PREPOSITION ASSIGNMENT IN ENGLISH 305
PURP, as identified in section 2.7.1, remain the same. In this context, as in
the benefactive interpretation, for introduces new semantic content. How
ever, the argument shared with the main clause is the entire verbal situation
described — [do' (x)] CAUSE [BECOME baked' (y)] — rather than a
single argument. Deputative for thus functions as an adjunct to the main
clause. Benefactive for with activity verbs, as in (66a) and (72), is likewise
an adjunct, but it does not have deputative semantics. Rather, it has a
benefactive meaning but not one of transfer as in (71a), since there is no
entity to be transferred. In these cases, the semantic content of LS 2
specifies some type of beneficial situation for some participant, the exact
nature of which will depend on the type of activity in particular and the
nature of the situation in general. The contrasting interpretations of sen
tence (73a) are based on the fact that the benefactive interpretation has to
do with the results of the verbal situation (who gets the cake), while the
deputative reading is concerned with the causing activity (who makes the
cake). The LSs for the two readings, (73b) and (75), explain the fact that
only (73), the benefactive relation, can be realized in a sentence such as
(76), where the benefactive recipient (locative), Rita, is positionally
defined.
(76) John baked Rita a cake.
The positional shift of patient {cake) and ultimate goal (Rita) is not possible
in an LS such as that specified in (75). That is, with the benefactive reading,
PURP and LS 2 are in the scope of CAUSE (as in (73b)), while in the
deputative interpretation, the entire LSi is in the scope of PURP. The reci
pient is a locative, a potential undergoer, while the deputative beneficiary
in (75) is an effector, not normally a potential undergoer argument in a sim
ple clause. Thus, it is only in the benefactive case that an argument from
the PURP clause can be mapped into the Undergoer slot.
2.7.3 Purposive for: non-predicative functions
The class one prepositional functions of for are illustrated by contexts
involving a sub-class of stative verbs: the cognition statives discussed at
length by Gawron (1986) describing structures of wanting or trying to get
(e.g., hope, long, yearn, etc.). In Gawron's analysis, for — with desire/cog
nition statives — is an argument PP inasmuch as it is semantically redun
dant with the verb. Our analysis labels these prepositional functions, as in
(77), non-predicative.