符號與傳媒:總第11輯:2015秋季號:11 節選

　　《符號與傳媒（2015秋季號11）》：　　All the while， I will be particularly interested in establishing if.there issuch a thing as pictorial metaphors （See Sonesson， 2003）.According to Lakoff， metaphoricity depends on cognitive， rather than linguistic，structures， and thus may well be identical for language and pictures as well asother visual s'igns.Indeed， this seems'to be taken for granted by， for instance， Charles Forceville （2004—2009）.Among students of pictorialsemiotics， on the contrary， the conclusion has been， as we will see， thatthere can be no pictorial metaphors， although other rhetorical figures are possible.I will suggest， however， that given the functional definition of themetaphor， inspired as much in Max Black （1962） as in Peirce， pictorial metaphors exist， but are differently manifested from those of verballanguage.On the other hand， the picture is an iconic sign， whether it contains metaphors or not， which is to say that all pictures share a property with metaphors， whether pictorial or not： that of being an iconic sign.　　Metaphor is， together with metonymy， one of the most abused terms ofsemiotics， starting from Jakobson's （1942） equation of the former with the syntagm， and the latter with the paradigm.This amounts to a confusion ofrelationships within sign systems with relationships between particular sign tokens or secondary relationships between signs.Indeed， metaphors， likemetonymies， are either created in a particular given text， or they are stock images， which relate signs， or at least sign contents， in stable relations，whereas syntagms and paradigms are constant primary relationships between signs.Moreover， the similarity present in a paradigm is often simply the position in the syntagm， whereas some more pregnant similarity relation is required in the metaphor.　　……