364 LAURA Α. MICHAELIS
object marking", and defines it as the process by which different case-
frames become "available for different semantic function patterns" (p. 9).
As an example of this, he points to the development of a competing case-
pattern for the verb doleo ("I grieve for") with an ablative rather than
accusative object. It is unfortunately not made clear here what distinct "se
mantic functions" are associated with these two case-patterns. The second
source he refers to as "satellite absorption" — a process whereby adjunct
arguments are incorporated into the "predicate frame of a verb," becoming
core arguments in the process. As an example of this process, he suggests
that the quirky verb faveo ("I favor") was originally a one-place predicate
with which a "dative satellite expressing "interest" became associated
habitually, to the extent that it gradually became part of the predicate
frame of the verb, thus yielding a two-place predicate" (p. 10). The selec
tion of non-accusative argument by such a predicate would thus be syn-
chronically unjustifiable in semantic terms. Although he does mention (p.
11) that there exist "certain intriguing verbs or groups of verbs" which may
"suggest a semantic justification for the use of specific cases," he neither
identifies these verb groups nor advances any semantic explanations for the
use of non-accusative objects. (One may presume he has in mind such clas
ses as verbs coding need and lack.)
This diachronic account of the development of non-accusative objects
is not antithetical to the present analysis, which, although not encompassing
the diachronic developments producing quirky case-patterns, incorporates
Pinkster's claim that these case-patterns are not amenable to a synchronic
semantic explanation and must thus be regarded as idiomatic. The present
analysis differs from Pinkster's in that it locates the idiomaticity of verbs
licensing quirky case-patterns primarily in their lack of transitivity rather
than in the case patterns themselves. This difference has, as will be shown,
important ramifications for the treatment of the quirky case/impersonal
passive correlation.
Although Pinkster's suggestions of diachronic sources for quirky case-
patterns are intriguing, his analysis fails to account for the facts of their syn
chronic distribution or for the behavior of predicates licensing such case
patterns. In fact, even the diachronic analysis itself seems somewhat
inadequate — although the diachronic process of benefactive satellite
absorption can apparently explain why fa veo sanctions a dative non-subject
argument, it does not appear that this explanation suffices for all or even
most verbs bearing dative objects. It does not seem plausible to suppose,