Modifiers Pronominal Introduction: The semantic classications proposed for relative clauses and nominal modiers have typically been formulated by looking at cases of the form (d)-(adj+)-n-{cp/adjp} (e.g. the Italian students who are linguists). The goal of this talk it to draw the attention of the linguistic community to the behavior of two nominal modier constructions which apply to pronouns, rather than common nouns: Pron-N (1), and Pron-RC ((2) for Italian cases). I will argue that these modiers do not t in any of the previously proposed classes (appositive viz restrictive, sortal-internal viz sortal-external, Partee 1973, Grosu and Landman 1998), and that they are best understood as presuppositional operators. Data and Problems: while the Pron-N construction has interesting and well-known gaps (3), common to English and Italian(/Romance), the latter lls the missing cases with the semantically similar Pron-RC construction, illustrated in (4) and (5). At least three questions arise: (a) what is the semantic role of these modiers; (b) what is the semantic dierence, if any, between Pron-N and Pron-RC; (c) why there are gaps in the Pron-N, but not in the corresponding Pron-RC. Semantics: Since neither you linguists are clever nor (5)b can be used to address a group which is only partially constituted by linguists, it is natural to assume that the post-pronominal modiers cannot be restrictive. According to the traditional classication, they must thus be appositive. However, Pron-N dier from canonical appositive constructions in several respects: (a) the pronoun and the modier belong to the same phonological phrase (there is no pause); (b) Neither Pron-RC or Pron-N can modify (conjoined) proper names (6a); Pron-N cannot even modify pronoun conjunctions like those in (6b), unless a pause (which I take to indicates genuine nominal apposition) is inserted; (c) Pron-RC is restricted to subject relatives ((7) vs. (2)b); (d) parasitic gaps inside Pron-RC are possible; (e) oating quantiers cannot appear in Pron-N (in the same phonological phrase with the pronoun, cf. (9)c. vs. a,b). Analysis: I have two main claims. First, neither Pron-N not Pron-RC are restrictive, but they dier semantically from regular appositives in that the content of the pronominal modier is presupposed. The purpose of the modiers is to indicate that a property of the referent of the pronoun which is part of the Common Ground (modulo accommodation, see Stalnaker 1974) is causally relevant for the communicative goals of the discourse. A crucial test case is (10), which is incoherent, since being both linguists cannot (?) be the explanation for a diversity (contrast with the men who are linguists are dierent from the women who are linguists). Note that if the RC was either restrictive or non-restrictive the sentence would have perfectly sensible meanings, paraphrased in (11a) and (11b) respectively. Pronominal modiers are thus very close to the qua-construction (Italian in quanto): (12) is analogous to (10), and similarly incoherent. The second claim is that in Pron-N the property which is presupposed and proposed as causally signicant is the fact that the pronoun is an instance of a kind; the nominal must thus be able to denote a kind. This derives (3)a, under the assumption that singular count nouns do not denote kinds; in turn, I take the plurality of the pronoun to be a form DP-internal of agreement: it does not entail that the speaker/hearer is part of a real plurality. Thus (1) can be said by a lone speaker to a single hearer (both acting as representatives), while this is not true of a genuine apposition such as we, who are linguists, are dierent from you guys, who are engineers: who we are you guys are must be clear from the context. The notion of internal agreement also accounts for *I and you linguists (a syntactic number clash analogous to *one and one people as an attempt to say two people). The ban on proper names (and possibly loro they) might come from their origin as common nouns, in a position which clashes with the modier. Finally, some of the peculiarities of Pron-RC can be accounted for in terms of the presence of a pro subject (cf. the pronoun visible in the German Pron-RC Wir, die wir Linguisten sind ).
(1) (2)
[We linguists] are dierent from [you engineers] a. b. [Noi che siamo linguisti] dovremmo capirlo. [we that are linguists] should get it [Voi che incontrate sempre Claudia], ditemi di lei. [you that meet always Claudia], tell me about her
Pron-N construction Pron-RC constructions
(3)
a. *{I / you / he} linguist b. {we / youplur / *they} linguists a. b. c.
distinct from the appositional I, Claudius
(4)
Io che sono un linguista (dovrei capirlo) I that am a linguist (should get it) Tu che sei un linguista you that are a linguist Lui che ` un linguista e he that is a linguist {Jack e Jim / voi e Jack {Jack and Jim / youplur and Jack {Io e te / io e voi / {I and yousing / I and youplur /
(5)
a. b. c.
Noi che siamo linguisti we that are linguists Voi che siete linguisti you that are linguists Loro che sono linguisti they that are linguists
(6)
a. b.
/ Jack e io} *(,) (che siete/sono) ingegneri / Jim and I} (,) (that are2pl/3pl ) engineers noi e te } *(,) linguisti we and yousing } *(,) engineers sempre. always
(7) (8) (9)
*{Io / tu / lui / noi / voi / loro} che Carla incontra *{I / you / he / we / youplur / they} that Carla meets
Una decisione...