12h00
en ligne
The foot strikes back: a new defense of headedness and constituency in phonology
Björn Köhnlein (The Ohio State University)
While segmental rules can sometimes be phonetically unnatural, metrical structure appears to be generally well-behaved phonologically, such that a rather small set of assumptions suffices to account for many patterns regarding, e.g., syllabification and stress. One core theoretical question is whether prosodic phenomena are best modelled by including a prosodic hierarchy into phonological representations, or whether relevant patterns can be captured without reference to constituents, as argued for in, e.g., Strict CV. For instance, one recent approach to word stress within Strict CV employs the notion of incorporation, whereby stressed nuclei can incorporate grid marks of adjacent empty nuclei (e.g., work by Ulfsbjorninn and Faust); it has by now been shown repeatedly that this approach can successfully model various intricate aspects of stress systems.
Somewhat akin to van Oostendorp’s (2013) arguments in defense of the syllable, I claim that foot structure remains a useful tool for understanding prosodic patterns, its role crucially extending beyond word stress. Here, I mainly focus on certain word-accentual systems where, as I claim, getting a better sense of the role of headedness and size of foot domains (roughly bimoraic vs. disyllabic) helps capture a variety of prosodic and/or segmental oppositions. I demonstrate in what ways metrical approaches to accent can be considered preferable over existing tonal and grid-based analyses, including a discussion of why I consider it non-trivial to translate emerging generalizations into models of phonology where relevant accentual patterns are not derived from constituency. If, as argued here, the metrical approach is indeed on the right track, relevant accentual systems should be captured in any theory of metrical phonology, including Incorporation Theory (a term that I just made up, I think).
