Mind the gap – from polyphony to text
How people, as individuals or collectives, are constructed and ascribed points of view through language, and how language can be used to apply different perspectives and thereby allow other voices besides that of the speaker to be heard, are questions that have been addressed within a number of theoretical paradigms. Here, we will take the theory of linguistic polyphony from the French tradition (Ducrot 1984, Nølke, Fløttum et Norén 2004) and text linguistics drawing from the Anglo-Saxon paradigm (Nyström 2001) as our starting point.
Proceeding from Ducrot’s work, Nølke et al. (2004) have developed a formalised model for analysing the voices that manifest themselves in discourse, the types of points of view which they advance, and the accountability relations that exist to those points of view. The voices are inscribed in the semantics of the language and their chief characteristic is that they are made accountable for points of view, that consist of a meaning that can be expressed as a propositional content, but which does not necessarily assume the form of an utterance (Norén 2000, Nølke et al. 2004, 31).
Polyphony theory has largely confined itself to the semantics of words and utterances. Attempts have been made, however, to make the model applicable at the level of the text (Adam 2002, Fløttum 2010, Gjerstad 2011, Nølke et al. 2004, 99–116). Fløttum (2000, 2003) proposes the terms linguistic person and polyphonic coherence to bridge the gap between utterance and text. We hope to push the analysis a step further by introducing the concept of a polyphonic collective of multiple voices supporting the same line of argumentation. This does not mean that their points of view need to be rewordings of each other, but it is essential that they are argumentatively aligned towards the same conclusion (Adam 2006, 77, Ducrot 1980, 25, Norén 1999, 43).
While the strength of polyphony theory lies in its ability to describe different points of view within a single utterance, text linguistic studies show how expressions in different utterances are bound together. Nyström (2001, 40–44) distinguishes between identity links and links based on a semantic relationship. The first category of links includes actual identity while the second includes hyponymy/hypernymy and specification/generalisation (ibid. 46–47). These differ in that identity links presuppose the same referent, while a semantic link involves a relationship between part and whole.
Whereas referential chains can to some extent be defined in terms of semantic extension, polyphonic collectives are seen as an argumentative orientation rather than reference. Likewise, the argumentative orientation of a text cannot be determined in relation to an established conclusion, but is to be understood from the causal or consecutive relationships.
Linking polyphony and text linguistics has proved both effective and capable of operationalisation. They can be employed to investigate voices and points of view, but at different levels. Whereas the former handles semantic instructions and the semantics of words and utterances, the latter is concerned with the links between words and groups of words. A systematic application of the two theories and the development of bridging functions to overcome the gap between utterance and text, such as polyphonic collective, polyphonic and argumentative coherence and argumentative alignment, it is possible to analyse how different voices emerge and thus to arrive at an overall understanding of the argumentation of a text.
Texte & discours en confrontation dans l’espace européen. Pour un renouvellement épistémologique et heuristique. Metz, France