09.07.2015 Views

Practical Information - Generative Linguistics in the Old World

Practical Information - Generative Linguistics in the Old World

Practical Information - Generative Linguistics in the Old World

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Unbounded Successive-Cyclic Rightward MovementJason Overfelt, University of Massachusetts Amherst1. Introduction. The existence of rightward movement faces at least two serious challenges.First, Ross (1967) argued that it is subject to stricter locality conditions than, for <strong>in</strong>stance, whmovement.Second, Akmajian (1975) argued that rightward movement is not obviously successivecyclic<strong>in</strong> <strong>the</strong> way that wh-movement is often taken to be (cf. Chomsky 1977). This paper arguesthat rightward DP-movement actually displays both of <strong>the</strong>se properties given appropriate licens<strong>in</strong>gconditions. In particular, rightward movement can apply successive-cyclically <strong>in</strong> a potentiallyunbounded fashion when licensed by a parasitic gap as <strong>in</strong> (1), adapted from Engdahl (1983).(1) John offended e by not recogniz<strong>in</strong>g pg immediately – my favorite uncle from Cleveland.2. The Nature of <strong>the</strong> Displacement Operation. Based on <strong>the</strong> presence of derived island effects(Wexler & Culicover 1980) and non-sensitivity to <strong>the</strong> Right Edge Restriction Wilder (Wilder1999), <strong>the</strong> paper argues that structures like (1) are not derived via Right Node Rais<strong>in</strong>g (cf. Postal1994, a.o.), but by rightward DP-movement. Start<strong>in</strong>g from <strong>the</strong> observation by (Larson 1989) thata parasitic gap is obligatory when a DP is displaced rightward over an adjunct clause (2), <strong>the</strong>movement operation is suggested to be licensed by <strong>the</strong> parasitic gap.(2) Sam stole e because Kim wouldn’t buy [pg/*anyth<strong>in</strong>g] for him – an autographed pictureof Jonathan Frakes.Diagnostics <strong>in</strong>clud<strong>in</strong>g VP-ellipsis, VP-front<strong>in</strong>g, and a form of antecedent-conta<strong>in</strong>ed deletionsuggest that only adjunct clauses adjo<strong>in</strong>ed above <strong>the</strong> locus of typical focus-driven Heavy-NP Shiftat <strong>the</strong> edge of vP require <strong>the</strong> parasitic gap. Thus, <strong>the</strong> parasitic gap <strong>in</strong> (1) and (2) is licens<strong>in</strong>gadditional movement beyond Heavy-NP Shift <strong>in</strong> violation of <strong>the</strong> Right Roof Constra<strong>in</strong>t (3).(3) RIGHT ROOF CONSTRAINT (adapted from McCloskey 1999)Rightward movement may move a DP to <strong>the</strong> right edge of <strong>the</strong> vP that most immediatelyconta<strong>in</strong>s X, but no fur<strong>the</strong>r.3. Type-Driven Rightward Movement. The paper proposes a <strong>the</strong>ory for <strong>the</strong> derivation ofrightward DP-movement and parasitic gaps that achieves <strong>the</strong> representation for parasitic gap licens<strong>in</strong>gproposed <strong>in</strong> Nissenbaum (2000) but allows a parasitic gap to license movement beyondvP accord<strong>in</strong>g to local economy considerations. The parasitic gap doma<strong>in</strong> is a null-operator structure,and thus a 〈et〉 predicate, which is merged cyclically to <strong>the</strong> type t matrix clause. Rightwardmovement is allowed because it converts <strong>the</strong> matrix clause to a derived predicate, which allowsit to compose via predicate conjunction with <strong>the</strong> parasitic gap doma<strong>in</strong>. This repair strategy ismade possible by a logical extension of <strong>the</strong> operation Merge based on <strong>the</strong> ideas that Merge canbe counter-cyclic (Lebeaux 1988, a.o.) and should be decomposed <strong>in</strong>to a number of smaller operations(Hornste<strong>in</strong> 2009). I propose a sub-type of Merge called Mixed Merge, which has beendecomposed <strong>in</strong>to <strong>the</strong> steps <strong>in</strong> (4).(4) XP : ?? =⇒ XP : ?? =⇒ XP : tvP : tAdjunctP : 〈et〉XP : ??DPXP : 〈et〉DPv 0VP Op... pg...vP : t AdjunctP : 〈et〉vP : 〈et〉AdjunctP : 〈et〉V 0DPv 0V 0VP Op... pg...ev 01VPV 0 e 1Op... pg...1

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!