Thesis (doctoral)--Rijksuniversiteit te Leiden, 1996. / Text in English and Kikuyu with English translation, summary in Dutch. Includes bibliographical references (p. [288]-308) and index.

Thesis (doctoral)--Åbo akademi, 2006. / Extra t.p. with thesis statement inserted. Includes bibliographical references (p. 259-300).

South Africa has eleven official languages, ten of which are considered “resource-scarce”. For these languages, even basic linguistic resources required for the development of speech technology systems can be difficult or impossible to obtain. In this thesis, the process of developing Spoken Language Identification (S-LID) systems in resource-scarce environments is investigated. A Parallel Phoneme Recognition followed by Language Modeling (PPR-LM) architecture is utilized and three specific scenarios are investigated: (1) incomplete resources, including the lack of audio transcriptions and/or pronunciation dictionaries; (2) inconsistent resources, including the use of speech corpora that are unmatched with regard to domain or channel characteristics; and (3) poor quality resources, such as wrongly labeled or poorly transcribed data. Each situation is analysed, techniques defined to mitigate the effect of limited or poor quality resources, and the effectiveness of these techniques evaluated experimentally. Techniques evaluated include the development of orthographic tokenizers, bootstrapping of transcriptions, filtering of low quality audio, diarization and channel normalization techniques, and the human verification of miss-classified utterances. The knowledge gained from this research is used to develop the first S-LID system able to distinguish between all South African languages. The system performs well, able to differentiate among the eleven languages with an accuracy of above 67%, and among the six primary South African language families with an accuracy of higher than 80%, on segments of speech of between 2s and 10s in length. AFRIKAANS : Suid-Afrika het elf amptelike tale waarvan tien as hulpbron-skaars beskou word. Vir die tien tale kan selfs die basiese hulpbronne wat benodig word om spraak tegnologie stelsels te ontwikkel moeilik wees om te bekom. Die proses om ‘n Gesproke Taal Identifisering stelsel vir hulpbron-skaars omgewings te ontwikkel, word in hierdie tesis ondersoek. ‘n Parallelle Foneem Herkenning gevolg deur Taal Modellering argitektuur word ingespan om drie spesifieke moontlikhede word ondersoek: (1) Onvolledige Hulpbronne, byvoorbeeld vermiste transkripsies en uitspraak woordeboeke; (2) Teenstrydige Hulpbronne, byvoorbeeld die gebruik van spraak data-versamelings wat teenstrydig is in terme van kanaal kenmerke; en (3) Hulpbronne van swak kwaliteit, byvoorbeeld foutief geklasifiseerde data en klank opnames wat swak getranskribeer is. Elke situasie word geanaliseer, tegnieke om die negatiewe effekte van min of swak hulpbronne te verminder word ontwikkel, en die bruikbaarheid van hierdie tegnieke word deur middel van eksperimente bepaal. Tegnieke wat ontwikkel word sluit die ontwikkeling van ortografiese ontleders, die outomatiese ontwikkeling van nuwe transkripsies, die filtrering van swak kwaliteit klank-data, klank-verdeling en kanaal normalisering tegnieke, en menslike verifikasie van verkeerd geklassifiseerde uitsprake in. Die kennis wat deur hierdie navorsing bekom word, word gebruik om die eerste Gesproke Taal Identifisering stelsel wat tussen al die tale van Suid-Afrika kan onderskei, te ontwikkel. Hierdie stelsel vaar relatief goed, en kan die elf tale met ‘n akkuraatheid van meer as 67% identifiseer. Indien daar op die ses taal families gefokus word, verbeter die persentasie tot meer as 80% vir segmente wat tussen 2 en 10 sekondes lank. Copyright / Dissertation (MEng)--University of Pretoria, 2010. / Electrical, Electronic and Computer Engineering / unrestricted

Die verhandeling ondersoek die taalbeheersing van studente wat Afrikaans as tweede addisionele taal neem en meer spesifiek, die woordeskatkennis van die ondersoekgroep, naamlik swart addisioneletaalleerders in die Limpopo-provinsie in Graad 11 en 12. Dit is duidelik dat een van die probleme wat aanleerders het, ‘n gebrekkige woordeskat is. Dit beïnvloed hulle begrip van Afrikaans sowel as hulle produksie (skriftelik en ook mondeling). Een van die redes hiervoor is dat daar nie op ‘n sistematiese manier aandag gegee word aan óf woordeskatverwerwing (of –aanleer) óf die onderrig van woordeskat in die kurrikulum en sillabus nie. Hierdie navorsing se doelwitte is om die stand van die woordeskatkennis van Graad 11- en 12-leerders te bepaal en tweedens om met behulp van insigte oor die aanleer en/of verwerwing van leksikonitems die sukses van bepaalde strategieë te bepaal. Navorsingsvrae is die volgende: Wat is die stand van woordeskatonderrig vandag wêreldwyd? Wat is die stand van woordeskatonderrig in die Suid-Afrikaanse sillabus? Hoe goed is Graad 11 en 12 leerderders se kennis van die Afrikaanse woordeskat? Wat is Graad11- en 12 leerders se houding teenoor Afrikaans? Hoe kan mens hierdie houding(s) verklaar? Watter rol speel (intrinsieke) motivering en houdings teenoor Afrikaans? Watter metodes en tegnieke kan gebruik word om leerders se woordeskat te verbeter? / Dissertation (MA (Afrikaans))--University of Pretoria, 2008. / Afrikaans / unrestricted

This dissertation examines the relation between embodiment and language, knowledge and memory, as explored in the poetry of South African poet Gabeba Baderoon. In her three published collections of poetry, namely, The Museum of Ordinary Life, The Dream in the Next Body and A Hundred Silences, she depicts seemingly trivial and everyday events or experiences with acute attention to detail, all of which are connected by her unique portrayal of their embodied nature. In doing so, her work illustrates that intellectual activities typically associated with the mind, such as language, knowledge and memory, in fact require the incorporation of the body. Therefore, this dissertation studies the mind-body relation represented in her work with regard to these thematic concerns, since it is a crucial aspect of her poetry and aids not only in understanding and interpreting her work, but also the discourse on embodiment in general. These concerns do, moreover, not remain on a thematic level, but are evident in her poetry itself; that is, her poems too act as a form of embodiment. Furthermore, Baderoon’s poems are able to transcend the supposed mind-body dichotomy in a way that shows much in common with phenomenology, and especially the perspective held by authors such as Maurice Merleau-Ponty. This dissertation incorporates phenomenological ideas on the body and embodiment, as these assist in interpreting Baderoon’s work, as well as for the reason that her poetry sheds new light upon the understanding of such phenomenological ideas, too. Thus, this dissertation seeks to elucidate the manner in which Gabeba Baderoon’s poetry transcends the mind-body dichotomy by means of her exceptional employment of the notion of embodiment on a thematic as well as formal level. / Thesis (MA (English))--North-West University, Potchefstroom Campus, 2012

Idiolect refers to an individual’s unique use of language. Therefore, the author of a text can be identified by his/her use of language. This study is focused on Joan Hambidge’s recognisable idiolect across the boundaries of genres. It is expected that Hambidge will have a unique and recognisable idiolect, regardless of the genre she writes in. By making use of forensic linguistic principles, methods and applications, it has been shown that it is possible to determine an individual’s idiolect. Even though forensic principles are specifically focused on identifying an author, the methodology used in the research field can be applied to a corpus linguistic study to determine how clearly an individual’s idiolect features across the boundaries of genres. By researching the research subject, explaining her oeuvre, creating a literary background, as well as discussing the literary approaches that Hambidge uses in her respective genres, and what she writes about, the necessary literary background was created, which contributes to the complete image of Hambidge and her influences. By creating this background, it is possible to determine which external factors have an influence on Hambidge's idiolect. Linguistic research was done to determine the origin and background of sociolinguistics; as well as factors that can influence an individual’s idiolect. The background of forensic linguistics was provided, as well as the various corpus linguistic methods that can be used in a study such as this one. After the background was provided, the empirical analysis was executed, in which both stylistic and stylometric analyses were performed by making use of inter- and intra-corpus linguistic research, according to which Hambidge’s idiolect was identified. To identify Hambidge’s idiolect, the Taalkommissie corpus was used as a reference corpus to determine whether the idiosyncratic characteristics that were found in the Hambidge corpus truly are a unique feature or whether they can also be found in the Taalkommissie corpus. The application and execution of the methods made it possible to determine to which extent, if at all, Hambidge has a unique idiolect, and how this idiolect features across the boundaries of genres. The research has determined that Joan Hambidge has a unique idiolect and that the idiolect is especially clear when research is done about her corpus in its entirety. When Hambidge’s separate genres were compared to each other, it was clear that genre influences idiolect, but also that Hambidge did not follow the prescribed genre conventions. Even though the two novels that were compared, did not match as was expected, the other, various genres did agree. Various categories were identified, from which it is clear that distinguishing characteristics can be found in Hambidge’s corpus. It can therefore be said without a doubt that Hambidge has a unique idiolect across the boundaries of genres. / MA (Afrikaans and Dutch), North-West University, Potchefstroom Campus, 2014

This dissertation examines the relation between embodiment and language, knowledge and memory, as explored in the poetry of South African poet Gabeba Baderoon. In her three published collections of poetry, namely, The Museum of Ordinary Life, The Dream in the Next Body and A Hundred Silences, she depicts seemingly trivial and everyday events or experiences with acute attention to detail, all of which are connected by her unique portrayal of their embodied nature. In doing so, her work illustrates that intellectual activities typically associated with the mind, such as language, knowledge and memory, in fact require the incorporation of the body. Therefore, this dissertation studies the mind-body relation represented in her work with regard to these thematic concerns, since it is a crucial aspect of her poetry and aids not only in understanding and interpreting her work, but also the discourse on embodiment in general. These concerns do, moreover, not remain on a thematic level, but are evident in her poetry itself; that is, her poems too act as a form of embodiment. Furthermore, Baderoon’s poems are able to transcend the supposed mind-body dichotomy in a way that shows much in common with phenomenology, and especially the perspective held by authors such as Maurice Merleau-Ponty. This dissertation incorporates phenomenological ideas on the body and embodiment, as these assist in interpreting Baderoon’s work, as well as for the reason that her poetry sheds new light upon the understanding of such phenomenological ideas, too. Thus, this dissertation seeks to elucidate the manner in which Gabeba Baderoon’s poetry transcends the mind-body dichotomy by means of her exceptional employment of the notion of embodiment on a thematic as well as formal level. / Thesis (MA (English))--North-West University, Potchefstroom Campus, 2012

Idiolect refers to an individual’s unique use of language. Therefore, the author of a text can be identified by his/her use of language. This study is focused on Joan Hambidge’s recognisable idiolect across the boundaries of genres. It is expected that Hambidge will have a unique and recognisable idiolect, regardless of the genre she writes in. By making use of forensic linguistic principles, methods and applications, it has been shown that it is possible to determine an individual’s idiolect. Even though forensic principles are specifically focused on identifying an author, the methodology used in the research field can be applied to a corpus linguistic study to determine how clearly an individual’s idiolect features across the boundaries of genres. By researching the research subject, explaining her oeuvre, creating a literary background, as well as discussing the literary approaches that Hambidge uses in her respective genres, and what she writes about, the necessary literary background was created, which contributes to the complete image of Hambidge and her influences. By creating this background, it is possible to determine which external factors have an influence on Hambidge's idiolect. Linguistic research was done to determine the origin and background of sociolinguistics; as well as factors that can influence an individual’s idiolect. The background of forensic linguistics was provided, as well as the various corpus linguistic methods that can be used in a study such as this one. After the background was provided, the empirical analysis was executed, in which both stylistic and stylometric analyses were performed by making use of inter- and intra-corpus linguistic research, according to which Hambidge’s idiolect was identified. To identify Hambidge’s idiolect, the Taalkommissie corpus was used as a reference corpus to determine whether the idiosyncratic characteristics that were found in the Hambidge corpus truly are a unique feature or whether they can also be found in the Taalkommissie corpus. The application and execution of the methods made it possible to determine to which extent, if at all, Hambidge has a unique idiolect, and how this idiolect features across the boundaries of genres. The research has determined that Joan Hambidge has a unique idiolect and that the idiolect is especially clear when research is done about her corpus in its entirety. When Hambidge’s separate genres were compared to each other, it was clear that genre influences idiolect, but also that Hambidge did not follow the prescribed genre conventions. Even though the two novels that were compared, did not match as was expected, the other, various genres did agree. Various categories were identified, from which it is clear that distinguishing characteristics can be found in Hambidge’s corpus. It can therefore be said without a doubt that Hambidge has a unique idiolect across the boundaries of genres. / MA (Afrikaans and Dutch), North-West University, Potchefstroom Campus, 2014

Most South African subtitles are produced and broadcast in English despite the fact that English is the first language of only 8.2% of the entire population (Statistics South Africa, 2004). Therefore, current English subtitles are predominantly received as second language text. This poses questions as to how people perceive these subtitles, and if and how their reading of English second language (L2) subtitles differs from their reading of L1 (non-English) subtitles. In recent years, eye tracking has proven to be a valuable method in observing and measuring the eye movements of people watching and reading subtitles. In order to explain the use of eye tracking and in order to answer the question at hand, this study comprises a literature review and an empirical study. The literature review gives an in-depth account of previous studies that used eye tracking to study reading and elaborates on the parameters used to account for various findings. The two empirical components of this study examined the accessibility and effectiveness of English L2 subtitles by presenting native speakers of Afrikaans and Sesotho with subtitles displayed (a) in their native language, Afrikaans or Sesotho, and (b) in L2 English, while monitoring their eye movements with an SMI iViewX™ Hi-Speed eye tracker and comparing the data with that of English L1 speakers reading English subtitles. Participants were also given static text to read (accompanied by a corresponding comprehension test) in order to see if there was a relation between participants’ first and second language reading of static text and that of subtitling. Additionally, participants were given a questionnaire on their reading behaviour, reading preferences, access to subtitled television programming and reading of subtitles in order to find explanations for occurrences in the data. The initial hypothesis was that there would be a difference in L1 and L2 subtitle reading and attention allocation as measured by key eye-tracking parameters. Using ANOVAs, statistically significant differences were indeed found, but the differences were much more significant for the Sesotho L1 than the Afrikaans L1 speakers. After excluding possible confounding factors that were analysed in attempt to refute this hypothesis, the conclusion was that participants inherently read L1 and L2 subtitles differently. The hypothesis is therefore supported. However, the difference in L1 and L2 subtitle reading was not the only significant finding – the Sesotho L1 speakers’ reading data revealed a greater underlying issue, namely literacy. The problem of low literacy levels can be attributed to the participants’ socioeconomic background and history, and needs to be addressed urgently. Recommendations for future research include that the current study be broadened in terms of scope, sampling size, representativeness and experimental material; that the focus be shifted to the rest of the languages spoken in South Africa for which the users do not have a shared sense of bilingualism and for which the L1 skills and levels of L1 literacy vary; and to further explore the relation between the reading of static text and subtitle reading in order to ensure adequate subtitle reading in terms of proportional attention allocation. However, the issue of low literacy levels will have to be addressed urgently; only then will the South African viewing public be able to gain full access to any form of broadcast communicative material or media, and only then will they be able to benefit from subtitling and all that it offers. / North-West University (South Africa). Vaal Triangle Campus.

Most South African subtitles are produced and broadcast in English despite the fact that English is the first language of only 8.2% of the entire population (Statistics South Africa, 2004). Therefore, current English subtitles are predominantly received as second language text. This poses questions as to how people perceive these subtitles, and if and how their reading of English second language (L2) subtitles differs from their reading of L1 (non-English) subtitles. In recent years, eye tracking has proven to be a valuable method in observing and measuring the eye movements of people watching and reading subtitles. In order to explain the use of eye tracking and in order to answer the question at hand, this study comprises a literature review and an empirical study. The literature review gives an in-depth account of previous studies that used eye tracking to study reading and elaborates on the parameters used to account for various findings. The two empirical components of this study examined the accessibility and effectiveness of English L2 subtitles by presenting native speakers of Afrikaans and Sesotho with subtitles displayed (a) in their native language, Afrikaans or Sesotho, and (b) in L2 English, while monitoring their eye movements with an SMI iViewX™ Hi-Speed eye tracker and comparing the data with that of English L1 speakers reading English subtitles. Participants were also given static text to read (accompanied by a corresponding comprehension test) in order to see if there was a relation between participants’ first and second language reading of static text and that of subtitling. Additionally, participants were given a questionnaire on their reading behaviour, reading preferences, access to subtitled television programming and reading of subtitles in order to find explanations for occurrences in the data. The initial hypothesis was that there would be a difference in L1 and L2 subtitle reading and attention allocation as measured by key eye-tracking parameters. Using ANOVAs, statistically significant differences were indeed found, but the differences were much more significant for the Sesotho L1 than the Afrikaans L1 speakers. After excluding possible confounding factors that were analysed in attempt to refute this hypothesis, the conclusion was that participants inherently read L1 and L2 subtitles differently. The hypothesis is therefore supported. However, the difference in L1 and L2 subtitle reading was not the only significant finding – the Sesotho L1 speakers’ reading data revealed a greater underlying issue, namely literacy. The problem of low literacy levels can be attributed to the participants’ socioeconomic background and history, and needs to be addressed urgently. Recommendations for future research include that the current study be broadened in terms of scope, sampling size, representativeness and experimental material; that the focus be shifted to the rest of the languages spoken in South Africa for which the users do not have a shared sense of bilingualism and for which the L1 skills and levels of L1 literacy vary; and to further explore the relation between the reading of static text and subtitle reading in order to ensure adequate subtitle reading in terms of proportional attention allocation. However, the issue of low literacy levels will have to be addressed urgently; only then will the South African viewing public be able to gain full access to any form of broadcast communicative material or media, and only then will they be able to benefit from subtitling and all that it offers. / North-West University (South Africa). Vaal Triangle Campus.