Generative analyses have standardly assumed severe constraints on the form of syntactic representations. This thesis explores consequences of relaxing the undermotivated constraints that prohibit discontinuous and converging configurations. A principal benefit of this revision is that it facilitates the assignment of a uniform constituent analysis to constructions that exhibit different constituent orders. This in turn permits a more general account of structure-sensitive phenomena involving anaphora and extraction, and provides a means of extending configurational definitions to derived constructions and languages with variable or otherwise problematic word order conventions. Chapter 2 presents an arboreal model of phrase structure which admits discontinuity and multidomination by partially dissociating order and structure, and relaxing the requirement that each node have a unique parent. Chapter 3 argues for a discontinuous constituent analysis of the Celtic languages Irish, Welsh and Breton, which conform to a dominant VSO pattern. Chapter 4 examines the word order patterns of the Polynesian language Niuean, which again instantiates a VSO order, and proposes a multidomination analysis of the Niuean raising constructions described by Seiter (1980). Chapter 5 examines subordinate Germanic constructions and suggests an analysis which assigns intercalated structural descriptions to cross-serial dependencies. Chapter 6 suggests a strategy for assigning an articulated hiearchical structure to free word order or 'nonconfigurational' languages, and reconsiders the question of Dyirbal's structural ergativity. Chapter 7 presents arguments, based on anaphoric and extraction domains, that unbounded dependency constructions in English instantiate a canonically discontinuous structure. Chapter 8 adumbrates strategies for admitting some of the discontinuous representations proposed in earlier chapters. Chapter 9 briefly summarizes previous arguments, examines some unresolved issues, and considers the correspondence between discontinuous structural analyses, and conventional representations containing 'gaps'.

Let us call a conjoined NP properly containing a plural "multiply plural." The boys and the girls is a multiply plural NP while John and Mary or the boys are "simply plural." The question addressed here is whether multiply plural NPs differ semantically from simply plural NPs. I contrast two theories which differ in their answer to this question as a result of a difference in the meaning they assign to and. According to one theory, the "union theory," every felicitous plural NP denotes a set-like entity all of whose members are individuals (individuals do not themselves have members). On the "sets theory," felicitous multiply plural NPs denote entities that have two or more members some or all of which are not individuals. For example, the boys and the girls, if felicitous, denotes an entity having two members, one being the entity denoted by the boys, the other the entity denoted by the girls. Each of these latter two entities have individuals as members. It follows on the sets theory that the boys and the girls refers to a different entity than does the NP the children. According to the union theory, the entity that the boys and the girls refers to has only individuals as members. It is the same entity that the children refers to. Syntactic complexity is mirrored in semantic complexity on the sets theory but not on the union theory. I will argue that English predicate extensions are not sensitive to the semantic distinctions present exclusively in the sets. The distinctions of the sets theory cannot therefore affect truth conditions and hence it should yield to the union theory. In the course of this argument, I introduce a new context dependent analysis of distributivity. The apparent coreference, in some contexts, of singular collective and plural terms, for example the Senate and the Senators, has been used to argue for the sets theory. In response, I argue that the referent of a singular collective NP is different in kind from the referent of a non-collective plural.

This dissertation is concerned with the role of the mora in the stress system and syllable structure of Spanish. Considerations of stress assignment, sonority and the syllabification of high vocoids point to the independence of morification from syllabification. It is shown that an analysis in which morification precedes syllabification accounts for both stress assignment and the syllabification of high vocoids. Chapter 1 introduces theoretical assumptions about stress, syllable structure and Lexical Phonology in addition to background information from Spanish phonology and morphology. In Chapter 2 I show that a moraic trochee rule provides a precise and explanatorily adequate account of word stress in nouns, adjectives and adverbs. I claim that when stress is assigned in Spanish, syllables are not present; only mora structure has been assigned. In order to assign stress to the mora tier alone, stress assignment (to moras) must be distinguished from stress realization (on syllable heads) via Head-Prominence Relation which matches up stresses with sonority peaks after Syllabification has applied. Chapter 3 places the analysis of stress within the framework of Lexical Phonology. The model adopted has two lexical levels which are governed by the Strict Cycle Condition, but which lack multiple cyclicity. I show that all cyclic effects can be derived from the interaction of levels. While Level 1 stress is mora-sensitive, Level 2-derived forms such as diminutives and verbs undergo a syllable-sensitive Level 2 stress rule. The general picture which emerges is one in which prosodic rules apply to whatever structure is present. In Chapter 4 I confirm the hypothesis that stress assignment precedes syllabification by showing that the syllabification of high vocoids is predictable from mora assignment and stress alone. The proposal is then implemented by formulating the morification and syllabification algorithms, which are sensitive to distinct sonority scales. Chapter 5 examines Spanish cooccurrence restrictions in light of the moraic approach to prosodic processes. The sensitivity of onsets and codas to distinct sonority hierarchies supports the overall approach in which morification and syllabification are likewise sensitive to the same two sonority scales.

This dissertation is a theoretical study of the Spanish verb paradigms' stress system. Its objective is to present a metrical analysis of Spanish verbs using the grammatical model proposed in Optimality Theory. Although this is mostly a synchronic study, a synopsis is included about the evolution of the Spanish verb from the Latin forms, as well as the prososdic alterations that took place as a consequence of the disappearance of the distinction in vocalic quantity and the innovation of prosodic stress in Proto-Romance. The stress pattern of the Spanish verb system has put forth a continuous challenge in the literature, as it is debated whether verbal stress is phonological or morphological in nature. It is proposed in some studies that stress is assigned to abstract representations; others argue for the need of diachronic considerations. Even though most derivational studies acknowledge the importance of phonology and morphology in stress assignment, there is no agreement about the way in which both components interact within the stress algorithm. In this study, the nature of this interface is defined. According to contemporary metrical stress theory, stress is understood as the result of the rhythmical structure of languages. Thus, it is found that Spanish unmarked stress is due, among other things, to the fact that Spanish parses its metrical constituents in syllabic trochees and shows a certain degree of sensitivity to syllable weight. On the other hand, some morphemes contain prespecified metrical information in their lexical entry, which alters the rhythmical structure computed by the general algorithm and is responsible for the stress otherwise unpredictable by strictly prosodic principles. Optimality Theory provides a model that allow us to describe the stress mechanism of Spanish verbs. While the derivational model bases itself on the ordered application of rewriting rules, this new approach proposes that individual grammars consist of the same universal principles or restrictions: individual languages only vary in their particular hierarchy of these violable restrictions. In this way, we establish a set of ranked prosodic restrictions, by whose interaction the stress patterns of the Spanish verb can be systematically obtained.

In this dissertation I investigate exceptions to island conditions that have been noted in the literature and propose a generalized binding theory account of the licensing of wh-ts that takes wh-ts to be anaphoric and therefore subject to some of the same conditions as lexical anaphors. In Chapter 2 I make essential use of the mechanism of Tns-binding and subsume Tns-binding and AGR-binding under INFL-binding as a central property that regulates the relation between a wh-phrase and its wh-t. In this way I account for contrasts in the acceptability of long wh-movement in several languages as well as certain contrasts in the acceptability of QR contructions in French. In Chapter 3 I show that long wh-phrase wh-t dependencies are, in addition to INFL-binding, regulated by the presence of intervening referential NPs. I show that quantificational intervening NPs are much weaker blockers of long distance dependencies than referential NPs and allow violations of even strong islands such as complex NPs. These phenomena show that another principle is required to account for the full range of island constraints and exceptions to island contraints. In Chapter 4 I extend the theory presented in Chapter 2 to account for contrasts in the acceptability of clitic climbing in Spanish in terms of INFL-binding.

The existence of distinctive varieties of second language English in Africa has, by now, been long recognized. Such L2 Englishes are known to arise in situations where the socioeconomic value of English is high, but where restricted access to native speaker varieties of the language results in the establishment and, eventually, the generational transmission of a new secon4:1anguage variety. These 'New Englishes' have been found to possess certain structural similarities across geographical boundaries, while still retaining distinctively local features. The New Englishes of Africa, in particular, have been observed by several authors to be sufficiently similar to warrant the possible use of 'African English' as a generalized cover term for the group. Nevertheless, the continued study of L2 English varieties in separate geographical and political areas within Africa is an indication of the existence of distinctive, if in many ways similar, local varieties. The object of this dissertation is a systematic comparison of the syntactic structure of varieties of sub-Saharan L2 English, taking as a basis Black South African English as a point of comparison. The syntactic structures of these varieties are examined in order to determine the nature and extent of the structural similarities between them, as well as the degrees of difference that occur. It is widely acknowledged that of those sets of features of the New Englishes which differ from Standard English, syntactic variation forms the smallest part. Nevertheless, such variation does exist, both in differences between the New Englishes and the standard(s), and between the New Englishes themselves. The syntactic features of Black South African English are discussed and compared with those of other African Englishes, in order to develop a means of describing such language varieties in relation to one another, and of assessing the extent to which certain of their syntactic features can be recognized as pan-African. A more detailed analysis of the structure of the relative clause in the varieties is given, drawing on theories regarding the origin of certain New English features, as a means of explaining the non-standard occurrence of resumptive pronouns within the relative clause. Finally, the need for corpus-based research into African Englishes is stressed, as a means of determining the frequency of occurrence of those features identified as typical of the varieties.

This study set out to examine the sociolinguistics of social change amongst a group ofyoung, middle-class coloured people who were educated in a predominantly white school environment. The demise of the apartheid system in the early 1990s led to a situation in which racial mixing in government-run schools was permitted for the first time. I conducted sociolinguistic interviews with 20 self-identified 'coloured' Cape Town residents, who attended schools that were formerly open only to white children. . I analysed the data on two levels. Firstly, an analysis of accent, focusing on three salient phonetic markers of South African English, namely the GOOSE, BA TH and PRICE lexical sets (Wells 1982). Acoustic analysis of these vowels was done using a computer software programme, Praat, to record a total of 4410 tokens for the 20 speakers. The second level of analysis investigated how the informants constructed social identities in those unprecedented educational circumstances. I used three theories of identity to do this: Speech Accommodation Theory (Giles 1973), Social Identity Theory (Tajfel 1972) and the Linguistic Market (Bourdieu and Boltanski 1975). Comparing the results of the phonetic and sociological analyses, I found that the two levels of analysis supported the same conclusion: the young, coloured people in the sample subscribe to a coloured social identity, but have clear links with the white community. This suggests that they occupy an intermediate space between the two race groups, which is not surprising given the significant contact they had with members of both communities. There is some evidence of a separation between the coloured community and the typical 'coloured' accent, however, suggesting that one does not need to sound 'typically coloured' in order to be part of the coloured community. It is likely that we are witnessing the formation of an upper middle-class within the coloured middle-class community.

The intertwined planetary crises of anthropogenic climate change and ecological degradation have already begun, and the effects are being felt around the world (IPCC, 2022, World Meteorological Organisation, 2021). However, it is still not being treated as a crisis by global policymakers or the media. Although Africa contributes only 4% of global greenhouse gas emissions, the continent is the most vulnerable to the impacts of the climate crisis. At the same time, South Africa is an outlier to the rest of Africa, being the 12th highest greenhouse gas emitter worldwide (Global Carbon Atlas, 2020). This is due in large part to South Africa's reliance on coal, both as the centre of the country's export economy and as a source of energy at home. However, South African policymakers have yet to take any significant climate change mitigation actions. This is the central social and ecological problem that serves as impetus for this study. This study investigates the climate change-related communications of South Africa's three major political parties, namely the ANC, the DA, and the EFF, from a critical perspective. It explores a wide variety of public political discourse genres in which the topic of climate change is mentioned. Utilising an ecolinguistic and critical discourse analytic framework, this study aims to uncover the “stories” (Stibbe, 2015:6) South African political parties are telling about climate change. Drawing on Fairclough's (1989) three stage approach to analysis, as well as tools from corpus linguistics and the emerging field of ecolinguistics (Stibbe, 2015), it examines both broad quantitative patterns across the data and specific instances of discourse in detail. In particular, the discursive strategies of naturalization, framing, and erasure and salience patterns are investigated. As a CDA study, the focus is on making visible the messages that are implicit in discourse, particularly those in service of power. As an ecolinguistic study, the focus is on the potential ecologically damaging impacts of these messages. Using the above analytical tools, this study shows the political parties' climate change communications to be sites of discursive conflict. It illustrates the ways all three parties, but particularly the ANC and DA, continuously undermine their ostensible commitment to combating the climate crisis with discourses that “tame” it and evade responsibility for addressing it. This is done through framing climate change as an economic issue and as an opportunity for technological innovation and entrepreneurship, through telling the story that individual actions are key to solving climate change, through naturalizing current levels of carbon emissions, and through presenting climate change as “manageable”. In addition, all three parties construct the natural world as existing for the benefit of humans. While the EFF does not exhibit the above taming discourses to the same extent, their dominant climate change story is one of erasure through the absence of discussion. Thus, all three parties naturalize the continued destruction of the planet's life support systems and obscure the central causes of this destruction

The study investigated the Namibian transitional language in education policy and the role of code-switching in achieving it. Firstly, the study identified the types, structures and roles of code-switching in the Grade 4 bilingual classroom. Secondly, the study established the challenges faced by educators and students regarding English as the medium of instruction during the transition phase. Thirdly, the study established the challenges that educators face in implementing the Namibian language in education policy during the transition phase in Grade 4. The data were collected at the Gabriel Mubita Primary School (pseudonym) of the Zambezi region of Namibia. The study adopted qualitative research methods and data were collected via classroom observations and in-depth interviews. Furthermore, the participants were two educators who were observed and interviewed, together with their Grade 4 students. The data of the study were audio and visual recorded, translated and transcribed. Thereafter, the data collected were linguistically and thematically analysed into themes and sub-themes. The findings of the study were that code-switching was predominately utilised by educators and students at the school under investigation. It was revealed that educators employed three different types of code-switching namely situational, educational and interpersonal codeswitching for pedagogical, teaching and social reasons. The research also found that three different structures of code-switching were utilised by educators, mainly intra-sentential and inter-sentential, while tag-switching was rarely utilised. Code-switching was utilised for roles such as: to encourage students to participate in the classroom, for explanations and clarification purposes, to maintain discipline in the classroom, to show solidarity, to reiterate, to translate and to explain the subject content that students could not grasp as they were beginners. The findings revealed that educators and students faced challenges such as lack of English proficiency, shortage of teaching and scholarship materials, students' linguistic under-preparedness and limited exposure to the target language. Lastly the findings showed that educators had insufficient training on the transition phase, lack of policy awareness, and that the transition to English as the medium of instruction takes place too early. The study recommends that the policymakers must involve educators when developing the policies and ensure that educators are trained in policymaking, orthography of two common languages such as Subiya and Sifwe must be developed since students in the Zambezi Region are not taught in their mother tongue but rather in Silozi, the lingua Franca of the Zambian origin, educators must be given a pre-service and in-service training by the Ministry of Education on how to deal with code-switching or formulate a language policy that includes code-switching since it is considered a powerful teaching tool during the transition phase, teaching and learning material resource shortages within schools should be addressed by the Ministry of Education.

No description available.