Abstract
Karen Sullivan explores the way in which metaphor is evoked by combining heretofore separately handled approaches to its analysis from within cognitive linguistics, including frame semantics, cognitive grammar (CG), construction grammar (CxG) and conceptual metaphor theory (CMT). Drawing largely on uncontentious examples of metaphor to assist in the development of a systematic and sustained account of how its meaning is put together, she provides in-depth semantic analyses which support her argument that it ‘piggybacks’ on non-metaphoric constructions. The illumination of such process of meaning-making presents the overriding aim of the book.
Central to Sullivan’s project of description is the identification of dependent and autonomous elements within metaphors which are classed as conceptual, namely those which convey one thing as being another, as well documented in the Master Metaphor List, a staple resource of CMT. The list comprises set items, such as SOCIETY IS A BODY, which Sullivan also applies as underpinning, ready-made conceptual arrangements to her analysis. As she explains in the introduction to her book, conceptual metaphors usually involve the mapping of a concrete and easily grasped concept (source domain, e.g. BODY) onto a more abstract one (target domain, e.g. SOCIETY), allowing for ease of processing of combinatorial formulations such as ‘economic health’ (economy < SOCIETY [target] = health < BODY [source]). A principal thesis by Sullivan is that the metaphor’s dependent elements evoke the source, while its autonomous ones point to the target domains. The greater part of the book is therefore dedicated to describing the independent and autonomous elements within metaphoric constructions, as well as the way in which the relevant conceptual source and target domains are mapped onto one another.
Sullivan’s use of autonomy and dependence is based on semantics, not syntax, with the concepts themselves derived from CG, which similarly incorporates semantics into its conceptualization of grammar. In brief, the autonomous element is argued to elaborate the dependent one – it ‘fills in’ the dependent element, which is hence understood as relative to it. In truth, both elaborate one another, as Sullivan herself concedes; yet the autonomous element is taken to elaborate the dependent one to a greater extent than vice versa. Despite rather a complex discussion in clarification of such elaboration, which also integrates additional concepts from CG, determining the autonomy–dependence relation appears to be largely centered on the actual ease with which one is able to conceptualize the one without the other, that is, the easier one being labeled autonomous.
While Sullivan sets out from a predefined base of assumptions about what constitutes conceptual metaphor, the CMT list of concept domains, she insightfully adds the component of frame semantics to her analysis of mapping. This allows her to define elements on the basis of generic roles pertaining to things, people, events or situations, which are seen to fit within a semantic nexus of associations. In doing so, she draws mainly on pre-existing examples, those of the FrameNet project, which utilizes data from the British National Corpus. However, it is not always clear why some roles are selected over others in illustration of mapping, casting some doubt on analytic systematicity.
Throughout, meaning is seen to combine with form in the analysis of what are correspondingly represented as constructions, in keeping with the premise of CxG. Following its initial theoretical overview, the book therefore primarily describes metaphoric domain mappings by means of such constructions, which are selected by Sullivan herself to include adjectives and adverbs, argument structure, closed class items, as well as combined and extended constructions (although she further considers metonymic inferencing, albeit in a somewhat digressional attempt to define metaphor more narrowly by means of its differentiation).
The process of piggybacking itself, in terms of moving on the back of constructions from one domain to another, appears to depend on the semantic frames which Sullivan has identified from within a corpus of non-metaphoric language use. As such, they allow for theoretical claims of invariance, that is, of non-metaphoric and metaphoric use. As this presents the defining thesis of the book, the interrelation of constructions and frames might be clarified a little better by Sullivan in the conclusions, which are somewhat scant in conveying their overall relevance. This would appear all the more necessary as she has selected some over other constructions in illustration of autonomy–dependence relations which support her core argument.
The book is part of an edited series on ‘Constructional Approaches to Language’, and it is probably fair to say that it would be best appreciated by the reader with some, or more, grounding in cognitive linguistics, and with consequent understanding of its theoretical premises, as well as the nitty-gritty of Sullivan’s fine-grained semantic analyses; nevertheless, it is arguably an interesting read for any linguist. While Sullivan’s work is lucidly presented, the cognitive linguist will be better placed to fully appreciate her feat of bringing together various and complex elements from their seemingly insular subfields, to present a workable analysis, which would appear to considerably extend the existing body of knowledge on metaphor.
For those who prefer to analyze metaphor from within its context of use, for example, discourse analysts, the semantic focus may appear somewhat narrow and deep, not to mention filtered, in so far as it draws on pre-existing conceptual metaphors (from the list) and frames (from FrameNet). However, this is precisely what allows Sullivan to draw generalizations about the way in which metaphor is constructed, with potentially widespread implications for our understanding of cognitive representations as systems of shared knowledge and belief. Most insightful, perhaps, is the degree to which metaphor, of the type which requires the least processing effort, is shown to be constrained in its articulation by non-metaphoric meaning structures. As Sullivan argues, it is the very boundedness of creativity which allows for its efficiency in communication.