Aspects of Methaphorical Definition in the Sciences

ASPECTS OF METHAPHORICAL DEFINITION IN THE SCIENCES Willian Frawley 1.Introduction In their classic essay on the structure of definition in natural languages, Manfred Bierwisch and Ferenc Kiefer (1970) remark that consideration of the types of sentences which occur in definitions might not only be instructive as to the character of ordinary language definitions, but "might also shed some light on the complex problem of the character and development of systems of theory-bound terms" (Bierwisch and Kiefer, 1970:76). That is, a systematic unpacking of the problem of ordinary language definitions may be revealing about the structure of meaning in specialized discourse: science in particular. It is the purpose of this paper to take Bierwisch and Kiefer's suggestion as imperative and to look at the structure of definitions in scientific lexicons, focusing especially on the character of metaphorical definitions in science. To do this, I first consider the nature and function of metaphorical expressions in science to justify scientific metaphors as objects worth studying and to see what sort of information might have to be codified by the lexicographer of science. Thereafter, I discuss the structure of scientific definitions, emphasizing the logic of metaphorical definition to demonstrate how metaphorical definitions represent the semantic functions of metaphor in scientific discourse in general. 2.Metaphor in Scientific Discourse One of the most important legacies of the demise of logical empiricist metascience is the realization that science is not a rigidly referential discourse which reduces all of its terms to ostensión. That scientific discourse is shot through with imprecision, ambiguity , approximate meaning, and metaphor has been the essential claim of numerous post-empiricist philosophers of science. Pepper (1942) articulated quite early the position that every explanatory system has at its base a "root metaphor" which organizes the hypotheses of the discipline seeking to explain the world. Seventeenth century mechanism and nineteenth century organicism 118 William Frawley1 19 The second problem of the definiens which is relevant here is its openness. If one looks cursorily through a specialized lexicon, one sees great variability in the length of the definientes. Why is this? The answer, I think, lies in the type of definition. Substitutive and reductive definientes are closed. Since they introduce no new information into the definition, the definientes have a limit. This is obvious for substitutive definientes, which function only to provide alternate nomenclature and which terminate after the introduction of the synonym. But it is not so obvious for reductive definientes. Consider the definiens of "bolometric correction" from the astrophysics glossary: "the visual (or photovisual) magnitude minus the bolometric magnitude of a star. It is always a positive number" (Hopkins, 1980:19). Why does the definiens stop there? Because a reductive definiens is found in a complete definition, the meaning of the definiendum can be expressed totally: the concept which the definiendum represents is completely known in astrophysics, and the reductive definiens reaches its limit when the definiendum is fully explicated (in fact, because it is fully explicated). This is not true for metaphorical definitions, whose definientes are open because of the fact that they introduce new information. Consider the term "black hole" in astrophysics; it is defined by Hopkins (1980:18) as follows: A black hole partitions three-demensional space into two regions: an inner region bounded by a smooth two-dimensional surface (called the horizon), and the region outside the horizon which is asymptotically flat. The essential characteristic of a black hole is that the inner region cannot communicate with the outer region: i.e., light cannot escape from it. It can be said that at the horizon the gravitational field is so strong that even light cannot escape. For a spherically symmetric black hole radius of the horizon is 2GM/c2 — sometimes called the Schwartzschild radius. If an object with the mass of the Sun has a radius of 2.5 km, it would be a black hole. Black holes represent one of the possible endpoints of stellar evolution for stars very much more massive than the Chandresekhar limit. Good black hole candidates: Cygnus X-I, Circinus X-I, and V861 Scorpii. 120Metaphorical Definition in the Sciences As is well known from popularized astronomy, a black...