Error-Driven Retrieval in Agreement Attraction Rarely Leads to Misinterpretation.

ABSTRACTDoes producing syntactic agreement rely on syntactic or memory-based retrieval processes? The present study investigated the extent to which syntactic processing deficits and working memory (WM) deficits predict susceptibility to agreement attraction [Bock, K., & Miller, C. A. (1991). Broken agreement. Cognitive Psychology, 23, 45–93], where speakers tend to erroneously produce plural agreement for a singular subject when another noun in the sentence is grammatically plural. Four brain-injured patients with varying degrees of grammatical and WM deficits completed sentences with local nouns that matched or mismatched in number with the head noun, and that were plausible or implausible subjects. Both aspects of grammatical deficits and the extent of WM deficits predicted the extent of agreement attraction effects. These data are consistent with the proposal that producing an agreeing verb involves a cue-based search in WM for an appropriate controlling noun, which is subject to interference from other elements in memory with similar properties [cf. Badecker, W., & Kuminiak, F. (2007). Morphology, agreement and working memory retrieval in sentence production: Evidence from gender and case in Slovak. Journal of Memory and Language, 56(1), 65–85. doi:10.1016/j.jml.2006.08.004].

Agreement attraction errors (such as the number error in the example “The key to the cabinets are rusty”) have been the object of many studies in the last 20 years. So far, almost all production experiments and all comprehension experiments looked at binary features (primarily at number in Germanic, Romance, and some other languages, in several cases at gender in Romance languages). Among other things, it was noted that both in production and in comprehension, attraction effects are much stronger for some feature combinations than for the others: they can be observed in the sentences with singular heads and plural dependent nouns (e.g.,“The key to the cabinets…”), but not in the sentences with plural heads and singular dependent nouns (e.g., “The keys to the cabinet…”). Almost all proposed explanations of this asymmetry appeal to feature markedness, but existing findings do not allow teasing different approaches to markedness apart. We report the results of four experiments (one on production and three on comprehension) studying subject-verb gender agreement in Russian, a language with three genders. Firstly, we found attraction effects both in production and in comprehension, but, unlike in the case of number agreement, they were not parallel (in production, feminine gender triggered strongest effects, while neuter triggered weakest effects, while in comprehension, masculine triggered weakest effects). Secondly, in the comprehension experiments attraction was observed for all dependent noun genders, but only for a subset of head noun genders. This goes against the traditional assumption that the features of the dependent noun are crucial for attraction, showing the features of the head are more important. We demonstrate that this approach can be extended to previous findings on attraction and that there exists other evidence for it. In total, these findings let us reconsider the question which properties of features are crucial for agreement attraction in production and in comprehension.

We report two parallel experiments conducted in French and in English in which we induced subject-verb agreement errors to explore the role of syntactic structure during sentence production. Previous studies have shown that attraction errors (i.e., a tendency of the verb to agree with an immediately preceding noun instead of with the subject of the sentence) occur when a preverbal local noun disagrees in number with the subject head noun. The attraction effect was accounted for either by the proximity of the local noun to the verb in the linearised sentence (linear distance hypothesis) or by the processing simultaneity of the head and local nouns situated in the same clause (clause packaging hypothesis). In the current experiments, speakers were asked to complete complex sentential preambles. Contrary to the predictions of these two hypotheses, we found that agreement errors were more frequent following an intermediate modifier (e.g., *The threat-S to the presidents-P of the company-S ARE serious) than an immediately preverbal modifier (e.g., *The threat-S to the president-S of the companies-P ARE serious). It is suggested that attraction is determined by the syntactic distance between the interfering noun and the head noun at a stage of the grammatical encoding of the sentence during which syntactic units are organised into a hierarchical structure.

Across four eye-tracking studies and one self-paced reading study, we test whether attraction in subject-verb agreement is affected by (a) the relative linear positions of target and distractor, and (b) the active dependency status of the distractor. We find an effect of relative position, with greater attraction in retro-active interference configurations, where the distractor is linearly closer to the critical verb (Subject…Distractor…V) than in pro-active interference where it is more distant (Distractor…Subject…V). However, within pro-active interference configurations, attraction was not affected by the active dependency status of the distractor: attraction effects were similarly small whether or not the distractor was waiting to complete an upcoming dependency at the critical verb, with Bayes Factor analyses showing evidence in favour of a null effect of active dependency status. We discuss these findings in terms of the decay of activation, and whether such decay is affected by maintenance of features in memory.

Determiner phrases (DPs), an overarching term, can be classified into two determiner types: referential determiner phrases (RDPs, e.g., the boy) and quantificational determiner phrases (QDPs, e.g., each boy). Using the event-related potential (ERP) technique, this study explored the modulation of RDP vs. QDP in the online processing of English subject-verb agreement with omission errors by Chinese learners of English, addressing the question of whether singular quantification increases or decreases Chinese learners' sensitivity to agreement violations. The experiment manipulated the determiner type, specifically RDP vs. QDP, and grammaticality (grammatical vs. ungrammatical). The results indicated that similar to previous studies, a P600 effect was elicited in response to subject-verb agreement violations with omission errors, demonstrating that Chinese L2 learners are sensitive to such agreement violations. Additionally, the ERP patterns exhibited variations due to D-linking and number specification of RDP and QDP. Regarding D-linking, subject-verb agreement violations in the QDP conditions, necessitating integration of discourse-related knowledge, elicited laterally and frontally distributed P600 effects associated with integration complexity at the discourse level; however, non-D-linked referential determiners elicited the posteriorly-distributed P600 effects. Differences in number specification resulted in the distinctive P600 latencies and whether P600 was preceded by N400 or not. While both the RDP and QDP conditions exhibited the P600 effects, the onset latency of this effect in the number-unspecified RDP condition was 300 ms later compared to the number-specified QDP condition. Furthermore, an additional N400 component observed in the RDP condition suggests that L2 learners acquire morphologically complex subject-verb agreements by rote, treating them as unanalyzed chunks. This N400 component was absent in the QDP condition. From these results, the conclusion can be drawn that L2 learners are sensitive to the subject-verb agreement violations with omission errors, and L2 processing patterns of subject-verb agreement vary with different features of determiners, providing further evidence for the cue-based retrieval model during comprehension of grammatical sentences. Pedagogical implications are provided, and the future research direction is suggested.

Theoretical accounts of the language production process have claimed that grammatical encoding steps during the formulation stage happen in a largely automatic fashion, unimpeded by other cognitive processes. By eliciting agreement attraction errors, our study tested the effect of external distractor noise on the generation of subject–verb agreement in spoken language. We modelled noisy environments with three different speech-free sounds with spectral, or both spectral and intensity modulation characteristics of speech. In silence and unmodulated noise, we found evidence for a plural mismatch effect, where a plural local noun attracts agreement away from a singular marked head noun. Under modulated noise the error patterns changed, and the number of errors increased in cases where the head noun of the preamble was marked for plural. In addition, background noise led to a reduction of speech rate and a reduction of utterance latency. Our results indicate that unspecific, speech-free noise can create a secondary task load which exerts an influence on the grammatical encoding stage of sentence formulation. We suggest that additional load leads to a slowing down of processing and subsequent difficulty to access the correct number information of the head noun when trying to retrieve an inflected verb form. Subjects overcome this difficulty by resorting to default singular marking on the verb. The results of this study challenge previous claims about the informational encapsulation of the grammatical encoding stage.

The processing of subject-verb number agreement has been extensively studied when the subject precedes the verb; by contrast, agreement processing in dependencies in which the subject follows the verb has been less investigated. To address this gap, the authors measured the processing of sentences with postverbal subjects in Italian adults, investigating the role of syntactic structure and discourse licensing on agreement processing. They tested whether number agreement violations in verb-subject orders were processed differently depending on whether they occurred in transitive structures with clitic-left dislocation versus unaccusative structures with locative inversion. Because transitive structures are temporarily ambiguous and more constrained by discourse conditions than unaccusative structures, the authors hypothesized that the former might increase comprehenders’ memory load and impair the detection of agreement violations. The results of speeded acceptability judgments and self-paced reading tasks failed to support this hypothesis. However, transitive structures were accepted less often in isolation and elicited higher reading times than unaccusative structures, consistent with an increased cognitive load. The authors propose that the two structures are processed differently, but that this does not affect the computation of morphosyntactic relations like agreement. Instead, measures targeting interpretative processes—like discourse integration—may be more promising for future research.

Psycholinguistically Speaking: Some Matters of Meaning, Marking, and Morphing

An active question in psycholinguistics concerns how we mentally encode and retrieve linguistic information in memory. In particular, it remains unclear what information (“cues”) guide retrieval. Previous work has extensively tested retrieval of noun phrases, but less is known about retrieval of other constituents, including verb phrases (VPs). This study examines retrieval for VP ellipsis to allow for a more comprehensive theory of cues. Four experiments (acceptability, self-paced reading) used an interference paradigm to examine voice information (active, passive) in retrieval. Results revealed a selective profile: passive ellipsis shows interference, but active ellipsis does not. These results are aligned with the markedness asymmetry observed for agreement attraction, where marked features (plural, passive) trigger interference, but unmarked features (singular, active) do not. This analysis motivates a unified account of verbal dependencies where markedness plays a more fundamental role than previously assumed. Lastly, I use ACT-R to demonstrate how markedness effects might arise in a cue-based retrieval architecture and discuss the current findings with respect to the leading theories of interference effects.

The puzzle of number agreement with disjunction

What makes syntactic processing of subject–verb agreement complex? The effects of distance and additional agreement features

Event-Related Brain Potentials Elicited by Failure to Agree

Pronominal dependencies have been shown to be more resilient to attraction effects than subject-verb agreement. We use this phenomenon to investigate whether antecedent-clitic dependencies in Spanish are computed like agreement or like pronominal dependencies. In Experiment 1, an acceptability judgment self-paced reading task was used. Accuracy data yielded reliable attraction effects in both grammatical and ungrammatical sentences, only in singular (but not plural) clitics. Reading times did not show reliable attraction effects. In Experiment 2, we measured electrophysiological responses to violations, which elicited a biphasic frontal negativity-P600 pattern. Number attraction modulated the frontal negativity but not the amplitude of the P600 component. This differs from ERP findings on subject-verb agreement, since when the baseline matching condition obtained a biphasic pattern, attraction effects only modulated the P600, not the preceding negativity. We argue that these findings support cue-retrieval accounts of dependency resolution and further suggest that the sensitivity to attraction effects shown by clitics resembles more the computation of pronominal dependencies than that of agreement.

Do learners who have mastered two gender systems prior to learning a third have an advantage in gender processing over those who have mastered only one gender system? To address this question, we recruited two groups of adult Polish learners of L3/Ln Swedish with and without prior knowledge of German. They completed a self-paced reading task in Swedish that involved two types of gender marking: the indefinite article, which is realised similarly in Swedish and German, and the definite suffix, which is uniquely present in Swedish. We found that while both groups were sensitive to gender agreement violations involving the indefinite article, only learners with knowledge of German showed sensitivity to gender agreement violations involving the definite suffix. Therefore, knowledge of non-native German conferred an advantage for the processing of dissimilar rather than similar gender marking. We interpret these findings in light of morphological transfer from German that consists in the employment of the existing morphemes and assigning them new grammatical functions. More broadly, the study indicates that the gender system in a non-native language rather than the native language, might be more helpful for the acquisition of a new type of gender marking in L3/Ln.

This study investigates the presence and latency of the P600 component in response to subject-verb agreement violations in spoken language in people with and without developmental dyslexia. The two groups performed at-ceiling level on judging the sentences on their grammaticality, but the ERP data revealed subtle differences between them. The P600 tended to peak later in the left posterior region in the dyslexic group compared with the control group. In addition, the group of dyslexic subjects did not show a P600 in response to sentences with a plural NP subject. These results suggest that brain activation involved in syntactic repair is more affected by linguistic complexity in developmental dyslexia compared with non-dyslexic individuals.