The aim of this thesis is, first, to describe the phonetic and phonological features of Neath English, secondly, to discuss the possible effect of non-linguistic factors on people's choice of speech forms and, finally, to analyse the intonation of Neath English. Based on the data collected in fieldwork, the accent description gives the vowel and consonant phonemes and their realisations. Neath English, which is the English spoken in Neath, South Wales, is compared and contrasted to RP as well as other Southern Welsh English accents. To illustrate the similarities and differences to RP and Welsh, vowel and consonant typologies are given. Additionally, this thesis looks at how and to what extent the non-linguistic factors affect people in their choice of speech forms. A detailed computer analysis of the data gives insight into how age, male/female and education have influenced people's speech forms. Six groups of informants are compared to each other in order to discuss their different use of certain variables in their speech. In the description of Neath English intonation, the nuclear tones, the tones in heads and tails and the grammatical, attitudinal, and discourse functions of Neath English intonation are described. Neath English intonation is also compared to RP and other Southern Welsh English accents.

This thesis investigates the concept of the Northern Middle English dialect area as largely homogenous, taking its core data from four Northern manuscripts of the devotional poem, the Pricke of Conscience. Due largely to a lack of localised manuscripts from this region, the NME area has been little-studied. A Linguistic Atlas of Late Mediaeval English is an invaluable resource for the dialectologist, but where anchor texts are lacking, it cannot form a sole basis for study. This thesis approaches dialect study in the region from a different angle, compiling and comparing linguistic data from its core manuscripts to determine what variation can be observed. Each manuscript is also considered in its codicological and sociolinguistic context, using scribal behaviour as a means of illuminating the possible provenances of manusctipts and their language types. Finally, the data is considered alongside similar information from edited texts from this region, to place it in its proper context and examine what patterns of variation emerge. The findings of this thesis show a greater range of variation across the NME region than traditional descriptions would indicate. The variation attested demonatrates that some features previously held to be 'non-northern' or anomolous are, in fact, organically and consistently in use in this area. This study identifies and describes five language varieties observable within NME. The combination of textual and codicological evidence enables suggestions to be made about the identity of several of the scribes examined, connecting language varieties to occupation and possible provenance. This thesis presents a more nuanced, heterogeneous picture of the NME region, which may serve as a basis for further study in the area.

The research presented in this thesis uses various instrumental techniques to determine the main phonetic characteristics of the Intervocalic stops in both British English and Korean. Pilot experiments were carried out using structured /VCV/ nonsense isolated items (chapter 2 and 3), and the results of the pilot experiments were confirmed with real speech items in connected speech In the main experiment (chapter 4). The results show that in intervocalic positions, the lax (phonologically voiced in BrEng and unaspirated lax in Korean) and tense (voiceless in BrEng and aspirated tense in Korean) stops are significantly differentiated by the duration of the preceding vowel, the duration of oral closure and maximum lingual contact and that the Korean phonologically unaspirated tense and lax stops are significantly differentiated by these measures also. The results also show that in intervocalic position, the so-called lax stops (phonologically voiced in BrEng and unaspirated lax in Korean) are rarely voiced throughout the closure phase but have a substantial period of devoicing and a measurable VOT. Also full voicing does not automatically occur as the duration of oral closure of a stop is reduced. The underlying physiological mechanism for voicing and devoicing during intervocalic stops in the two languages was discussed in the light of data on glottal area from a photo-electric glottograph. Some physiological explanations for the effect on the phonetic variables of the place of articulation, stress placement and tempo variation are discussed.

Both predator and prey have evolved to maximise reproductive success by balancing food intake with risk. There has been a bias in predator-prey studies, where prey behaviour has been examined in detail, yet predators are assumed to follow simplistic rules. I use three-spined sticklebacks predating upon invertebrate prey to test a range of ways in which prey risk was hypothesised to be affected by predator behaviour. The relationship between encounter rate and prey density has been recently shown not to be directly proportional, and theoretical arguments have been made that predator search behaviour can explain this trend. I test these arguments, and show acceleration of a predator's search path can in fact lead to the observed less-than-directly proportional relationship between prey density and encounter rate. The perceptual constraints of predators can have major impacts on prey risk. Once encountered, an attack was more likely when prey were encountered late in a search, probably due to a decrease in anti-predator vigilance as the fish became more habituated to the arena. In a subsequent study, larger groups of prey were more quickly found, as were larger numbers of groups. This led to the conclusion that the field of attention is a subset of the total visual field, and this is also supported by denser prey being more conspicuous. Although the predator responded to increased prey group size and density with a reduced time to detect and attack prey, attacks on such groups were less successful due to the confusion effect. Interestingly, I show the effect of prey density to be sensitive to spatial scale, where a large-scale measure of density affected conspicuousness and a small-scale measure affected attack success. This was explained by a reduction in the total number of prey in the visual field as a group of prey is approached and attacked. In the final chapter, I turn my attention to differences in temperament within a predator population, and how this affects prey risk. As expected, bolder fish represented a greater risk to the prey. However, as larger fish tended to be more bold, suggesting boldness was driven by their perception of predation risk, a within community behaviourally-mediated trophic cascade may occur. where the risk to prey is driven by their predators' own perceived risk of predation. This shows that optimal foraging decisions under the threat of predation, as well as perceptual constraints, can mediate the effect of predator behaviour on prey risk.

The systems of spelling found in the extant manuscripts of a late Middle English text are rarely considered in discussions of the transmission of that text. If they are, scholars have typically used the occurrence of dialectal spellings to allocate manuscripts to geographical areas or the occurrence of identical spellings, often unusual spellings, in corresponding locations across the manuscripts to recover the usage of the presumed archetype. The basis for much of this scholarship has been profiles which rely upon text samples or which list what spellings are found in a manuscript but do not reveal the internal distribution of these spellings in that manuscript. This study considers the spelling and codicology of nine complete, textuallyimportant manuscripts of Geoffrey Chaucer's Canterbury Tales, each of which was copied by a different scribe. The author semi-automatically indexes the electronic transcripts prepared by the Canterbury Tales Project from the spellings registered in the Project spelling databases for one of the tales. He extracts a comprehensive spelling profile for each manuscript from this index. The profiles correspond to the questionnaire used for Angus McIntosh, M.L. Samuels, et aI., A Linguistic Atlas of Late Mediaeval English (Aberdeen, 1986).

The present study investigates the phonological productive and perceptual competence of a group of Rwandan learners of English and the effect that phonological deviations have on their intelligibility and comprehension in spoken English. In order to discover the hierarchy and degree of difficulty these subjects have in the segmental and suprasegmental features of English, productive and perceptual tests of words and sentences were designed and administered to a group of 60 subjects. The study also attempts to explain the effect of various interlanguage phenomena which occur in the production and perception of the pronunciation of English by Rwandan speakers. The results of this study support many of the claims of CA, EA and phonological interlanguage. Chapter One gives background sociolinguistic information on the roles of Kinyarwanda, French, Kiswahili and English in Rwanda. Chapter Two discusses a number of theoretical key issues involved in language learning and acquisition. Chapter Three defines the topic of the study, i.e. intelligibility and comprehension, to gain insight into the study and to provide a framework for the research design and methodology. Chapter Four is a description of the Kinyarwanda and English phonological systems, which is the basis of the predictions of the difficulties and the design of data for Chapters Five and Six. Chapter Five analyses, categorizes and explains the source of deviations in the data gathered from subjects' pronunciation of words and sentences as interpreted by native English speakers. Chapter Six analyzes the effect of phonological deviations on the subjects' comprehension of spoken English. Chapter Seven concludes the whole study with a discussion of the major findings, and suggests some useful steps towards more effective teaching of the pronunciation for better intelligibility and comprehension in English.

No description available.

The thesis presents an empirical socio-phonetic investigation of the acoustics of the three short vowels in the DRESS, TRAP and STRUT lexical sets (Wells 1982) in London. The vowels have been reported by a number of phoneticians and variationists to have shifted in particular directions in Received Pronunciatioin (RP) and London English during the course of the 20th century; the directions of the movements, however, seem to be rather complicated. Moreover, there have been relatively fewer instrumental studies for these vowels in London. The main purpose of this research, therefore, is to provide detailed patterns of recent vowel shifts involving these three vowels in London English in relation to internal and external factors. Acknowledging RP and Cockney as referential accents on a multidimensional accent continuum in London, it is presumed that Londoners closer to the upper and lower ends of social continuum are distinguished as `London Upper Middle Class (UMC)' speakers and `London Working Class (WC)' speakers respectively. Social class classification is made on the basis of speakers' occupational information. The application of the vowel formant normalisation technique called S-procedure (Watt & Fabricius 2002) allows direct visual and statistical comparisons for multiple speakers regardless of their physical differences. Investigations are made not only by traditional descriptions of relative placements of vowels in a visual two-dimensional FI /F2 vowel space but also by a recent innovative `angle and Euclidean distance calculations' procedure (Fabricius 2007) with thorough statistical analyses. Results show complicated but interesting correlations between the movements of these vowels and the social and phonological characteristics. One of the most interesting findings is an ongoing vowel change process called 'TRAP/STRUT rotation' (Fabricius 2006: 3,2007: 310) among (female) London UMC speakers who show a well progressed anticlockwise chain shift involving DRESS, TRAP and STRUT, whereas there is no evidence for this process among London WC speakers who show a rather moderate vowel shift involving only TRAP and DRESS. In this respect, the most innovative group is discussed to be the female young London UMC speakers, followed by the male young London UMC speakers as far as the data in the current study are concerned. The finding of a clear difference between two different accent groups in the realisations of the vowels is discussed to suggest a correlation between social class and accent variation in London, as well as to enhance the validity of occupation as a single indicator for people's social class. Observing a great number of general and minute patterns from the statistical results, the thesis attempts to provide possible explanations for the vowel changes in London, as well as extends its discussions for possible implications with regard to internal and external factors.

This thesis is intended to provide a unified and coherent theoretical analysis of phonological weakening processes of vowels and consonants in English within the framework of Optimality Theory (Prince and Smolensky 1993). The analysis of weakening phenomena may vary according to the theory you adopt and the language you choose, but in this thesis, vowel reduction and consonant lenition in English will be explored in a constraint-based approach. In addition, most importantly, I seek to show which generalisations can equally be applied to both consonant lenition and vowel reduction in terms of a phonological theory. The key questions to be investigated in this thesis are as follows. 1) How do we represent phonological weakening phenomena in terms of segmental features or elements? 2) How can these representational elements be integrated into the constraint ranking and evaluation mechanisms in Optimality Theory? 3) Do the historical data such as the initial fricative voicing and vowel reduction in Old and Middle English give us any insight in this regard? There seems to be a similarity between consonant lenition and vowel reduction in terms of their phonological behaviour. For instance, both consonant lenition and vowel reduction can be treated as loss of some element or reduction of complexity in Element Theory (e.g. Harris 1994). This is an interesting point of my PhD project because this kind of representational approach to weakening phenomena has rarely been applied in Optimality Theoretic analysis. Therefore, what is intended to do in this thesis is that melodic representation will be used for modelling weakening phenomena within the framework of Optimality Theory. iv In this regard, I suggest a combined theoretic account of weakening phenomena involving the combination of two approaches namely Element Theory and Optimality Theory, which differentiates this account from previous analyses. I argue that the constraint *COMPLEX[Element], where ‘element’ refers to one of the primitives of Element Theory, plays a central role in analysing phonological weakening processes in this thesis. In addition, it will be shown throughout the thesis that these processes can be accounted for within the constraint interaction between positional faithfulness constraints such as IDENT[Element] and the integrated constraint *COMPLEX[Element] which I propose in this thesis.

No description available.