430 MARY L. NUNES
(1990) for a discussion of inverse-form distribution in the Georgian perfect series
(1990:243-247).- As future discussion will demonstrate, the vNP treats affectedness — as defined
by independently-motivated principles of RRG theory — even more topically
than animacy in the case of a clearly-delimited subset of English nominals which,
in the absence of post-vN arguments, retain the result-state process reading of
their V sources (cf. discussion following Table 2). Deane's failure to include some
notion of affectedness in his analysis of topicality undoubtedly reflects the fact
that his concern was with the topic-comment relation of possessive NPs and Ns,
not with the topic-comment relation of argument NPs and verbally-derived Ns. - The EXPs of the ACM experiential state vNs are actually "doubly topical" in that
they are by definition animate and, in terms of the definition of affectedness which
will be presented in later discussion, they are affected. - Although agents, as the volitional controllers of an action, are by definition ani
mate, they may be linked to the direct argument position only in constructions
capable of receiving a strict ACT interpretation. In such nominals, only the agent
bears the single macrorole (A) available to ACTs. In vNPs receiving a strict ACT
interpretation, then, the vN does not fall into the group of nominals defined here
as "capable of taking both macroroles." - The ACT examples included here are all represented in their delimited ACM
interpretation. To emphasize that the argument taken as the delimiter is treated
"as if" it were an U, the notation for U is placed in quotes in these examples: "U." - In the RRG framework, PATs are by definition affected. Consequently, to
include the phrase "is effected" with the term "PAT" is redundant. The redun
dancy is used here to help distinguish Type PATs, which come into a new state
(the city'spat destruction), from Type PATs, which either come into a state of
existence — i.e. are effected — (the light bulb's invention) or come into a reinstan-
tiated state of existence (the poem's translation). - Analysis also has a "concrete" performance-object interpretation: Anderson's
analysis is on Professor Harmon's desk. Whether it is correct to regard the "PAT-
U" — the focus of the completed performance-object activity — as being interpre-
tationally incorporated into the vN is not clear. That is, in this result sense,
analysis could possibly be analyzed either as a "concretized" accumulated action
or as the (analyzed) instantiation of the thing analyzed (cf. Type BACM perfor
mance objects).
In the strictest "thing" sense of "result," the result forms of nominals seem
invariably to incorporate the U into the meaning of the vN. (nominals having
EXP Us are an exception, in that it is the A which is incorporated into the mean
ing of the result vN: John's amusements are many, where amusements and the
things which amuse are equivalents.) Determining precisely how the "concrete"
result forms of ACT performance objects like analysis should be handled will have
to await a more extensive investigation of a broader data base.