342 LAURA Α. MICHAELIS
verbs and verb classes licensing deviant case-patterns also stray rather far
from the transitive canon. No one of these predicate classes reflects the
sense of dynamic interaction critical to the transitive prototype.
There are, however, a number of problems for the semantically based
model of the transitive case-pattern. First, it does not appear that the
nominative-accusative pattern necessarily reflects close adherence to the
transitive prototype. For example, as observed by Pinkster (to appear:4),
emotion verbs like doleo ("I grieve for") and gaudeo ("I rejoice at"),
although they do not form personal passives, regularly govern accusative
arguments. Such verbs are no closer to the transitive prototype than are, for
example, the inverse verbs of (3); one would not, therefore, expect them to
sanction accusative objects. Secondly, as shown in (6), several verbs taking
dative objects, notably noceo ("I harm") and such "verbs of opposition" as
obsto ("I oppose"), repugno ("I fight against") and resisto ("I resist")
appear to adhere very closely to the transitive prototype (at least in their lit
eral uses), but fail to license the predicted nominative-accusative pattern.
To this second counterargument, one might give the following objec
tion: insofar as surface case-patterns can be taken as evidence of "underly
ing transitivity", some cross-linguistic data suggest that verbs in the class of
noceo and resisto (let us call this class "verbs of harming and opposing") do
not constitute good examples of the transitive prototype. In German, for
example, the verbs widerstreben ("to oppose") and schaden ("to hurt")
sanction dative objects, as do a number of Greek verbs coding opposition,
e.g., άντιτάσσμαι ("I oppose in battle") and άντιτίθημι ("I set against"). It
appears, however, that certain facts of Latin impugn the claim that verbs of
this class have a special affinity for dative objects: as shown in (7), many of
the predicates which license dative objects have syntactically transitive
synonyms. Thus, the "quirky" noceo has the "normal" synonym laedo and
the "quirky" resisto the "normal" synonym impedio. If syntactic transitivity
(i.e., the nominative-accusative case pattern) reflects semantic transitivity,
apparently synoymous lexical items should not license both transitive and
intransitive case-patterns, nor, for that matter, should those verbs which
adhere closely to the transitive prototype fail to sanction the transitive case-
pattern.
A third argument against the view that the allocation of case patterns
in Latin is semantically governed (and in favor of the view that such alloca
tion is idiosyncratic) involves apparent regularization of quirky case in
Latin. As Pinkster (to appear:2) observes, "Apart from the "deviant" non-