Locality, Cyclicity, and Resumption: At the Interface between the Grammar and the Human Sentence Processor

Abstract

We present an experimental investigation of the role of resumptive pronouns. We investigate object extraction in WH-questions for a range of syntactic configurations (nonislands, weak islands, strong islands) and for multiple levels of embedding (single, double, and triple). In order to establish the crosslinguistic properties of resumption, parallel experiments were conducted in three languages, viz. English, Greek, and German. Three main experimental results are reported. First, resumption does not remedy island violations: resumptive pronominals are at most as acceptable as gaps, but not more acceptable. This result disconfirms claims in the literature that resumptives can 'save' island violations. Second, embedding reduces acceptability even in extraction out of nonislands and declaratives, structures standardly assumed to be fully grammatical. Third, nonislands and weak islands pattern together and contrast with strong islands in terms of the effect of resumption and embedding. Our experimental findings show a remarkable consistency across the three languages we investigate; crosslinguistic variation appears confined to quantitative differences in crosslinguistically identical principles. We argue that these experimental results can be explained by the interaction of grammatical principles with resource limitations of the human parser. In particular, extraction from nonislands and weak islands imposes increased demands on the computational resources of the parser. We extend Gibson's (1998) syntactic prediction locality theory in order to formalize this intuition and account for the processing complexity of A-bar dependencies.*

Data from eye-tracking corpora as evidence for theories of syntactic processing complexity

“…Integration cost is standardly defined in terms of the number of discourse referents intervening between a head and its dependents, but alternatives have been proposed in the literature. For example, Alexopoulou and Keller (2007) show that two types of extraction from wh-phrases can differ in processing complexity, even though they involve the same number of intervening discourse referents. Based on this result, they argue that the number of intervening syntactic heads (rather than discourse referents) is the crucial factor for determining integration cost.…”

Resumption and gaps in English relative clauses: Relative acceptability creates an illusion of ‘saving’

“…Because McDaniel and Cowart (1999) tested only structures in which a gap simultaneously creates an ECP effect and violates a wh-island, structures in which a gap would create an ECP effect but not violate an island constraint should be tested. Additionally, as noted in Alexopoulou and Keller (2007), future experiments should test resumptives and gaps in relative clause structures.…”

“…Surprisingly, the expectation that English speakers should find resumptive pronouns to be more acceptable than island-violating gaps was not met in Alexopoulou and Keller (2007), which tested the acceptability of resumptive pronouns and gaps in English, German, and Greek. Using Magnitude Estimation, Alexopoulou and Keller tested sentences with wh-extraction out of object position in non-islands (bare clauses and 'that'-clauses), weak islands ('whether'-clauses), and strong islands (relative clauses) (Alexopoulou and Keller 2007:117 For all conditions, Alexopoulou and Keller found that resumptives were at most as acceptable as the corresponding gaps, but, critically, were never more acceptable.…”

A comparison of informal and formal acceptability judgments using a random sample from Linguistic Inquiry 2001–2010

“…This development means that there are two methods for collecting acceptability judgments currently in widespread use in the field of syntax: the relatively informal traditional methods that have largely established the foundation of the field for the past 60 years (henceforth informal methods), and the more formal experimental methods that have been gaining popularity over the past 15 years (henceforth formal methods). This methodological dichotomy has led a number of researchers to ask which method is empirically superior (e.g., Bard et al, 1996;Keller, 2000;Edelman and Christiansen, 2003;Phillips and Lasnik, 2003;Featherston, 2005aFeatherston, , 2005bFeatherston, , 2007Featherston, , 2008Featherston, , 2009Ferreira, 2005;Sorace and Keller, 2005;Wasow and Arnold, 2005;den Dikken et al, 2007;Alexopoulou and Keller, 2007;Bornkessel-Schlesewsky and Schlesewsky, 2007;Fanselow, 2007;Grewendorf, 2007;Haider, 2007;Newmeyer, 2007;Sprouse, 2007;Culbertson and Gross, 2009;Myers, 2009aMyers, , 2009bPhillips, 2010;Bader and Häussler, 2010;Dąbrowska, 2010;Gibson and Fedorenko, 2010;Culicover and Jackendoff, 2010;Gross and Culberton, 2011;Sprouse, 2011b;Weskott and Fanselow, 2011;Gibson et al, 2011;Almeida, 2012, 2013;Gibson and Fedorenko, 2013). Our goal in this paper is to substantially increase the empirical basis of this line of research by comparing the results of informal and formal methods for a very large and random s...…”