Quadros lexicais em editoriais, reportagem jornalística e revista de notícias: uma abordagem da Linguística de Corpus
Data
2021-08-31
Autores
Odízio, Íris Cristine 
Orientadores
Veirano Pinto, Marcia
Tipo
Dissertação de mestrado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
O objetivo principal deste trabalho é descrever a linguagem dos registros escritos editoriais, reportagem jornalística e revista de notícias por meio da identificação, classificação discursivo-funcional e análise da variação sistemática de seus quadros lexicais. Quadros lexicais são sequências de palavras semifixas, isto é, sequências de palavras compostas por duas ou mais palavras fixas e uma variável (BIBER; GRAY, 2013). A abordagem da Linguística de Corpus, que se ocupa primordialmente da linguagem em uso, será o norteador teórico deste trabalho. O corpus investigado é um subcorpus do Corpus Brasileiro de Variação de Registro (CBVR), este último composto por 48 registros em sua totalidade. Cada um dos três registros que compõe o corpus deste estudo contém 20 textos, totalizando, portanto, 60 textos, apresentando 10.047 ocorrências (palavras) e 3.227 tipos (palavras distintas) em editoriais; 10.737 ocorrências e 3.242 tipos em reportagem jornalística e 18.309 ocorrências e 4.755 tipos em revista de notícias. A identificação dos quadros lexicais do corpus foi efetuada com o auxílio do software AntGram 1.2.3 (ANTHONY, 2020). Os quadros foram identificados de acordo com os seguintes critérios: (1) pertenciam exclusivamente a um dos registros em estudo; (2) eram comuns aos três registros em estudo; (3) estavam presentes em pelo menos sete textos de cada registro e (4) tinham frequência mínima de 20 ocorrências em cada registro. Esses cuidados foram tomados para que os quadros lexicais fossem efetivamente característicos dos registros em questão e não apenas frequentes em poucos textos. Após a identificação e seleção dos quadros lexicais, as ferramentas Concordance e Concordance Plot, pertencentes à suíte de ferramentas de análise linguística AntConc 3.5.8 (ANTHONY, 2019), foram utilizadas para que os quadros lexicais fossem classificados funcionalmente. Tal classificação funcional permitiu uma melhor compreensão do papel da linguagem na elaboração de editoriais, reportagem jornalística e revista de notícias e foi embasada pelas categorias propostas por Biber (2006). Os resultados mostram que o registro editorial é o único abertamente marcado por posicionamento e que o registro revista de notícias parece ser mais padronizado e informal do que o registro reportagem jornalística, por ter um número maior de quadros lexicais e apresentar expressões idiomáticas em sua linguagem.
The aim of this study is to profile the language of editorials, newspaper reportage, and magazine news by means of (1) identifying their lexical frames; (2) classifying these frames according to their discourse functions; (3) determining the extent of variation in these frames in terms of frequency and discourse functions across registers. Lexical frames are discontinuous sequences in which two or more words form a ‘frame’ that surround a variable slot (BIBER, GRAY, 2013). The study is based on the corpus linguistics approach, which investigates how language is used by speakers and writers. The corpus used in this study is a subset of the Brazilian Register Variation Corpus (CBVR), which, in its entirety, is composed of 48 registers. Each of the three registers selected for this study—i.e., editorials, newspaper reportage, and magazine news—is comprised of 20 texts, therefore totaling 60 texts and 10,047 tokens and 3,227 types (editorials), 10,737 tokens and 3,242 types (newspaper reportage), and 18,309 tokens and 4,755 types (magazine news), totalling 39,093 tokens and 11,224 types. The lexical frames were identified with the help of the software AntGram 1.2.3 (ANTHONY, 2020). The frames were identified according to the following criteria: (1) they belonged to only one of the registers in the corpus; (2) they belonged to all of the registers in the corpus; (3) they were present in at least seven texts in each register and (4) they had a minimum frequency of 20 occurrences per register. These criteria were adopted to ensure that the lexical frames were representative of the registers being studied rather than of just few texts. Once the lexical frames were identified, the Concordance and Concordance Plot, tools in AntConc 3.5.8 (ANTHONY, 2019), were used to aid the functional classification of the frames. Such functional classification, which was based on the categories proposed by Biber (2006), allowed for a better understanding of language use in the writing of editorials, newspaper reportage, and magazine news. Results show that the register editorials is the only register overtly marked by stance and that the register magazine news seem to be more patterned and informal than the register newspaper reportage. Such conclusion is based on a greater number of lexical frames identified for the register magazine news as well as the presence of idiomatic expressions in its discourse.
The aim of this study is to profile the language of editorials, newspaper reportage, and magazine news by means of (1) identifying their lexical frames; (2) classifying these frames according to their discourse functions; (3) determining the extent of variation in these frames in terms of frequency and discourse functions across registers. Lexical frames are discontinuous sequences in which two or more words form a ‘frame’ that surround a variable slot (BIBER, GRAY, 2013). The study is based on the corpus linguistics approach, which investigates how language is used by speakers and writers. The corpus used in this study is a subset of the Brazilian Register Variation Corpus (CBVR), which, in its entirety, is composed of 48 registers. Each of the three registers selected for this study—i.e., editorials, newspaper reportage, and magazine news—is comprised of 20 texts, therefore totaling 60 texts and 10,047 tokens and 3,227 types (editorials), 10,737 tokens and 3,242 types (newspaper reportage), and 18,309 tokens and 4,755 types (magazine news), totalling 39,093 tokens and 11,224 types. The lexical frames were identified with the help of the software AntGram 1.2.3 (ANTHONY, 2020). The frames were identified according to the following criteria: (1) they belonged to only one of the registers in the corpus; (2) they belonged to all of the registers in the corpus; (3) they were present in at least seven texts in each register and (4) they had a minimum frequency of 20 occurrences per register. These criteria were adopted to ensure that the lexical frames were representative of the registers being studied rather than of just few texts. Once the lexical frames were identified, the Concordance and Concordance Plot, tools in AntConc 3.5.8 (ANTHONY, 2019), were used to aid the functional classification of the frames. Such functional classification, which was based on the categories proposed by Biber (2006), allowed for a better understanding of language use in the writing of editorials, newspaper reportage, and magazine news. Results show that the register editorials is the only register overtly marked by stance and that the register magazine news seem to be more patterned and informal than the register newspaper reportage. Such conclusion is based on a greater number of lexical frames identified for the register magazine news as well as the presence of idiomatic expressions in its discourse.