The effect of filler complexity across main- and embedded clauses on the acceptability of wh-island extraction in English.

Keywords
Loading...
Thumbnail Image
Issue Date
2023-08-22
Language
en
Document type
Journal Title
Journal ISSN
Volume Title
Publisher
Title
ISSN
Volume
Issue
Startpage
Endpage
DOI
Abstract
The acceptability of syntactic island violations is generally taken to ameliorate when wh-filler complexity is increased. This effect is exclusively found in weak islands. English, for example, has weak islands, because it has filler complexity amelioration. Beljon et al. (2021) posit that Dutch has strong islands, because they found no convincing evidence for a filler complexity effect in Dutch wh-islands. In their analysis, they conclude that filler-complexity amelioration effects do not appear in Dutch main clauses, because a featural Relativized Minimality analysis shows that the Dutch grammar does not distinguish between complex or bare fillers in main clauses. The Dutch main clause verb-second (V2) constraint is suggested as the reason, because V2 ostensibly causes the C-position in Dutch main clauses to always carry some extra feature F, in addition to a Q feature. When the filler takes this position, it has full featural overlap with the embedded subject, which entails the highest degree of intervention effect, meaning strong islands. This analysis makes the clear prediction that sentences without V2 constraints (e.g. Dutch embedded clauses) should lack the feature that V2 adds to the C-position. Then, Inclusion can be avoided, leading to a weaker intervention effect, and consequently, weaker islands. Theoretically, this suggests that Dutch embedded clauses could show complexity-amelioration. The other side of this prediction is that languages without V2, such as English must be shown to have a complexity effect that stays consistent in both main and embedded clauses, because otherwise the factor V2 cannot be singled out as the cause for strong islands in Dutch. The consistency of the complexity effect between main and embedded clauses in English is what the present study aimed to empirically test. To ascertain this, a graded acceptability judgment task was conducted. The effect of complexity was found to be significant and consistent across both types of clauses, which aligns with the predictions made by the analysis in Beljon et al. that non-V2 island extractions should show consistent complexity effects across both main and embedded clauses. Keywords: syntactic islands, wh-movement, filler complexity, Relativized Minimality embedded questions, graded judgments
Description
Citation
Faculty
Faculteit der Letteren