Abstract: Nominals in Nguni languages have two prefixes (see (1) and (2)). The outer prefix, called the augment, is absent on NPIs and optional on post-verbal wh-words ((3)a,b). This talk explores the distribution and structure of nominals with and without augments ([+/-A]).
- u-mu-ntu /a-ba-ntu [Zulu]
AUG1-1-person /AUG2-2-person ‘a/the person/(the) people’
- i-si-tsha /i-zi-tsha AUG7-7-bowl /AUG8-8-bowl ‘a/the bowl/(the) bowls’
- A-ngi-bon-i mu-ntu [Zulu]
NEG-1sSA-see-FV 1-person ‘I don’t see anybody.’
- U-bon-e (u)-bani? [Zulu]
2sSA-see-PST (AUG1)-1.who
‘Who did you see?’
I will argue that despite compatibility with both definite and indefinite interpretations in (1) and (2), augments are Ds. Thus I support analyses along the lines of (4)a, and reject (4)b.
- DP (De Dreu 2008, Gambarage 2019 a.o.) b. KP (Halpert 2015; Pietraszko 2020)
D4…NP/fP K2DP
|
AU ENT N
| m!u-ntu
u mu-ntu AUG 1-person 1-person
Restrictions on preverbal subject position in Nguni languages (henceforth PVSP) bear on this analytical choice. Nominals in PVSP must be [-focus] and must bear the augment.
- *[U-John kuphela] u-fik-ile. [Zulu]
AUG1-1John only 1SA-arrive-DISJ.PST * SonlyVO
[Intended: Only John arrived]
- *(U)bani u-fik-ile? *Swh VO
(AUG1)-1.who 1SA-arrive-DISJ.PST
[Intended: Who left?]
- A-ngi-fun-i [ukuthi *(u)mu-nthu a-pheke lutho] ✓ [+Foc] [-A] OB NEG-1SA-want-PRES that *(AUG)-1-person 1SA-cook-SUBJ 11thing
‘I don’t want anybody to cook anything.’
Carstens & Mletshe (2016) argue that the PVSP augment requirement is a subcase of the
[-focus] requirement, because as wh-phrases and NPIs, augmentless nominals are types of [+focus] expressions. But Pietraszko (2020) shows that in Ndebele subjunctive, participial, and relative clauses, [+focus] material is tolerated in PVSP by some speakers, while augmentless nominals are not – a circumstance that Carstens & Mletshe’s analysis cannot capture.
- Subjunctive clause [Ndebele]
- U-fun-a ukuthi [u-bani a-buy-e]? [+A] wh-SU OK
2sSA-want-FV that AUG-1who 1SA-come-SUBJ ‘Who do you want to come?’
- Ngi-funa ukuthi [u-Sipho kuphela a-buy-e] [+A only-SU OK]
1sSA-want-FV that AUG-Sipho only 1SA-come-SUBJ
‘I want only Sipho to come.’
- *U-fun-a ukuthi [bani a-buy-e] *[-A] in same environment
2sSA-want-FV that 1who 1SA-come-SUBJ [Intended: who do you want to come?]
In my talk I will relate these Nguni PSVP facts to prohibitions on bare nominal preverbal subjects in a range of other languages (Contreras 1986, Longobardi 1994, Deprez 2000, Li 1998, Cheng & Sybesma 1999). The cross-linguistic parallels argue that the exclusion of augmentless nominals as preverbal subjects should be addressed in a way that works for unrelated languages, precluding an account based on an idiosyncratic system of structural Case hypothesized just for Nguni (Halpert 2016, Pietraszko 2020). A long tradition has attributed constraints on bare nominal subjects in various languages to problems arising from their having empty or missing D heads. I suggest that EPP-sensitivity or labelling-theoretic issues may explain these problems, and relate them to additional restrictions in Nguni, namely that CPs and PPs are barred from PVSP as well, in all clause types (Halpert 2015, Buell 2011).
Speaker bio: Vicki Carstens completed her BA in African Languages and Linguistics at University of Wisconsin-Madison, and her MA and PhD in Linguistics at UCLA. Much of her work is concerned with agreement phenomena and nominal syntax in Bantu languages. In recent years, she has focused on the Nguni languages of South Africa. Her work has appeared in volumes and journals including Linguistic Inquiry, Natural Language and Linguistic Theory, Language, Lingua, and The Linguistic Review. She teaches courses in linguistic field methods, syntax, and general linguistics.
