Advances in Role and Reference Grammar

ON DEVIANT CASE-MARKING IN LATIN 367

third person singular inflection. In explaining the impersonal passives
sanctioned by verbs manifesting deviant case-patterns, Jensen states (p.
33):

[DJative object verbs...likewise lack a subcateogorized complement

unmarked for case — the case features of the complement of such verbs

are marked in the lexicon, thus blocking [the passive rule] in these cases.

Passives of such verbs are generated in the lexicon, but they do not assign

a theta-role to their subjects, nor is there a trace to perform this function,

and so they too are necessarily impersonal.

Jensen's analysis, like the present analysis, can account for the fact that
while verbs having accusative objects form personal passives, verbs lacking
them, whether having one or two places, can form only impersonal pas­
sives. Both analyses recognize deviant case-patterns as lexically idiosyncra­
tic, although each locates that markedness in a different conceived compo­
nent of the verb's lexical entry. The present analysis regards "quirky verbs"
as sanctioning an argument position unlinked to a macrorole; Jensen's
analysis regards such verbs as assigning a case feature to their objects.
Hence, both Jensen's analysis and the present analysis provide a plaus­
ible explanation for the relationship between quirky case and impersonal
passivization. And yet Jensen's analysis, concentrating almost exclusively
upon the passivization properties of verbs sanctioning deviant case-pat­
terns, fails to examine the range of deviant case-patterns sanctioned by
these verbs. As it focuses upon dative-object verbs, it does not attempt to
explain the distributions of dative, ablative, and genitive objects — that the
latter two are relatively uncommon; that they appear, unlike dative objects,
to be associated with particular semantic subclasses of verbs; that they most
commonly encode a theme argument; and that in this function they often
appear to vary freely with one another. Further, Jensen's analysis lacks an
account of the case-pattern parallelisms between two-place verbs bearing
quirky case and three-place predicates. Treating the case assigned the
former type of verb as a lexical idiosyncracy, he seems to overlook a gen­
eral pattern of case assignment exemplified by both transitive three-place
verbs and "quirky" two-place verbs. In the present analysis, the indirect-
object and non-accusative-object coding function of the dative are seen as
manifestations of the same case-marking principle (24c), as are the "beweg­
liches object" and non-accusative-object coding functions of the ablative
and genitive (26). Jensen's analysis, however, has no means by which to
account for the evident case-pattern parallelisms obtaining between intrans-