This essay argues that Percy Bysshe Shelley’s (1792–1822) analogical poetics at once anticipates and challenges contemporary cognitive-scientific models of conceptual structure. Section 1 outlines unresolved logical and motivational issues that limit the explanatory power of conceptual metaphor and blending theories with respect to conceptual conflict and creativity. The deficiency may be supplied by recourse to English Romantic theories of poetic imagination, which are centrally occupied with the logic of conceptual conflict and the motive for creativity. Particularly pertinent are Shelley’s suggestive hypotheses about the projective processes that drive novel metaphoric conceptualization, which he (in company with cognitive theorists) posits as the engine and outpost of creativity. To demonstrate the plausibility of Shelley’s hypotheses, section 2 marshals as evidence early twentiethcentury literarycritical responses to Shelley’s poetics, especially as enacted in his 1820 ode “To a Sky-Lark.” These diverge in their evaluations of the poem, but they nevertheless converge in their descriptive accounts of the peculiar cognitive effect primed by Shelley’s complexly metaphoric verse, which orients attention not to emergent meanings or achieved mental representations but rather to underlying processes of meaning-making and representation that precede, produce, and ceaselessly replace any such products of (literary) cognition. Section 3 attributes this peculiarly dynamic effect to the poem’s insistent violations both of consistent conceptual structure and of “directionality” constraints on metaphoric projections from one conceptual domain to another. These violations upset deeply ingrained habits of conceptualization, frustrating the normally automatic processes that generate more or less consistent mental representations and thereby rendering those processes perceptible. The analysis thus illustrates a reciprocal exchange between poetics and cognitive science: the systematic deviances of Shelley’s verse can be exactly characterized in terms developed by cognitive metaphor theory; so characterized, those deviances may in turn be systematically manipulated to test and improve blending theory’s account of what it itself describes as “the mind’s hidden complexities” (Fauconnier and Turner 2002).
This essay has been much improved by the generous and exacting criticism of Meir Sternberg, to whom I am most grateful.