ON DEVIANT CASE-MARKING IN LATIN^341To the view that intransitivity is an idiosyncratic syntactic feature of
some two-place predicates in Latin, one can object, as has Rice (1987), that
"transitivity is something above and beyond the lexical or logical definition:
a verb taking a direct object or one sustaining two arguments" (p. 422), and
that in fact syntactic transitivity is merely the reflex of a semantic structure
to which verbs licensing the nominative-accusative pattern adhere closely
and to which "quirky" two-place predicates do not. Rice claims that transi
tive verbs instantiate, to a greater or lesser degree, a non-linguistic transi
tive prototype, which she characterizes thus: "two entities, which are usu
ally conceived of as being asymmetrically related, are involved in some
activity; the interaction between them is unidirectional; because there is
movement and effect, contact between the two entities is presumed to take
place, with the second entity being directly affected by the contact insti
gated by the first; finally, the two entities are taken to be distinct from each
another..." (p. 423).
Hence, such force-dynamic predicates as caedo, "I strike down" (la),
whose semantic structures closely fit the prototype, receive nominative-
accusative case patterns. The extent to which a given predicate must
deviate from this model before it receives "quirky case" is, of course, a
complex problem; the important issue from our point of view is, however,
the following: do the "quirky" two-place predicates in Latin represent sig
nificant deviations from the transitive prototype adhered to by the "nor
mal" verbs? It appears that the majority of "quirky" verbs certainly do. The
manner in which many of the "quirky" Latin verbs deviate from the transi
tive prototype is rather apparent. Most of these verbs, including the inverse
verbs (3) and the verbs of recollection (8), denote cognitive activities of var
ious kinds. Such activities clearly depart from Rice's transitive-event
scenario: the percept and cognizer arguments are not easily viewed as dis
tinct from one another, their interaction does not involve movement (di
rected or otherwise), and neither entity is apparently contacted or directly
affected by the other. The verbs of need and lack shown in (11) and (13)
also deviate from the transitive prototype: the nominatively coded argu
ment neither has contact with the item lacked nor affects it in any way. The
"quirky" verbs of abundance shown in (9) and (12) differ from the verbs of
need and lack in that they presumably require contact between locative and
theme arguments; both of these verb classes, however, denote the type of
static, configurational relationship between entities that, according to Rice,
is correlated with low transitivity. Hence, it is apparent that many of the