The Lexical Semantics of Parenthetical-as and Appositive- Which

Despite their superficial similarities, nonrestrictive relatives and asparentheticals show contrasting behavior in a range of apparently disparate domains, including (i) equative constructions (section 4); (ii) selective island contexts (section 5); and (iii) clause-internal ‘‘niching’’ (section 7). Additionally, as-parentheticals allow a wider range of interpretations relative to their antecedents than do appositive relatives (section 6). This paper offers a unified account of these differences and others based largely on the respective semantic types of the gaps these clauses define: as-clause traces are propositional; nonrestrictive relative traces are individual denoting (i.e., nominalized propositions). The type distinction follows from the lexical denotations of these morphemes (section 2) and combines with independently motivated principles to predict the clauses’ divergent behavior. The analysis also unifies the various kinds of appositive-relatives and similarly informs our understanding of predicate-type as-clauses (e.g., Sue hates parties, as does Ali); see section 8. 1. Remarks In his classic 1984 paper ‘‘Inner islands,’’ Ross proposes that certain puzzling contrasts between the appositives italicized in (1) are due to the ‘‘adverbiality’’ of as. (1) a. Americans should get cheap oil, as the whole world knows t. b. Americans should get cheap oil, which the whole world knows t. I call appositives of the (1a) sort as-parentheticals or as-clauses; (1b) involves a (clausal) nonrestrictive relative (NRR). Ross’s characterization of as as ‘‘adverbial’’ allows him to link extraction failures like those in (2), since they can be attributed to the general markedness of extracting an adverb across a syntactic island boundary; compare (3) in which the extractees lack ‘‘adverbiality.’’ Syntax 5:1, April 2002, 55–88 ß Blackwell Publishers Ltd, 2002. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA * My thanks to Daniel Büring, John Frampton, Chris Kennedy, Vera Lee-Schoenfeld, Jim McCloskey, Bill Ladusaw, Jason Merchant, Line Mikkelsen, Paul Postal, and Geoff Pullum for their invaluable advice and assistance. I thank also the Syntax reviewers for their numerous challenging questions and helpful comments; this paper was substantially improved by their insights. Any remaining mistakes are my responsibility. 1 See also Ross 1967/1984 (sect. 6.1.1.4) and Ross 1973 (p. 151, n. 21), the latter a squiblength footnote on as-clauses. 2 Ross motivates his proposals using data involving extraction across negation. As I show in section 5.2, the situation regarding such extraction is more complicated empirically than it is for selective islands like the adjunct in (2) and (3). (2) a. *Aldrich stole the files, as Op1 the investigator hesitated before announcing t1. b. *[How confidently]1 did the investigator hesitate before announcing t1 that Aldrich stole the files? (3) a. ?Aldrich stole the files, which Op1 the investigator hesitated before announcing t1. b. ?What1 did the investigator hesitate before announcing t1? Although the ungrammaticality of (2a,b) receives a uniform treatment here, I do not follow Ross in attributing them to ‘‘adverbiality.’’ ‘‘Nonindividualhood,’’ though lacking in Rossian zip, is more accurate. This paper’s basic claim is that all NRRs involve extraction of an individualdenoting phrase (type hei), even those that appear to pronominalize something propositional, as in (1b) or something predicative, as in (4a). (4) a. Ali could see the jackalope, which Joan also (said that she) could t. b. Ali could see the jackalope, as Joan also (said that she) could t. In contrast, (1a) contains a proposition-type trace, whereas (4b) contains a property-type trace (type hs,he,tii). The status of these various variables follows from the respective morphemes’ lexical denotations, which are identical up to variable type (see section 2). The minimal difference combines with independently motivated principles to predict not only extraction contrasts but a host of other semantic and syntactic points of divergence. At the same time, the similarity of the lexical entries jibes with the many distributional parallels of these constructions. A theme of this paper, and the larger research program of which it is a part, is that wide ranging and apparently unconnected restrictions (and freedoms) on structures can be made to follow from the lexical denotations of particular morphemes. Importantly, these denotations are not specific to English. Ross’s observations about the appositives in (1) extend not only to the clause types in (4) (see section 8) but also to German pairs like (5), which diverge as the English cases do. (5) a. Amerikaner sollten minderwertiges Benzin kriegen, was die Americans should low-quality oil get which the ganze Welt t weiss. whole world knows ‘Americans should get cheap oil, which the whole world knows.’ b. Amerikaner sollten minderwertiges Benzin kriegen, wie die Americans should low-quality oil get as the ganze Welt t weiss. whole world knows ‘Americans should get cheap oil, as the whole world knows.’ ß Blackwell Publishers Ltd, 2002 56 Christopher Potts