Abstract：Abstract: Over the past decades, two de facto hypotheses about translation have been upheld: One is that translated text is composed of only linguistic code, and the other is that interlingual translation is perceived of as translation per se. The two hypotheses, an epitome of verbocentrism, are no longer valid in handling the problems raised by different semiotic interchange within visual texts. This paper, from multimodal perspective, proposes three new hypotheses about translation based on a critical analysis of the above two old ones. Accordingly, studies of visual texts translation need to give its top priority to the following four problematics: the study of “lexis” of visual texts, the study of “grammar” of visual texts, the study of transformation mechanism of visual texts and the study of methodology.
Key words: multimodal; verbocentrism; visual texts; translation
