In this article we investigate the role of semantic relations in grammatical alternations. The specific alternation we look at is that between the proper name modifier construction, e.g. the Obama government, and the determiner genitive, e.g. Obama's government. Through the use of an experimental study in which participants were asked to rate the naturalness of the two constructions in 20 attested natural language contexts and provide paraphrases of the semantic relations in question, we tested when the two constructions alternate and whether either construction expresses semantic relations that block alternation. Our initial finding is that none of the relations we studied is categorically associated with only one of the constructions, but that certain relations – notably possession and name – are far more preferentially associated with determiner genitives and proper name modifiers respectively. Despite these ‘default’ associations, participants nevertheless identified a range of possible interpretations for many of the examples, meaning that our study simultaneously supports the opposing theoretical views of default relations and semantic underspecification. Further, our study validates the inclusion of semantic relations in genitive alternation studies as a major factor despite the notorious difficulties in their operationalisation. Animacy distinctions, although more straightforward to codify, appear to be of lesser importance. Methodologically, our study shows the value of an experimental approach as a corrective to researcher intuitions about the identification of semantic relations in context.

This article studies the emergence of a grammatical pattern, the proper noun modifier construction shown in the Obama administration, an Edinburgh restaurant. The only dedicated historical corpus study, by Rosenbach (2007, 2010), is limited in terms of time depth and data included, and suggests that only proper noun modifiers denoting places such as Edinburgh are found in the early seventeenth century. Using corpus data that span the full history of English, we trace the construction back to two Old English precursors, genitival modifiers without inflectional marking, e.g. Jericho feldes ‘the fields of Jericho’, and compounds, e.g. Easter æfen ‘Easter eve’. We combine macro-level visualisations of distributions and qualitative micro-analyses to show how these source constructions developed into the present-day English construction. The development defies simplistic views on grammatical change, but illustrates that grammatical patterns develop out of multiple sources under the influence of a multiplicity of factors. New patterns only emerge gradually and exploit existing ambiguities in the language.

My central concern is the special use of proper names in the English noun phrase first discussed by Rosenbach (2006, 2007, 2010; Koptjevskaja-Tamm & Rosenbach 2005): proper names which are used as modifiers with an identifying function, e.g. theBushadministration (‘Which administration does the noun phrase refer to? The one headed by Bush’). On the basis of a corpus study, I argue that existing analyses of Rosenbach (2007) and Schlücker (2013) fail to account for all cases; they also fail to capture the seemingly contradictory syntactic and functional properties of these proper names in a unified way. My alternative analysis is framed within Halliday's (1994) functional model of the English noun phrase, but radically thinks beyond the typical association of functions with word classes (see also Rijkhoff 2009). My proposal is that the majority of these proper names can be analysed as epithets, a function typically associated with adjectival modifiers such as theredcar. A smaller set, proper name modifiers such as aKerrysupporter, are analysed as complements (Payne & Huddleston 2002). I end by discussing the implications of this dual analysis for another open question, whether proper name modifiers are morphosyntactically phrasal modifiers or part of compounds.

On the basis of extensive corpus analysis, we reconstruct the history of very and the paths of change along which it acquired new meanings. We propose an analytical model that, firstly, assigns general semantic functions to the (sub)modifier relations in the English noun phrase and, secondly, identifies subsenses of these functions on the basis of collocational, semantic and pragmatic distinctions observed in different contexts. Thus, we arrive at a comprehensive description of the various (sub)modifier relations in which very has functioned in its history. Having been borrowed into English as part of fixed collocations such as very Ihus (‘the true Jesus’) and croice verra (‘the true cross’), very successively acquired the functions of descriptive modifier, noun-intensifier, focus marker, adjective-intensifier, classifier, postdeterminer-intensifier, quantifier-intensifier and postdeterminer. This description allows us to interpret the history of very as a paradigm case of progressive grammaticalization and (inter)subjectification involving leftward movement in the English NP (Adamson 2000). Our analytical model allows us to capture finer mechanisms of change such as collocational extension, pragmatically driven host class expansion, invited inferences and analogy.

In this article, it is proposed that processes of grammaticalization are determined and constrained not only by the source semantics of the grammaticalizing item, i.e. lexical persistence in the sense of Hopper (1991), but also by the original structure the item occurs in. This previously unrecognized feature of grammaticalization is referred to as structural persistence. The need to distinguish a structural equivalent to lexical persistence is argued on the basis of a particularly exemplary case, viz. the grammaticalization processes found with one lexically specific set of grammaticalizing elements in English, adjectives of difference such as other, different, various, etc. Before their grammaticalization, these adjectives occur in two different structural configurations, viz. (1) external comparison, in which the adjective describes a relation of difference between the referent of the noun phrase and a second, separately coded, entity, and (2) internal comparison, in which the entities that are said to be different are all denoted by the noun phrase containing the adjective. Even though they undergo the same general semantic process of grammaticalization and delexicalization in both structures, the adjectives acquire a different grammatical function in each of them. The different outcomes of the grammaticalization process can only be explained by relating them to the specific properties of the two source structures.

In this article we make a case for recognizing deictification as a type of grammaticalization and semantic shift in the NP analogous to auxiliarization in the VP. The specific analogy we point out is between lexical verbs that grammaticalize into secondary auxiliaries bound by the finite, as in is going to, has to + verb, and lexically full adjectives that grammaticalize into postdeterminers bound by the primary determiner, as in a different, the same + noun. We present five case studies of the development of postdeterminer meanings, based on the analysis of diachronic and synchronic data. The adjectives studied are opposite, complete, old, regular and necessary, whose postdeterminer uses relate to the basic deictic systems of space, quantity, time and modality. Our analysis of the data shows that the mechanism of secondary deictification can be given a unified characterization as the semantic shift by which a general relation expressed by the adjective is given a subjective reference point in or relative to the speech event.