Relational information in sentence memory

This paper is concerned with the coding and retrieval of relational or case information in sentences. Two experiments test the strong dependence hypothesis that relational information is checked in a secondary stage following the retrieval of facts from memory, where facts are retrieved on the basis of concept-overlap alone. In particular, memory for agent and object relational information in simple reversible sentences was investigated. The first experiment showed that relational information affected the reaction time and accuracy of the falsification of sentences that could have been rejected on the basis of content information alone. The second experiment approached the same issues using a recall procedure. It demonstrated the existence of some relational memory, even in the absence of any content recall. Both sets of data argued against the dependence hypothesis, but were compatible with the hypothesis that relational information is neither dependent on, nor secondary to, content information. The current findings and previous results concerning the impact of relational distinctions on interference effects are discussed in the context of the desirable properties of information retrieval systems.     

Individual differences in sentence memory

Results from an experiment with two parts are presented in this paper. In part one, participants listened to sentences containing two, three, four, or five clauses, and were asked questions about the content of the sentences. The results of part one demonstrate that an important unit of representation in sentence memory is the clause, and not some other component of discourse structure. In part two, the same group of participants performed eight different short-term storage/working memory tasks. A composite complex span score was computed for each participant based on three working memory tasks closely based on Daneman & Carpenter's (1980) reading span task. This working memory measure was significantly correlated with the participants' performance on the sentence memory task in part one. A second working memory measure—N-back—was also significantly correlated with the participants' performance on the sentence memory task, and there was no correlation between their performance on the complex span task and the N-back task. It is therefore concluded that (i) working memory consists of a number of dissociable components; and (ii) memory for sentences taps into more than one of these working memory components. Furthermore, the high correlations of sentence memory with the complex span and the N-back tasks (neither of which are language processing tasks) suggests that memory for sentences is not simply a result of linguistic experience; rather, it is likely that an independent working memory component contributes to participants' performance on the sentence memory task.     

Storage and retrieval processes in sentence memory

Two experiments investigated how much time Ss take to verify a sentence with respect to a set of subject-verb-object (SVO) propositions they had learned. Three triplets of SVOs provided differing degrees of equivocation, including (1) unique SVOs, (2) a V with 3 SOs, and (3) an SV with 3 Os. A general class of models was proposed regarding the representation of prepositional trees in memory and an algorithm by which a probe tree might be verified against a set of trees in memory. In Expt I, verification times were least for Condition (1), slower and equal for Conditions (2) and (3). In Expt II, a fourth condition, an SO with 3 Vs, produced verification times as slow as those in Condition (3). Of several models considered, the closest-fitting one assumed parallel search processes proceeding simultaneously from S, V, and O positions of the probe, scanning through memory trees structered according to the HAM (Human Associative Memory) theory of Anderson and Bower.     

Contextual imagery in meaning and memory

Three hundred homonyms were selected, and sentence fragments were written to emphasize two meanings of each. The words were rated on image-arousing capacity both in and out of context and on frequency of occurrence in context. Imagery values for the words out of context were predicted quite well by an average of the contextual imagery ratings, weighted by their relative frequencies. The finding is consistent with the hypothesis that words presented in isolation are interpreted in specific senses according to a frequency bias, with imagery ratings reflecting those senses. In a memory experiment. words were selected to vary orthogonally on both contextual and out-of-context imagery. Recall of words in isolation was a function of out-of-context imagery, while recall of words presented in context was a function of contextual imagery, further supporting the hypothesis.     

Verbal working memory and sentence comprehension

This target article discusses the verbal working memory system used in sentence comprehension. We review the concept of working memory as a short-duration system in which small amounts of information are simultaneously stored and manipulated in the service of accomplishing a task. We summarize the argument that syntactic processing in sentence comprehension requires such a storage and computational system. We then ask whether the working memory system used in syntactic processing is the same as that used in verbally mediated tasks that involve conscious controlled processing. Evidence is brought to bear from various sources: the relationship between individual differences in working memory and individual differences in the efficiency of syntactic processing; the effect of concurrent verbal memory load on syntactic processing; and syntactic processing in patients with poor short-term memory, patients with poor working memory, and patients with aphasia. Experimental results from these normal subjects and patients with various brain lesions converge on the conclusion that there is a specialization in the verbal working memory system for assigning the syntactic structure of a sentence and using that structure in determining sentence meaning that is separate from the working memory system underlying the use of sentence meaning to accomplish other functions. We present a theory of the divisions of the verbal working memory system and suggestions regarding its neural basis.     

Mental imagery and pictorial memory

Previous research has failed to show consistent effects upon memory performance of individual differences in the use of mental imagery. A test of pictorial memory was devised which varied the ease with which the stimulus patterns could be verbally encoded. A factor analysis related performance in this task to a variety of other individual characteristics. The recall of material which could not be readily verbalized defined a factor within the solution which showed significant loadings of age and performance in the Progressive Matrices, but not of subjective ratings of the vividness of experienced imagery. The recall of material which could be readily verbalized did not contribute to the factorial structure of the results. It was concluded that the two sorts of pictorial material are remembered in quite different ways.     

Working memory and planning during sentence production

Speakers retrieve conceptual, syntactic and lexical information in advance of articulation during sentence production. What type of working memory (WM) store is used to hold the planned information before speaking? To address this question, we measured onset latencies when subjects produced sentences that began with either a complex or a simple initial noun phrase, while holding semantic, phonological or spatial information in WM. Although we found that subjects had longer onset latencies for sentences beginning with a complex noun phrase, showing a phrasal scope of planning, the magnitude of this complexity effect was not affected by any type of WM load. However, subjects made more syntactic errors (but not lexical errors) for sentences beginning with a complex noun phrase, suggesting that advance planning for these phrases occurs at a syntactic rather than lexical–semantic level, which may account for the lack of effect with various types of WM load in the current study.     

Working memory components in written sentence generation

College students wrote either simple or complex sentences using 2 prompt nouns while components of working memory were distracted with a concurrent task. Loads on the visual and spatial components of working memory (retain a shape) and verbal component (retain 3 or 6 digits) were compared with a no-load control. Only the 6-digit load reliably reduced sentence length relative to the control, suggesting that unimpeded sentence generation requires verbal working memory. The sentence length effect may arise from a failure to retrieve and maintain lexical representations during grammatical encoding. Memory load had no effect on grammatical and spelling errors, implying that syntactic and orthographic processing were undisturbed. Other possibilities locate the difficulty in planning conceptual content or in phonological encoding, but some evidence speaks against them.     

Imagery effects and semantic similarity in sentence recognition memory

The idea that subjects often use imagery to discriminate semantically similar sentences was tested in three experiments. In the first experiment, subjects heard subject-verb-object sentences in the context of either a comprehension task or an image-generation task. Their memory for the sentences was tested using a two-alternative forced-choice recognition test in which different types of distractor sentence were used. A sentence semantically similar to the target sentence was one type; a sentence with the same subject and object nouns as the target sentence, but dissimilar in meaning, was another type; and a sentence similar in meaning to one of the stimulus sentences, but not to the target sentence, was a third type. The results showed that the image-generation instructions enhanced later recognition performance, but only for semantically similar test items. A second experiment showed that this finding only holds for high-imagery sentences containing concrete noun concepts. A third experiment demonstrated that the enhanced recognition performance could not be accounted for in terms of a semantic model of test-item discrimination. Collectively, the results were interpreted as providing evidence for the notion that subjects discriminate the semantically similar test items by elaborating the sentence encoding through image processing.     

Modality of presentation and blocking in sentence recognition memory

In contrast to the findings of Bransford and Franks (1971) that no specific information is retained about acquisition sentences in their recognition memory paradigm, it was shown that subjects can consistently distinguish OLDs from NEWs when visual presentation is used, but not when auditory presentation is used. This is in agreement with the speculation of Katz and Gruenewald (1974) as to why such a split was observed in their study. This difference also has a bearing on the typical finding of no modality effects in long-term memory, as compared to the consistent superiority of auditory presentation in short-term memory. It was also found that blocking semantically related sentences during acquisition heightened the perceived "OLDness" of test sentences compared to nonblocked presentation.     

Computational principles of working memory in sentence comprehension

Understanding a sentence requires a working memory of the partial products of comprehension, so that linguistic relations between temporally distal parts of the sentence can be rapidly computed. We describe an emerging theoretical framework for this working memory system that incorporates several independently motivated principles of memory: a sharply limited attentional focus, rapid retrieval of item (but not order) information subject to interference from similar items, and activation decay (forgetting over time). A computational model embodying these principles provides an explanation of the functional capacities and severe limitations of human processing, as well as accounts of reading times. The broad implication is that the detailed nature of cross-linguistic sentence processing emerges from the interaction of general principles of human memory with the specialized task of language comprehension.     

The role of the verb in sentence memory

Investigators have frequently found that nouns are more important than verbs in the memory representation of sentences. This pattern of results is incompatible with recent theories of sentence memory that have emphasized the verb and with the results of sentence comprehension studies that have demonstrated the importance of the verb. The four experiments reported here were designed to reexamine the role of the verb in sentence memory. Linguistic variables other than sentence function were held constant in three studies. Verbs were recalled as well and were as effective retrieval cues as nouns in these experiments. The to-be-remembered sentences were presented in the context of paragraphs in the fourth experiment. This reduced the difference in level of recall between subjects, verbs, and objects. The conclusion was drawn that the verb is no less important than sentence nouns in the memory representation.     

Working memory and the vividness of imagery

The working memory framework was used to investigate the factors determining the phenomenological vividness of images. Participants rated the vividness of visual or auditory images under control conditions or while performing tasks that differentially disrupted the visuospatial sketchpad and phonological loop subsystems of working memory. In Experiments 1, 2, and 6, participants imaged recently presented novel visual patterns and sequences of tones; ratings of vividness showed the predicted interaction between stimulus modality and concurrent task. The images in experiments 3, 4, 5, and 6 were based on long-term memory (LTM). They also showed an image modality by task interaction, with a clear effect of LTM variables (meaningfulness, activity, bizarreness, and stimulus familiarity), implicating both working memory and LTM in the experience of vividness.     

SenseCam, imagery and bias in memory for wellbeing

Identifying and modifying the negative interpretation bias that characterises depression is central to successful treatment. While accumulating evidence indicates that mental imagery is particularly effective in the modification of emotional bias, this research typically incorporates static and unrelated ambiguous stimuli. SenseCam technology, and the resulting video-like footage, offers an opportunity to produce training stimuli that are dynamic and self-relevant. Here participants experienced several ambiguous tasks and subsequently viewed SenseCam footage of the same tasks, paired with negative or positive captions. Participants were trained to use mental imagery to inter-relate SenseCam footage and captions. Participants reported increased levels of happy mood, reduced levels of sad mood, and increased task enjoyment following SenseCam review with positive versus negative captions. This shift in emotional bias was also evident at 24-hour follow-up, as participants recollected greater task enjoyment for those tasks previously paired with positive captions. Mental imagery appears to play an important role in this process. These preliminary results indicate that in healthy volunteers, SenseCam can be used within a bias modification paradigm to shift mood and memory for wellbeing associated with performing everyday activities. Further refinements are necessary before similar methods can be applied to individuals suffering from subclinical and clinical depression.     

Self-rated imagery and encoding strategies in visual memory.

The value of self-rated vividness of imagery in predicting performance was investigated, taking into account the mnemonic strategies utilized among subjects performing a visual-memory task. Subjects classified as 'good' or 'poor' imagers, according to their scores in the Vividness of Visual Imagery Questionnaire (VVIQ; Marks, 1972), were to detect as rapidly as possible differences between pairs of similar pictures presented consecutively. No coding instructions were given and the mnemonic strategies used were analysed by studying subjective reports and objective performance measurements. The results indicated that the subjects utilized two main strategies--a detail or an image strategy. The detail strategy was the more efficient. In accordance with a previous study (Berger & Gaunitz, 1977), it was found that the VVIQ did not discriminate between performance by 'good' and 'poor' imagers. However, among subjects who used the image strategy, 'good' imagers performed more rapidly than 'poor' imagers. Self-rated imagery may then have some value in predicting performance among individuals shown to have utilized an image strategy.     

Test anxiety,sentence comprehension,and recognition memory

Abstract

The present study explored the relationships between trait and state test anxiety of students, and comprehension and recognition memory for sentences. The results demonstrated that test anxiety effects could be found in comprehension-based recognition memory for sentences under conditions that minimized complex retrieval operations. Comprehension was related to the test anxiety deficit, supporting the hypothesis that anxiety-recognition relations were mediated by anxiety-comprehension and comprehension-recognition relations. However, a partial correlation analysis suggested that this comprehension-deficit hypothesis is not adequate to explain the total deficit. Elaborative processes were suggested as a potential contributor to the findings.     

Effects of elaboration and relational distinctiveness on sentence memory

Memory for normal and distinct target sentences in elaborated and isolated contexts was examined in two experiments. Distinctiveness was operationalized as the uniqueness of the stated relation among the elements (words) of the sentence. In both experiments distinct target sentences were recalled significantly better than normal target sentences. Robust elaboration effects emerged only in Experiment 2, however, when care was taken to construct elaborations that were causally related to the target sentences. Further, the positive mnemonic effects of elaboration in Experiment 2 combined additively with the distinctiveness effects. These results demonstrate that text manipulations emphasizing both relational distinctiveness and causal elaboration facilitate memory performance. Several possible theoretical mechanisms underlying this facilitation are outlined.     

Short-term memory and sentence processing: Evidence from neuropsychology

Traditional models of memory assume that short-term memory, as measured by memory span, plays an important role in linguistic processing and the learning ofverbal information. Contradicting this view are findings from a brain-damaged patient, E.A., who, despite a verbal memory span of about two items, demonstrated normal sentence comprehension in a variety oftasks. She was, however, impaired whenever verbatim phonological information had to be maintained or learned. These results and those from other patients with reduced span suggest that the phonological storage capacity that is critical to memory span plays only a limited role in language processing, specifically in the maintenance and learning of phonological forms. Implications for models of short-term memory are discussed. It is argued that short-termmemory should be seen as deriving from the processing and retentive capacities of language processing modules, with span tasks drawing on only a subset of these modules.     

The role of working memory in inferential sentence comprehension

Existing literature on inference making is large and varied. Trabasso and Magliano (Discourse Process 21(3):255–287, 1996) proposed the existence of three types of inferences: explicative, associative and predictive. In addition, the authors suggested that these inferences were related to working memory (WM). In the present experiment, we investigated whether WM capacity plays a role in our ability to answer comprehension sentences that require text information based on these types of inferences. Participants with high and low WM span read two narratives with four paragraphs each. After each paragraph was read, they were presented with four true/false comprehension sentences. One required verbatim information and the other three implied explicative, associative and predictive inferential information. Results demonstrated that only the explicative and predictive comprehension sentences required WM: participants with high verbal WM were more accurate in giving explanations and also faster at making predictions relative to participants with low verbal WM span; in contrast, no WM differences were found in the associative comprehension sentences. These results are interpreted in terms of the causal nature underlying these types of inferences.