Herman Kolk

Structure and limited capacity in verbal working memory: A study with event-related potentials ☆

In order to test recent claims about the structure of verbal working memory, two ERP experiments with Dutch speaking participants were carried out. We compared the ERP effects of syntactic and semantic mid-sentence anomalies in subject and object relative sentences. In Experiment 1, the participants made acceptability judgments, while in Experiment 2 they read for comprehension. Syntactic anomalies concerned subject-verb disagreement, while semantic anomalies were related to implausible events (e.g., *The cat that fled from the mice ran through the room). Semantic anomalies did not elicit an N400 effect. The semantic as well as syntactic anomalies elicited P600 effects, with similar centro-parietal scalp distributions. For both kinds of anomaly, the P600 effects were modulated by syntactic complexity: they were either relatively small (Experiment 1) or absent (Experiment 2) in object relative sentences. Taken together, our results suggest that: (a) verbal working memory is a limited capacity system; (b) it is not subdivided into an interpretative and a post-interpretative component (); (c) the P600 can reflect the presence of a semantic bias in syntactically unambiguous sentences; (d) the P600 is related to language monitoring: its function is to check upon the veridicality of an unexpected (linguistic) event; (e) if such a check is made, there is no integration of the event and hence no N400 effect.

Agrammatism as a variable phenomenon

Abstract With a group of 11 Dutch agrammatic speakers, the studies on production and comprehension of word order by Saffran et al. (1980) and Schwartz et al. (1980) were replicated. We found a pattern of errors that was qualitatively the same as observed by these authors. However, the absolute number of errors was much lower: as a group our agrammatics performed highly above chance. In a second study, we asked seven patients to order various types of sentences in which the position of the verb varied. Again the level of performance was highly above chance. There was no effect of sentence type with respect to verb position (VSO or SOV). Another effect of sentence type, however, was apparent. If the position of the verb could be determined by processing a cue that was within the clause to be ordered, the sentences elicited few errors. Sentences with out-of-clause cues were much more difficult. It seems that there is variability, both between patients and between sentences. A series of possible accounts of t...

When Heuristics Clash with Parsing Routines: ERP Evidence for Conflict Monitoring in Sentence Perception

Monitoring refers to a process of quality control designed to optimize behavioral outcome. Monitoring for action errors manifests itself in an error-related negativity in event-related potential (ERP) studies and in an increase in activity of the anterior cingulate in functional magnetic resonance imaging studies. Here we report evidence for a monitoring process in perception, in particular, language perception, manifesting itself in a late positivity in the ERP. This late positivity, the P600, appears to be triggered by a conflict between two interpretations, one delivered by the standard syntactic algorithm and one by a plausibility heuristic which combines individual word meanings in the most plausible way. To resolve this conflict, we propose that the brain reanalyzes the memory trace of the perceptual input to check for the possibility of a processing error. Thus, as in Experiment 1, when the reader is presented with semantically anomalous sentences such as, The fox that shot the poacher, full syntactic analysis indicates a semantic anomaly, whereas the word-based heuristic leads to a plausible interpretation, that of a poacher shooting a fox. That readers actually pursue such a word-based analysis is indicated by the fact that the usual ERP index of semantic anomaly, the so-called N400 effect, was absent in this case. A P600 effect appeared instead. In Experiment 2, we found that even when the word-based heuristic indicated that only part of the sentence was plausible (e.g., that the elephants pruned the trees), a P600 effect was observed and the N400 effect of semantic anomaly was absent. It thus seems that the plausibility of part of the sentence (e.g., that of pruning trees) was sufficient to create a conflict with the implausible meaning of the sentence as a whole, giving rise to a monitoring response.

Working memory constraints on syntactic processing: An electrophysiological investigation

Event-related potentials (ERPs) and reaction times (RTs) were used to study how the processing of sentences with morphosyntactic violations is constrained by working memory (WM) capacity. The available WM capacity was varied by three orthogonal manipulations: (1) syntactic complexity; (2) additional WM load; and (3) verbal WM span. The processing of the morphosyntactic violations was reflected in longer RTs in ungrammatical compared with grammatical sentences, and in an anterior negativity and a centroparietal positivity in the ERPs. While the behavioral grammaticality effect was not influenced by the WM manipulations, the ERP effects were. The amplitude of the anterior negativity was modulated by the combination of complexity and load, and by WM span. The onset of the centroparietal positivity was delayed in the high-load condition, and for the low-span group. ERPs over the course of the sentences showed a frontal negative slow wave under high WM load, largest for the low-span group. The finding that online syntactic processing is related to WM span and to additional WM load does not support the theory that there is a WM capacity specific for syntactic processing.

Syntactic Persistence in Dutch

Three experiments are reported that showed effects of “structure priming,” the tendency to repeat syntactic structure across successive sentences. These effects were demonstrated in a previously untested language, Dutch. All experiments studied spoken sentence production. Importantly, pre-experimental baselines were measured for all target structures in order to assess possible effects of frequency on the magnitude of priming effect. We obtained priming with dative sentences, including datives with medially placed prepositional phrases, a sentence type not tested before. In one experiment we obtained priming effects with passives, including passives with sentence-final verbs, which also have never been tested before. However, we failed to obtain priming effects with active transitives. A comparison with the baseline data suggested that priming is not related to baseline frequency. Further, the results allowed us to draw an empirical generalization: Structure priming is a relatively long-term event, lasting at least several trials. The implications of these findings are discussed.

On Parallelism between Production and Comprehension in Agrammatism

Publisher Summary This chapter discusses the parallelism between production and comprehension in agrammatism. Until 1916, the term agrammatism was used to refer to any disorder in the production of sentence. Since then, agrammatism has been referred to the simplification of sentence form, chiefly reflected in the omission of function words and inflectional endings, and paragrammatism to a failure to produce a correct sentence form, manifesting itself in the incorrect use of sentence form elements. On the basis of different studies, it is now taken for granted that in agrammatism, the productive deficits are always accompanied by parallel deficits in comprehension. The three most recent approaches to agrammatism have all assumed that some processing component shared by production and comprehension processes is disrupted.

A Time-Based Approach to Agrammatic Production

A time-based approach to agrammatic speech is presented. The paper consists of three parts. In the first part, the literature which deals with agrammatic comprehension as a problem of disrupted timing, that is, as a slow-down of syntactic computation and/or a rapid decay of the results of syntactic processing, is reviewed. In a second part, the hypothesis that similar timing problems cause difficulties in production as well is discussed. Two possible ways in which this can happen are described. First, slow down or rapid decay can lead to desynchronization within the process of syntactic tree formation. Second, a slow down of syntactic processing can cause asynchrony between the production of a syntactic slot and the retrieval of the proper grammatical morpheme from the mental lexicon. This hypothesis predicts that morphemes which are dependent on a relatively complex part of the syntactic tree will elicit relatively many errors. Results from the literature which seem to confirm this prediction are discussed. In the third part of the paper, the possible ways in which a patient can adapt to the reduced temporal window that would result from a timing deficit are discussed. Message simplification will reduce the size of the required temporal window and will therefore have a beneficial effect on the error rate. Restart of the computational process will profit from previously reached activation levels so that synchrony is easier to reach and error rate is reduced. Empirical work which appears to support these hypotheses is reviewed.

One or More Labels on the Bottles? Notional Concord in Dutch and French

Three experiments, the first two in Dutch and the other in French, in which subject-verb agreement errors were induced, are reported. We investigated the effects of the number of tokens in the conceptual representation of the to-be-uttered subject noun phrase (i.e. distributivity). Previous studies have failed to show an effect of this variable in English (Bock & Miller, 1991; Vigliocco, Butterworth, & Garrett, in press). However, Vigliocco, Butterworth and Semenza (1995) and Vigliocco et al. (in press) did find an effect of distributivity in Italian and Spanish. In an attempt to account for this difference across languages, three structural differences between English and Spanish/Italian have been considered: (1) richness of verbal morphology; (2) possibility of post-verbal subjects; (3) possibility of null subjects. In the present study, we tested French and Dutch, which share some but not all of these properties with Italian and Spanish. In both languages, a distributivity effect was obtained, a result...

Agrammatism, Paragrammatism and the Management of Language.

Abstract Two studies are reported in which the following theory is tested: The agrammatic sentence form which is observed in the spontaneous speech of Brocas aphasics is due to the selection of elliptical syntactic structures in which the slots for many of the closed-class words that occur in complete sentences are lacking. The selection is strategic: Its purpose is to prevent the computational overload that would result if a complete sentence form were attempted. Paragrammatic output, as observed in the spontaneous speech of Wernickes aphasics, results from a lack of such strategic adaptation - the computational overload now causes morphological errors to occur. For eight properties of normal ellipsis in German, it was predicted that they would be more characteristic of agrammatic than of paragrammatic spontaneous speech. These properties were frequent omission of function words; infrequent omission of inflections; infrequent substitution of function words or inflections; frequent use of the infinitive...

Late positivities in unusual situations

Language processing leaves a large number of differentand variable traces in the EEG. Two ERP effects how-ever are generally seen as highly reliable. Manipulationsof syntactic structure elicit a late positivity, startinground 600 ms after the critical event and lasting untilabout 800–900 ms. These positivities are elicited by allsorts of grammaticality violations, but also by sentenceambiguity and by a high degree of sentence complexity.Manipulations in the meaning domain on the other handelicit a negativity, peaking at about 400 ms after the crit-ical event. This is the N400 effect, perhaps the mostintensively studied of all ERP effects, and generally seenas an index of semantic integration, in that it reflects theease with which a word is integrated into its context, beit another word, a sentence or a piece of discourse. How-ever, this reassuring picture of a P600 effect reflectingsyntactic and an N400 effect reflecting semantic process-ing, has been brutally disturbed by a number of studiesin which semantic violations failed to elicit an N400effect but triggered a P600 effect instead. What are wegoing to make out of this confusing state of affairs? Itseems that there are two ways to go. On the one hand,we may need to think differently about the relationshipbetween syntax and semantics. This is the point of viewdefended by the authors of the two papers devoted tothis phenomenon in this issue. On the other hand, aswe will argue, the results may also indicate that we needto develop ways to think about the relationship betweenthe language system and the executive system.A first approach to the finding that semantic anomalieslead to a P600 rather than to an N400 effect is taken in thestudy by Kemmerer et al. Perhaps the semantic anomalieswhich elicited the ‘syntactic’ P600 effect in the previousstudies represent a violation of a grammatical-semanticconstraint, because they involve rearranged thematic roles,as in at breakfast the eggs would eat..., a type of sentenceemployed by Kuperberg et al. As Kemmerer et al. argue,there are other examples of such grammatical-semanticconstraints, in particular the order of descriptive prenomi-nal adjectives. Behavioural data indicate that languageusers find particular orders unacceptable. For instance,an attribute such as size should precede an attribute suchas colour. In support of this hypothesis, the authors foundthat sequences like a * brown big dog elicited a P600 effectand not an N400 effect. So the dividing lines between syn-tax and semantics are perhaps less clear than conventionallinguistic theories suggest. The fruitfulness of this approachalso became evident from earlier studies by these authorsreporting specific difficulties with grammatical-semanticconstraints on prenominal adjective order in a group ofbrain damaged patients. Unexpectedly, however, a P600effect was also observed after sequences involving twoadjectives contradicting each other, as in * a small bigdog. That both types of semantic anomaly elicit a P600effect can be explained by assuming that in both a processof restructuring takes place to make sense out of the anom-aly. By construing the sequence of the two adjectives andthe noun as a hierarchical—[Adj + [Adj + N]]—ratherthan as a flat—[Adj + Adj + N]—structure, the sequencescan be meaningfully interpreted as ‘a big dog that is brown’and ‘a small dog that is—relatively—big’ respectively.Think of the well-known movie in which Dustin Hoffmanplayed the role of Little Big Man, where little and bigcan describe the same person because ‘little’ indicates phys-ical size, whereas ‘big’ relates to character and courage.The process of restructuring makes it understandablewhy the ‘syntactic P600’ was evoked in both cases. Interest-ingly, both structures elicited not only a P600 effect but alsoa ‘reversed’ N400 effect in that the N400 amplitude wasreduced relative to the control condition. Only in the case