Practical Information - Generative Linguistics in the Old World
Practical Information - Generative Linguistics in the Old World
Practical Information - Generative Linguistics in the Old World
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Rajesh Bhatt & Stefan Ke<strong>in</strong>e(University of Massachusetts, Amherst)Verb Clusters and <strong>the</strong> Semantics of Head MovementBackground: Bobaljik & Wurmbrand (2005) argue for an analysis of German long passivesaccord<strong>in</strong>g to which <strong>the</strong> embedded object undergoes rais<strong>in</strong>g to <strong>the</strong> matrix clause to receivecase from matrix T. The crucial piece of evidence <strong>in</strong> support of this view is <strong>the</strong> emergence ofobligatory wide scope of <strong>the</strong> object <strong>in</strong> this configuration. Under B&W’s analysis, <strong>the</strong> embeddedclause <strong>in</strong> (1a) is a vP and <strong>the</strong> object receives accusative case <strong>in</strong>side it. It may hence have lowscope with regard to <strong>the</strong> matrix verb. In <strong>the</strong> long passive <strong>in</strong> (1b), by contrast, <strong>the</strong> embeddedclause is a VP and <strong>the</strong> object must move to <strong>the</strong> matrix clause to get case. B&W correlate thisobligatory case-driven rais<strong>in</strong>g with wide scope.(1) a. weil alle Fenster zu öffnen vergessen wurdes<strong>in</strong>ce all w<strong>in</strong>dows.acc to open forgotten was‘s<strong>in</strong>ce it was forgotten to open all <strong>the</strong> w<strong>in</strong>dows’ [forget ≫ ∀]b. weil alle Fenster zu öffnen vergessen wurdens<strong>in</strong>ce all w<strong>in</strong>dows.nom to open forgotten were‘s<strong>in</strong>ce it was forgotten to open all <strong>the</strong> w<strong>in</strong>dows’ [*forget ≫ ∀; ∀ ≫ forget]The pervasiveness of matrix scope: Closer scrut<strong>in</strong>y reveals that wide scope <strong>in</strong> long passives ismuch more widespread. The contrast <strong>in</strong> (2) demonstrates that <strong>the</strong> <strong>in</strong>direct object allen Studenten‘all students.dat’ can have low scope if <strong>the</strong> direct object receives accusative (<strong>in</strong> (2a)) but <strong>in</strong>variablytakes scope over vergessen ‘forget’ if <strong>the</strong> direct object receives nom<strong>in</strong>ative (<strong>in</strong> (2b)).(2) a. weil den Fritz allen Studenten vorzustellen vergessen wurdes<strong>in</strong>ce <strong>the</strong>.nom Fritz all.dat students.dat to.<strong>in</strong>troduce forgotten was‘s<strong>in</strong>ce it was forgotten to <strong>in</strong>troduce Fritz to all students’ [forget ≫ ∀]b. weil der Fritz allen Studenten vorzustellen vergessen wurdes<strong>in</strong>ce <strong>the</strong>.nom Fritz all.dat students.dat to.<strong>in</strong>troduce forgotten was‘s<strong>in</strong>ce it was forgotten to <strong>in</strong>troduce Fritz to all students’ [*forget ≫ ∀; ∀ ≫ forget]This pattern even generalizes to adjuncts. The scope of <strong>in</strong> jedem Zimmer ‘<strong>in</strong> every room’correlates with <strong>the</strong> case of <strong>the</strong> embedded object (<strong>in</strong>dicated by agreement):(3) a. weil <strong>in</strong> jedem Zimmer Äpfel zu essen vergessen wurdes<strong>in</strong>ce <strong>in</strong> every room apples.acc to eat forgotten was‘s<strong>in</strong>ce it was forgotten to eat apples <strong>in</strong> every room’ [forget ≫ ∀]b. weil <strong>in</strong> jedem Zimmer Äpfel zu essen vergessen wurdens<strong>in</strong>ce <strong>in</strong> every room apples.nom to eat forgotten were‘s<strong>in</strong>ce it was forgotten to eat apples <strong>in</strong> every room’ [*forget ≫ ∀; ∀ ≫ forget]Thus, every element of <strong>the</strong> embedded clause takes wide scope <strong>in</strong> long passives. B&W’s analysisis unable to capture <strong>the</strong> contrasts <strong>in</strong> (2) and (3) precisely because it attributes <strong>the</strong> wide scope <strong>in</strong>(1b) to case-driven rais<strong>in</strong>g. Nei<strong>the</strong>r <strong>the</strong> <strong>in</strong>direct object nor adjuncts depend on case-assignmentfrom <strong>the</strong> matrix clause. We conclude that <strong>the</strong> wide scope <strong>in</strong> long passives is not due to case.Adjacency and scope: Not all <strong>in</strong>stances of long passives lead to wide scope of embeddedquantifiers. In particular, it only does if <strong>the</strong> two ma<strong>in</strong> verbs are adjacent to each o<strong>the</strong>r, as <strong>the</strong>yare <strong>in</strong> (1)–(3). If <strong>the</strong> <strong>in</strong>f<strong>in</strong>itival verb is topicalized, low scope is possible. Extraposition patterns<strong>in</strong> <strong>the</strong> same way.(4) [Allen Studenten vorzustellen] wurde der Fritz schon wieder vergessenall.dat students to.<strong>in</strong>troduce was <strong>the</strong>.nom Fritz yet aga<strong>in</strong> forgotten‘it was yet aga<strong>in</strong> forgotten to <strong>in</strong>troduce Fritz to all <strong>the</strong> students’ [forget ≫ ∀]All <strong>the</strong> data discussed so far can be summarized <strong>in</strong> terms of <strong>the</strong> generalization <strong>in</strong> (5).