Towards A Genre-based Analysis Model For Multimodal Newspaper Texts

Multimodality and multimodal discourses or texts have become part of contemporary life. Newspaper texts as an important form of mass communication in social life have also undergone a shift from texts presented dominantly in language with graphics as supporting or decorating illustrations, to the texts presented both in language and images at present. The change of modes of expression in news discourse has brought new demands for theoretical description and interpretation. Studies have to be conducted to describe and interpret those changes and the meaning multiplication by the combination of language and image in these multimodal newspaper texts.Multimodal newspaper texts in the present study refer to news stories and commentary articles in printed newspapers which are represented both in written language and image. They are important components of visual multimodal texts, which are created with the static combination of written language and image. Image is used as a cover term to include photographs and graphic artworks such as maps, diagrams, cartoons, comic strips or drawn symbols.Seeing a multimodal newspaper text as both an object of perception and an object of interpretation, we attempt to build a genre-based model which not only presents a systematic way to decompose a multimodal newspaper text into analyzable parts and to make it possible to analyze the pictorial and verbal elements separately, but also captures the mechanisms that make the different parts of the multimodal text into a coherent whole. In the genre-based model, compositional features of multimodal newspaper texts are identified and analyzed at three layers:the Base layers which deal with the intravisual interface, the Layout layer and the Rhetorical layer with the intervisual interface. The Base layer is to identify components or unit composites that compose a visual multimodal, and the different representation systems and meaning potentials of each mode can be dealt with separately here. At the Layout layer and the Rhetorical layer, the interaction between the two modes is discussed. The Layout layer deals with the spatial arrangement of unit composites and layout structure of the multimodal text as a whole, which can reveal the importance and salience of these units and be an important feature of multimodal newspaper texts as an object of perception. The Rhetorical layer deals with the logic-semantic organization or rhetorical relations that make the units into a coherent whole.However, the features identified at these layers cannot be accounted for from within. There are broader, more ’global’ decisions that influence the choices in the three layers. Therefore, we map the notion of genre to the study of multimodal texts. We propose that multimodal genre should be categorized by two parameters:one which takes into account features arising from the conditions of production and consumption; and the other which, defined according to the linguistic traditions of genre, deals with only the inherent features of the multimodal text. Print genre, online genre and mobi genre are identified by the first parameter. To make more delicate genre memberships among the print newspaper genre, multimodal newspaper texts are again categorized according to their communicative or generic purposes and layout pattern of verbal text and image. With these two parameters, multimodal newspaper texts can be grouped as image-nuclear news story, image-text complementary news story, image-nuclear opinion, and image-text complementary opinion.To reveal the relationship between genre and multimodal newspaper texts, we carry of a comparative study of two genres of multimodal newspaper texts. For this, we select3opinion articles on politics and then select another3news stories with the same or similar topics. The analysis shows that multimodal news texts have certain regularities in the three layers:1) At the Base layer, the features are presented by the consistent choices of typography (type face and size, etc.), graphics (photograph or cartoon), base units and unit composites.2) At the Layout layer, analyses with the area model show that multimodal news texts can be represented visually and spatially either by image-nuclear or image embedded layout design.3) At the Rhetorical layer, unit composites are related to each other logically and semantically, usually with the unit composite of Headline as nucleus. However, there are salient differences between the news texts analyzed:1) the multimodal news stories texts choose photograph whereas the opinion articles choose cartoons to make multimodal texts (a former study of50multimodal news stories and50multimodal opinion articles shows the same result and makes this finding more credible). The two types of image are saliently different in the realization of modality-in the photographs, there is no exaggeration of colour saturation, differentiation and modulation; the details of the participants and the circumstantial elements clearly, and thus makes the event it presents more contextualized; whereas in the cartoons, the elements are presented in black and white, the represented objects, entities or actions are abstracted via caricature or other ways of drawing, the main features of the represented participants have been emphasized to present them from particular attitudinal viewpoints and details are more reduced, so makes the presented event more de-contextualized.2) Transitivity analysis shows that the two texts are different in the choices of process types-in the news story, material process is used with the highest frequency and it is followed by verbal process which projects speeches, whereas in the opinion genre text, though material process is still the most frequently used type, it has a lower percentage, and relational process, instead of verbal, comes second.3) Analysis of mood and modality shows that the verbal bodies of the texts also are saliently different-the exclusive choice of declarative Mood structure to realize the unmarked speech function of making a statement in contrast with the joint use of declarative mood and interrogative mood in the opinion genre, and the high frequency of choice of past tense as primary tense in contrast with the high frequency of choices of present tense, and rare use of modality markers in the news story in contrast with more frequent use of polarity and modality makers in the opinion genre.4) In the opinion genre, there are repeated uses of first personal singular and description of the writer’s personal experiences, and there is a relatively high concentration of conditional clauses to discuss hypothetical, predicting events if particular actions are or are not followed. Such choices are not present in the news story genre. It is then found that genre considerations constrain the choices of unit composites and their relations, their layout patterns and rhetorical relations in image-text complementary news story genre and opinion genre in print newspaper. The newspaper genre determines that texts on newspaper pages share similarities in typographical, graphical and layout features, for newspapers at a time are produced with the same or similar technology and canvas. The typographical, graphical and layout choices enhance the severity and credibility of a qualify newspaper. Rhetorical relations show that rhetorical organization and layout organization are compatible with each other. And relations of expansion are more often used in these newspaper texts for expansion relations relate one unit composite to another as a direct representation of world experience, not as a representation of another (semiotic) representation. Sub-genres of newspaper genre constrain choices of unit composites and their realizations. The differences of news photograph and cartoon in realizing interactional meanings, and the different ways to interpersonal meaning in their verbal texts (Headline, Deck, Verbal Body) differentiate the opinion genre and the news story genre, and work together to achieve the communicative purpose of each genre. The photograph presents a real scene in standard color saturation, differentiation and modulation and attempts to show objectively a reality, whereas the emphasis of cartoon is on presenting a subject in such a way that main features are caricatured, and so drawing attention to the attitudes that the drawers with to express Cartoons are therefore more subjective and opinion-salient, and more suitable to be in the opinion genre of newspaper texts. The different choices to realize interpersonal meanings are also contributive to the communicative purposes of the two genres:the exclusive choice of declarative Mood, the most frequent use of past tense and rare use of modality markers in the news story show that the news story is just to recount objectively the events happed, rather than to make personal inferences or judgments. Choices of declarative and interrogative Mood structures, the high frequency of choices of present tense and polarity and modality makers, the choices of first person singular pronoun and conditionals for discussion of hypothetical situations or possibilities for the future help to express the writer’s opinions overtly. Genre expectations are reflected and shaped by texts that are instances of those multimodal newspaper genres.The present study is significant for it proposes an integrated model which not only offers sufficient and comprehensive description of the features in multimodal newspaper texts but also can investigate the correlation between multimodal genre and textual features-how the choices of verbal and pictorial elements contribute to the communicative purposes of different genres. Generic regularities can be identified and interpreted more effectively when we have a sufficiently detailed characterization of the features at the different stratum in our model. With the analysis of the correlations between genre and linguistic and pictorial features, we contribute to the line of profile research on mutual predictability between genre and texts, and to the line of profile research on news discourse. And undoubtedly, the study will help readers to gain better understanding of multimodal newspaper texts. It is hoped that the model can be applicable for analyzing all visual multimodal texts.