Accessing integrated and nonintegrated propositional structures in memory

This paper investigates the access properties associated with different propositional structures. Two memory experiments are reported, in which the underlying structures of sentences were integrated or not. Some sentences tested had the same concept repeated across the propositions (integrated), whereas other sentences had no explicit repeated arguments (non-integrated). Accessibility to the memory traces of the sentences was manipulated through the acquisition and the testing conditions. In Experiment 1, subjects received either immediate or delayed recall tests, under free or cued conditions. Integrated sentences were recalled better than nonintegrated ones under conditions of high accessibility (immediate recall or delayed cued recall). In contrast, under the low-access condition (delayed free recall), nonintegrated sentences were recalled slightly better than the integrated ones. Experiment 2 confirmed and extended the results for delayed free recall. Here again, under conditions of low sentence access, nonintegrated sentences were recalled better. These results were interpreted according to theory dealing with the lag effect in list learning.     

A propositional theory of recognition memory

This paper modifies the Anderson and Bower (1972) theory of recognition memory for words. A propositional representation is outlined for the contextual information underlying word recognition. Logical arguments are offered for preferring this representation over the undifferentiated associative representation used earlier. The propositional representation is used to interpret effects of verbal context upon recognition memory. The implications of these context effects are considered for two-process models of recall and recognition.     

Retrieval of propositional information from long-term memory

Three experiments are reported in which subjects learn propositions like A hippie is in the park. The experiments manipulate the number of such propositions involving a particular person (e.g., hippie) or a particular location (e.g., park). After learning the material, subjects are asked to judge whether particular probe propositions are from the study set. Times to make these judgments about probe propositions increase with the number of study propositions involving the person or location used in the probe proposition. A model is presented which assumes a subject simultaneously accesses memory from all concepts in a probe proposition and serially searches through all study propositions involving each concept. Search of memory terminates as soon as one search process from a concept finds the probe proposition or exhausts the study propositions attached to that concept.     

Spacing effects in picture memory

Three experiments were conducted to capitalize on the conclusion of Shaffer and Shiffrin (1972) that complex visual scenes are not rehearsed in testing the hypothesis that the effect of spacing on memory is due to rehearsal. In Experiment I, a list of vacation slides was presented in which both the number of repetitions and the spacing of repetitions were varied. Subsequent frequency judgments showed an effect of spacing much like that found using verbal materials. In Experiments II and III, effects of filled and unfilled spacing intervals were compared, and it was concluded that the spacing effect is primarily a function of the duration of the spacing interval. No evidence was found to support the notion that pictures are rehearsed. Rehearsal apparently cannot play the key role in an adequate, completely general explanation of the spacing effect.     

Organizational factors in picture recognition memory

Since humans have been shown to have a large capacity for dealing with pictures but not for more abstract forms of visual information, it follows that humans must rely on existing mental structures in picture recognition tasks. A series of experiments were designed to examine some characteristics of that structure. The results indicate that something analogous to categories exists for pictures and that when subjects view pictures they tag these categories. The decision in a recognition task that they have or have not seen a picture appears to be based on whether or not an existing category has been tagged. For example, the results indicate subjects can detect a new picture as new with a high degree of certainty. Further, the results show that subjects use information contained in or associated with the new member of the pair in a forced-choice procedure for deciding which member of the pair is old. Since other research has shown that figure-ground ratings made by one group of subjects is a reliable predictor of errors for subjects ranging in age from five to adult, there is evidence that whatever the exact nature of these categories they are not only common between individuals but are established at a very early age.     

Initial codes for order in memory

The roles of positional and associative information in preserving the order of short sequences of items were investigated either by presenting an item in the same serial position on successive trials (position transfer), or by presenting the same pair of items in adjacent but shifted positions on successive trials (association transfer). While Experiment I found that positive transfer of position information occurred after just one repetition of an item in the same serial position, Experiment II found no evidence for association transfer after a single trial. If an item is presented in the same position, or if a pair of items are presented in adjacent positions on several successive trials, position transfer effects built up more rapidly than association transfer effects (Experiments III and IV). However, the rate at which association transfer effects are established depends on the type of item used, with lists of concrete words showing rapid establishment of associations (Experiment V) but lists of digits showing only slow development of associations (Experiment IV). It was concluded that serial order may be preserved both by positional and associative codes, but the establishment of an associative code is linked to the extraction of information about the items themselves.     

Effectiveness of color in picture recognition memory

Investigations were conducted into the effectiveness of color in picture recognition memory. In the study phase, half of the pictures were presented in color and the other half in black and white. In the test phase, half were presented in the same color mode as the study pictures and the other half in the other mode. In immediate and 1-week-delayed tests, the recognition performance was highest when color pictures were used in both the study and test phases. The recall for the color mode of the study pictures, however, was not as good, even with the color pictures. This suggests that the effectiveness of color in picture recognition is not necessarily due to the memory for colors in the pictures themselves, but is probably due to the distinctiveness of features highlighted by the colors. We also found that in the recall performance for the color mode it was more difficult to detect the deletion of colors than to detect their addition.     

An auditory analog of the picture superiority effect

Previous research has found that pictures (e.g., a picture of an elephant) are remembered better than words (e.g., the word "elephant"), an empirical finding called the picture superiority effect (Paivio & Csapo. Cognitive Psychology 5(2):176-206, 1973). However, very little research has investigated such memory differences for other types of sensory stimuli (e.g. sounds or odors) and their verbal labels. Four experiments compared recall of environmental sounds (e.g., ringing) and spoken verbal labels of those sounds (e.g., "ringing"). In contrast to earlier studies that have shown no difference in recall of sounds and spoken verbal labels (Philipchalk & Rowe. Journal of Experimental Psychology 91(2):341-343, 1971; Paivio, Philipchalk, & Rowe. Memory & Cognition 3(6):586-590, 1975), the experiments reported here yielded clear evidence for an auditory analog of the picture superiority effect. Experiments 1 and 2 showed that sounds were recalled better than the verbal labels of those sounds. Experiment 2 also showed that verbal labels are recalled as well as sounds when participants imagine the sound that the word labels. Experiments 3 and 4 extended these findings to incidental-processing task paradigms and showed that the advantage of sounds over words is enhanced when participants are induced to label the sounds.     

Sensorimotor prediction and memory in object manipulation.

When people lift objects of different size but equal weight, they initially employ too much force for the large object and too little force for the small object. However, over repeated lifts of the two objects, they learn to suppress the size-weight association used to estimate force requirements and appropriately scale their lifting forces to the true and equal weights of the objects. Thus, sensorimotor memory from previous lifts comes to dominate visual size information in terms of force prediction. Here we ask whether this sensorimotor memory is transient, preserved only long enough to perform the task, or more stable. After completing an initial lift series in which they lifted equally weighted large and small objects in alternation, participants then repeated the lift series after delays of 15 minutes or 24 hours. In both cases, participants retained information about the weights of the objects and used this information to predict the appropriate fingertip forces. This preserved sensorimotor memory suggests that participants acquired internal models of the size-weight stimuli that could be used for later prediction.     

Semantic and phonetic memory codes in beginning readers

In two experiments a group (N=15) of poor beginning readers and of good readers (N=15) were auditorily presented with continuous item lists. The children were asked to indicate whether each item had occurred previously in the list. In Experiment 1, using real words, later items were related either semantically or as rhymes to earlier ones. False positives to each item type were taken as indices of memory coding and showed that good readers encoded both semantic and surface aspects of items. In contrast, poor readers made a large number of meaning-based confusions (saying “old” tohouse whenhome had been presented earlier) but almost none based on rhyme (home/comb confusions). In a second experiment, with phonetically legal nonwords as items, poor readers made a significant number of phonetic false positives, although the good reader controls made more. The results are interpreted as confirming that poor readers are relatively insensitive to surface features of language but that this weakness is most marked when sound and meaning are both available as memory codes.     

Conceptual memory for picture and prose episodes

In two experiments, subjects decided whether probes were true or false of one of some previously studied picture and prose episodes. The results indicated that response times to probes depicting explicit events were faster when they were in the same modality (picture or prose) as their episodes than when they were in the opposite modality, but response times to probes depicting implicit events were as fast when they were in the opposite modality as when they were in the same modality. These results suggest that the conceptual representations of picture and prose episodes are identical in form. The second experiment indicated that probes could be evaluated as fast from long as from short episodes. This result is consistent with the notion that retrieving information from episodes is done, not by scanning all the events of an episode, but by restricting the search only to the relevant slot in the schema into which the episode is assimilated.     

Two types of information in picture memory.

It is assumed that recognition memory for pictures is based on two types of information. The first is information about specific details in a picture. The process of encoding this type of information is identified with what N. H. Mackworth and others have termed looking at "informative areas" in pictures. The second informational component is designated as "general visual information." Two experiments were carried out investigating (a) the extent to which recognition responses to pictures are based on specific detail vs. general visual information, (b) whether the amount of specific detail information may be manipulated by varying the complexity of a target picture, and (c) the rate at which the two types of information are acquired. The results indicate that the rate of encoding specific details varies with the number of potential informative areas in a pictures and, given that a detail is encoded, memory performance is not substantially affected by target complexity, exposure time, or presence or absence of a mask.     

Coordinating visual and kinaesthetic memory codes.

Four experiments examined the Connolly & Jones (1970) model which postulates that translation between modalities in the cross-modal paradigm occurs before storage in short-term memory. In general, the results provided no support for the translation notion. Delaying until the end of the retention interval knowledge of the reproduction mode failed to produce a matching performance decrement. Subjects were able to maintain the code of original presentation through the retention interval even when they did not expect reproduction to be in this mode. In addition, the asymmetry in the cross-modal matching of visual (V) and kinaesthetic (K) information, whereby K-V performance is more accurate than V-K performance, was found to occur only under certain visual display conditions. The implications of these findings for general models of cross-modal translation were discussed.     

Working memory capacity and dual-task interference in picture naming

Researchers have found no agreement on whether dual-task interference in language performance, such as dual-task interference from tone discrimination on picture naming, reflects passive queuing or active scheduling of processes for each task. According to a passive-queuing account, while a central response-selection bottleneck is occupied by the tone discrimination task, picture naming is held in a passive queue until the bottleneck is freed. In contrast, according to an active-scheduling account, participants determine the order in which the tasks proceed, monitor progress on the tasks, suspend picture naming and hold it in working memory, and determine when to resume picture naming. Here, we report a study that assessed the relative merits of the queuing and scheduling accounts by examining whether the magnitude of dual-task interference in picture naming is associated with individual differences in the capacity of monitoring and updating of working memory representations, as assessed by the operation-span task. We observed that the updating/monitoring ability correlated with the speed of picture naming and with the magnitude of the interference from tone discrimination on picture naming. These results lend support to the active-scheduling account of dual-task interference in picture naming.     

The form of analog size information in memory

The information used to choose the larger of two objects from memory was investigated in two experiments that compared the effects of a number of variables on the performance of subjects who either were instructed to use imagery in the comparison task or were not so instructed. Subjects instructed to use imagery could perform the task more quickly if they prepared themselves with an image of one of the objects at its normal size, rather than with an image that was abnormally big or small, or no image at all. Such subjects were also subject to substantial selective interference when asked to simultaneously maintain irrelevant images of digits. In contrast, when subjects were not specifically instructed to use imagery to reach their decisions, an initial image at normal size did not produce significantly faster decisions than no image, or a large or small image congruent with the correct decision. The selective interference created by simultaneously imaging digits was reduced for subjects not told to base their size comparisons on imagery. The difficulty of the size discrimination did not interact significantly with any other variable. The results suggest that subjects, unless specifically instructed to use imagery, can compare the size of objects in memory using information more abstract than visual imagery.     

Semantic and visual memory codes in learning disabled readers

Two experiments investigated whether learning disabled readers' impaired recall is due to multiple coding deficiencies. In Experiment 1, learning disabled and skilled readers viewed nonsense pictures without names or with either relevant or irrelevant names with respect to the distinctive characteristics of the picture. Both types of names improved recall of nondisabled readers, while learning disabled readers exhibited better recall for unnamed pictures. No significant difference in recall was found between name training (relevant, irrelevant) conditions within reading groups. In Experiment 2, both reading groups participated in recall training for complex visual forms labeled with unrelated words, hierarchically related words, or without labels. A subsequent reproduction transfer task showed a facilitation in performance in skilled readers due to labeling, with learning disabled readers exhibiting better reproduction for unnamed pictures. Measures of output organization (clustering) indicated that recall is related to the development of superordinate categories. The results suggest that learning disabled children's reading difficulties are due to an inability to activate a semantic representation that interconnects visual and verbal codes.     

Phonetic analysis of speech and memory codes in beginning readers

Two experimental tasks, a speech segmentation and a short-term memory task, were presented to children who began to learn to read following either the “phonic” or the “wholeword” method. The segmentation task required the child to reverse two segments (either two phones or two syllables) in an utterance. The phonic group performed significantly better than the whole-word group in the “phonic reversal” task, but no difference appeared in the “syllable reversal” task. This indicated (1) that most children by the age of 6 years are ready to discover that speech consists of a sequence of phones and (2) that the moment at which they do it is influenced by the way they are taught to read. In the memory task, the children recalled series of visually presented items whose names either rhymed or did not. The difference in performance for the rhyming and nonrhyming series was significant in both groups. It was no greater for the phonic than for the whole-word group and was uncorrelated with the “phonic reversal” task. These results are discussed in connection with the distinction between ways of lexical access and ways of representing verbal information in short-term memory.     

Analysis of memory codes and cumulative rehearsal in observational learning

The present study examined the influence of memory codes varying in meaningfulness and retrievability and cumulative rehearsal on retention of observationally learned responses over increasing temporal intervals. Symbolic codes combining meaningfulness with retrievability produced superior memory performances, but responses were poorly retained in symbolic representations containing only one of these properties. Individual response analysis further confirmed that the more meaningful the codes incorporating retrieval guides the better are modeled responses learned and retained. Cumulative rehearsal had differential effects on memory performances depending upon the serial input position of the responses and the form into which the modeled behavior was encoded. Code rehearsals facilitated retention of early and intermediate responses which were repeated more than later ones, but this was true mainly for codes vulnerable to loss. The overall findings provide further corroborative evidence that memory performances are governed more by information coding than by associative strengthening processes.