The idea of devoting a special issue to the question of future time reference in English began with the seminar we jointly convened on ‘What Future for the Future Tense in English?’ for the ESSE-8 conference held at the Institute of English Studies, University of London, between 29 August and 2 September 2006. This in turn raises the question as to why we should have thought of future time reference in English as a topic of interest in the first place. We had in fact both been interested in questions relating to tense, aspect and modality for several years, and we had long been struck by the fact that (much to the despair of many non-native speakers) English is particularly well endowed with ways of expressing future time reference by comparison with many other European languages.

This article provides the first comprehensive analysis of the development of the future time expressions will, ’ll, shall, be going to, progressive with future time reference, and be to in the course of the late modern period. The article focuses on possible reasons for the considerable changes that have taken place in the past few centuries. To what degree can the changes be described as certain forms having been (partially) replaced by others? To what degree have general or register-specific changes in discourse affected the use of future time expressions? These questions are investigated on the basis of the British part of ARCHER (A Representative Corpus of Historical English Registers).

The analysis reveals that it is a complex interaction of both types of processes that is responsible for the recent evolution of future time expressions. Redistribution processes turn out to be highly complex in themselves, going far beyond the frequently described replacement of shall by will and probably proceeding in chains. With respect to discourse change, one result is an unexpected overall decrease in the tendency of writers (and speakers) to refer to their own plans, intentions, etc. Partly responsible for this development is a discourse change in science writing, where the author has increasingly disappeared from the text, so that text structure is much less frequently expressed in terms of the author's intention. A further register-specific discourse change that the investigation brings to light is a development in diaries from an earlier restriction to reporting past events to the expression of more personal views, including hopes and fears for the future.

Most grammarians refuse to treat will as a marker of future tense in English. We examine the arguments against treating will as a tense and find them weak; the arguments in favour of treating it as a modal also turn out to be poor. We argue that will should be treated as a marker of future tense, and that its so-called modal uses are either not modal or have independent explanations. The one exception is the volitional use of will: to account for this, we propose that willingness is a semantic relic from an earlier meaning of the word.

This article describes and analyses five different ways of expressing futurity in English (shall/will, be going to, be to, the simple present and the present progressive) in a Construction Grammar framework. It suggests that the different expressions can be captured as an onomasiologically motivated family of constructions in which the single constructions are differentiated by complex co- and contextual configurations. The latter can be elegantly captured in a Construction Grammar framework since constructions by definition can include pragmatic features. Also, this article claims that constructions may be equipped with an additional ‘context slot’, in which co- and contextual information can be stored. In a final section, this article turns to the issue of tense as a grammatical phenomenon and its genesis in grammaticalisation processes. It is suggested that a Construction Grammar account can make the age-old debate about a future tense in English redundant. Instead, it complements studies in grammaticalisation and opens up some interesting perspectives on parallel developments in the onto- and phylogenesis of constructions.

This article discusses the synchronic status and diachronic development of will be -ing and shall be -ing (as in I'll be leaving at noon).2 Although available since at least Middle English, the constructions did not establish a significant foothold in standard English until the twentieth century. Both types are also more prevalent in British English (BrE) than American English (AmE).

We argue that in present-day usage will/shall be -ing are aspectually underspecified: instances that clearly construe a situation as future-in-progress are in the minority. Similarly, although volition-neutrality has been identified as a key feature of will/shall be -ing, it is important to take account of other, generally richer meanings and associations, notably ‘future-as-matter-of-course’ (Leech 2004), ‘already-decided future’ (Huddleston & Pullum et al. 2002) and non-agentivity. Like volition-neutrality, these characteristics appear to be relevant not only in contemporary use, but also in their historical expansion. We show that the construction has evolved from progressive aspect towards more subjectivised evidential meaning.

This article argues that be to is primarily a modal auxiliary expressing the necessity of future actualization of the ‘residue-situation’ (= the situation referred to by the clause minus be to). Eight possible ‘M-origins’ (= origins of the necessity) are identified. The ‘futurish’ use of be to in present-day English is closely related to these modal uses, especially to the use in which the M-origin is an official arrangement.

The modal interpretation shifts to a futurish interpretation when the emphasis shifts from the present existence of the necessity to the future actualization of the residue-situation. This shift of emphasis is accompanied by a loss of doubt about this future actualization.

In other words, the futurish reading comes to the fore when the (strong or weak) origin of the necessity is bleached, so that the hearer's attention is directed to the future actualization of the residue-situation. Various cases of such bleaching are treated. In some cases (e.g. when be to collocates with still or yet, as in He is still to keep the first of his promises), the bleaching of the M-origin is complete, so that only the sense of futurity (and hence of ‘not-yet-factuality’) is left. In some examples there is no clear difference between be to and will any more, so that the two are interchangeable within the same sentence.