Phonologists have made lasting contributions to the study of metrical verse, but have been largely absent from discussions of prosody in non-metrical ('free') verse, despite interest from practitioners of literary criticism in incorporating insights from phonology (see e.g. Rumsey 2000, Gerber 2015). I will argue that this is a niche ripe for phonological exploitation, and that the failure to do so thus far is not arbitrary, but related to the analogy Generative Metrics makes between verse and grammar (Halle and Keyser 1966, Kiparsky 1975, 1977, Hayes 1983, 1988). Proceeding on this analogy, Generative Metrics has greatly improved our understanding of the limits to metrical variation in the corpora of individual poets and poetic traditions. Taken too literally, however, it obscures the 'artfulness' of language art — a point made in one of the earliest critiques of Halle and Keyser's seminal work by W. K. Wimsatt (1970). 

This talk is an attempt to reimagine how linguistically informed scholarship on language art might develop in the light of Wimsatt's objections, which have never been answered, and point to new research opportunities for phonologists interested in literary applications. What is needed, I argue, is a reorientation of linguistic poetics to Aristotle's notion of mimesis (see e.g. Thompson 1961 for a mimetic understanding of the relation between metre and prosody, and Meltzoff 2012 for a contemporary view of its psychological and evolutionary underpinnings). Since mimesis, in the sense relevant to understanding artful expression, is a type of ostension or gesture, the required theoretical adjustments are of a 'structure-adding', rather than a 'structure-changing' type: it has a natural place as part of an ostensive-inferential pragmatics in Kiparsky's (1987) revised version of Jakobson's (1960) linguistic poetics program. (See also Sperber and Wilson 1996 [1986], and Pilkington 1992 for applications of ostensive-inferential pragmatics to literary devices.)

I will give empirical demonstrations of artful linguistic mimesis on two levels. First, I will consider how typographic arrangement may mimetically cue prosodic structure and intonation, illustrating with well known poems by William Carlos Williams, Gwendolyn Brooks, and Robert Creeley, and drawing on evidence from the analysis of recitations by the poets themselves (made publically available through the PennSound Archive and Woodberry Poetry Room at Harvard). Second, I will address prosodic structure and intonation in recitations of selected nonmetrical verse, showing how Ezra Pound's incantatory style can be described phonologically, laying the foundations for further, phonologically informed, critical work. An important aim here will be to show how prosody and intonation mimic (evoke similar responses to) biologically evolved signals that have been the focus of recent evolutionary accounts of the human capacity for rhythm and discrete tonality, which also underlie musical ability (e.g. Mehr and Krasnow 2017). I will conclude by considering how the view of language art that emerges from these explorations may help shape our understanding of the kind of ability that human artfulness is, linking to contemporary discussions of where art fits into human evolution (e.g. Davies 2015).