Search

A notorious challenge in the study of the diachrony of English is to determine whether developments in syntax, including changing frequencies of a particular construction, or word-order changes as suggested by perceived patterns in extant texts, represent genuine linguistic changes or are due to changes in conventions of writing. What is intuitively clear, however, even to a casual eye, is that a piece of English prose from, say, the 16th-century differs markedly from texts from the 18th-century. Yet such judgements cannot be based on syntactic changes alone, since essential grammatical features of Present-Day English are in place already by the end of the Late Middle English period. As a result, these differences are often simply ascribed to the notoriously elusive domain of style. The current study attempts to come to grips with the issue of period-specific conventions of writing by focusing on features of discourse structure and textual organisation as of the Early Modern English period. It can be positioned at the meso-level between large-scale quantitative approaches of sentence-level linguistic features and detailed, small-scale discourse-analytic studies of individual texts. Texts selected for the current purpose, manuals for equine care, derive from a sub-domain of instructional writing with a long history in the vernacular. As these texts share similar communicative purposes and deal with the same "global" topics of feeding and looking after a horse, any differences between them cannot be attributed to different genres or differences in subject matter. This permits us to zoom in on 'agnates', different ways of expressing the same meanings, and allows us to see how the stylistic options selected by authors achieve the various communicative goals that have to be negotiated, such as discourse coherence or the transition to new topics. The three main sections in this dissertation offer different ways to identifying developments in discourse organisation. The first section explores the traditional corpus-based approach that is frequently used to measure the parameter of "personal involvement", an indicator of periodic style. Initially, this approach restricts itself to measuring the contribution of frequencies of individual lexical items like first and second person pronouns. Next, this section will focus on the presence and linguistic realisation of the interlocutors of these instructional texts, i.e. the writer and the reader. The second main section will try to diagnose such varying styles by employing a completely data-driven, quantitative methodology which offers a linguistically unbiased and theory-independent perspective on the data in the corpus. This second approach offers cues as to how `subliminal' patterns of grammar may affect perceptions of style, and how quantitative measures may aid in assessing whether the texts in our corpus cluster in expected or unexpected ways. The third section draws on theories of referential coherence and textual progression. By charting the variation with which texts from different periods in the history of English apply conventions for discourse organisation, it offers an insight into developments of hierarchical discourse structures (i.e., coordinated versus subordinated discourse relations) and practices of co-reference. Taken together, these three independent measures offer a novel, multi-angled approach to stylistic developments in prose writing. Combining features `above the sentence' level which involve discourse and information structural changes, this dissertation affords a glimpse into the emergence of written textual conventions, or 'grammars of prose', in the history of English.

In Middle English and Scots, instances of gan and can behave differently from etymologically related beginnen in that they are mainly, or exclusively, found with the plain infinitive and with a non-ingressive meaning. They also occur in narrative verse (rhymed and non-alliterative), where they have a metrical, intensive-descriptive or textual function. All of this suggests that gan and can are more advanced in the divergence of their development towards auxiliation than the verb beginnen. Earlier studies mainly concentrate on the meaning and/or function of gan and can in verse (Wuth 1915, Beschorner 1920, Funke 1922, Mustanoja 1960, Kerkhof 1966, Visser 1969 and Brinton 1981; 1983; 1988 amongst others), whereas investigations by Brinton (1981; 1988; 1996), Ogura (1997; 1998; 2013) and Sims (2008; 2014) address the divergence in the development of this verb and its variant in terms of grammaticalization, but with references to Middle English in general. Studies by Los (2000; 2005), on the other hand, deal with the grammaticalization of onginnan and beginnan with the plain infinitive in Ælfric’s works. However, no studies have been carried out on whether gan and can, as well as beginnen develop differently in terms of grammaticalization in the ‘English’ of the six northern counties of England and of Scotland in the late 14th and the 15th centuries, conventionally referred to as Northern Middle English and Early Scots, respectively. With the aid of Northern Middle English and Early Scots texts from computerised corpora (The Helsinki Corpus of English Texts, The Innsbruck Corpus of Middle English Prose and The Teaching Association for Medieval Studies, as well as The Helsinki Corpus of Older Scots and A Linguistic Atlas of Older Scots), this study looks into whether: a) gan and can, as well as beginnen differ with respect to their morphological paradigms, in view of what we know about grammaticalization and the development of invariant forms? b) these verbs differ with respect to their complements, in view of claims in the literature that the more grammaticalized variant takes the plain infinitive; and c) gan and can are a development from onginnan and aginnan, originally expressing ingression but shown in the literature to have undergone semantic bleaching in Old English and in early Middle English period? This study shows that in Northern Middle English and Early Scots, gan and can display characteristics of grammaticalization, while beginnen participates in global language changes affecting the category of the verb in ME and Scots.

The evolution of the genitive noun phrase in English has been the subject of numerous studies, yet some aspects of this evolution have received less attention than others. In this study I address two of these less studied aspects: the evolution of the plural genitive noun phrase in Middle English (1150-1500), and the decline of the overtly case-marked genitive modifiers (singular and plural) in the same period. The former has generally been presented as following the same path of the singular genitive noun phrase; the latter has been all but ignored, with only a single study (Thomas 1931) which explicitly examines the use of the genitive definite article and strong adjective. The study uses text samples from two electronic corpora, the Linguistic Atlas of Early Middle English and the Penn-Helsinki Parsed Corpus of Middle English, Second Edition, as well as samples from printed editions. The texts used in the present study have been selected with the aim of covering as wide a geographical and chronological range as possible. The thesis examines how and why the number of endings for the genitive plural inflection first increased (in the period up to about 1350) and then decreased (from 1350 onward), a fluctuation not found in the singular inflected genitive noun. The number of available inflectional endings increased due to the morphophonological weaknesses of the -V ending type – the dominant ending type inherited from OE – leading to instabilities in the inflectional system which allowed alternatives to arise. However, the number of genitive plural inflectional endings then decreased, apparently affected not only by the phonological strength/weakness of the ending types but also the type of noun phrase that these were associated with. The inflectional ending which survives, -Vs, is most commonly found with genitive noun phrases in which the genitive noun is animate and the noun phrase has one of the genitive functions labelled POSSESSIVE in this study. This distribution of the various inflectional endings according to animacy and function is related to the rise of the periphrastic genitive plural noun phrase. The initial preferred environment for the periphrastic genitive construction is noun phrases with those functions which will be referred to as NONPOSSESSIVE. As the inflected genitive becomes increasingly restricted to a single noun phrase type, the periphrastic construction expands, to become the default genitive construction by the end of the period. The thesis examines the decline of overtly case-marked genitive modifiers in Middle English, both adjective and determiners. In general, the trend is that morphologically more conservative texts are more likely to preserve case-marked modifier forms, although some marked forms are more widespread due to the development of fixed expressions. Where case-marked modifiers are maintained, historical grammatical gender agreement and the strong/weak adjective distinction are often preserved. Factors which play a role in the survival of marked modifiers are chronological distribution, impact of Old English exemplars, and the development of certain fixed expressions with the adjectives. Thomas (1931) considered the loss of case-marked definite articles and strong adjectives to be the principal factor in the shift from inflected to periphrastic genitive constructions, but the evidence from the present study shows that this is not the case for all texts.

The intricacies and exceptions of high vowel deletion in Old English have been the subject of much debate in recent historical phonology. Traditional philological handbooks such as Campbell (1959) describe the process within the assumptions of the Neogrammarian tradition. As such, high vowel deletion has been described as a phonological process that removes historically high and synchronically unstressed vowels after a heavy syllable, or two light syllables. However, the descriptions in these handbooks also reveal that exceptions are common, and as per the Neogrammarian tradition, these are usually assumed to be the result of analogy. In contrast, recent studies have sought to account for the exceptions in a way that lends more explanatory power (e.g. Stratal Optimality accounts including Bermúdez-Otero 2005). Such accounts have shown that there is more to the exceptions than analogy, and that phonological rules, as their synchronic activity declines, can become entangled with other morphological and phonological conditioning, due to the high levels of surface opacity that causes them to become unlearnable. Many of the accounts of high vowel deletion have focused on the West Saxon of Alfred (Early WS) and Ælfric (late WS), and recent descriptions of high vowel deletion have largely focused upon the noun declensions (e.g. Bermúdez-Otero in prep) and the weak verb preterites (Minkova 2012). In this study, I focus in particular upon the behaviour of high vowel deletion in the strong and weak verbs; including the past participles and both the present and preterites. The selected data represent the Early West Saxon dialect and also the Late Northumbrian dialect found in the Lindisfarne Gospel gloss. Discussion of the process as found in nouns and adjectives will also be incorporated. The study has two larger aims: 1. To provide an analysis of syncope for newly collected data sets from Early West Saxon and Lindisfarne verbs, and 2. To contribute to the debate surrounding how to account for morphophonological interaction within inflexional paradigms. The data reveal evidence to show that high vowel deletion is indeed suffering from the demise of its original phonological conditions in the verbs. It is not argued however that full lexicalization has yet taken place throughout the verbs. Instead, the data present a range of degrees of morphologization, within which the original phonological conditions have become supplemented by additional morphological conditions. Additional phonological conditioning is also in evidence. The Lindisfarne strong past participles, it is argued, represent a morphological category within which weight-based syncope is synchronically blocked. The wider question of how and why morphological and phonological conditions come to be added to existing phonological processes is addressed, and I argue that such phenomena result from unsustainable levels of opacity in the grammar (Anderson 1989), and that a theoretical framework that allows for the interaction of phonology and morphology within the grammar is necessary. The Optimality Theoretic analyses proposed in this study have the benefit of accounting for instances of phonologization through constraint interaction. It is also argued that the ways in which morphological category determines a) the way in which a phonological condition applies, and b) whether it applies at all, is best analysed using cophonological analyses (Anttila 2002a etc.).

It is widely accepted that Old English personal pronouns often turn up in ‘special’ positions, i.e. positions in which functionally equivalent nominals rarely, if ever, appear. Leading theories of Old English syntax (e.g. van Kemenade 1987, Pintzuk 1991, 1996, Hulk & van Kemenade 1997, Kroch & Taylor 1997) account for the syntax of specially placed pronouns in different ways, but all treat special placement as a freely available option. Focusing on pronominal objects of prepositions in particular, this thesis shows, firstly, that current theories fail to account for the variety of special positions in which these pronouns appear and argues that at least three special positions must be recognised. The central concern of this thesis, however, is whether special placement is the freely available option that leading theories assume. Drawing on evidence from a number of descriptive studies of the syntax of pronominal objects of prepositions (e.g. Wende 1915, Taylor 2008, Alcorn 2009), statistical evidence is presented to show that, in a number of contexts, the probability of special placement is either too high or else too low to be plausibly ascribed to free variation. The thesis explores the linguistic basis of each of the statistically significant parameters identified, finding answers in some cases and intriguing puzzles in others.