ON DEVIANT CASE-MARKING IN LATIN 359(30) Hic ager colono est paenitendus.
this(M) field(M) colonist(D) is shaming(N)
"This field ought to shame the colonist." Col. 3,2In (30), the field (ager) should not be ashamed, as one might predict from
the passive-periphrastic form (cf. 5), but should instead be a source of
shame. Hence, the participle paenitendus, in concert with the copula,
means not "ought to be ashamed", but "ought to shame." The passive of
such verbs is simply the deontic version of the active form. The fact that the
passive of such verbs does not promote the cognizer argument to subject
might seem to impugn the claim that the accusative argument coding this
thematic role is here linked to the undergoer macrorole. One might also
argue, however, that such verbs have no genuine passive forms merely
because they are impersonal. The presence of an accusative object might be
a necessary but not sufficient precondition upon personal passivization; a
finite active is perhaps another prerequisite for the personal passive.
Because we cannot determine here whether the lack of the personal passive
reflects the lack of the requisite macrorole or the lack of a finite form, we
must for the time being remain agnostic with respect to this issue.
Returning to the genitive argument sanctioned by this predicate class,
it appears that there is an additional issue that must be investigated with
respect to its SmP status. As discussed above,where a language allows both
SmPs and PrPs, a particular construction may be tied to one or both pivot
types. Are both pivot types critical to the "raising to object" construction,
or only one? The typical instance of this construction, exemplified in sen
tences (23b,d,f), involves the "raising to object" of an argument that is both
PrP and SmP, i.e., that both bears a macrorole and outranks any other
argument with respect to the actor end of the A/U hierarchy. But, as we
have seen in (23a,c,e), SmPs which are not PrPs can also be "raised to
object", with preservation of their deviant case. These facts would compel
one to conclude that it is the SmP, rather than the PrP which is crucial to
this construction. This conclusion would, of course, have to be modified
were it discovered that PrPs lacking SmP status could also be raised to
object. A critical test might then involve the subjects of inverse verbs,
which, as mentioned in fn. 5, have PrP status but, as themes, are outranked
for SmP status by the locative argument receiving dative case-marking. If
such subjects could serve as subjects of accusative-infinitive object comple
ments, then this fact would indicate that this construction is sensitive to