(99) B/C/S zna
&
m to``ga sta
&
ro
¯
ga c
ˇ
o
`
veka ‘I know that old man ’
Rus my
´
uvı
´
deli vyso
´
kogo stro
´
gogo milicione
´
ra
‘we saw a tall stern-looking policeman’
Pol widze˛ ładnego kota ‘I (can) see a handsome cat’
In the personal pronouns all languages have common forms for accusative and
genitive for 1 Person and 2 Person. For 3 Person, some retain some distinction, either
in relation to animacy (Czech, Slovak), the clitic/non-clitic forms (Polish, Sorbian),
or in the feminine only (B/C/S, Slovenian); the rest make no distinction (see 5.5.2.3).
There are several proposed explanations for this phenomenon. One approach is
phonological: with the rearrangement of the phonological system in relation to
hard and soft consonants, the consequent effect on /i/ and /y/, and the reorganiza-
tion of the declensions as hard/soft variants, a form like PSl NomPl rabi was
perceived as out of place in a hard declension and pressured to become raby, that
is, to merge with the accusative plural and partner the soft NomPl koni. In turn, the
soft equivalent of the accusative plural should be koni, and not kone
ˇ
(5.5.1.1). Slavic
had therefore to face the decision of resisting these pressures so as to maintain the
nominative accusative distinction, giving in to them and losing the distinction, or
finding a morphological solution. This account overlaps with arguments from
morphology, since marked categories are expected to hold or acquire desinences,
while unmarked categories like the nominative do not (Greenberg, 1969; Bethin
1984). Or there are more functional-syntactic arguments which refer to syntactic
ambiguity and the prominence of animate actors and patients (Comrie, 1978).
On the other hand, the (personal) pronouns had no such problems, yet had
clearly gone in this direction already in Proto-Slavic, as evidenced by OCS forms,
where only the enclitic forms retained separate accusatives (see 5.5.3), which means
that the motive and solution were in the first place syntactic, with the nouns making
use of the existing pattern as a solution to their newly developed homophony.
As this was a Proto-Slavic problem, all languages became involved in the genitive–
accusative change, even those which did not subsequently develop the hard soft
phonological or morphological opposition (5.5).
Why should this development not affect the neuters, at least in the nouns?
Comrie (1978) argues plausibly that a major factor was the foregrounding of
animate agents and patients in the structure of the noun phrase: particularly
since so many verbs can take both animate subjects and animate objects, and
since word order was relatively free and driven by pragmatic rather than gramma-
tical needs, it was important to separate out the grammatical roles of (animate)
agents and patients. There was no homophony in the singular forms of the other
major class of animates, the feminines of the a/ja-declension (5.5.1.3).
338 6. Syntactic categories and morphosyntax