There is an interesting slippage in Jakobson's theory. If we remember his terminology we have context (referential function), addressor (emotive), addressee (conative), contact (phatic; gotta love that word), code (metalingual), and message (poetic).
In other words, he identifies the message not with referential content but with the phonetic and prosodic characteristics of that particle piece of language used to communicate. So my post below is wrong: in a translation the message will be totally transformed. What will remain constant is the contextual or referential function of language.
So here are two problems: context does not (necessarily) equal "referential." There is the context of communication, say 19th century French realism with Balzac and his French readers of the time. And there is the referential aspect of this same communication: the imagined world of the comédie humaine.
Secondly, there are two meanings of message: the message as a particular piece of language, with its phonetics, its prosody, its morphology, etc... and the message as understood as semantic content extractable for the purposes of translation. RJ clearly means the first with "message" because he defines it as poetic and not, say, as the "ideological function" of language. I don't mean to over think this, but there is a sort of murkiness in definition. A need for a few new terms, maybe.
Yes, RJ's idiosyncratic terminology has always been a problem, esp. in classroom situations, where the sense of "message" as having to do directly with meaning is always so strong in students' minds. The easiest way to break it down is to think of message in his sense of corresponding to Saussure's signifier, and context to signified. This, in turn, requires further clarification of the difference between signified and referent (which is always needed anyway).
ResponderEliminar