This article addresses the question of why certain epistemic verbs, particularly (I) think, have very high rates of zero complementizer in their clausal complements. I consider and reject the influential proposal from Thompson & Mulac (1991b) that the occurrence of zero complementizer in these contexts is associated with syntactic reanalysis of the epistemic verb and its subject to an adverbial-like epistemic phrase. Instead, I propose that zero complementizer signals a shift in the informational structure of the utterance, in which the embedded clause conveys the main assertion, but the syntactic structure of the sentence need not be altered. Further routinization is evident with I think but not with other epistemic verbs. Results from a study of newspaper prose are presented in support of the analysis.

In this article, I examine the ‘modal’ or ‘empathetic’ (Lyons, 1977: 677) use of the distal (or ‘nonproximal’) determiner/pronoun that: namely, where the intended referent may have just been evoked in the immediately prior discourse, but where the distal pronoun that, not the ‘in-focus’ it or the ‘activated’, proximal this is used. The rationale behind the choice of this particular type of indexical seems to be that the speaker is distancing him/herself from the referent, not wishing to ascribe actuality to it in the way that would be the case if either it or this were used instead. Examination of this particular value of that leads to the hypothesis that the principles underlying the choice of that as opposed to this or it generally are not derived ‘objectively’, as it were, from their situational use in terms of degrees of proximity of a referent or demonstratum to the speaker or hearer, nor primarily in terms of attention focus. They are, rather, social and cognitive, and play an important interactional role in the construction of discourse.

This article discusses the plausibility of a correlation or even a causal relation between two phenomena that can be observed in the history of English ditransitives. The changes concerned are: first, the emergence of the ‘dative alternation’, i.e. the establishment of a link between the double object construction (DOC) and its prepositional paraphrase, and second, a reduction in the range of verb classes associated with the DOC, with the construction's semantics becoming specialised to basic transfer senses. Empirically, the article is based on a quantitative analysis of the occurrences of the DOC as well as its prepositional competitors in the Penn–Helsinki Parsed Corpus of Middle English, 2nd edition (PPCME2). On the basis of these results, it will be argued that the semantic narrowing and the increasing ability of ditransitive verbs to be paraphrased by a to-prepositional construction (to-POC) interacted in a bi-directional causal manner.

This article traces the complement selection properties of the adjective accustomed from the eighteenth century to the present. The adjective has frequently selected sentential complements, but the study illustrates a major change affecting the form of such complements. In the eighteenth century they were regularly of the to infinitive type, but today they are almost as regularly of the to -ing type. Both syntactic and semantic factors are identified in the article as having an impact on the change. The study also compares the pace of the change in British and American English, arguing that incipient change was discernible in both varieties as early as the nineteenth century. It is argued further that, at the present time, the change has been completed more fully in American English than in British English. This conclusion is reached by considering sentences with extraction or filler–gap dependencies. The data in the article come from large electronic corpora.

This article presents arguments against a uniform head-initial analysis of Old English clause structure. Three analyses that have been proposed for Old English – variation in the headedness of underlying structure, uniform head-initial structure with object movement, and uniform head-initial structure with pied piping – are presented and evaluated in terms of the Old English data that they are able to account for. In particular, it is argued that the positions of verbs and their complements in constructions with quantified and nonquantified objects, pronominal objects, particles, and double objects cannot be derived without stipulations within uniform head-initial accounts, but can be derived unproblematically within a framework that incorporates a headedness parameter, requiring only a stipulation to block V–O–Aux order.

Developing an idea first articulated by Luick in 1896, Halle & Keyser (1971) posit the introduction of a new accentuation rule in Middle English (ME), the weight-sensitive Romance Stress Rule (RSR). All post-1971 accounts of English stress take the syllable weight principle of the RSR as their starting point. For twenty-five years there has been no scrutiny of the assumption that syllable weight became relevant for stress assignment in ME. It has been claimed, and the claims have not been addressed, that the RSR is part of the phonology of late OE (O'Neil, 1973), that by late ME the RSR had completely replaced the Germanic accentuation patterns (Nakao, 1977), and that the existence of the RSR in ME justifies the reconstruction of analogical pronunciations such as hardi, holi, riding(e), a real shift from the Germanic to the Romance category (Halle & Keyser, 1971; Luick, 1896). The paper refutes these claims on the basis of new evidence and analysis in terms of generalized prosodic constraints.

The present article contrasts two processing principles, the Domain Minimization Principle (e.g. Hawkins 1999, 2004) and the Complexity Principle (e.g. Rohdenburg 1996, 2007b) in structures involving the extraction of postverbal elements out of (competing) complement clauses. The Domain Minimization Principle may be described as a processing tendency which consists in minimizing the size and complexity of various domains including the filler–gap domain in cases like the following: This is a problem (that) they had promised (that) they would tackle. By contrast, the Complexity Principle represents a correlation between processing complexity and grammatical explicitness. It stipulates that more explicit (and typically more bulky) constructional options are favoured in cognitively more complex environments.

This article deals with two sets of rival complements. The behaviour of the first group of clausal alternatives is in line with Hawkins’ prediction, though incompatible with the Complexity Principle. However, there is an even larger group of complement pairs whose distribution inside and outside extraction contexts is predicted by the Complexity Principle but unaccounted for by Domain Minimization. Thus, in extraction contexts, the visible effects of the two antagonistic principles are found with virtually complementary ranges of complement types. The article concludes by attempting to account for the kind of division of labour observed between the two principles under scrutiny. It will be suggested among other things that, in English, the marked infinitive (on its own or with an associated NP) enjoys a privileged or target status in extraction contexts.

This paper expands the analogy between events and count nouns, and between states/activities and mass nouns in English to include other situation types, including iteratives, habits, and multiple situations. It explores the evidence used to make such analogies, namely, the quantificational features of deverbal nouns, as each of the deverbalizing devices in English has its own aspectual qualities. It shows further that a number of parallels can be drawn between different types of bounding and debounding in the nominal and verbal domains. Finally, the paper presents a schema of the cross-categorial analogies relating to inherent meaning (mass/count features in nouns, situation types in verbs) and the semantic operations of (de)bounding and (de)collectivizing (quantificational and aspectual modification).

This article presents a reanalysis of the syntax, semantics, pragmatics, and phonology of the word actually. It endeavours to show, among other things, that in addition to the distinction between Greenbaum'’s ‘disjunct’ and his ‘adjunct’, we seem to require here a distinction between two kinds of ‘disjunct’: (i) a ‘propositional modifier’ and (ii) a ‘discourse modifier’ (or ‘discourse transition marker’). An examination of constraints on the use of actually as a propositional modifier suggests the need for a feature ‘assertive/nonassertive’, not just on root sentences but on all structures that denote propositions, whatever their syntactic form.

This paper examines the polysemy in the English verb get, which can denote possession, movement, causation, obligation, and change of state among other senses. The analysis builds on a decomposition of get (based on the Benveniste/Freeze/Kayne analysis of possessive constructions) into [ingressive + ‘be’ + preposition]; this lexical entry allows the current polysemy to be derived from a number of reanalyses within different syntactic contexts. Using diachronic data, I show that possession leads to movement as well as stative uses (possession and obligation), movement develops into the causative and inchoative, from which the passive develops, and the infinitival causative gives rise to permission and ingressive aspect. The appearance of each new meaning-construction is motivated by context-dependent mechanisms of reanalysis which account for language change as the result of the language learner reanalyzing the correspondences between syntactic and semantic elements.

In Present-day English (PDE), the to-dative construction refers to clauses like John sold/offered/mentioned/gave the books to Mary, in which a ditransitive verb takes a Recipient that is expressed as a to-Prepositional Phrase (to-PP). This study examines the to-dative construction in Old English (OE). I show, first of all, that this construction was not rare in OE, in contrast to what has been suggested in the literature. Second, I report on two corpus studies in which I examined the ordering behaviour of the NP and the to-PP. The results of the first study suggest that the same ordering tendencies already existed in OE as in PDE: both the NP-to-PP and the to-PP-NP orders were grammatical, but the NP-to-PP was the most frequently used one. However, in OE, the to-PP-NP was more common than in PDE, where its use is heavily restricted. My second corpus study is informed by the multifactorial approach to the English dative alternation and uses a mixed-effects logistic regression analysis to evaluate the effects of various linguistic (verbal semantics, pronominality, animacy, definiteness, number, person and length) and extra-linguistic variables (translation status, time of completion/manuscript) on the ordering of NP and to-PP. The main finding is that, generally speaking, the same factors that motivate the dative alternation in PDE were involved in OE as well. No evidence was found for the influence of verbal semantics nor of the extra-linguistic variables. Finally, I argue against the view that to was semantically reanalysed from a Goal to a Recipient marker from OE to ME. Building on evidence that the Recipient use of to was already embryonically present in OE, I make the case that this semantic change was far more gradual than traditionally assumed.

The so-called invariant tags, such as eh, okay, right and yeah, are extremely frequent in general English speech and have been studied extensively in recent years, especially in the spoken expression of teenagers, where they are a very common feature.

In this article I focus on innit, as in She love her chocolate innit? and It was good innit? For this purpose, I analyse and discuss data extracted from two comparable corpora of teen speech: the Bergen Corpus of London Teenage Language (COLT), compiled in 1993, and the Linguistic Innovators Corpus (LI), created in 2004. The analysis considers the syntactic, semantic and pragmatic features of the expression, and asks three fundamental questions: has innit remained stable in use over time or has it increased in frequency? Has innit preserved its original grammatical features, or has it undergone notable changes? To what extent is innit especially associated with teentalk, at least in London speech, rather than with the speech of adults? Contrasts are also drawn with the standard tag isn't it.

Findings confirm that innit is typical of the language of London teenagers and has not gone out of use; on the contrary, its frequency has increased over the last few years. In contrast, the proportion of tokens found in the language of their adult counterparts is rather marginal. At present, innit conserves syntactic features of its own: it does not follow the regular question tag formation rules and can represent not only the verb be but also do, have and most of the modal verbs. Furthermore, it continues to show a high degree of flexibility in the sentence, occurring not only in final but also in initial and medial positions. Finally, it appears that innit should no longer be regarded as a simple invariant tag. It tends to behave more and more like a pragmatic marker serving to express the speaker's attitude to the content of the message, thus often reflecting the relationship between the participants in the interaction, and also contributing to the organisation of the discourse. In this respect, two new discourse functions of innit are identified and described: emphatic and text organiser.

In previous work, we have found that New Zealand listeners who produce merged tokens of near and square can accurately distinguish between the vowels in perception even though they report that they are guessing. The ability to distinguish the vowels is affected by a variety of factors for these listeners, including the likelihood that the speaker and experimenter maintain the distinction (Hay et al. 2006b; Hay et al. 2010). In this article, we report on experiments that examine the production and perception of real and nonsense words in the context of two mergers: the Ellen/Allan merger in New Zealand English and the lot/thought merger found in American English. The results demonstrate that speakers’ degree of merger depends at least partially on whether the word is a real or nonsense word. Additionally, the results indicate that a token's real word status affects the merger differently in production and perception. We argue that these results provide evidence in favour of a hybrid model of speech production and perception, one with both abstract phoneme-level representations and acoustically detailed episodic representations.

This article investigates the omission of the that-complementizer in extraposed subject clauses (e.g. It is obvious (that) she did it), which has so far received very little attention in the literature. Using corpus data from the British component of the International Corpus of English (ICE-GB) various potential conditioning factors are analysed. Of these the following are found to have a major impact on the choice of zero that: informality of the text category, type of matrix predicate, and information value of the complement clause. On the basis of these corpus results the article then proposes a unified explanation for the use of that or zero by positing an underlying abstract feature of ‘distance’; for the that-complementizer, a semantic residue of its original demonstrative use, which gives rise to different pragmatic interpretations, depending on the actual use of that in context.

Basing myself on synchronic and diachronic data analysis, I argue in this article that size nouns (SNs) such as bunch/load(s)/heap(s) of within binominal NPs display synchronic variation which can be hypothesized to be the result of grammaticalization processes. Synchronically, I propose that SNs have two major non-head uses, a quantifier use, e.g. a bunch of people walked in, and a valuing(-quantifying) use, in which the referent is evaluated rather than quantified. The latter is restricted mainly to bunch/load of, e.g. What a bunch of gobbledygook. The semantic and syntactic reanalysis of SNs as quantifiers has recently been acknowledged (e.g. Traugott forthcoming), but the valuing use of SNs remains largely unrecognized (see Brems 2007). On a theoretical level, it will be argued that head, quantifier and valuing(-quantifier) SN-uses synchronically have to be studied as collocationally constrained constructions in that the semantico-syntactic parsing of each SN-use links up with specific collocational patterns (Sinclair 1991). Head uses are restricted to sets of (un)count concrete nouns, whereas quantifier uses team up with all sorts of (un)count concrete as well as abstract nouns. Valuing uses show restrictions to concrete animate and abstract nouns, which they typically evaluate negatively, and have negative semantic prosody patterns, in which the SNs themselves come to predict negative collocates (see Louw 1993; Stubbs 1995; Bublitz 1996). The grammaticalization of SNs will be hypothesized to involve not only processes of semantic generalization and collocational extension, but also collocational reclusterings characterized by particular semantic prosody constraints. The latter are not traditionally associated with processes of grammaticalization and hence offer new insights into the semantic changes that may accompany grammaticalization.

This article investigates a variety of ways in which prosodic factors influence blend structure in English. Recent approaches no longer consider blends unpredictable, but the role of stress in blend formation has not been investigated in detail yet. This article addresses this problem, focusing on the role of stress in determining the switchpoint of the two bases in the blend, and on the question of what determines the stress pattern of the blend. We investigate these questions using experimentally derived forms, coined by native speakers on the basis of carefully controlled word pairs as stimuli. The results demonstrate that the length of the blend, the location of the switchpoint, and the stress of the blend are crucially determined by stress properties of the two base words of the blend, above all by those of the second word. At a theoretical level, the most important single finding is that preservation of the stress of the second word may happen independently of preservation of segmental material of the stressed syllable (e.g. préstitant from prestígious + dóminant). In contrast to stress, and contrary to earlier claims, syllabic constituency is shown to be of minor importance for switchpoint location. The theoretical implications of these findings are discussed. On a methodological level, our results show that experimentally elicited blends constitute a valid and highly useful resource for research on blend structure.

Using data drawn from twenty American plays written since 1889, this paper analyzes the usage of the weak obligation modals should and ought, which previous researchers (e.g. Coates, 1983; Leech, 1987; Palmer, 1987) have regarded as essentially synonymous. It is shown that there is a clear distinction between these words, with should expressing individual opinions and ought emphasizing a common opinion regarding the obligation in question. This use of obligation should does not occur in five plays written between 1889 and 1911, appearing for the first time in this database in a play written in 1926. The development of obligation should parallels other changes in the modal system which have taken place at the same time – as should has increased in frequency, so have the other individually-oriented modals got to and gonna, and as ought has declined in frequency, so have the other group-oriented modals must and will.

This article reports the results of a study based on a corpus of simple (i.e. only written) and complex (both written and oral) shortenings, that is, initialisms (a category comprising acronyms, e.g. laser, and alphabetisms, e.g. BBC), as well as clippings (lab), blends (motel), and abbreviations (Mr). The study intends to confirm the validity of the tools provided by the revised version of prototype theory for a systematic account of these metalinguistic categories, which are often neglected in lexical morphology. A set of defining parameters and possible values was devised and applied to the items of the corpus. As a result, these items were distributed along a continuum of central, peripheral, and borderline cases of all the categories considered. The body of the article is an in-depth account of the centre–periphery structure of the superordinate category of initialisms, with particular attention being paid to the fuzzy boundaries featuring clippings and blends. In the last section, shortenings are categorized and represented by means of a radial polycentric network of the type devised in Lakoff (1987).

The article tests the generalisation of the curvilinear hypothesis and the tendency of females to lead linguistic change in vocalic mergers on the basis of two mergers currently in progress in Charleston, South Carolina: the low-back merger and the pin–pen merger. It is based on sociolinguistic interviews with 100 informants, aged 8–90, covering the socioeconomic spectrum of the city. The speech of 90 of the informants is analysed acoustically; it is supplemented with minimal-pair tests and word list reading. F1/F2 measurements and minimal-pair test results are subjected to a series of multiple linear regression analyses, with social class, gender, age and style as independent variables. While the low-back merger is a change from below showing a female lead and a curvilinear effect of social class, the pin–pen merger shows a decreasing monotonic relationship with social class and no female advantage. The difference is argued to be due to the two mergers being at different levels of social awareness.

Variation within grammars is a reflection of variation between grammars. Subject agreement and synthetic negation for the verb be show extraordinary local variation in the Survey of English Dialects (Orton et al., 1962–71). Extracting partial grammars of individuals, we confirm leveling patterns across person, number, and negation (Ihalainen, 1991; Cheshire, Edwards & Whittle, 1993; Cheshire, 1996). We find that individual variation bears striking structural resemblances to invariant dialect paradigms, and also reflects typologically observed markedness properties (Aissen, 1999). In the framework of Stochastic Optimality Theory (Boersma & Hayes, 2001), variable outputs of individual speakers are expected to be constrained by the same typological and markedness generalizations found crosslinguistically. The stochastic evaluation of candidate outputs in individual grammars reranks individual constraints by perturbing their ranking values, with the potential for stable variation between two near-identical rankings. The stochastic learning mechanism is sensitive to variable frequencies encountered in the linguistic environment, whether in geographical or social space. In addition to relating individual and group dialectal variation to typological variation (Kortmann, 1999; Anderwald, 2003), the findings suggest that an individual grammar is sensitively tuned to frequencies in the linguistic environment, leading to isolated loci of variability in the grammar rather than complete alternations of paradigms. A characteristic of linguistic variation that has emerged in distinct fields of enquiry is that variation within a single grammar bears a close resemblance to variation across grammars. Sociolinguistic studies, for instance, have long observed that ‘variation within the speech of a single speaker derives from the variation which exists between speakers’ (Bell, 1984: 151). In the present study, individual patterns of variation in subject–verb agreement with affirmative and negative be extracted from the Survey of English Dialects(SED, Orton et al., 1962–71) show striking structural resemblances to patterns of interdialectal, or categorical, variation.