Recent media interest in rhetorical figures (e.g. NPR’s program “On the Media,” 9/19/08) prompted by campaign speeches that exploit them has centered on antimetabole, which is (pace Janet Lapidos on Slate.com, 9/12/08) a species of chiasmus, defined as any structure in which the constituents are repeated in reverse, yielding the pattern ABBA. An oft-cited example of antimetabole is Quintilian’s Non ut edam vivo, sed ut vivam edo “I do not live that I may eat, but eat that I may live” (Institutio oratoria 9.3.85). This figure is related to such others in the nomenclature as polyptoton and antanaclasis. What unites them, besides their being patterned repetitions, is the master trope, PARONOMASIA, alias the pun.
What puns in ordinary discourse do–like any paronomasia–is (inter alia) call attention to themselves by exposing a formal resemblance (including complete identity) that undergirds semantic difference. Paronomasia rises to loftier heights as the stock in trade of poetic language. In the tradition of European verse, even rhyme is a kind of paronomasia. What is important to understand in rhyme, moreover, is the fact of its functioning to establish a semantic equivalence between rhyme-fellows despite their difference in meaning. Any two words that rhyme are ipso facto likened to each other in meaning by the very fact of their form as such. As with all elements of poetic language, form thereby becomes a part of content.
This condition is important to understand in assaying the impact of rhetorical figures like antimetabole. Beyond the simple fact of their calling attention to themselves as formal entities (the so-called poetic function), they have the further effect of calling the very fixity of meaning into question (what the Russian Formalists termed ostranenie ‘making it strange’ and made the foundation of their theory of modern aesthetics). Paronomastic figures like antimetabole tend to undermine this fixity of meaning. Nowhere is this more potent than in the unmasking of clichés or fixed phrases.
When politicians maunder about “change” while resorting to figures like antimetabole–presumably because the figure recalls other (and more illustrious) politicians’ use of it (Roosevelt, Churchill)–there may be some half-conscious sense on the speechwriters’ part that weakening the fixity of meaning in this way lends rhetorical support to the message of “change.” Form thus enters content. But there is also a cost to this rhetorical strategy, and it is not just that prominent use of figures of speech tends to detract from the message by underscoring “rhetoric” at the expense of “substance.” Rather, it is that focus on the message for its own sake (the ‘how,’ alias the poetic function) always tends to abrade the validity of the referential function (the ‘what’) of the same message. Form as part of content thus poses an epistemological danger, one which Plato detected long ago when he called poets liars in the Republic.

MICHAEL SHAPIRO