Empty nouns in Bantu locatives
Vicki Carstens
January 1997

In this 1997 article I argued that locative phrases in many Bantu languages are headed by silent PLACE nouns meaning roughly INSIDE, SURFACE, and VICINITY. The approach was based on several factors: (i) locative phrases can appear in argument positions and control verbal agreement, (ii) they can contain expressions bearing locative concord and modifying PLACE noun semantic content, but nonetheless (iii) the overt locative markers (commonly ku, pa, and mu) fail diagnostics for nominal status. I argued that ku, pa, and mu are gender-particular Case-markers akin to ‘of’ and other items commonly referred to as linkers, thus answering a question recently raised in Caha & Pantcheva 2014 as to why the hypothetical silent PLACE nouns in such analyses do not trigger the presence of linkers. My analysis built on a proposal in my 1991 dissertation that so-called secondary prefixes in Bantu are noun class morphology for other silent nominal elements. For example, the Chichewa ka-mu-nda/ti-mi-nda – ‘little field/s’ bear two prefixes because ka- and ti- are number prefixes for a silent diminutive noun of a particular gender (see also Taraldsen 2013) while the inner prefixes mark the number of the diminutivized nouns. But locative class “prefixes” differ from other secondary prefixes in that, at least in some Bantu languages, they exhibit behavior characteristic of independent words (see Bresnan & Mchombo 1995). This difference is reflected in the account of the locative morphemes that I provide. Here is the link to the published version in The Linguistic Review: http://www.degruyter.com/view/j/tlir.1997.14.issue-4/tlir.1997.14.4.361/tlir.1997.14.4.361.xml?format=INT
Format: [ pdf ]
Reference: lingbuzz/002224
(please use that when you cite this article)
Published in: The Linguistic Review 14:361-410
keywords: locatives, place, noun class, bantu, chichewa, morphology, syntax
Downloaded:1904 times


[ edit this article | back to article list ]