Abstraction Processes in Artificial Grammar Learning

Four experiments explored the extent to which abstract knowledge may underlie subjects' performance when asked to judge the grammaticality of letter strings generated from an artificial grammar. In Experiments 1 and 2 subjects studied grammatical strings instantiated with one set of letters and were then tested on grammatical and ungrammatical strings formed either from the same or a changedletter-set.Evenwith a change ofletter-set, subjects were found to be sensitive to a variety of violations of the grammar. In Experiments 3 and 4, the critical manipulation involved the way in which the training strings were studied: an incidental learning procedure was used for some subjects, and others engaged in an explicit code-breaking task to try to learn the rules of the grammar. When strings were generated from a biconditional (Experiment 4) but not from a standard finite-state grammar (Experiment 3), grammaticality judgements for test strings were independent of their surface similarity to specific studied strings. Overall, the results suggest that transfer in this simple memory task is mediated at least to some extent by abstract knowledge.     

The relationship between the structural mere exposure effect and the implicit learning process

Three experiments are reported that investigate the relationship between the structural mere exposure effect (SMEE) and implicit learning in an artificial grammar task. Subjects were presented with stimuli generated from a finite-state grammar and were asked to memorize them. In a subsequent test phase subjects were required first to rate how much they liked novel items, and second whether or not they thought items conformed to the rules of the grammar. A small but consistent effect of grammaticality was found on subjects' liking ratings (a "structural mere exposure effect") in all three experiments, but only when encoding and testing conditions were consistent. A change in the surface representation of stimuli between encoding and test (Experiment 1), memorizing fragments of items and being tested on whole items (Experiment 2), and a mismatch of processing operations between encoding and test (Experiment 3) all removed the SMEE. In contrast, the effect of grammaticality on rule judgements remained intact in the face of all three manipulations. It is suggested that rule judgements reflect attempts to explicitly recall information about training items, whereas the SMEE can be explained in terms of an attribution of processing fluency.     

Artificial grammar learning in children: abstraction of rules or sensitivity to perceptual features?

We examined sensitivity to grammatical sequences of colors in an artificial grammar learning task in a sample of 120 children aged between 5 and 8 years. The aim of the experiment was to test whether the children would preferentially learn the specific salient features of the items they were exposed to or the rules that generated these items. The children were divided into two experimental groups (identical grammar but training items differing in their surface features) and a control group (random items). The results showed that regardless of age, participants learned the most frequent salient features of the items, as well as some kind of abstract relational information. However, the 8-year-olds presented a more complex result profile, with one of the experimental groups apparently developing sensitivity to grammatical rules. These results are discussed with reference to the main current models of implicit learning. Overall, the results provided more support for stimulus-specific processing models than for rule-based models.     

The knowledge acquired during artificial grammar learning: Testing the predictions of two connectionist models

An artificial grammar learning experiment is reported which investigated whether three types of information are learned during this kind of task: information about the positions of single letters, about fragments of training strings, and about entire training strings. Results indicate that participants primarily learned information about string fragments and, to a lesser extent, information about positions of letters. Two connectionist models, an autoassociator and a simple recurrent network (SRN), were tested on their ability to account for these results. In the autoassociator simulations, similarity of test items to entire training items had a large effect, which was at variance with the experimental results. The results of the SRN simulations almost perfectly matched the experimental ones. Received: 22 September 1998 / Accepted: 22 February 1999     

Recency effects in visual memory

A number of studies were conducted, in each of which a series of abstract visual patterns was presented, and the subject was then asked to choose which of two test items was in the list. The items contained specifiable visual features, and similarity could therefore be varied in a relatively known way. As in earlier studies by other workers with randomly generated patterns, a recency effect was obtained. However, this effect did not depend on similarity between the items in the list, or between them and an intervening activity. Such factors do in some cases affect the average level of performance, but not the magnitude of recency. Nor was recency abolished by tasks interposed between presentation and test. These findings suggest a general mechanism of short-term memory, rather than a specifically sensory one. However, the recency effect did depend on the similarity of location of items in the visual display. Thus there is some evidence for a specific sensory store, with items arriving more recently over-writing those which came earlier and which were similar in location.     

Comparison and the development of knowledge

This paper considers the role of comparison in the development of knowledge. Results show that comparing similar objects makes them appear more similar. Comparing dissimilar objects, on the other hand does not make them appear more similar, and in some circumstances may make them appear less similar. The effect of comparison on similar items was especially striking since participants judged items to be more similar after comparison even if the comparison task was to list differences between the two items. Further, this effect appears specific to comparison and does not appear to be simply due to a "fleshing out" of object representations (listing properties of two objects without comparing the objects themselves served to increase the objects' similarity regardless of whether the objects were similar or dissimilar to start). This suggests that comparison may play a special role in partitioning bits of experience into categories, sharpening categorical boundaries, and otherwise helping us create conceptual structure above and beyond that offered by the world.     

The use of control groups in artificial grammar learning

Experimenters assume that participants of an experimental group have learned an artificial grammar if they classify test items with significantly higher accuracy than does a control group without training. The validity of such a comparison, however, depends on an additivity assumption: Learning is superimposed on the action of non-specific variables—for example, repetitions of letters, which modulate the performance of the experimental group and the control group to the same extent. In two experiments we were able to show that this additivity assumption does not hold. Grammaticality classifications in control groups without training (Experiments 1 and 2) depended on non-specific features. There were no such biases in the experimental groups. Control groups with training on randomized strings (Experiment 2) showed fewer biases than did control groups without training. Furthermore, we reanalysed published research and demonstrated that earlier experiments using control groups without training had produced similar biases in control group performances, bolstering the finding that using control groups without training is methodologically unsound.     

Transfer in artificial grammar learning: the role of repetition information

In this article, the authors report 2 experiments that investigated the sources of information used in transfer and nontransfer tasks in artificial grammar learning. Multiple regression analyses indicated that 2 types of information about repeating elements were crucial for performance in both tasks: information about the repetition of adjacent elements and information about repetition of elements in the whole item. Similarity of test items to specific training items and chunk information influenced participants' judgments only in nontransfer tasks.     

The ineluctable modality of the audible: perceptual determinants of auditory verbal short-term memory

Classical cognitive accounts of verbal short-term memory (STM) invoke an abstract, phonological level of representation which, although it may be derived differently via different modalities, is itself amodal. Key evidence for this view is that serial recall of phonologically similar verbal items (e.g., the letter sounds b, c, g, and d) is worse than that of dissimilar items, regardless of modality of presentation. Here we show that the effect of such phonological similarity in STM can be fully accounted for by the joint action of articulatory similarity, leading to errors in speech planning processes, and acoustic similarity within auditorily presented lists, which modulates their perceptual organization. The results indicate that key evidence used to argue for the existence of abstract phonological representation can in fact be fully accounted for by reference to modality-specific perceptual and motor planning mechanisms.     

Abstractionist and processing accounts of implicit learning

Five experiments evaluated the contributions of rule, exemplar, fragment, and episodic knowledge in artificial grammar learning using memorization versus hypothesis-testing training tasks. Strings of letters were generated from a biconditional grammar that allows different sources of responding to be unconfounded. There was no evidence that memorization led to passive abstraction of rules or encoding of whole training exemplars. Memorizers instead used explicit fragment knowledge to identify the grammatical status of test items, although this led to chance performance. Successful hypothesis-testers classified at near-perfect levels by processing training and test stimuli according to their rule structure. The results support the episodic-processing account of implicit and explicit learning.     

Sequential learning in individuals with agrammatic aphasia: evidence from artificial grammar learning

We examined sequential learning in individuals with agrammatic aphasia (n = 12) and healthy age-matched participants (n = 12) using an artificial grammar. Artificial grammar acquisition, 24-hour retention, and the potential benefits of additional training were examined by administering an artificial grammar judgment test (1) immediately following auditory exposure-based training, (2) one day after training, and (3) after a second training session on the second day. An untrained control group (n = 12 healthy age-matched participants) completed the tests on the same time schedule. The trained healthy and aphasic groups showed greater sensitivity to the detection of grammatical items than the control group. No significant correlations between sequential learning and language abilities were observed among the aphasic participants. The results suggest that individuals with agrammatic aphasia show sequential learning, but the underlying processes involved in this learning may be different than for healthy adults.     

Impaired artificial grammar learning in agrammatism

It is often assumed that language is supported by domain-specific neural mechanisms, in part based on neuropsychological data from aphasia. If, however, language relies on domain-general mechanisms, it would be expected that deficits in non-linguistic cognitive processing should co-occur with aphasia. In this paper, we report a study of sequential learning by agrammatic aphasic patients and control participants matched for age, socio-economic status and non-verbal intelligence. Participants were first exposed to strings derived from an artificial grammar after which they were asked to classify a set of new strings, some of which were generated by the same grammar whereas others were not. Although both groups of participants performed well in the training phase of the experiment, only the control participants were able to classify novel test items better than chance. The results show that breakdown of language in agrammatic aphasia is associated with an impairment in artificial grammar learning, indicating damage to domain-general neural mechanisms subserving both language and sequential learning.     

Sources of confidence judgments in implicit cognition

Subjective reports of confidence are frequently used as a measure of awareness in a variety of fields, including artificial grammar learning. However, little is known about what information is used to make confidence judgments and whether there are any possible sources of information used to discriminate between items that are unrelated to confidence. The data reported here replicate an earlier experiment by Vokey and Brooks (1992) and show that grammaticality decisions are based on both the grammatical status of items and their similarity to study exemplars. The key finding is that confidence ratings made on a continuous scale (50%—100%) are closely related to grammaticality but are unrelated to all of the measures of similarity that were tested. By contrast, confidence ratings made on a binary scale (high vs. low) are related to both grammaticality and similarity. The data confirm an earlier finding (Tunney & Shanks, 2003) that binary confidence ratings are more sensitive to low levels of awareness than continuous ratings are and suggest that participants are conscious of all the information acquired in artificial grammar learning.     

Want to block earworms from conscious awareness?B(u)y gum!

In a recent article, Jamieson and Mewhort (2009) proposed a novel account of artificial grammar learning (AGL), which is based on a multitrace model of episodic memory, the Minerva 2 model. According to this account, test performance in AGL is based on an assessment of global similarity of the test strings to the memory traces of the training strings. In this article, simulation studies are presented, showing for three different AGL experiments that the predictions of the Minerva 2 model strikingly deviate from participants' performance. It is argued that participants' test performance is not generally based on general similarity.     

Effects of category diversity on learning,memory, and generalization

In this study, we examined the effect of within-category diversity on people's ability to learn perceptual categories, their inclination to generalize categories to novel items, and their ability to distinguish new items from old. After learning to distinguish a control category from an experimental category that was either clustered or diverse, participants performed a test of category generalization or old-new recognition. Diversity made learning more difficult, increased generalization to novel items outside the range of training items, and made it difficult to distinguish such novel items from familiar ones. Regression analyses using the generalized context model suggested that the results could be explained in terms of similarities between old and new items combined with a rescaling of the similarity space that varied according to the diversity of the training items. Participants who learned the diverse category were less sensitive to psychological distance than were the participants who learned a more clustered category.     

The subjective experience of remembering in artificial grammar learning

The question of what processes are involved in artificial grammar learning has been the subject of a great deal of debate for nearly four decades. Neuropsychological and some behavioural data have found evidence for episodic memory processes such as recollection and familiarity in artificial grammar learning. To date all of this evidence has been found using objective techniques that do not rely on subjective reports. However, recollection and familiarity are associated with distinct phenomenal states that can be measured using the well-known “remember”–“know” procedure. This paper reports three experiments in which evidence was found for recollection and familiarity-based influences in artificial grammar learning using subjective reports of the experience of remembering. Subjective reports were not related to the similarity of the test items to the study items, suggesting that they do not merely reflect levels of confidence. The data are discussed with reference to multiple process models and a modified signal-detection model.     

汉字字形关联对错误记忆的影响

采用DRM范式,研究汉字字形关联对错误记忆的影响。实验以占汉字主体的形声字为材料。学习阶段使用了两类字表:在一半字表中,各字表内汉字及相应的关键诱饵具有相同形旁;另一半字表中,各字表内汉字及相应关键诱饵具有相同声旁。再认测试中,被试对两类关键诱饵均表现出了明显的错误记忆,并且其效应大小与字表汉字相似度成正比。这些结果说明,汉字的字形关联可以诱发错误记忆现象,并且这种错误记忆效应由汉字间字形相似度决定。     

The subliminal mere exposure effect does not generalize to structurally related stimuli.

R.F. Bornstein (1994) questioned whether subliminal mere exposure effects might generalize to structurally related stimuli, thereby providing evidence for the existence of implicit learning. Two experiments examined this claim using letter string stimuli constructed according to the rules of an artificial grammar. Experiment 1 demonstrated that brief, masked exposure to grammatical strings impaired recognition but failed to produce a mere exposure effect on novel structurally related strings seen at test. Experiment 2 replicated this result but also demonstrated that a reliable mere exposure effect could be obtained, provided the same grammatical strings were presented at test. The results suggest that the structural relationship between training and test items prevents the mere exposure effect when participants are unaware of the exposure status of stimuli, and therefore provide no evidence for the existence of implicit learning.     

Artificial grammar learning by 1-year-olds leads to specific and abstract knowledge.

Four experiments used the head-turn preference procedure to assess whether infants could extract and remember information from auditory strings produced by a miniature artificial grammar. In all four experiments, infants generalized to new structure by discriminating new grammatical strings from ungrammatical ones after less than 2 min exposure to the grammar. Infants acquired specific information about the grammar as demonstrated by the ability to discriminate new grammatical strings from those with illegal endpoints (Experiment 1). Infants also discriminated new grammatical strings from those with string-internal pairwise violations (Experiments 2 and 3). Infants in Experiment 4 abstracted beyond specific word order as demonstrated by the ability to discriminate new strings produced by their training grammar from strings produced by another grammar despite a change in vocabulary between training and test. We discuss the implications of these findings for the study of language acquisition.     

Inference and classification learning of abstract coherent categories

Category learning research has primarily focused on how people learn to classify items using simple observable features. However, classification is only 1 way to learn categories. In addition, many concepts have an underlying coherence that explains the featural similarity among exemplars, such as abstract coherent concepts whose instances differ greatly on their observable features. In 3 experiments, the authors investigated how abstract coherent categories are acquired through 2 common means of category learning, classification and inference. Because inference promotes more focus on within-category information than does classification, they hypothesized that inference learning would lead to a better understanding of the underlying coherence of abstract coherent categories. All 3 experiments support this prediction.