Thesis (MA)--Stellenbosch University, 2013. / ENGLISH ABSTRACT: This thesis reports on the ways in which rape perpetrated by males on females is constructed in news stories and the advice column, Dear Dolly, published in the South African publication, Drum magazine. The data collected for the study spans from 1984 to 2004, encompassing both 10 years before and 10 years after a democracy. The paper uses critical discourse analysis (Fairclough 2003) as main analytical tool and but also draws on critical feminist theory (Bourke 2007) and other strands of discourse analysis such as Van Dijk‟s (1998) socio-cognitive approach. The findings suggest that there is on the one hand a decrease in explicit victim blaming after 1994, but that subtle and opaque victim blaming is still evident in the news stories, letters to the advice column, and the responses from the columnist. These rape discourses presented in Drum magazine after 1994 are as Bakhtin (1981) suggests made up of multiple voices articulating different gendered discourses. Discourses that make women responsible for their safety and protection against rape are prevalent while at the same time rape is constructed as a “horror story” and the perpetrator as the “monster”. In this thesis, I argue that even though the use of less explicit victim blaming might seem like a positive move in the representation of rape and gender, this is not always the case. The more subtle forms of victim blaming avoid contestation and consequently often go unchecked (Fairclough 2003: 58). This makes the manufacturing of consent easier and makes it more difficult to counteract dominant discourses. I subsequently call for more studies on this underrepresented topic in discourse analysis in South Africa. / AFRIKAANSE OPSOMMING: Hierdie tesis doen verslag oor die maniere waarop verkragting met mans as oortreders en vroue as slagoffer gekonstrueer word in nuus stories en in die advies kolom Dear Dolly in die Suid-Afrikaanse publikasie, Drum magazine. Die data verteenwoordig die tydperk vanaf 1984 tot 2004 (insluitend tien jaar voor en 10 jaar na demokrasie). Die tesis gebruik kritiese diskoers analise (Fairclough 2003) as hoof analitiese instrument maar leen ook van kritiese feministiese teorie (Bourke 2007) en ander tipes diskoersanalise soos Van Dijk (1998) se sosiokognitiewe benadering. Die bevindinge van die tesis stel voor dat daar aan die een kant ‟n afname in is in die eksplisiete blamering van slagoffers na 1994, maar dat subtiele en ondeursigtige blamering van slagoffers nog steeds voorkom in die nuusstories, briewe na die advies kolom en in die antwoorde van die kolomskrywer. Die diskoerse wat in Drum magazine na 1994 gevind word bestaan soos Bakhtin (1981) voorstel uit vele verskillende stemme wat verskillende diskoerse oor geslagsverhoudinge verteenwoordig. Diskoerse wat vroue verantwoordelik hou vir hul eie veiligheid en beskerming kom wyd voor, terwyl verkragting ook gekonstrueer word as ‟n “erotiese riller” en die oortreders gekonstrueer word as monsters. In hierdie tesis stel ek voor dat hoewel die gebruik van minder eksplisiete slagoffer blamering lyk soos ʼn positiewe beweging in die representasie van verkragting en geslagsgelykheid, is dit nie noodwendig die geval nie. Subtiele vorme van slagoffer blamering is moeiliker om te bevraagteken en word dikwels nie krities beskou nie (Fairclough 2003: 58). Dit maak die produksie van konsent makliker en maak dit moeiliker om dominante diskoerse teë te gaan. Gevolglik stel ek voor dat baie meer studies oor hierdie onderverteenwoordige onderwerp in diskoersanalise in Suid-Afrika gedoen moet word.

The purpose of this study is to examine the rhetorical function of mental illness regarding the way it constructs the public view of individuals with mental illness. The study is based on the following research questions: What is the rhetorical function of mental illness during the time period 1951-2016? How has the language used to describe mental illness changed over time? How does language use construct and deconstruct the public view on mental illness? Methods based on theories about ideology, genre and discourse have been used to disclose the view on mental illness in medical discourse. The results indicate that the rhetorical function of mental illness is to construct what is normal or deviant. The results also suggest that language use changes the public perception of mental illness. This study shows that the medical discourse is a powerful tool to decrease stigma related to mental illness.