Abstract Analysing the centres of orientation in a literary text, narratologists since Genette have always worked with the distinction between voice (who speaks?) and viewpoint (who sees?). In contrast, nearly all lyricologists have remained entirely focussed on the category of voice, trying to identify, apart from the speaking voice, at best further speech instances in the poems under discussion. The present article challenges this voice-restricted tradition. It criticizes the deficiencies of various traditional one-voice-models (such as the identification of the speaking voice with the poet, a persona or the ›lyric I‹) as well as those of more recent voice-oriented approaches (by Hempfer and by Culler). It also discusses the blind spots of the multi-voice models developed in the course of lyricology’s ›narrative turn‹ (by Hühn and by Burdorf) and rejects Zymner’s recent suggestion to replace the notion of ›voice‹ by the purely textual concept of an ›addresser‹ as a ›structure of signs‹. Subsequently, the framework of an alternative, viewpoint-oriented lyricology is sketched by adopting from Cognitive Linguistics such key-concepts as viewpoint-multiplicity, mental spaces and deictic markers as well as from Cognitive Grammar the notions of ›viewing arrangement‹ and ›split conceptualiser‹. Further instruments that can help to account for those poems in which the centre of perception seems to unite in itself the viewpoints of different ›voices‹ are such concepts as ›viewpoint-blends‹ and ›viewpoint-zooms‹. More akin to cognition, in which there are no strict boundaries, but only transitions, they can help to overcome long-lasting, hardly successful attempts to establish rigid divisions in the lyric (e. g. between personal and impersonal, monologic and dialogic, biographical and non-biographical or lyric and narrative poems). Instead, the viewpoint-framework does justice to the fact that such binaries constitute no polar oppositions but mere scales of difference. Moreover, a viewpoint-approach makes possible a much more fine-grained analysis of poetry – in a way better even than an earlier attempt by Petzold (2012) to extend lyricology’s voice-centred methodology by integrating into it a Genette-inspired distinction of three types of focalization. The last section of this article is devoted to the complex problem of viewpoint-taking by the readers, which is aggravated by the underdetermined deixis and situational generality of many poems. As is finally illustrated by drawing on a poem with a particularly complicated deictic set-up (Eichendorff’s »Lockung« [»Lure«]), readers will admittedly not always be able to process all viewpoints contained in a text. Nevertheless, it can be shown with the help of this example that viewpoint-perception is often an indispensable prerequisite for a full understanding of a poem’s meaning.
Read full abstract