A New Definition of 'Simile': Aristotle Reconsidered

John E. Ziolkowski (George Washington University)

1. The Problem: Distinguishing Similes From Metaphors.

A. Statement of confusion: “The interpenetration of simile and metaphor has proven to be a most confusing issue in discussions of these figures of speech, and the confusion also seems to go all the way back to Aristotle.” (Steve Nimis, "Aristotle's Analogical Metaphor," Arethusa 21.2 [1988] 215)

B. Similar constructions given different names: "Thus strife . . . should perish / and wrath . . . / which rises up much sweeter than dripping honey (polu glukiwn melitoV kataleibomenoio) in the chests of men or like smoke (hute kapnoV)." A traditional scholar (Lucas, Style [1962] 186) called these two figures a metaphor followed by a simile. In my opinion the two expressions are both similes introduced by different words, a comparative adjective implying 'than' in the first expression and a ‘particle’ equivalent to 'like' in the second.

C. The same construction called both simile and metaphor: at Critias 109b5-7 a passage comparing gods to shepherds ("just as a shepherd tends his flock") is called a simile on one page of a recent book and a metaphor on the next ("the established shepherd metaphor”). Then 'the helmsman metaphor' at Critias 109 c2-4 is labeled 'an appropriate simile.'

2. The traditional definition and my proposal.

Aristotle says that a simile is a weakened metaphor, distinguished only by the presence of a word that makes it explicit (a prothesis - the words 'as' or 'like'). The problem with this definition is that it makes the distinction so minuscule that there is no useful reason to make it. The notion of explicitness is fine but I propose revising the definition of explicitness. It is not the expressed prothesis that should define a simile, but rather the expressed tenor. Tenor is the subject of a comparison and vehicle is the comparison introduced: "Your eyes (tenor) are like (prothesis) diamonds (vehicle)." In Aristotle there is no equivalent for what we call 'tenor’ and ‘vehicle', but if we employ these useful terms we can create new definitions with more sharply defined borders. Let us define a simile as "a figurative comparison in which both tenor and vehicle are expressed: 'tenor + vehicle' (+/- prothesis)" while a metaphor would be "an implied comparison in which the vehicle is expressed but the tenor is not: '(tenor?) + vehicle'.” For example, at Odyssey 20.13-15 Odysseus’ heart ‘growls’ metaphorically in anger at the behavior of the Suitors’ mistresses and then a simile of a dog growling follows immediately; but the action of Odysseus' heart that is described as a growl (vehicle) is only implied: "beat" (the tenor). In the simile, on the other hand, both tenor and vehicle are stated: “As a dog (vehicle) growls . . . so he (tenor) growled.” Furthermore, it is not merely a comparative conjunction (‘like’ or ‘as’) that calls attention to the construction but also verbs, nouns, adjectives and adverbs. Also even this word need not be expressed when the tenor is clear (in predicate constructions like the example above).

3. Discussion of recent scholarship.

Scholars (Blondell, Boys-Stones, Innes, Kittay, Pender, Silk) have produced a proliferation of terms ('metaphorical transference', 'extension by synecdoche') and distinctions (illustrative vs modeling, simple vs complex, ordinary vs deviant, normal vs abnormal). My proposal perhaps would not affect their general conclusions, but it would simplify the recognition of similes, expand the number of comparisons that should be considered similes, and eliminate excessive terminology.