Forgotten but not gone: savings for pictures and words in long-term memory

Five experiments examined the relearning of words, simple line-drawing pictures, and complex photographic pictures after retention intervals of 1 to 10 weeks. For those items that were neither recalled nor recognized, the identical item was relearned better than an unrelated control item, as measured by a recall test following relearning. This relearning advantage in recall held for all three classes of material and extended to the cross-modality case (i.e., picture-word and word-picture) and the same-referent case (i.e., two pictures of the same object). However, recognition tests of relearning failed to detect this same relearning advantage for apparently forgotten items. Taken together, these findings conflict with the existing account of savings. Most fundamental, the classic argument that relearning serves a trace-strengthening function is undetermined by the observed recall-recognition contrast. An alternative explanation of savings is suggested wherein relearning assists retrieval of information, thereby affecting recall in particular.     

Non-declarative sequence learning does not show savings in relearning

Researchers have utilized the savings in relearning paradigm in a variety of settings since Ebbinghaus developed the tool over a century ago. In spite of its widespread use, we do not yet understand what type(s) of memory are measurable by savings. Specifically, can savings measure both declarative and non-declarative memories? The lack of conscious recollection of the encoded material in some studies indicates that non-declarative memories may show savings effects, but as all studies to date have used declarative tasks, we cannot be certain. Here, we administer a non-declarative task and then measure savings in relearning the material declaratively. Our results show that while material outside of awareness may show savings effects, non-declarative sequence memory does not. These data highlight the important distinction between memory without awareness and non-declarative memory.     

Distinct contributions of the basolateral amygdala and the medial prefrontal cortex to learning and relearning extinction of context conditioned fear

We studied the roles of the basolateral amygdala (BLA) and the medial prefrontal cortex (mPFC) in learning and relearning to inhibit context conditioned fear (freezing) in extinction. In Experiment 1, pre-extinction BLA infusion of the NMDA receptor (NMDAr) antagonist, ifenprodil, impaired the development and retention of inhibition but post-extinction infusion spared retention. Pre-extinction infusion of the GABA(A) agonist, muscimol, depressed freezing and impaired retention as did post-extinction infusion. In Experiment 2, pre-extinction mPFC infusion of ifenprodil spared the development of inhibition whereas muscimol depressed freezing. Both impaired retention when infused pre- or post-extinction. Thus, the development of inhibition involves NMDAr activation in the BLA, whereas its consolidation involves both NMDAr activation in the mPFC and NMDAr-independent mechanisms in the BLA. In Experiment 3, BLA infusion of ifenprodil impaired relearning and retention of inhibition when infused before but did not impair retention when infused after re-extinction. BLA infusion of muscimol depressed freezing but did not impair retention when infused before or after re-extinction. In Experiment 4, mPFC infusion of ifenprodil impaired relearning when infused before re-extinction, whereas muscimol depressed responses. Both drugs impaired retention when infused into the mPFC before or after re-extinction. Thus, relearning to inhibit fear responses involves NMDAr activation in both the BLA and mPFC and consolidation of the inhibitory memory involves NMDAr activation in the mPFC. However, relearning and consolidation occur in the absence of neuronal activity within the BLA. We propose that NMDAr in the mPFC supports relearning inhibition when the BLA is inactivated.     

Long-term retention of a motor skill: Implicit sequence knowledge is not retained after a one-year delay

Experimental work shows that some motor skills are retained well over a long delay and others are not. It is possible that motor tasks that are primarily implicit are retained over long intervals, whereas those with an explicit component are subject to more rapid forgetting. To test whether implicit motor skills are retained, 24 (experienced) participants were retested after a one-year delay for their retention of an implicit motor sequence. They had been trained on the sequence in a button-pushing task. Although the experienced participants responded faster than they had a year earlier, they showed no retention of the sequence compared to 24 participants who had received no training a year earlier, and no savings in relearning the sequence. The overall speed effect may indicate that the amount of training is crucial to the degree of retention of a skill. The results indicate that it is not the case that implicit motor skills are well retained whereas those with an explicit component are not.     

Inference making and linking both require thinking: Spontaneous trait inference and spontaneous trait transference both rely on working memory capacity

Past research suggests that spontaneous trait inference (STI) and spontaneous trait transference (STT) may reflect different cognitive processes, the former being inferential and the latter associational. The present research was designed to explore whether either or both of these processes involve thinking that occupies cognitive capacity. Four studies suggest that reductions in available cognitive capacity reduce both STI and STT effects, both on measures of savings in relearning (which reflect the strength of trait associations with a person) and on trait ratings measures (which reflect the strength of trait inferences made about a person). Similar results were obtained using an individual difference measure of cognitive capacity. Although these results suggest that STI and STT are similar, in that both exhibit interference from reductions in cognitive capacity, other results, such as halo effects in trait ratings, support previous assertions that their underlying processes are distinct.     

Trace resistance and the decay of long-term memory

A theory of storage in long-term memory is described which characterizes the trace by two properties: strength and resistance. The theory makes four assumptions: (a) The rate of decay of strength equals the force on the trace divided by the resistance. (b) Force is proportional to trace strength and to the similarity of current traces to the previously established trace. (c) Resistance increases as the square root of trace age. (d) The resistance of a trace transfers completely to subsequent increments. The theory accounts for long-term retention functions over delays from 1 min to 2 yr, long retrograde amnesia, unlearning, effects of multiple learning trials and spacing of practice, and possibly recency judgments and retention functions following relearning.     

Differential effects of muscarinic receptor blockade in prelimbic cortex on acquisition and memory formation of an odor-reward task

The present experiments determined the consequences of blocking muscarinic cholinergic receptors of the prelimbic (PL) cortex in the acquisition and retention of an odor-reward associative task. Rats underwent a training test (five trials) and a 24-h retention test (two retention trials and two relearning trials). In the first experiment, rats were bilaterally infused with scopolamine (20 or 5 microg/site) prior to training. Although scopolamine rats showed acquisition equivalent to PBS-injected controls, they exhibited weakened performance in the 24-h retention test measured by number of errors. In the second experiment, rats were injected with scopolamine (20 microg/site) immediately or 1 h after training and tested 24 h later. Scopolamine rats injected immediately showed severe amnesia detected in two performance measures (errors and latencies), demonstrating deficits in retention and relearning, whereas those injected 1 h later showed good 24-h test performance, similar to controls. These results suggest that muscarinic transmission in the PL cortex is essential for early memory formation, but not for acquisition, of a rapidly learned odor discrimination task. Findings corroborate the role of acetylcholine in consolidation processes and the participation of muscarinic receptors in olfactory associative tasks.     

Contracting,equal, and expanding learning schedules: The optimal distribution of learning sessions depends on retention interval

In laboratory and applied learning experiments, researchers have extensively investigated the optimal distribution of two learning sessions (i.e., initial learning and one relearning session) for the learning of verbatim materials. However, research has not yet provided a satisfying and conclusive answer to the optimal scheduling of three learning sessions (i.e., initial learning and two relearning sessions) across educationally relevant time intervals. Should the to-be-learned material be repeated at decreasing intervals (contracting schedule), constant intervals (equal schedule), or increasing intervals (expanding schedule) between learning sessions? Different theories and memory models (e.g., study-phase retrieval theory, contextual variability theory, ACT-R, and the Multiscale Context Model) make distinct predictions about the optimal learning schedule. We discuss the extant theories and derive clear predictions from each of them. To test these predictions empirically, we conducted an experiment in which participants studied and restudied paired associates with a contracting, equal, or expanding learning schedule. Memory performance was assessed immediately, 1 day, 7 days, or 35 days later with free- and cued-recall tests. Our results revealed that the optimal learning schedule is conditional on the length of the retention interval: A contracting learning schedule was beneficial for retention intervals up to 7 days, but both equal and expanding learning schedules were better for a long retention interval of 35 days. Our findings can be accommodated best by the contextual variability theory and indicate that revisions are needed to existing memory models. Our results are practically relevant, and their implications for real-world learning are discussed.     

Contrasting effects of lateral striate and superior colliculus lesions on visual discrimination performance in rhesus monkeys.

Rhesus monkeys with lesions of lateral striate cortex, monkeys with superior colliculus lesions, and unoperated monkeys were tested for retention of a preoperatively acquired pattern discrimination. The three groups of monkeys were then tested in two-choice, color, color discrimination tests, one involving varying degrees of stimulus-response (S-R) separation and the other, administered several months later, involving various directions of S-R separations. The monkeys were also tested in a series of two-choice pattern discriminations, following each of which they were tested for relearning when the patterns were masked with bars or circles. The monkeys with lateral striate lesions were moderately retarded in retention of the pattern discrimination, whereas those with superior colliculus lesions were not. The monkeys with colliculus lesions, but not those with lateral striate lesions, were impaired in both S-R separation tests, which demonstrates that their deficit was not transient or solely due to a difficulty in shifting the gaze in one direction. The lateral striate monkeys, unlike those with colliculus lesions, were dificient in relearning discriminations between masked patterns. These findings suggest that superior colliculus and striate cortex may be involved in two different aspects of attention: respectively, shifting attention (and orientation) from one spatial locus to another and maintaining attention on fixated stimuli. Alternative interpretations of the effects of the lesions, based on their retinotopic loci, are discussed.     

Effects of Practice upon the Metabolic Cost of Mental Work

Two experiments with isopods, Armadillidium vulgare, were reported. In Experiment I, two groups of five Ss each learned to run a T maze to escape light and heat. Both original learning (OL), and relearning (RL) the next day, were to a criterion 9/10 correct. Between OL and RL one group was immobilized (I), and the other group was given the opportunity for activity (A). The I treatment resulted in greater savings on RL than did the A treatment, and it was thus concluded that A is a retroaction treatment. Experiment II was like I except that all Ss were extinguished immediately following OL, and then given either A or I until RL. In this experiment, the A treatment led to no savings, but the I treatment produced negative savings (more trials on RL relative to OL). It was concluded that the A treatment had retroactively interfered with extinction, and this result was discussed as analogous to differential spontaneous recovery in favor of the A group. It was concluded that Liberman's (4, 5) theory (that spontaneous recovery is due to the forgetting of extinction) is supported by the result of Experiment II.     

Spontaneous Trait Associations and the Case of the Superstitious Banana

We explored whether spontaneous trait transference (STT) occurs in cases where inanimate objects, rather than human bystanders, are presented in the context of trait-implicative behavior. The results provide a vivid demonstration of the associability of primed trait constructs by demonstrating STT to inanimate objects to which one does not normally attribute traits (e.g., superstitious being associated with a banana). This finding is consistent with the notion that the trait associations involved in STT are formed independent of person perception processes. The implications of this finding for the types of associative processes exposed by the popular savings in relearning methodology are discussed.     

内隐记忆研究的实验方法探索：重学──分离法及其测量效应

为客观计量初学后潜留的内隐记忆效应和探索内隐记忆叠加效应的内在机制,创用了“重学──分离法”。结果发现：“重学──分离法”能有效地测得初学后内隐记忆的保存量;重学后内隐记忆的叠加效应明显,增值为６３．４％;内隐记忆效应与年龄、学历（知识水平）之间无确定的对应关系。     

What happens when we relearn part of what we previously knew? Predictions and constraints for models of long-term memory

 Part-set relearning studies examine whether relearning a subset of previously learned items impairs or improves memory for other items in memory that are not relearned. Atkins and Murre have examined part-set relearning using multi-layer networks that learn by optimizing performance on a complete set of items. For this paper, four computer models that learn each item additively and separately were tested using the part-set relearning procedure (Hebbian network, CHARM, MINERVA 2, and SAM). Optimization models predict that part-set relearning should improve memory for items not relearned, while additive models make the opposite prediction. This distinction parallels the relative ability of these models to account for interference phenomena. Part-set relearning provides another source of evidence for choosing between optimization and additive models of long-term memory. A new study suggests that the predictions of the additive models are broadly supported. Received: 29 April 1999 / Accepted: 26 July 1999     

Sequential variables and the Kamin effect

The results of two experiments demonstrated that retention deficits occurred in a within-subjects design only when the intermediate interval test was either the first or second test of the retention of prior aversive conditioning experience; when two relearning tests preceded a 1-hr test, no performance deficits were observed. It was suggested that internal cues may serve as memory attributes of avoidance behavior during early stages of avoidance learning, while other stimuli may facilitate the retrieval of the memory of prior avoidance training with repeated testing.     

Working memory load in the initial learning phase facilitates relearning: a study of vocabulary learning

In this study, the effect of working memory load in the initial learning phase on the relearning phase was examined. Exp. 1 examined the effect of articulatory suppression in paired associative learning with a relearning method. In the learning phase, 28 participants (M age=21.8 yr., SD=2.1) learned all word-nonword associations under conditions of articulatory suppression or simple tapping. After a delay, they answered a cued recall task in a nonsuppressed condition and a relearning task in a simple tapping condition. The subjects who learned under the simple tapping condition in the learning phase required significantly more trials in the relearning phase. In Exp. 2, 28 participants (M age=20.8 yr., SD=1.5) participated, and the result was replicated. These results suggested that working memory load facilitates relearning. Results are best explained by contextual interference (Battig) rather than by the less-is-more hypothesis of Newport.     

Levels of association theory

Three levels of association theory are distinguished: (a) empirical laws relating operationally definable units; (b) theoretical concepts of association, supported by converging lines of evidence; and (c) theories assimilating concepts of association into more elaborate structures. These levels correspond roughly to stages in the evolution of association theory from the mid-19th century to the present. Ebbinghaus's major contribution as a theorist was to accomplish the transition from the first level to the second. The analysis of serial learning in terms of his conception of multiple types of associations may prove to have greater generality than has yet been realized. However, to account for many phenomena of practical and theoretical interest, this model requires augmentation by a control concept that provides a basis for organization beyond serial linkages of units.     

Developments in declarative memory

The second year of life is marked by changes in the robustness of recall memory. Both retrieval and storage processes have been implicated as the major source of age-related improvements in recall. Children 13 to 20 months of age were matched for levels of learning of laboratory events (thereby eliminating encoding as a source of developmental difference) and tested for recall after delays as long as 6 months. In Experiment 1, 16-month-olds evidenced less loss of information and more relearning than 13-month-olds. In Experiment 2, 20-month-olds evidenced less loss of information and more relearning than 16-month-olds. Patterns of performance across test trials and in relearning implicate a decline in susceptibility to storage failure as the primary source of the observed developmental trend.     

Cognitive Load Theory,Element Interactivity,and the Testing and Reverse Testing Effects

Summary: The testing effect arises when learners who are tested rather than relearning material obtain superior scores on a final test than those who relearn. Based on cognitive load theory, six experiments examined whether the effect was evident under low or high element interactivity (a measure of complexity) conditions. Students learning to write types of text were participants. In all experiments, effects on a final common test of two presentations were compared with a single presentation followed by a preliminary test. The testing effect on immediate tests was larger and more likely using lower element interactivity materials. A reverse testing effect was likely on immediate tests tapping higher element interactivity material but possibly eliminated by using a delayed test. Copyright © 2017 John Wiley & Sons, Ltd.     

The first memory drum.

For most of the 20th century, the memory drum was the standard American apparatus for memory research. The first memory drum dates from 1887 in the work of G. E. Müller and Frederich Schumann, and an illustration of their device appeared 16 years later. This device apparently was invented to resolve the difficulties created by Ebbinghaus's method. The memory drum was a kymograph used to control the display of learning materials. Cattell used something similar for his dissertation with Wundt, though not for verbal material. From Müller's writings, images of this device and kymographs contemporary to Müller's innovation, characteristics, aspects, and difficulties of the development of this drum are presented in the context of its use to resolve the difficulties created by Ebbinghaus's method. An improved device was offered by instrument makers Diederichs in G?ttingen and Zimmermann in Leipzig, both in 1894. Surprisingly, from 1903 Zimmermann offered only the original, unimproved device, mistakenly portraying it as appropriate for paired-associate learning procedures.