Labiovelars and the labial-velar hypothesis: Phonological Headedness in Bare Element Geometry

Abstract

This paper begins by presenting and strengthening the labial-velar hypothesis with a novel interpretation of the allophonic distributions of labials and velars in Lusoga (Hyman 2017). The labial-velar hypothesis is an Element Theory proposal that labial (Lab) and dorsal (Dors) consonants are marked by same ‘dark’ place feature |U| (akin to Jakobson & Halle’s (1956) grave). According to this hypothesis, the only difference between Lab and Dors members of any class (say stops) is that Lab is the headed manifestation of |U|, while Dors is the headless version of the same element (Backley & Nasukawa 2009). This, hypothesis however, leads to the strange statement that labial-velar segments such as the labial-velar ‘double’-stop (KP), which would appear to be headless and headed at the same time. What follows is an exploration of the term ‘headedness’ in phonology. Building on the positives of previous approaches, and pairing back the assumptions to the first principles of combination (Merge), I come up with a novel theory of headedness and segment-internal structure: Bare Element Geometry (BEG). The labial-velar hypothesis in light of BEG will then be applied to KP. In order to obtain the right representation for labial-velar ‘double’-stops (KP), it is necessary to precisely understand KP’s distribution and behaviour. Especially because I will reveal some previously unappreciated facts about these segments. In opposition to the classical literature (and all previous accounts known to me), I will conclude that KP is composed of two ordered root nodes, each with a separate place feature. This is based on a detailed typological study and phonological interpretation of attested and unattested nasal + KP assimilation patterns, as well as a novel interpretation of some obscure nasalization facts of KP in Guere (Western Kru). This conclusion that KP has two ordered root nodes is what allows BEG’s model of headedness to interpret a feature as simultaneously headed and headless in different parts of a segment. Moreover, it correctly predicts Cahill (1999)’s empirical observation that KP is universally headed by Labial.

Elements, Government, and Licensing

“…As pointed out by Ulfsbjorninn (2021) and argued in this chapter, templatism doesn't appear to align well with the research agenda of Government Phonology and ET, which among other things have always sought to explore cognitive homology between (morpho-)syntax and phonology and to reduce where possible that which is supposed to be 'special' about phonology vis-à-vis related cognitive faculties. It has been proposed that the structure building process underlying both syntax and melodic phonology may be essentially the same, namely simple symmetric set-merge.…”

“…Example (8) illustrates some correspondences between MRs in set-and conventional ET-notation. Much like Breit (2013) and Ulfsbjorninn (2021), this suggests that the basic mechanism of structure building underlying melody (vis-à-vis phonology) and syntax may show deep homology. The most striking and fundamental difference is that melodic structure building (at least under standard assumptions) is limited to second merge:…”

“…I propose to revise Breit's (2013) account in two important ways. First, I follow Ulfsbjorninn (2021) in assuming only set-merge as the underlying structure building operation (in line with other work proposing to eliminate pair-merge from syntax, cf., for example, Oseki, 2015). Second, I propose that relations in ET uniformly map MRs to MRs.…”

“…Kaye (2000Kaye ( , 2001, following , calls these the operators (the term dependents is also used) and means all the elements in a representation that are not the head. Breit (2013Breit ( , 2017 and Ulfsbjorninn (2021) argue for a view of a (structural) complement which explicitly includes the embedded copy of the head. I will side with the latter view.…”

Metrical exceptionality and stress shift in Romanian nouns and adjectives

“…The root node, a segmental level organisation of Elements that attaches to a C or V is making a resurgence in Strict CV representations (Ulfsbjorninn 2021a(Ulfsbjorninn , 2021bScheer to appear).…”