Issues in the Phonology-S yntax Interface in African Languages

Prosodic phenomena in African languages, and most notably Bantu languages, have played – and continue to play – an important role in the development of theories of the phonology-syntax interface. (See, e.g., Cheng & Downing (2009, 2012), Dobashi (2004, 2009, 2010); Hyman & Byarushengo (1984), Kandybowicz (2009), Kanerva (1990), Odden (1987, 1990, 1995), Pak (2008), Seidl (2001), Selkirk (1986, 2000, 2011), Truckenbrodt (1995, 1999, 2007) and several papers in Inkelas & Zec (1990).) A central question for these theories is, as Chen (1990) so neatly puts it, “What must phonology know about syntax?” One finds two leading approaches, which provide two very different answers to this question. Indirect reference theories (e.g., Nespor & Vogel 1986, Selkirk 1986, Kanerva 1990, Truckenbrodt 1995) propose that phonology is not directly conditioned by syntactic information. Rather, the interface is mediated by phrasal prosodic constituents, which need not match any syntactic constituent. Only very limited syntactic parameters can be referred to in defining the prosodic constituents. Direct reference theories (e.g., Kaisse 1985; Odden 1995; Pak 2008; Seidl 2001), in contrast, argue that phrasal prosodic constituents are superfluous, as phonology can – indeed, must – refer directly to syntactic structure. In this paper, I first present a brief survey of phrasal processes (mostly tonal) in selected Bantu languages. Then I briefly review how the indirect and direct reference approach account for these syntactic contexts. As we shall see, a challenge for both of these approaches is that reference to nonsyntactic factors is necessary to account for phrasing in some languages. A final issue that I will address is the growing importance of experimental methodology in phonology and its implications for work on the phonology-syntax interface.