Nederlandse Taalkunde - Volume 18, Issue 3, 2013

This article deals with the cross-linguistically rare phenomenon of past tense imperatives and, more precisely, with the past perfect imperative in Dutch. With this construction, the speaker typically conveys the undesirability of a state of affairs that cannot be changed anymore and reprimands the addressee. The article examines the two functional accounts that have been proposed in the literature. Proeme (1984) provides a definition of the imperative that is so broad that no a priori restrictions of verb types, aspect and tense can be assumed. It is argued that his starting point is debatable from a usage-based perspective and that his account cannot explain the construction’s standard interpretation, the dubious acceptability of other interpretations and its scarcity in the world’s languages. Duinhoven (1995) regards the past perfect imperative as the outcome of the interaction of the counterfactual conditional inversion construction and the (conditional) imperative construction. His fairly idiosyncratic account is evaluated positively in the light of construction grammar and the typological literature on directive strategies and insubordination. The assumed processes of analogy and conventionalization are argued to be at odds with the construction’s infrequency in corpora, though.

In this paper we study three frozen imperative constructions in Dutch (geef op!, opgehoepeld!, and opgepast!) and inquire their highly specific reference properties: all arguments disallow for displaced reference, similar to what we see in animal communication (‘forum-readings’). We show that these seemingly idiosyncratic, frozen, and isolated properties can be derived from principles drawn from productive language competence. We show that the block on displaced reference realizes an endpoint on a scale of forum-readings. The extreme properties are compositionally imported into the construction by a resultative particle (e.g. op) that projects a specifier that is non-thematic. Dutch imperative adverbs (toch, maar, eens) are classified according to their quantificational properties. They import distinct possibleworld semantics of which the forum reading is a special case. The null spellout of arguments in the op-construction is tied to the one-phase nature of these utterances.

This article reviews Binary Tense theory as developed by Verkuyl (2008) on the basis of three oppositions proposed by Te Winkel (1866): present—past, synchronous—posterior and imperfect—perfect. Verkuyl followed Te Winkel in assuming that the verb zullen (‘will’) can be used as a temporal auxiliary expressing posteriority, but here it will be argued that zullen should be considered a purely modal verb and thus does not contribute to the temporal meaning of the clause. The notion of posteriority is, in fact, an integral part of the meaning of past and present, and can be brought to the fore by the use of temporal adverbs like morgen ‘tomorrow’ or by pragmatic considerations. The article discusses several proposals in the Dutch linguistic literature in which the modal nature of zullen is recognized, but they fail to be convincing in the absence of a crucial innovation: the rejection of the point of speech n as the present in favor of assuming a present tense domain in which n is a moving point that splits the actualized part of the present from its non-actualized (modal) part.

This contribution discusses three problems with the analysis of Dutch zullen (‘will’/’shall’) proposed by Verkuyl & Broekhuis. First, they do not distinguish systematically between instances in which the finite form is constituted by the main verb and instances in which the finite form is an auxiliary such as zullen; the semantics they assume for PRES is too vague to capture the interpretation of tensed forms, especially that of epistemic modals. Second, both diachronic and synchronic evidence suggests that there cannot be an absolute distinction between the temporal and the modal reading of zullen. Third, tense itself may also be used to convey modality. This is particularly clear for the past tense of zullen and it cannot be captured by a temporal definition of PAST.

This reply addresses Ronny Boogaart’s comments on our contribution Temporaliteit en modaliteit. A general problem with these comments is that Boogaart does not seem to appreciate that we make a strict distinction between semantics (the meaning of the sentence) and pragmatics (the use of the sentence in a given context). The former determines the situations in which a given sentence is true, whereas the latter determines a subset of these situations that are ‘intended’ by the speaker. The central role we attribute to pragmatics in determining the actual location of the eventuality in the relevant present/past tense domain immediately refutes RB’s claim that we do not explicitly refer to contextual factors.