The role of parallelism in the real-time processing of anaphora

Abstract

Parallelism effects refer to the facilitated processing of a target structure when it follows a similar, parallel structure. In coordination, a parallelism-related conjunction triggers the expectation that a second conjunct with the same structure as the first conjunct should occur. It has been proposed that parallelism effects reflect the use of the first structure as a template that guides the processing of the second. In this study, we examined the role of parallelism in real-time anaphora resolution by charting activation patterns in coordinated constructions containing anaphora, Verb-Phrase Ellipsis (VPE) and Noun-Phrase Traces (NP-traces). Specifically, we hypothesised that an expectation of parallelism would incite the parser to assume a structure similar to the first conjunct in the second, anaphora-containing conjunct. The speculation of a similar structure would result in early postulation of covert anaphora. Experiment 1 confirms that following a parallelism-related conjunction, first-conjunct material is activated in the second conjunct. Experiment 2 reveals that an NP-trace in the second conjunct is posited immediately where licensed, which is earlier than previously reported in the literature. In light of our findings, we propose an intricate relation between structural expectations and anaphor resolution.

Keywords:

• Parallelism
• Anaphora
• Ellipsis
• Unaccusative verbs

Acknowledgements

This research was supported by NIH-NIDCD grant DC000494 to Lewis P. Shapiro and by the following fellowships to Josée Poirier: Joseph-Armand Bombardier Foundation Internationalist Fellowship, William Orr Dingwall Neurolinguistics Fellowship and the Sheila & Jeffrey Lipinsky Family Fellowship in Language and Communicative Disorders. The authors gratefully thank Dr. Gina Taranto for her valuable assistance with verb selection and the research assistants for their help with data collection.

Notes

¹Only participants who showed accuracy levels above 70% on the lexical decision task and on comprehension questions were considered in data analysis. Ten participants were excluded based on this criterion. It is assumed that a poorer performance on either task is indicative of the participant's failure to attend to or understand the experimental paradigm.

²Probes were taken from a database built over the years in the Language Processes Laboratory (SDSU).The probes in this database have been pre-tested for baseline reaction times and semantic relatedness to their prime. The frequency of occurrence of the control (mean: 74) and related probes (mean: 97) did not significantly differ (t ₃₅ = 0.004, p > .05, two-tailed.)

³We note that Shapiro et al. (2003, Exp.3) did not observe a reliable priming effect (17 ms, ns) at their pre-elision position in “and”-coordinated elliptical constructions. Although caution must be taken in interpreting a null effect, their results raise the interesting possibility that a prosodic break (marking an upcoming relative clause) could have terminated the activation of first-clause material by the time of the pre-elision position. In effect, the presence of a relative clause breaks the parallelism between the antecedent and elliptical clauses, and information available to the processor well before the pre-elision position. The processor's strategy to use the first-conjunct structure to guide second-conjunct parsing thus becomes ill-advised, and activation of first-conjunct material would become obsolete.

⁴As in Experiment 1, only participants who showed accuracy levels above 70% on the lexical decision task and on comprehension questions were considered in data analysis. Seven participants were excluded based on this criterion. It is assumed that a poorer performance on either task is indicative of the participant's failure to attend or understand the experimental paradigm.

⁵We note that UE verbs yielded a reliable difference between NP2-related and unrelated probes at the downstream position. This effect was not one of facilitation: decisions on NP2-related probes were slower than on unrelated probes. Such “negative effects” are difficult to interpret but could be attributable to subject-specific inhibition or processing difficulty.