Speaker: Mitya Privoznov (MIT) and Neil Banerjee
Title: On Russian doctors and Estonian numbers, or nominal concord as allomorphy
Time: Thursday, April 27th, 5pm – 6:30pm
Location: 32-D831 (changed)
Abstract: Here’s a puzzling paradigm from Russian based on Zaliznyak (1967, 1980) and Pesetsky (2013):
(1)
a. vchera nash-a div-a opjat’ skandali-l-a.
yesterday our-F.SG.NOM diva-SG.NOM again make.fuss.IMP-PST-F
‘Yesterday our diva (woman) was making a fuss again.’
b. vchera nash div-a opjat’ skandali-l.
yesterday our.M.SG.NOM diva-SG.NOM again make.fuss.IMP-PST.M
‘Yesterday our diva (man) was making a fuss again.’
c. vchera nash-a div-a opjat’ skandali-l.
yesterday our-F.SG.NOM diva-SG.NOM again make.fuss.IMP-PST.M
‘Yesterday our diva (man) was making a fuss again.’
d. *vchera nash div-a opjat’ skandali-l-a.
yesterday our.M.SG.NOM diva-SG.NOM again make.fuss.IMP-PST-F
The noun diva ‘diva’ can trigger either feminine (1a) or masculine (1b) agreement on adnominal modifiers and on a finite verb. In addition, there is a possibility of an agreement mismatch: feminine on the adnominal modifiers and masculine on the finite verb (1c). The opposite mismatch is out (1d). This is a mirror image of the vrach ‘doctor’ paradigm from Pesetsky (2013, 39-50). Only in the vrach ‘doctor’ case, the direction of the possible mismatch is the opposite: masculine on the adnominal modifiers and feminine on the finite verb. Interestingly and quite surprisingly, in non-nominative cases the only available agreement for diva ‘diva’ is feminine (while for vrach ‘doctor’ it’s masculine):
(2)
a. ja priexa-l k nash-ej div-e.
I arrive.PFV-PST.M to our-F.SG.DAT diva-SG.DAT
‘I came to visit our diva.’
b. ?/*ja priexa-l k nash-emu div-e.
I arrive.PFV-PST.M to our-M.SG.DAT diva-SG.DAT
In this talk, we will propose an account for this phenomenon, based on a modification of the idea from Pesetsky (2013) and a modification of the feature hierarchy from Bayırlı (2017). We will also extend our analysis to mismatches in Estonian number concord (see, e.g., Norris 2014) and develop a theory of concord as allomorphy of adnominal modifiers, assuming a replacive top-down mechanism of Lexical Insertion (see, e.g. Halle and Marantz 1993).