as bizarre or humorous. Consider this apparently real question asked in a US courtroom trial:

Dies could be taken as a change-of-state verb, presupposing that the person is not alive, just as passes ... away could be taken as a change-of-state phrasal verb meaning that the person has moved away from life, thus presupposing that life is not now their state. Until introduces a temporal clause presupposing that he did know something about it the next morning. Of course, in our normal world view we assume that people who die are no longer sentient and thus could not know anything about it the next morning. The humour in this question then, and the reason it was posted on the web, is precisely the bizarre presuppositional incongruity created by the different presuppositional triggers.

Stalnaker (1972, [1974] 1991) was concerned with how presuppositions are understood in context. His work is the first to elaborate a full pragmatic account of presuppositions (see also Gazdar 1979). His definition of a presupposition is as follows: “To presuppose a proposition in the pragmatic sense is to take its truth for granted, and to presume that others involved in the context do the same” (Stalnaker 1972: 387–388). He explains the difference between assertion and presupposition:

The distinction between presupposition and assertion should be drawn, not in terms of the content of the propositions expressed, but in terms of the situations in which the statement is made – the attitudes and intentions of the speaker and his audience. Presuppositions, on this account, are something like the background beliefs of the speaker – propositions whose truth he takes for granted, or seems to take for granted, in making his statement. (Stalnaker 1991: 472)

Note two things: (1) presuppositions are not simply the properties of sentences but involve the beliefs of speakers; and (2) presuppositions are part of the presumed common background beliefs of participants. Regarding (1), it is this fact that means a pragmatic account is required. This is not to say that semantics has no role to play, but rather that the interpretation of the semantics is constrained by the pragmatic context, as in fact we were suggesting in the previous paragraph. This is neatly illustrated by Stalnaker (1991: 475):

For example, D.T. Langendoen points out in a paper on presupposition and assertion that normally, if one said “my cousin isn’t a boy anymore” he would be asserting that his cousin had grown up, presupposing that he is male. But one might, in a less common context, use the same sentence to assert that one’s cousin had changed sexes, presupposing that she is young. If a semantic account of presupposition is given of this case, then one would say that the sentence is therefore ambiguous. On the pragmatic account, one just points to two different kinds of situations in which a univocal sentence could be used.

Regarding (2), it is obviously more economic not to have to spell out all one’s assumptions, and presuppositions are one way of achieving that. However, whether presuppositions involve common ground is more controversial (see, for example, Abbott 2000). Stalnaker was aware of some of the difficulties, pointing out that defining presupposition in terms of common ground works “in normal, straightforward serious conversational contexts where the overriding purpose of the conversation is to exchange information”, but that “difficulties” “come with contexts in which other interests besides communication are being served by the conversation” (1991: 474).