Language and Literature: International Journal of Stylistics

The essential structural property of metaphor is that it represents a double-layered conceptualisation of the target domain (the ‘literal’ or ‘tenor’ language) in terms of the source domain (the ‘figurative’ or ‘vehicle’ language). Most linguistic approaches to metaphor provide sentence-level accounts of the phenomenon. But literary metaphor is frequently discursive: there is an entire metaphorical ‘undercurrent’ running through a whole text, which may manifest itself in a large number and variety of ‘single’ metaphors. What is needed, therefore, is (i) a way of accounting for metaphor discursively, rather than sententially; (ii) a way of dealing with the resolving ‘undercurrent’ stratum rather than the superficial ‘single metaphor’ stratum; and (iii) a way of representing the double-layered conceptual structure of metaphors. A new conceptual discourse model of text-worlds is presented here, which naturally captures the conceptual layering inherent in language generally (and not just in metaphor), which treats the ‘undercurrent’ aspect as being equivalent to ‘gist’ or ‘macrostructure’ in text linguistics (using the concept of the ‘megametaphor’), and which automatically provides a discursive account of the phenomenon of extended, or sustained, metaphor.

Prosodists have never developed an adequate yet coherent definition of rhythm. The most popular definition, ‘periodic recurrence,’ is coherent but inadequate. Elaborations of this definition have been more adequate but incoherent. Like musical rhythms, linguistic rhythms are best defined as recursive hierarchies of prominence. Complex rhythms are componential, organising three prominence hierarchies (metre, grouping, and prolongation) into tightly co-ordinated forms. These co-ordinated hierarchical forms create our impression of processive time.

‘Point of view’ in fiction has been a much debated concept ever since the time of Henry James, but unfortunately this term has never been defined with the required precision. As a result, there has always been some confusion in the critical discussion of this subject. Seymour Chatman (1990), however, has recently addressed himself to the difficult task of clarifying the issue.² His theory, in my view, offers an excellent model for a systematic description of narrative ‘point of view’. By way of demonstration, I will test the effectiveness of this model through a detailed analysis of the narration of D.H. Lawrence's short story ‘The blind man’.