The Editors and Cambridge University Press are delighted to announce that commencing with the present issue English Language and Linguistics is formally associated with the International Society for the Linguistics of English (ISLE; www.isle-linguistics.org) as its official journal. We believe that ELL and ISLE are a perfect match in that both aim to promote research on the English language worldwide. We are confident that this exciting collaboration will be beneficial to the field of English linguistics and all its practitioners. We are working with ISLE on benefits for members; details will be made available on the ISLE website during 2012.

It is claimed in van Kemenade (2000: 62) that clauses with initial negative constituents are a context in which subject–verb inversion occurs throughout the history of English. However, different patterns of negative inversion are seen at different periods of English. I argue that changes in the availability of negative inversion reflect changes in the way sentential scope for negation is marked in negative concord constructions. Thus, negative concord involving Middle and Early Modern English not does not co-occur with negative inversion, but negative concord involving Middle English ne does. Changes to negative inversion can be seen to parallel changes in the way sentential scope negation is expressed at successive stages of the Middle English Jespersen Cycle. I propose that the changes to negative inversion and Jespersen's Cycle should both be analysed as changes in the ability of negative items to mark sentential scope for negation. This observation can be formalised within a Minimalist framework as variation in the LF-interpretability of negative features, following the account of Jespersen's Cycle proposed by Wallage (2008).

Brinton & Traugott (2005) and Brinton (2008) have suggested that light verb constructions of the type take a look (at) are instances of grammaticalisation. This article shows that this is because the emphasis has been on the verb take. Exploring the light verb construction take prisoner, we see that one and the same construction involves both lexicalisation and grammaticalisation processes. For grammaticalisation, the focus will be on the semantic bleaching of take and the productivity of the pattern take + NP. For the lexicalisation of the construction, we will focus on the increasing fixedness of the collocation take prisoner, evident from the decreasing acceptability of the pattern make prisoner, and on the decategorialisation of the original NP prisoner, which is manifested in the loss of plural -s inflection in prisoner. The article further investigates the decategorialisation of prisoner, revealing that the word order of prisoner(s) relative to its complement NP (e.g. take the men prisoner(s) vs take prisoner(s) the men) has a considerable effect on the speed of plural s-deletion.

Definite article reduction (DAR) is a dialectal variable confined to the North of England. In DAR dialects, the standard article alternates with a reduced one, which is mostly realised as a glottal stop and sometimes as a voiceless fricative before vowels. DAR is a salient sociolinguistic marker in the sense of Labov (1972) and Trudgill (1986). This article argues that its salience is derived from its status as a good boundary marker that listeners can utilise in order to segment the speech stream.

Salience is the property of a variable which makes it cognitively or perceptually prominent both for speakers of the dialect and for speakers of other dialects. DAR is a salient marker inasmuch as it shows variation and style shifting, can be an identity marker, and has long been recognised by layperson and linguist alike as a typical feature of Northern speech.

It will be argued that DAR is salient since it is a good word-boundary marker. The reduced article constitutes domains of low transitional probability, which listeners exploit to segment the speech signal. It has been recognised that word segmentation plays a major role in speech processing and that listeners use statistical inferences (besides other kinds of information) to locate word boundaries. The conclusion is that the salience of DAR can be derived from its distributions, as these distributions result in the variable's perceptual prominence.

The factors governing word order in binomials, i.e. coordinated word pairs of the same word class, have been the subject of discussion for a long time in linguistics. For example, why do we say law and order but not order and law? The article tests seventeen different potential ordering constraints that have previously been suggested, from the areas of semantics, phonology and word frequency, by checking over 500 high-frequency binomials extracted from the BNC against them. A clear hierarchy of constraints is suggested following the analysis of their success in predicting binomial order. In addition, however, attention is drawn to the reversibility status of binomials. The vast majority of English binomials is reversible to a smaller or larger degree. Reversibility scores were computed for all binomials in the sample so that the relationship between reversibility on the one hand and the adherence to the ordering constraints on the other could be analysed, finding that a number of semantic and metrical ordering constraints indeed increase their predictive success towards the frozen end of the reversibility cline. Complying with these constraints, then, increases the likelihood of a binomial to be less reversible. Claims for the influence of certain factors on the freezing process are thus substantiated for the first time.

This article discusses deletion in spoken and written English. It notes that subjects are frequently dropped both in informal spoken English (Napoli 1982; Zwicky & Pullum 1983b) and in certain registers of written English such as diaries (Haegeman 1990, 1997, 2007; Haegeman & Ihsane 1999, 2001). The article argues in favour of Napoli's phonological analysis of left-edge deletion in spoken English, and provides a formalisation of Napoli's account in the framework of Selkirk's (1995, 2001, 2011) optimality-theoretic analysis of syntax–phonology mapping. A comparison is drawn with the case of subject drop in the diary register. Due to the difference in surface distribution of the phenomenon between the spoken and written cases, the analysis cannot transfer directly. However, I suggest that, combined with arguments made by Haegeman (2002) for a sentence-medial position for modifiers in written English, the phonological analysis can account for a large subset of the diary drop cases.

This article attempts to repair the neglect of the qualitative uses of some and to suggest an explanation which could cover the full range of usage with this determiner – both quantitative and qualitative – showing how a single underlying meaning, modulated by contextual and pragmatic factors, can give rise to the wide variety of messages expressed by some in actual usage. Both the treatment of some as an existential quantifier and the scalar model which views some as evoking a less-than-expected quantity on a pragmatic scale are shown to be incapable of handling the qualitative uses of this determiner. An original analysis of some and the interaction of its meaning with the defining features of the qualitative uses is proposed, extending the discussion as well to the role of focus and the adverbial modifier quite. The crucial semantic feature of some for the explanation of its capacity to express qualitative readings is argued to be non-identification of a referent assumed to be particular. Under the appropriate conditions, this notion can give rise to qualitative denigration (implying it is not even worth the bother to identify the referent) or qualitative appreciation (implying the referent to be so outstanding that it defies identification). The explanation put forward is also shown to cover some's use as an approximator, thereby enhancing its plausibility even further.

Much of the research that has been carried out into the functions of actually and – to a lesser extent – really has focused on their so-called ‘discourse functions’. However, when they appear medially both actually and really are usually classified as intensifiers, and it has been argued that they are often interchangeable (see for example Lenk 1998; Oh 2000; Taglicht 2001). The purpose of this article is to test current thinking on this question by casting further light on the way medial actually and really are used in spoken discourse. Two complementary approaches are taken. Firstly, the interchangeability hypothesis is assessed on the basis of quantitative analyses of data from the British National Corpus. Secondly, the question of the extent to which actually and/or really function as intensifiers in preverbal position is addressed via a detailed qualitative analysis of data from a small corpus of recent BBC radio broadcasts of the panel-based political discussion programme Any Questions. The analyses presented here suggest that the interchangeability hypothesis is untenable and that the two adverbs have different core meanings, with any intensifying function being largely the result of interplay between the distinct semantic properties of each adverb and the discourse context.