How Conceptual Metaphors are Productive of Spatial-Graphical Expressions

Abstract

How Conceptual Metaphors are Productive of Spatial-Graphical Expressions Timothy C. Clausner (Clausner@HRL.Com) HRL Laboratories, LLC Human Centered Systems Department, 3011 Malibu Canyon Road Malibu, CA 90049 USA Abstract The theory of conceptual metaphors is adopted in which conceptual relations are productive of linguistic meta- phorical expressions. Conceptual metaphors vary in their degree of productivity according to semantic principles. Spatial-graphical expressions of non-spatial concepts are investigated providing evidence that they are instantia- tions of metaphors. For three cases of differing productivity it is argued that the same semantic princi- ples which result in metaphor productivity for linguistic expressions also result in spatial-graphical expressions. 1 Background Language gives us words, and constructions made of words, to talk about abstract concepts. We find in space, conventional shapes and organizations of shapes which also convey abstract concepts. These representa- tions in space are typically experienced visually, but not exclusive of other experiential modalities. This paper addresses the problem of how spatial-graphical repre- sentations convey abstract meanings by means of metaphors, which allow us to understand or express abstract concepts in terms of concrete expressions, par- ticularly ordinary, relatively static, conventional devices (e.g., map legends, key pads, and clocks). Fourceville’s (1996) analysis of abstract concepts conveyed by creative images and language in advertis- ing, aims toward a theory of ‘pictorial metaphor’. Tversky (2001) treats depictions, such as maps, graph- ics, and icons as involving spatial metaphor derived from concrete world experience, across languages and cultures. Zacks, & Tversky (1999) argue that systematic correspondences between graph forms and interpreta- tion are naturally derived, not due to knowledge of explicit conventions. This paper takes a similar treat- ment of metaphor, adopting cognitive semantic theory (Clausner, 1993, 1994; Clausner & Croft, 1997; Grady; 1997; Lakoff, 1993; Lakoff & Johnson, 1980; Lakoff & Turner, 1987), which treats metaphor as conventional schemas expressive of ordinary conventional language. The author thanks the three anonymous reviewers of this paper for their helpful comments. An earlier version of this research was presented at the Seventh International Cognitive Linguistics Conference, 2001, Santa Barbara, California. In this theory of metaphor, knowledge is organized into experientially based domains; e.g., SPACE , TIME , LIVING THINGS (see Clausner & Croft, 1999, for an overview of the theory of domains in cognitive seman- tics). A conventional metaphor is a stored relation between two domains. Concepts from an abstract (tar- get) domain are systematically comprehended or expressed in terms of concepts from a different, often concrete, (source) domain. For example, MORE IS UP AND LESS IS DOWN is a conventional metaphor whose source domain UP - DOWN stands in relation to the target domain MORE - LESS . This metaphor is a semantic struc- ture which can be instantiated as linguistic expressions; e.g., rising prices, fell ill, high esteem, fell unconscious. Language expresses abstract concepts metaphorically by means of spatial and other basic perceptual concepts (Grady, 1997). Metaphors that relate spatial source do- mains to non-spatial target domains can be productive of linguistic expressions about non-spatial abstract meanings by using words having spatial meanings. The metaphor MORE IS UP is strongly implicated by investi- gations of graphs as expressions in space. Tversky, Kugelmass & Winter (1991) found that subjects as- signed interpretations to the axes of graphs, such that increasing quantity was preferentially assigned to the vertical axis, and temporal concepts were preferentially assigned to the horizontal axis. Gatis & Holyoak (1996) investigated subjects’ interpretation of graphs, finding a significant advantage when the variable being queried was assigned to the vertical axis. They argue that graphing increasing quantity in terms of vertical spatial increase is based on the metaphor MORE IS UP . Given that there is evidence for conventional metaphor being expressed in the construal of spatial graphs, this paper proposes the following hypothesis: The same cognitive principles which determine metaphor productivity for linguistic expressions also determine metaphor produc- tivity for spatial expressions. This hypothesis will be tested with respect to a specific technical characteriza- tion of metaphor productivity. Productivity in Metaphors Clausner & Croft (1997) argue that just as phonological schemas vary in their productivity of base-derived relations, so semantic schemas (i.e., meta-