Corpus-based Lexicography for Sesotho

Show simple item record

dc.contributor.advisor Prinsloo, Danie J. (Daniel Jacobus), 1953-
dc.contributor.postgraduate Setaka, Mmasibidi
dc.date.accessioned 2019-07-04T12:12:39Z
dc.date.available 2019-07-04T12:12:39Z
dc.date.created 2019
dc.date.issued 2018
dc.description Dissertation (MA)--University of Pretoria, 2018. en_ZA
dc.description.abstract For centuries, dictionaries were compiled based upon the knowledge of the lexicographer and information retrieved from manually consulted sources, mainly through a process of reading and marking. This approach meant that much of the information used in the dictionary relied upon the knowledge of the lexicographer. It is vital to rely on the lexicographer’s knowledge of the language but this has its shortcomings, since there is no single individual who knows all the words or terms, their meanings and usage, the words they combine with, and so on, in a specific language. The utilization of this method left room for errors and omissions because the lexicographer could easily overlook some words due to factors like time, fatigue, limited knowledge of the lexicographer, etc. Important words, for example words likely to be looked for by the target users of the dictionary, could accidentally be omitted. In the 1980s, the corpus era was born and the lexicography field changed forever. Collins COBUILD in Birmingham spearheaded this era with the publication of the first corpus-based dictionary, the Collins COBUILD Dictionary in 1987. Since the corpus era began, lexicographers no longer rely solely on their knowledge of the language, intuition, or the limited information gathered from available written sources, which are very limited for African languages. The corpus allows the lexicographer to have access to huge volumes of authentic data from written texts and transcribed oral data. This research will therefore critically discuss dictionary compilation for Sesotho and spearhead the use of corpora in the compilation of Sesotho dictionaries, so that lexicographers do not compile dictionaries as if they are compiling the first dictionary for the language. In addition, they should take into account tasks like lexicographic planning, amongst other factors required to compile a good user-friendly dictionary. Key words Corpora, collocations, concordances, lexicography, lexicographical planning, microstructure, macrostructure, lemmatisation. en_ZA
dc.description.availability Unrestricted en_ZA
dc.description.degree MA en_ZA
dc.description.department African Languages en_ZA
dc.identifier.citation Setaka, M 2018, Corpus-based Lexicography for Sesotho, MA Dissertation, University of Pretoria, Pretoria, viewed yymmdd <http://hdl.handle.net/2263/70367> en_ZA
dc.identifier.other S2019 en_ZA
dc.identifier.uri http://hdl.handle.net/2263/70367
dc.language.iso en en_ZA
dc.publisher University of Pretoria
dc.rights © 2019 University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria.
dc.subject UCTD en_ZA
dc.subject Lexicography en_ZA
dc.subject.other Humanities theses SDG-04
dc.subject.other SDG-04: Quality education
dc.title Corpus-based Lexicography for Sesotho en_ZA
dc.type Dissertation en_ZA


Files in this item

This item appears in the following Collection(s)

Show simple item record