Category Locality Theory: A unified account of locality effects in sentence comprehension

In real-time sentence comprehension, the comprehender is often required to establish syntactic dependencies between words that are linearly distant. Major models of sentence comprehension assume that longer dependencies are more difficult to process because of working memory limitations. While the expected effect of distance on reading times (locality effect) has been robustly observed in certain constructions, such as relative clauses in English, its generalizability to a wider range of constructions has been empirically questioned. The current study proposes a new metric of syntactic distance that capitalizes on the flexible constituency of Combinatory Categorial Grammar (CCG), and argues that it offers a unified account of the locality effects. It is shown that this metric correctly predicts both the presence of the locality effect in English relative clauses and its absence in verb-final languages, without assuming language- or dependency-specific differences in the sensitivity to the locality effect. It is further shown that the CCG-based distance is a significant predictor of the self-paced reading times from an English corpus, even when other known predictors such as dependency-based locality and surprisal are taken into account. These results suggest that human sentence comprehension involves rapid integration of input words into efficiently compressed syntactic representations, and CCG is a plausible theory of the grammar that subserves this process.

The Effect of Syntactic Similarity on Intra-Sentential Switching Costs: Evidence from Chinese-English Bilinguals

In order to better understand the role of syntactic similarity in a code-switched sentence, the current study explored the effect of similar and different syntactic structures on Chinese-English bilinguals' intra-sentential switching costs. L2 proficiency and switching directions as factors that potentially intervene in bilingual performance were together explored to see if there was any interaction. We manipulated the degree of syntactic similarity by utilizing clauses in active voice (greater similarity) and passive voice (lesser similarity). The study conducted a self-paced reading paradigm as a more natural language reading processing. Results showed overall longer reading times for active sentences than passive counterparts, which supported a syntactic similarity impediment rather than facilitation. The impediment seemed to be predominant irrespective of L2 proficiency. Furthermore, syntactic similarity modulated the asymmetry of switching costs between forward (L1-L2) and backward (L2-L1) direction: word RTs for the 1st and the 2nd switched word yielded greater costs in L2-L1 condition, while greater costs in L1-L2 condition was observed in 3rd switched word RTs and average RTs. The present study observed syntactic similarity impediment rather than facilitation for Chinese-English bilinguals. Notably, syntactic similarity plays a predominant role compared to L2 proficiency, and modulates the asymmetry between switching directions.

Processing and Comprehension of Locally Ambiguous Participial Relative Clause Sentences in Russian

The study tested how the Recency Preference and Predicate Proximity model (Gibson et al. in Cognition 59(1):23-59, 1996, https://doi.org/10.1016/0010-0277(88)90004-2 ) plays out by examining the attachment preferences of native Russian speakers when processing locally ambiguous participial relative clause sentences with three potential NP attachment sites in Russian. Using a self-paced reading task, reading times and noun phrase selection responses were collected. Results showed significantly shorter reading times at the disambiguating region and higher accuracy rate of selection in the high-attaching condition than in the middle- and low-attaching conditions. No significant differences were found between the middle- and low-attaching conditions. We argue that Predicate Proximity is a much stronger factor than Recency Preference in Russian.

Capturing individual differences in sentence processing: How reliable is the self-paced reading task?

Advances in research on language processing have originally come from group-level comparisons, but there is now a growing interest in individual differences. To investigate individual differences, tasks that have shown robust group-level differences are often used with the implicit assumption that they will also be reliable when used as an individual differences measure. Here, we examined whether one of the primary tasks used in psycholinguistic research on language processing, the self-paced reading task, can reliably measure individual differences in relative clause processing. We replicated the well-established effects of relative clauses at the group level, with object relative clauses being more difficult to process than subject relative clauses. However, when using difference scores, the reliability of the size of the relative clause effect was close to zero because the self-paced reading times for the different relative clause types were highly correlated within individuals. Nonetheless, we found that the self-paced reading task can be used to reliably capture individual differences in overall reading speed as well as key sentence regions when the two types of relative clause sentences are considered separately. Our results indicate that both the reliability and validity of different sentence regions need to be assessed to determine whether and when self-paced reading can be used to examine individual differences in language processing.

Discourse coherence modulates use of predictive processing during sentence comprehension

Context has been shown to be vitally important for comprehension. Lexical processing is facilitated when words are highly predictable given their local sentence context, suggesting that people pre-activate likely upcoming words to aid comprehension. However, this facilitation is affected by knowledge about the global context in which comprehension takes place: people predict less when in an environment where expectations are frequently violated. The current study investigated whether discourse coherence is an additional cue that comprehenders use to modulate lexical prediction. In a series of online, self-paced reading experiments, participants read target sentences preceded by short contextual preambles. Local facilitation effects were manipulated through the cloze probability of a critical word within the target sentence and discourse coherence was manipulated by varying the degree to which the target sentence was consistent with the information presented in the preamble. In the first two experiments, target sentences were read more slowly when they occurred in less coherent discourses, but no local facilitation effects were observed. In the third experiment, we strengthened the predictability manipulation by using semantically anomalous critical words. In this experiment, predictable words were processed more quickly and anomalous words more slowly when they occurred in highly coherent discourse. Our results suggest that comprehenders are sensitive to shifts in the topic of discourse and that they downregulate predictive processing when they encounter incoherence in the discourse. This is consistent with recent theoretical accounts suggesting that comprehenders flexibly engage in predictive processing, pre-activating semantic and lexical information less when their expectations are less likely to be reliable.

FAB: A "Dummy's" program for self-paced forward and backward reading

The self-paced reading paradigm has been popular and widely used in psycholinguistic research for several decades. The tool described in this paper, FAB (Forward and Backward reading), is a tool created to hopefully and maximally reduce the coding demands and simplify the operation costs for experimental researchers and clinical researchers who are doing experimental work, in their designing, coding, implementing, and analyzing self-paced reading tasks. Its basis in web languages (HTML, JavaScript) also promotes experimental implementation and material sharing in our era of open science. In addition, FAB has a unique forward-and-backward mode that can track regressive-like behaviors that are usually only recordable using eye-tracking or mouse-tracking equipment. In this paper, the specific application and usage of FAB is demonstrated in one laboratory and two online validation experiments. We hope this free and open-sourced tool can benefit research in a diverse range of contexts where self-paced reading is desirable.

Exploring factors influencing young children's learning from storybooks: Interactive and multimedia features

Electronic storybooks are increasingly popular with preschoolers. The purpose of our research was to investigate the effects of interactive and multimedia features in electronic storybooks on preschoolers' learning. We assigned 4- to 6-year-old children to different reading conditions in two experiments. Children were required to complete tests for learning outcomes and answer questions about the reading experience. In Experiment 1, children in the interactive (self-paced) group needed to turn the pages by a button on the page, whereas the pages were turned automatically in the non-interactive (system-controlled) group. We found that children in the system-controlled condition performed better in inference making than children in the self-paced condition. In Experiment 2, we used a 2 (Animation: present or non-present) × 2 (Background Music: present or non-present) between-participants design. We found that children's scores of learning and interest in groups with animations were higher than those in groups without animations. This research suggests that electronic books with animations congruent with the learning content promote learning for both adults and young children. Thus, we offer suggestions for designers of electronic books. Moreover, the study provides implications for educators and parents, and we suggest that multi-featured electronic storybooks for preschoolers should be carefully selected.

Robust Processing Advantage for Binomial Phrases with Variant Conjunctions

Prior research has shown that various types of conventional multiword chunks are processed faster than matched novel strings, but it is unclear whether this processing advantage extends to variant multiword chunks that are less formulaic. To determine whether the processing advantage of multiword chunks accommodates variations in the canonical phrasal template, we examined the robustness of the processing advantage (i.e., Predictability) of binomial phrases with non-canonical conjunctions (e.g., salt and also pepper; salt as well as pepper). Results from the cloze study (Experiment 1) showed that there was a high tendency of producing the canonical conjunct (pepper), even in the binomials that contained non-formulaic conjunctions. Consistent with these findings, results from two eye tracking studies (Experiments 2a and 2b) showed that canonical conjuncts were read faster than novel conjuncts that were matched on word length (e.g., paprika), even in the binomials with variant conjunctions. This robust online processing advantage was replicated in a self-paced reading study that compared all three Conjunction Types (Experiment 3). Taken together, these findings show that binomials with variant function words also receive facilitated processing relative to matched novel strings, even though both types of strings are neither conventional nor relatively frequent. Exploratory analyses revealed that this processing speed advantage was driven by the lexical-semantic association between the canonical conjuncts (salt-pepper), rather than lexical and phrasal frequency. Overall, these results highlight flexibility in the processing of multiword chunks that current models of multiword storage and processing must take into account.

Fixation-related fMRI analysis reveals the neural basis of parafoveal processing in self-paced reading of Chinese words

While parafoveal word processing plays an important role in natural reading, the underlying neural mechanism remains unclear. The present study investigated the neural basis of parafoveal processing during Chinese word reading with the co-registration of eye-tracking and functional magnetic resonance imaging (fMRI) using fixation-related fMRI analysis. In the gaze-contingent boundary paradigm, preview conditions (words that are identical, orthographically similar, and unrelated to target words), pre-target word frequency and target word frequency were manipulated. When fixating the pre-target word, the identical preview condition elicited lower brain activation in the left fusiform gyrus relative to unrelated and orthographically similar preview conditions and there were significant interactions of preview condition and pre-target word frequency on brain activation of the left middle frontal gyrus, left fusiform gyrus and supplementary motor area. When fixating the target word, there was a significant main effect of preview condition on brain activation of the right fusiform gyrus and a significant interaction of preview condition and pre-target word frequency on brain activation of the left middle frontal gyrus. These results suggest that fixation-related brain activation provides immediate measures and new perspectives to understand the mechanism of parafoveal processing in self-paced reading.

Determiner-Number Specification and Non-Local Agreement Computation in L1 and L2 Processing

The present study employed a self-paced reading task in conjunction with concurrent acceptability judgements to examine how similar or different English natives and Chinese learners of English are when processing non-local agreement. We also tested how determiner-number specification modulates number agreement computation in both native and non-native processing by manipulating number marking with demonstrative determiners (the versus that/these). Results suggest both groups were sensitive to non-local agreement violations, indexed by longer reading times for sentences containing number violations. Furthermore, we found determiner-number specification facilitated processing of number violations in both native and non-native groups in an acceptability judgement task only, with stronger sensitivity to violations with demonstrative determiners than those with bare determiners. Contrary to some theories that predict qualitative differences between native and non-native processing, we did not find any significant differences between native and non-native speakers, despite the fact that the Chinese speakers of English had to process a novel linguistic feature absent in their native language.

Stimulus data and experimental design for a self-paced reading study on emoji-word substitutions

This data paper presents the experimental design and stimuli from an online self-paced reading study on the processing of emojis substituting lexically ambiguous nouns. We recorded reading times for the target ambiguous nouns and for emojis depicting either the intended target referent or a contextually inappropriate homophonous noun. Furthermore, we recorded comprehension accuracy, demographics and a self-assessment of the participants’ emoji usage frequency. The data includes all stimuli used, the raw data, the full JavaScript code for the online experiment, as well as Python and R code for the data analysis. We believe that our dataset may give important insights related to the comprehension mechanisms involved in the cognitive processing of emojis. For interpretation and discussion of the experiment, please see the original article entitled “The processing of emoji-word substitutions: A self-paced-reading study”.

Negation Cancels Discourse-Level Processing Differences: Evidence from Reading Times in Concession and Result Relations

Seminal studies on negation revealed that negative sentences are difficult to process, as they require an extra mental step. Similarly, at the discourse level, concession has been repeatedly shown to be more complex than other relations such as result because it implies the denial of an inference. The affinity between negation and concession prompted the present study to test whether overt verb polarity would affect the processing of upcoming discourse relations. In particular, it investigated whether negation can act as a cue to help process concessive relations. Results from four self-paced reading experiments indeed show a robust facilitation effect of negation on concession that cancels the baseline difference between concessive and result relations, thus nuancing existing context-blind categorizations of concession as a highly complex relation. This study furthers our understanding of how various types of cues interact in discourse processing and switches the focus from "what makes negation easier to process" to "what is made easier thanks to negation".

Subject Filled-gap Effects: A Self-paced Reading Study with Mandarin Learners of English

This study examined how L1 Mandarin L2 English learners with different L2 proficiency levels processed the syntactic structure of English sentences. Using a self-paced reading paradigm, this study investigated three issues: first, whether Mandarin learners of English were able to predict the syntactic structure, subject filled-gap, in online English sentence reading; second, whether the parser would commit more strongly to the subject gap analysis when there is a longer distance between the filler and the potential gap; third, how L2 proficiency levels would influence this process. The results showed that regardless of their English proficiency levels, Mandarin learners of English were capable to predict a subject filled-gap when processing English sentences. Besides, the subject filled-gap effect was not found in the sentences with adjuncts between the filler and the potential gap. This indicated that L2 learners' predictive ability in syntactic processing did not need to be activated by the increase of processing time or difficulty. Due to the efficiency of sentence processing, there is no need to increase the processing burden for L2 learners.

Processing cataphors: Active antecedent search is persistent

Cataphors precede their antecedents, so they cannot be fully interpreted until those antecedents are encountered. Some researchers propose that cataphors trigger an active search during incremental processing in which the parser predictively posits potential antecedents in upcoming syntactic positions (Kazanina et al., Journal of Memory and Language, 56[3], 384-409, 2007). One characteristic of active search is that it is persistent: If a prediction is disconfirmed in an earlier position, the parser should iteratively search later positions until the predicted element is found. Previous research has assumed, but not established, that antecedent search is persistent. In four experiments in English and Norwegian, we test this hypothesis. Two sentence completion experiments show a strong off-line preference for coreference between a fronted cataphor and the first available argument position (the main subject). When the main subject cannot be the antecedent, participants posit the antecedent in the next closest position: object position. Two self-paced reading studies demonstrate that comprehenders actively expect the antecedent of a fronted cataphor to appear in the main clause subject position, and then successively in object position if the subject does not match the cataphor in gender. Our results therefore support the claim that antecedent search is active and persistent.

Modeling Misretrieval and Feature Substitution in Agreement Attraction: A Computational Evaluation

We present computational modeling results based on a self-paced reading study investigating number attraction effects in Eastern Armenian. We implement three novel computational models of agreement attraction in a Bayesian framework and compare their predictive fit to the data using k-fold cross-validation. We find that our data are better accounted for by an encoding-based model of agreement attraction, compared to a retrieval-based model. A novel methodological contribution of our study is the use of comprehension questions with open-ended responses, so that both misinterpretation of the number feature of the subject phrase and misassignment of the thematic subject role of the verb can be investigated at the same time. We find evidence for both types of misinterpretation in our study, sometimes in the same trial. However, the specific error patterns in our data are not fully consistent with any previously proposed model.

Complement Coercion in Mandarin Chinese: Evidence from a Self-paced Reading Study

The study aims to explore the processing pattern of Mandarin Chinese sentences with complement coercion. Complement coercion is a known linguistic phenomenon in which some verbs, semantically requiring an event-denoting complement, are combined with an entity-denoting complement, as in Mary began the book. The combination (i.e., event-selecting verb + entity-denoting noun) has been reported to involve type mismatch, and thus elicits processing difficulty. While the phenomenon has been extensively studied in Indo-European languages, such as English and German, it is debatable if the phenomenon exists in a typologically distinct language from English (e.g., in structural complexity of words), such as Mandarin. To provide empirical evidence, the study conducted a self-paced reading experiment to compare the processing patterns of coercion sentences and non-coercion controls in Mandarin. The results showed longer reading times for the coercion sentences than the non-coercion counterparts, which supported previous findings about the processing difficulty of complement coercion.

Online representations of non-canonical sentences are more than good-enough

Proponents of good-enough processing suggest that readers often (mis)interpret certain sentences using fast-and-frugal heuristics, such that for non-canonical sentences (e.g., The dog was bitten by the man) people confuse the thematic roles of the nouns. We tested this theory by examining the effect of sentence canonicality on the reading of a follow-up sentence. In a self-paced reading study, 60 young and 60 older adults read an implausible sentence in either canonical (e.g., It was the peasant that executed the king) or non-canonical form (e.g., It was the king that was executed by the peasant), followed by a sentence that was implausible given a good-enough misinterpretation of the first sentence (e.g., Afterwards, the peasant rode back to the countryside) or a sentence that was implausible given a correct interpretation of the first sentence (e.g., Afterwards, the king rode back to his castle). We hypothesised that if non-canonical sentences are systematically misinterpreted, then sentence canonicality would differentially affect the reading of the two different follow-up types. Our data suggested that participants derived the same interpretations for canonical and non-canonical sentences, with no modulating effect of age group. Our findings suggest that readers do not derive an incorrect interpretation of non-canonical sentences during initial parsing, consistent with theories of misinterpretation effects that instead attribute these effects to post-interpretative processes.

Single-Stage Prediction Models Do Not Explain the Magnitude of Syntactic Disambiguation Difficulty

The disambiguation of a syntactically ambiguous sentence in favor of a less preferred parse can lead to slower reading at the disambiguation point. This phenomenon, referred to as a garden-path effect, has motivated models in which readers initially maintain only a subset of the possible parses of the sentence, and subsequently require time-consuming reanalysis to reconstruct a discarded parse. A more recent proposal argues that the garden-path effect can be reduced to surprisal arising in a fully parallel parser: words consistent with the initially dispreferred but ultimately correct parse are simply less predictable than those consistent with the incorrect parse. Since predictability has pervasive effects in reading far beyond garden-path sentences, this account, which dispenses with reanalysis mechanisms, is more parsimonious. Crucially, it predicts a linear effect of surprisal: the garden-path effect is expected to be proportional to the difference in word surprisal between the ultimately correct and ultimately incorrect interpretations. To test this prediction, we used recurrent neural network language models to estimate word-by-word surprisal for three temporarily ambiguous constructions. We then estimated the slowdown attributed to each bit of surprisal from human self-paced reading times, and used that quantity to predict syntactic disambiguation difficulty. Surprisal successfully predicted the existence of garden-path effects, but drastically underpredicted their magnitude, and failed to predict their relative severity across constructions. We conclude that a full explanation of syntactic disambiguation difficulty may require recovery mechanisms beyond predictability.

Context and Literality in Idiom Processing: Evidence from Self-Paced Reading

In a self-paced reading study, we investigated how effects of biasing contexts in idiom processing interact with effects of idiom literality. Specifically, we tested if idioms with a high potential for literal interpretation (e.g., break the ice) are processed differently in figuratively and literally biasing contexts than idioms with a low potential (e.g., lose one's cool). Participants read sentences that biased towards a figurative or literal reading of idioms and continued with resolutions that were congruent or incongruent with these biases (e.g., [The new schoolboy/the chilly Eskimo] just wanted to break the ice [with his peers/on the lake]…). While interpretations of high-literality idioms were strengthened by supporting contexts and showed costs for incongruent resolutions, low-literality idioms did not show this effect. Rather, interpreting low-literality idioms in a literal manner showed a cost regardless of context. We conclude that biasing contexts are used in a flexible process of real-time idiom processing and meaning constitution, but this effect is mediated by idiom literality.

Memory Load Effect in the Real-Time Processing of Scalar Implicatures

This study examines effects of memory load on the processing of scalar implicature via a dual-task paradigm using reading span and self-paced reading. Results indicate that participants showed online sensitivity to underinformative sentences (e.g., Some birds have wings and beaks) at the end of the sentence. This online sensitivity disappeared when participants were under increased memory load. Moreover, participants in the memory-load condition did not show sensitivity to semantically false sentences (e.g., All books have pictures and drawings). These results pose important conceptual and methodological questions of (1) whether the processing cost associated with scalar implicatures can be attributed to general proposition evaluation rather than scalar implicature derivation per se (Bale et al. in Semant Linguist Theory 20:525-543, 2010), and (2) to what degree memory load affects implicature computation only. I conclude with a discussion of these two issues for future research.

Speedy Metonymy, Tricky Metaphor, Irrelevant Compositionality: How Nonliteralness Affects Idioms in Reading and Rating

It is widely acknowledged that fixed expressions such as idioms have a processing advantage over non-idiomatic language. While many idioms are metaphoric, metonymic, or even literal, the effect of varying nonliteralness in their processing has not been much researched yet. Theoretical and empirical findings suggest that metonymies are easier to process than metaphors but it is unclear whether this applies to idioms. Two self-paced reading experiments test whether metonymic, metaphoric, or literal idioms have a greater processing advantage over non-idiomatic control sentences, and whether this is caused by varying nonliteralness. Both studies find that metonymic and literal idioms are read significantly faster than controls, while the advantage for metaphoric idioms is only tenuous. Only experiment 2 finds literal idioms to be read fastest of all. As compositionality of the idioms cannot account for these findings, some effect of nonliteralness is suggested, together with idiomaticity and the sentential context.

Delayed Application of Binding Condition C During Cataphoric Pronoun Resolution

Previous research has shown that during cataphoric pronoun resolution, the predictive search for an antecedent is restricted by a structure-sensitive constraint known as 'Condition C', such that an antecedent is only considered when the constraint does not apply. Evidence has mainly come from self-paced reading (SPR), a method which may not be able to pick up on short-lived effects over the timecourse of processing. This study investigates whether or not the active search mechanism is constrained by Condition C at all points in time during cataphoric processing. We carried out one eye-tracking during reading and a parallel SPR experiment, accompanied by offline coreference judgment tasks. Although offline judgments about coreference were constrained by Condition C, the eye-tracking experiment revealed temporary consideration of antecedents that should be ruled out by Condition C. The SPR experiment using exactly the same materials indicated, conversely, that only structurally appropriate antecedents were considered. Taken together, our results suggest that the application of Condition C may be delayed during naturalistic reading.

Testing the online reading effects of emotionality on relative clause attachment

Previous research has shown the impact of the emotional dimension of nouns (i.e., valence and arousal) on the completion of relative clauses (RC) that are preceded by a double antecedent [e.g.,: Someone shot the servant (the first noun phrase, NP1) of the actress (the second noun phrase, NP2) who was on the balcony] (Fraga et al. in Q J Exp Psychol 65:1740-1759, 2012). The present study explored for the first time the role of emotional valence, specifically emotional positive nouns, on RC disambiguation in a self-paced reading experiment. Two types of NP1-NP2 relationships were compared: emotional-neutral vs. neutral-emotional. Results showed NP1 preferences in the emotional-neutral condition, whereas no preferences were found in the neutral-emotional condition. We conclude that during reading, the emotional properties of nouns play a role in disambiguation preferences: RC attachment preferences can be neutralized when emotional factors are manipulated. The results are discussed within the framework of current models of sentence processing and with reference to the controversial differences between comprehension and production.

Single-Word Recognition Need Not Depend on Single-Word Features: Narrative Coherence Counteracts Effects of Single-Word Features that Lexical Decision Emphasizes

Research on reading comprehension of connected text emphasizes reliance on single-word features that organize a stable, mental lexicon of words and that speed or slow the recognition of each new word. However, the time needed to recognize a word might not actually be as fixed as previous research indicates, and the stability of the mental lexicon may change with task demands. The present study explores the effects of narrative coherence in self-paced story reading to single-word feature effects in lexical decision. We presented single strings of letters to 24 participants, in both lexical decision and self-paced story reading. Both tasks included the same words composing a set of adjective-noun pairs. Reading times revealed that the tasks, and the order of the presentation of the tasks, changed and/or eliminated familiar effects of single-word features. Specifically, experiencing the lexical-decision task first gradually emphasized the role of single-word features, and experiencing the self-paced story-reading task afterwards counteracted the effect of single-word features. We discuss the implications that task-dependence and narrative coherence might have for the organization of the mental lexicon. Future work will need to consider what architectures suit the apparent flexibility with which task can accentuate or diminish effects of single-word features.

Metonymy as Referential Dependency: Psycholinguistic and Neurolinguistic Arguments for a Unified Linguistic Treatment

We examine metonymy at psycho- and neurolinguistic levels, seeking to adjudicate between two possible processing implementations (one- vs. two-mechanism). We compare highly conventionalized systematic metonymy (producer-for-product: "All freshmen read O'Connell") to lesser-conventionalized circumstantial metonymy ("[a waitress says to another:] 'Table 2 asked for more coffee."'). Whereas these two metonymy types differ in terms of contextual demands, they each reveal a similar dependency between the named and intended conceptual entities (e.g., Jackendoff, 1997; Nunberg, 1979, 1995). We reason that if each metonymy yields a distinct processing time course and substantially non-overlapping preferential localization pattern, it would not only support a two-mechanism view (one lexical, one pragmatic) but would suggest that conventionalization acts as a linguistic categorizer. By contrast, a similar behavior in time course and localization would support a one-mechanism view and the inference that conventionalization acts instead as a modulator of contextual felicitousness, and that differences in interpretation introduced by conventionalization are of degree, not of kind. Results from three paradigms: self-paced reading (SPR), event-related potentials (ERP), and functional magnetic resonance imaging (fMRI), reveal the following: no main effect by condition (metonymy vs. matched literal control) for either metonymy type immediately after the metonymy trigger, and a main effect for only the Circumstantial metonymy one word post-trigger (SPR); a N400 effect across metonymy types and a late positivity for Circumstantial metonymy (ERP); and a highly overlapping activation connecting the left ventrolateral prefrontal cortex and the left dorsolateral prefrontal cortex (fMRI). Altogether, the pattern observed does not reach the threshold required to justify a two-mechanism system. Instead, the pattern is more naturally (and conservatively) understood as resulting from the implementation of a generalized referential dependency mechanism, modulated by degree of context dependence/conventionalization, thus supporting architectures of language whereby "lexical" and "pragmatic" meaning relations are encoded along a cline of contextual underspecification.

Encouraging prediction during production facilitates subsequent comprehension: Evidence from interleaved object naming in sentence context and sentence reading

Many studies have shown that a supportive context facilitates language comprehension. A currently influential view is that language production may support prediction in language comprehension. Experimental evidence for this, however, is relatively sparse. Here we explored whether encouraging prediction in a language production task encourages the use of predictive contexts in an interleaved comprehension task. In Experiment 1a, participants listened to the first part of a sentence and provided the final word by naming aloud a picture. The picture name was predictable or not predictable from the sentence context. Pictures were named faster when they could be predicted than when this was not the case. In Experiment 1b the same sentences, augmented by a final spill-over region, were presented in a self-paced reading task. No difference in reading times for predictive versus non-predictive sentences was found. In Experiment 2, reading and naming trials were intermixed. In the naming task, the advantage for predictable picture names was replicated. More importantly, now reading times for the spill-over region were considerable faster for predictive than for non-predictive sentences. We conjecture that these findings fit best with the notion that prediction in the service of language production encourages the use of predictive contexts in comprehension. Further research is required to identify the exact mechanisms by which production exerts its influence on comprehension.

Antecedent Frequency Effects on Anaphoric Pronoun Resolution: Evidence from Spanish

High-frequency words are usually understood and produced faster than low-frequency words. Although the effect of word frequency is a reliable phenomenon in many domains of language processing, it remains unclear whether and how frequency affects pronominal anaphoric resolution. We evaluated this issue by means of two self-paced reading experiments. Native speakers of Spanish read sentences containing the anaphoric noun or pronoun at the subject syntactic position (Experiment 1) or at the object syntactic position (Experiment 2) while the antecedent of the anaphor was either a high-frequency or a low-frequency word. Results showed that nominal anaphors were read faster when referring to high-frequency than to low-frequency antecedents, and faster when referring to subjects than to objects. Critically, pronoun reading times were unaffected by the frequency and by the syntactic position of the antecedent. These results are congruent with theories assuming that syntactic information of the words is not frequency sensitive.

Keep it local (and final): Remnant preferences in "let alone" ellipsis

The let alone construction (John can't run a mile, let alone a marathon) differs from standard coordination structures (with and or but) by requiring ellipsis of the second conjunct--for example, a marathon is the remnant of an elided clause [[see text] a marathon]. In support of an ellipsis account, a corpus study of British and American English finds that let alone exhibits a Locality bias, as the second conjunct preferentially contrasts with the nearest lexical item of the same syntactic type. Two self-paced reading studies show that the Locality bias is active during online processing, but must be reconciled with indicators of semantic contrast and discourse information. Further, a sentence-rating study shows that the Locality bias interacts with a Finality bias that favours placing the let alone phrase at the end of a clause, which sometimes necessitates a nonlocal contrast. Together, the results show how a general bias in ellipsis for local contrasts is affected by discourse demands, such as the need for scalar contrast imposed by let alone, thereby offering a window into how possibly divergent syntactic and discourse constraints impact sentence processing.

Cross-Linguistic Differences in Processing Double-Embedded Relative Clauses: Working-Memory Constraints or Language Statistics?

An English double-embedded relative clause from which the middle verb is omitted can often be processed more easily than its grammatical counterpart, a phenomenon known as the grammaticality illusion. This effect has been found to be reversed in German, suggesting that the illusion is language specific rather than a consequence of universal working memory constraints. We present results from three self-paced reading experiments which show that Dutch native speakers also do not show the grammaticality illusion in Dutch, whereas both German and Dutch native speakers do show the illusion when reading English sentences. These findings provide evidence against working memory constraints as an explanation for the observed effect in English. We propose an alternative account based on the statistical patterns of the languages involved. In support of this alternative, a single recurrent neural network model that is trained on both Dutch and English sentences is shown to predict the cross-linguistic difference in the grammaticality effect.

The thematic hierarchy in sentence comprehension: A study on the interaction between verb class and word order in Spanish

Linking is the theory that captures the mapping of the semantic roles of lexical arguments to the syntactic functions of the phrases that realize them. At the sentence level, linking allows us to understand "who did what to whom" in an event. In Spanish, linking has been shown to interact with word order, verb class, and case marking. The current study aims to provide the first piece of experimental evidence about the interplay between word order and verb type in Spanish. We achieve this by adopting role and reference grammar and the extended argument dependency model. Two different types of clauses were examined in a self-paced reading task: clauses with object-experiencer psychological verbs and activity verbs. These types of verbs differ in the way that their syntactic and semantic structures are linked, and thus they provide interesting evidence on how information that belongs to the syntax-semantics interface might influence the predictive and integrative processes of sentence comprehension with alternative word orders. Results indicate that in Spanish, comprehension and processing speed is enhanced when the order of the constituents in the sentence mirrors their ranking on a semantic hierarchy that encodes a verb's lexical semantics. Moreover, results show that during online comprehension, predictive mechanisms based on argument hierarchization are used rapidly to inform the processing system. Our findings corroborate already existing cross-linguistic evidence on the issue and are briefly discussed in the light of other sentence-processing models.

Uncertainty reduction as a measure of cognitive load in sentence comprehension

The entropy-reduction hypothesis claims that the cognitive processing difficulty on a word in sentence context is determined by the word's effect on the uncertainty about the sentence. Here, this hypothesis is tested more thoroughly than has been done before, using a recurrent neural network for estimating entropy and self-paced reading for obtaining measures of cognitive processing load. Results show a positive relation between reading time on a word and the reduction in entropy due to processing that word, supporting the entropy-reduction hypothesis. Although this effect is independent from the effect of word surprisal, we find no evidence that these two measures correspond to cognitively distinct processes.

Verb Transitivity Bias Affects On-line Sentence Reading in People with Aphasia

BACKGROUND

Studies of sentence comprehension in non-disordered populations have convincingly demonstrated that probabilistic cues influence on-line syntactic processing. One well-studied cue is verb argument structure bias, which refers to the probability that a verb will occur in a particular syntactic frame. According to the Lexical Bias Hypothesis, people with aphasia have difficulty understanding sentences in which the verb's argument structure bias conflicts with the sentence structure (e.g., a transitively biased verb in an intransitive sentence). This hypothesis may provide an account of why people with aphasia have difficulty understanding both simple and complex sentences.

AIMS

The purpose of this study was to test the Lexical Bias Hypothesis using an on-line measure of written sentence comprehension, self-paced reading.

METHODS PROCEDURES

The participants were ten people with aphasia and ten non-brain-damaged controls. The stimuli were syntactically simple transitive and intransitive sentences that contained transitively- or intransitively-biased verbs. For example, the transitively-biased verb "called" appeared in sentences such as "The agent called (the writer) from overseas to make an offer." The intransitively-biased verb "danced" appeared in sentences such as "The couple danced (the tango) every Friday night last summer."

OUTCOMES RESULTS

Both groups' reading times for critical segments were longer when the verb's transitivity bias did not match the sentence structure, particularly in intransitive sentences.

CONCLUSIONS

The results were generally consistent with the Lexical Bias Hypothesis, and demonstrated that lexical biases affect on-line processing of syntactically simple sentences in people with aphasia and controls.