Advances in Role and Reference Grammar

448 LINDA SCHWARTZ

3.3 NE-cliticization and attributives and identificationals

As noted above, attributive and identificational predicates pattern subjec­

tively with respect to ne-cliticization, as shown in (18)-(19), repeated here

as (34)-(35).

(34) a. Molti ragazzi sono buoni.

many children are good

"Many children are good."

b. *?Ne sono buoni molti.

of-them are good many

"Many of them are good."

(35) a. Molti uomini sono professori.

many men are professors

"Many men are professors."

b. *Ne sono professori molti

of-them are professors many

"Many of them are professors."

Centineo's account of constructions which allow ne-cliticization is not too

clear, but it appears that she characterizes them as having an affected pivot

— i.e. a pivot which is the argument of a state predicate. However, this

characterization would incorrectly place attributives and identificationals in

the class which allows ne. I argue in this paper that the subjects of these

constructions either (i) strongly, are not themes, or (ii) weakly, are not typ­

ical themes. Under assumption (i) the statement in (36) will correctly

characterize the class of arguments which can take ne-cliticization. Under

assumption (ii), the statement in (37) will do likewise.

(36) Condition for n-cliticization (assumption (i))

Ne-cliticization may occur only if the quantified NP is theme/

patient of a stative predicate.

(37) Condition for ne-cliticization (assumption (ii))

Ne-cliticization may occur only if the quantified NP is a typical

theme/patient of a stative predicate.

Under assumption (ii) it is claimed that the appropriate charcterization of

ne-cliticization refers to a strict characterization of theme for which the sub­

jects of attributive and identificational predicates do not qualify due to their

ambiguous theme/location status.