When nothing goes right, go left: A large-scale evaluation of bidirectional self-paced reading

In two web-based experiments, we evaluated the bidirectional self-paced reading (BSPR) paradigm recently proposed by Paape and Vasishth (2021). We used four sentence types: NP/Z garden-path sentences, RRC garden-path sentences, sentences containing inconsistent discourse continuations, and sentences containing reflexive anaphors with feature-matching but grammatically unavailable antecedents. Our results show that regressions in BSPR are associated with a decrease in positive acceptability judgments. Across all sentence types, we observed online reading patterns that are consistent with the existing eye-tracking literature. NP/Z but not RRC garden-path sentences also showed some indication of selective rereading, as predicted by the selective reanalysis hypothesis of Frazier and Rayner (1982). However, selective rereading was associated with decreased rather than increased sentence acceptability, which is not in line with the selective reanalysis hypothesis. We discuss the implications regarding the connection between selective rereading and conscious awareness, and for the use of BSPR in general.

Frozen by context: Pragmatic factors of syntactic freezing

Syntactic freezing has mainly been approached from a structural point of view, recently though, more cognitive approaches in terms of processing costs have been proposed. One such processing account is the additive account. According to this approach, the freezing effect is best explained as an additive effect of two syntactic processes coming together, rather than being a phenomenon on its own. Another processing account argues that the freezing effect is the result of a prosodic garden path according to which extraction can only take place from a prosodically focused constituent. The current study provides empirical evidence for a less discussed factor contributing to the freezing effect, namely a pragmatic one. The pragmatic account requires frozen sentences to have contextually given referents. If no such referent is present, the sentence becomes less acceptable. The need for such a referent comes from the non- default word order associated with frozen sentences, which often highlights/focuses a certain constituent. Several experiments were run to test the pragmatic account. Based on the results it was concluded that pragmatic factors play a significant role in explaining the apparent freezing effects. Other factors however, seem to contribute to this effect as well since this effect cannot be fully explained in terms of pragmatic factors solely.

What causes lingering misinterpretations of garden-path sentences: Incorrect syntactic representations or fallible memory processes?

A key question in research on sentence processing concerns how sentences that have been misanalyzed are reinterpreted, and to what extent the parser’s attempts at revision are successful. Past work has shown that misinterpretations associated with a syntactic misparse linger even after the entire sentence has been processed (Christianson, Hollingworth, Halliwell, & Ferreira, 2001; Slattery, Sturt, Christianson, Yoshida and Ferreira, 2013). In two reading experiments, we sought to evaluate the level of representation that is responsible for misinterpretations of garden-path sentences. We combined reading measures with an offline comprehension task, which enabled us to conditionalize reading time analyses on correct versus incorrect question-answering performance. Our results suggest that reanalysis does not always result in a correct interpretation, either because the final interpretation does not always reflect the global structure or because reanalysis processes result in the creation of licit local trees but fail to generate a complete global parse for the entire sentence.

EXPRESS: When readers fail to form a coherent representation of garden-path sentences

Various studies within the Good-Enough Approach observe that people often make errors in answering comprehension questions after reading garden-path sentences such as While Anna dressed the baby played in the crib (e.g. Christianson, Hollingworth, Halliwell, & Ferreira, 2001). Recently (Slattery, Sturt, Christianson, Yoshida, & Ferreira, 2013), it has been claimed that readers form a full syntactic analysis of these sentences, but they do not completely prune the original misanalysis. This paper presents evidence that these findings do not hold for all garden-path sentences. The main finding of the Good-Enough Approach – that the comprehension questions targeting the initial misanalysis yield significantly higher rates of incorrect answers after garden-path sentences, in comparison to after control sentences – was replicated here in three self-paced reading experiments on Czech. However, these experiments show a similar pattern of results for other comprehension questions, such as questions targeting an analysis that is not syntactically licensed at any point of processing. These results point out that certain garden-path structures may be very hard to process and that the process of garden-path repair might not be successful at all. Based on these results and the results of previous studies, the idea of a range of difficulty levels for garden-path structures is proposed.

Parsing Model and a Rational Theory of Memory

This paper explores how the rational theory of memory summarized in Anderson (1991) can inform the computational psycholinguistic models of human parsing. It is shown that transition-based parsing is particularly suitable to be combined with Anderson's theory of memory systems. The combination of the rational theory of memory with the transition-based parsers results in a model of sentence processing that is data-driven and can be embedded in the cognitive architecture Adaptive Control of Thought-Rational (ACT-R). The predictions of the parser are tested against qualitative data (garden-path sentences) and a self-paced reading corpus (the Natural Stories corpus).

Bilinguals on the garden-path: Individual differences in syntactic ambiguity resolution

Syntactic parsing plays a central role in the interpretation of sentences, but it is unclear to what extent non-native speakers can deploy native-like grammatical knowledge during online comprehension. The current eye-tracking study investigated how Chinese–English bilinguals and native English speakers respond to syntactic category and subcategorization information while reading sentences with object-subject ambiguities. We also obtained measures of English language experience, working memory capacity, and executive function to determine how these cognitive variables influence online parsing. During reading, monolinguals and bilinguals showed similar garden-path effects related to syntactic reanalysis, but native English speakers responded more robustly to verb subcategorization cues. Readers with greater language experience and executive function showed increased sensitivity to verb subcategorization cues, but parsing was not influenced by working memory capacity. These results are consistent with exposure-based accounts of bilingual sentence processing, and they support a link between syntactic processing and domain-general cognitive control.