ATTRIBUTIVES AND IDENTIFICATIONALS 461
- The same pattern is cited in Foley & Van Valin (1984) for Lakhota.
- Burzio (1986) claims that no adjectives allow ne-cliticization; however, my native
Italian consultant felt that rce-cliticization was marginal with adjectives if the con
text was right but much worse with predicate nominals.
- Jackendoff (1983) assumes that such a schematization is possible, though he con
siders this schematization marginal for English. He observes that it would be con
sistent with his localistic account of thematic structure for this to be a major pat
tern in another language.
- This is apparently true of perception predicates and cognitive predicates in the
languages considered here, but no claims are made here about subject selection
cross-linguistically.
- In English, there is paraphrase evidence for the existence of both analyses:
(i) a. The crowd is quiet.
b. Sam is completely happy,
(ii) a. The crowd is in a state of quietness.
b. Sam is in a state of complete happiness,
(iii) a. Quietness has descended on the crowd.
b. Sam is completely filled with happiness.
In the examples in (ii), the argument corresponding to the predicate of the attribu
tive constructions in (i) is overtly marked as a locative, supporting a theme-subject
schema, while in (iii), the argument corresponding to the subject of the attributive
constructions in (i) is overtly marked as a locative, supporting a location-subject
schema.
- Russian generally does not use the copula in present tense in any of these con
structions.
- RRG doesn't use the notion "subject" but rather distinguishes a syntactic and/or
semantic pivot around which a construction is built. I will continue to use the term
"subject" for the languages discussed here, though a more definitive analysis
should tease this notion apart into various aspects.
- This was suggested to me by Robert Van Valin and is also given in Van Valin
(1990).
- If the locative-subject analysis is assumed, then the rule would be one of theme
incorporation.
- This is not to say that ne-cliticization makes reference to undergoer. It cannot, due
to the presence of motion accomplishment predicates among the constructions
which allow it. Rather, it refers to theme/patient, which is a prototypical under
goer but not prototypical actor.
- I assume that change-of-state is a kind of motion event.
