In October 2020 Peter Wolodarski, the editor-in-chief of the Swedish newspaper Dagens Nyheter (DN), announced that he wanted to improve the news coverage of the climate crisis. The aim of this study is to investigate whether DN has changed their way of reporting on the climate crisis through text and pictures within six months, and if so what those changes are. Framing theory is central in this study as it identifies the usage of generic and theme specific frames in DN’s articles. This is done by comparing all articles in September 2020 and March 2021 that contain the topic “climate crisis” published on their online service. The usage of pictures was also studied for additional context. In order to study this the methods qualitative text analysis and semiotic image analysis were used. To provide results the data then underwent a quantitative analysis as well as qualitative analysis.The findings of the study show that DN has shifted their usage of generic frames, meaning a change in which scale the climate crisis is portrayed, and an increased and widened usage of specific frames in their articles. However, the usage of pictures has not changed notably. In conclusion, does DN in fact seem to have changed their overall reporting on the climate crisis within the studied period. The study contribution is an added understanding of papers reporting on the clime crisis online as well as media’s usage of frames and how those can change in a relatively short period of time.

It has been more than ten years since the 9/11 attacks in 2001, but the events related to the attacks are still a focus for the whole world. This study examined the news coverage of the 9/11 tenth anniversary from Phoenix TV and CNN, which are among the most influential news media in China and the U.S., respectively. A systematic content analysis was performed using latest news, opinion articles, photographs, and videos as classified by CNN and Phoenix TV on their commemorative 9/11 tenth anniversary websites. Framing theory guided this thesis project. The results offer some evidence regarding differences in the selection and use of frames in the U.S. and Chinese media. Results also suggest that global media share news resources for global events. / Master of Arts

A content analysis and post hoc content analysis of 362 news articles in national newspapers, regional newspapers, and Internet news Web sites investigated the prevalence of issue-specific and generic frames, frame valence, and the personalization of media coverage of the steroid scandal in Major League Baseball. Research guided by framing theory found 2,353 frames present in the initial analysis and 2,834 frames present after the post hoc analysis. Generic frames were more prevalent than issue-specific frames in coverage in initial analysis. The post hoc analysis, however, indicates that issue-specific frames were more prevalent than generic frames in terms of times present. Frames are valenced negatively more frequently than neutrally or positively in coverage. Media coverage was focused on the individuals more often than on the organization, however, both the individuals and organization were treated similarly in terms of valence of frames. The findings of the analyses supported scholarship calling for more analysis of generic and issue-specific frames, the presence of valence in frames, and the personalization of media coverage in the political communication context that in this case is present in the sports media context as well. Findings merit further scholarship on broader source comparison in coverage of this scandal, agenda-setting in various forms, and further frame analysis in the sports media contexts and other contexts outside of the political communication context as well. / Master of Arts

The work in this thesis presents and evaluates a number of strategies by which English-learning children might discover the major open-class parts-of-speech in English (nouns, verbs and adjectives) on the basis of purely distributional information. Previous work has shown that parts-of-speech can be readily induced from the distributional patterns in which words occur. The research reported in this thesis extends and improves on this previous work in two major ways, related to the constructional status of the utterance contexts used for distributional analysis, and to the way in which previous studies have dealt with categorial ambiguity. Previous studies that have induced parts-of-speech from word distributions have done so on the basis of fixed “windows” of words that occur before and after the word in focus. These contexts are often not constructions of the language in question, and hence have dubious status as elements of linguistic knowledge. A great deal of recent evidence (e.g. Lieven, Pine & Baldwin, 1997; Tomasello, 1992) has suggested that children’s early language may be organized around a number of lexically-specific constructional frames with slots, such as “a X”, “you X it”, “draw X on X”. The work presented here investigates the possibility that constructions such as these may be a more appropriate domain for the distributional induction of parts-of-speech. This would open up the possibility of a treatment of part-of-speech induction that is more closely integrated with the acquisition of syntax. Three strategies to discover lexically-specific frames in the speech input to children are presented. Two of these strategies are based on the interplay between more and less frequent words in English utterances: the more frequent words, which are typically function words or light verbs, are taken to provide the schematic “backbone” of an utterance. The third strategy is based around pairs of words in which the occurrence of one word is highly predictable from that of the other, but not vice versa; from these basic slot-filler relationships, larger frames are assembled. These techniques were implemented computationally and applied to a corpus of child-directed speech. Each technique yielded a large set of lexically-specific frames, many of which could plausibly be regarded as constructions. In a comparison with a manual analysis of the same corpus by Cameron-Faulkner, Lieven and Tomasello (2003), it is shown that most of the constructional frames identified in the manual analysis were also produced by the automatic techniques. After the identification of potential constructional frames, parts-of-speech were formed from the patterns of co-occurrence of words in particular constructions, by means of hierarchical clustering. The resulting clusters produced are shown to be quite similar to the major English parts-of-speech of nouns, verbs and adjectives. Each individual word token was assigned a part-of-speech on the basis of its constructional context. This categorization was evaluated empirically against the part-of-speech assigned to the word in question in the original corpus. The resulting categorization is shown to be, to a great extent, in agreement with the manual categorization. These strategies deal with the categorial ambiguity of words, by allowing the frame context to determine part-of-speech. However, many of the frames produced were themselves ambiguous cues to part-of-speech. For this reason, strategies are presented to deal with both word and context ambiguity. Three such strategies are proposed. One considers membership of a part-of-speech to be a matter of degree for both word and contextual frame. A second strategy attempts to discretely assign multiple parts-of-speech to words and constructions in a way that imposes internal consistency in the corpus. The third strategy attempts to assign only the minimally-required multiple categories to words and constructions so as to provide a parsimonious description of the data. Each of these techniques was implemented and applied to each of the three frame discovery techniques, thereby providing category information about both the frame and the word. The subsequent assignment of parts-of-speech was done by combining word and frame information, and is shown to be far more accurate than the categorization based on frames alone. This approach can be regarded as addressing certain objections against the distributional method that have been raised by Pinker (1979, 1984, 1987). Lastly, a framework for extending this research is outlined that allows semantic information to be incorporated into the process of category induction.

[pt] A pesquisa tem como objetivo estudar a variação temporal de enquadramentos genéricos e específicos sobre saúde no contexto da pandemia de covid-19 no Brasil. Se trata de um estudo de caso realizado no Twitter sobre o tema da saúde (n = 31.339.922) entre 15 de março e 15 junho de 2020. Como categorias analíticas, se estudaram 3 frames genéricos e 3 específicos sobre saúde em contextos de epidemias. Os frames genéricos foram operacionalizados de forma dedutiva: conflito, atribuição de responsabilidade e moralidade (SEMETKO; VALKENBURG, 2000). Os frames específicos foram operacionalizados com o método indutivo (DE VREESE, 2005): consequências da pandemia, medidas de contenção e métodos de tratamento. Os tweets foram classificados automaticamente mediante um método computacional dictionary based, garantindo a confiabilidade, a validez e a reprodutibilidade (KRIPPENDORFF, 2011; SAMPAIO; LYCARIÃO, 2018). Foi realizada uma série temporal para observar a variação diária da evocação de cada quadro nos 93 dias estudados. Foi construída também uma rede temporal de usuários conectados mediante menções, retweets e respostas, sobre a qual foi calculada a métrica PageRank para medir a sua influência diária sobre a rede. Foram selecionados os dez atores mais proeminentes segundo o seu PageRank na data de maior pico de cada frame da série temporal. Finalmente, foram sistematizadas as informações sobre o contexto de análise e sobre o clima de opinião no Brasil mediante surveys representativos da população brasileira com frequência diária (n = 1.800, ME = mais ou menos 2 por cento, IC = 95 por cento). Os resultados da pesquisa apontam que a ordem de evocação dos frames, do mais ao menos proeminente, foi: conflito, atribuição de responsabilidade, consequências da pandemia, moralidade, medidas de contenção e métodos de tratamento. Em especial, os quadros do conflito, da atribuição de responsabilidade e das consequências da pandemia estiveram fortemente relacionados a um enquadramento negativo, episódico e de interesse humano dos tweets. Por outro lado, os demais frames (moralidade, medidas de contenção e métodos de tratamento) priorizaram enquadramentos temáticos, cujas implicações eram preponderantemente de natureza mais social e menos individual. Se destaca também a significativa presença de perfis anônimos entre os Top10 usuários de cada frame, assim como militantes, especialistas em saúde, influenciadores digitais, jornalistas, órgãos de mídia, políticos e perfis de outra natureza. Por fim, no que se refere ao estudo das condições que estão associadas aos picos mais altos de evocação dos frames genéricos e específicos da série temporal, foi observado um fenômeno que se definiu como sincronização do enquadramento. Se define a sincronização do enquadramento como um fenômeno de ajuste coletivo da frequência ativação em rede de um determinado frame por meio da interação entre os indivíduos e influenciado por quatro fatores: contexto, sucessão de eventos associados, clima de opinião e combinação entre frames e temas. / [en] The research aims to study the temporal variation of generic and specific frames about health in the context of the Covid-19 Pandemic in Brazil. It is a case study conducted on Twitter about health (n = 31,339,922) between March 15 and June 15, 2020. As analytical categories, 3 generic and 3 specific frames about health in epidemic contexts were studied. The generic frames were operationalized deductively: conflict, attribution of responsibility and morality (SEMETKO; VALKENBURG, 2000). The specific frames were operationalized with the inductive method (DE VREESE, 2005): consequences of the pandemic, containment measures, and treatment methods. The tweets were automatically classified using a dictionary-based computational method, ensuring reliability, validity, and reproducibility (KRIPPENDORFF, 2011; SAMPAIO; LYCARIÃO, 2018). A time series was performed to observe the daily variation of the evocation of each frame in the 93 days studied. A temporal network of users connected through mentions, retweets, and replies was also performed, on which the PageRank metric was calculated to measure their daily influence on the network. The ten most prominent actors were selected according to their PageRank on the peak date of each frame of the time series. Finally, information on the context of analysis and on the climate of opinion in Brazil was systematized through representative surveys of the Brazilian population with daily frequency (n = 1,800, ME = plus–minus 2 percent, CI = 95 percent). The results indicate that the order of evocation of the frames, from most to least prominent, was conflict, attribution of responsibility, consequences of the pandemic, morality, containment measures, and treatment methods. In particular, the frames of conflict, attribution of responsibility, and consequences of the pandemic were strongly related to a negative, episodic, and human interest framing of the tweets. On the other hand, the other frames (morality, containment measures and treatment methods), prioritized thematic framings, whose implications were preponderantly of a more social and less individual nature. Also noteworthy is the significant presence of anonymous profiles among the Top10 users of each frame, as well as activists, health experts, digital influencers, journalists, media organizations, politicians, and profiles of another nature, such as fandoms and satirical. Finally, regarding the study of the conditions that explain the highest peaks of evocation of the generic and specific frames in the time series, a phenomenon that was defined in this thesis as framing synchronization was observed. Framing synchronization is defined as the phenomenon of collective adjustment of the frequency of network activation of a given frame through interaction between individuals and influenced by five factors: context, succession of associated events, climate of opinion, and the combination of frames, and issues.