Implicit Memory: How It Works and Why We Need It

David W. L. Wu1

1University of British Columbia

Correspondence to: [email protected]

ABSTRACT Since the discovery that amnesiacs retained certain forms of unconscious learning and memory, implicit memory research has grown immensely over the past several decades. This review discusses two of the most intriguing questions in implicit memory research: how we think it works and why it is important to human behaviour. Using priming as an example, this paper surveys how historic behavioural studies have revealed how implicit memory differs from explicit memory. More recent neuroimaging studies are also discussed. These studies explore the neural correlate of priming and have led to the formation of the sharpening, fatigue, and facilitation neural models of priming. The latter part of this review discusses why humans have evolved with highly flexible explicit memory systems while still retaining inflexible implicit memory systems. Traditionally, researchers have regarded the competitive interaction between implicit and explicit memory systems to be fundamental for essential behaviours like habit learning. However, recently documented collaborative interactions suggest that both implicit and explicit processes are potentially necessary for optimal memory and learning performance. The literature reviewed in this paper reveals how implicit memory research has and continues to challenge our understanding of learning, memory, and the brain systems that underlie them.

Keywords: implicit memory, priming, neural priming, habit learning

INTRODUCTION several days despite having no recollection of the learning trials. For a time it was Not until Scoville and Milner believed that the acquisition of motor studied the amnesic patient H.M. in 1957 skills was the only type of learning H.M. did investigations into the biological and was capable of (Corkin, 1968). It would neurological basis of memory begin in take another decade until it was earnest. Among the many contributions to discovered that other non-motor skills science that H.M. provided, one of the could be learned as well. What was known most important and fascinating was the about learning skills retained during first convincing evidence of multiple amnesia was again broadened when it was memory systems. Milner (1962) showed demonstrated that amnesiacs had the that in ten trials, H.M. was able to learn to ability to learn perceptual skills like trace a star only by using its mirror image. mirror-reading (Cohen & Squire, 1980). Impressively, he retained the skill across These early investigations of H.M. and 1

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

other amnesiacs formed the foundation habit learning becoming addiction—but for research involving the relationship also in enhancing beneficial behaviours. between memory and awareness. This For example, understanding motor paper will look into the implicit memory memory may lead to the development of research that has been conducted since optimal strategies in skill acquisition. It then, with a focus on two key questions: may also lead to better teaching strategies how it works, and why we need it. for people who do not learn effectively with conventional teaching methods, like First, an examination of the amnesiacs or Alzheimer’s patients, or even behavioural and neuroimaging evidence of an ill-behaved household pet. priming will lead to a discussion on neural models which attempt to explain how MECHANISMS OF IMPLICIT MEMORY implicit memory may work. Historic behavioural priming studies of both The term implicit memory is a amnesic and non-amnesic subjects general label which refers to the provide convincing evidence that implicit observation that past experiences can memory and explicit memory systems are modify subsequent task performance, in fact distinct. More recent neuroimaging despite the learner lacking conscious studies have discovered the neural recollection of the learning episode correlate of priming, which has led to (Schacter, Chiu & Ochsner, 1993). Many sharpening, fatigue, and facilitation neural skills are acquired and maintained models of priming. Secondly, interactions implicitly, for example typing. Most between implicit and explicit memory people do not remember when or how systems will be used as evidence in the they learned to type, nor do they need to discussion of the need for an implicit consciously monitor their typing. Implicit memory system for human behaviour. memory is thus expressed through Behaviours like habit learning require a performance as opposed to explicit competitive interaction between implicit memory, which is expressed through and explicit memory systems, but more recollection (Squire, 2009a). While the recently theorized collaborative previous example illustrated procedural interactions may be necessary in memory, the most extensively studied producing optimal learning and memory. implicit memory paradigm is priming. (Schacter et al., 1993). Implicit memory encompasses a wide range of phenomena, including Early Priming Studies procedural and motor memory, habit learning, classical conditioning, and Priming refers to the phenomenon emotional memory (Squire, 2009b). where a recent encounter with a stimulus Because of this, implicit memory research creates an improvement in the ability to is important not only because it increases identify or process a related stimulus our basic understanding of the human (Tulving & Schacter, 1990). One of the brain, but also because of it role behind a main differences between priming and wide range of behaviours. Such an other types of implicit skill learning is that understanding is not only be necessary in priming takes only takes a single or a developing solutions where such small number of learning trials, whereas behaviours become maladaptive—like skill learning requires numerous learning 2

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

trials (Schacter et al. 1993). In the 1980s, samples began to paint a more detailed visual word priming was the most widely picture of how these systems differed. studied form of priming. Some of the most popular experimental paradigms One such difference revealed by included stem or fragment completion, early priming studies was the endurance where subjects are given the first three of priming effects. Fragment completion letters or fragments of the word and tasks have been shown to persist from as choose between multiple possible long as one week (Tulving, Schacter, & completions (i.e., for___ or f_re_t Stark, 1982) to over a year (Sloman, (forest)); and word identification, where Hayman, Ohta, Law, & Tulving, 1988). subjects are shown a word for a very brief While initial studies of amnesic patients duration (i.e., 35ms) and asked to identify failed to find priming effects at long delays it (Schacter et al. 1993). (Squire, Shimamura & Graf, 1987), it was later shown that the amnesiac K.C. The key principle that emerged retained priming effects over the course of from early studies of visual word priming a year (Tulving, Hayman & MacDonald, was the fact that processes involved in 1991). These long term effects are thought priming are distinct from ones pertaining to be related to the fact that manipulations to explicit memory. Strong support for of interference that normally impair this principle came from studies of explicit memory exhibited no effect on amnesic patients who had comparable performance on fragment and stem levels of priming as non-amnesic subjects completion tasks (Graf & Schacter 1987; in stem completion tasks (Graf, Squire & Sloman et al., 1988). The endurance of Mandler, 1984), provided that the priming effects added to the growing instructions did not suggest an explicit picture of how implicit memory was strategy. It was also shown that normal different from explicit memory. subjects had similar explicit/implicit dissociations. Manipulations to depth of Lastly, visual word priming was encoding (i.e., semantic study tasks which established to be relatively specific. focus on the meaning of a word, compared Priming has little cross-modality to non-semantic tasks which focus on the capabilities, a fact evident when priming physical properties of a word) produced was diminished with auditory—as opposed equal levels of priming despite large to visual—target stimuli in stem differences in explicit memory (Graf, completion (Graf, Shimamura, & Squire, Mandler & Hayden, 1982; Graf & Mandler, 1985), and word identification 1984). Post-experiment questionnaires (Hashtroudi et al., 1988). When subjects revealed that subjects who were not aware studied mixed list of words and pictures in of the relationship between the completion a word or a picture-fragment completion task and the learning task still showed task, priming was nearly abolished in equivalent priming effects as subjects who picture completion tasks despite free recall were aware (Bowers & Schacter, 1990). of the pictures being better than words Amnesiacs provided the first piece of the (Weldon & Roediger, 1987). Bilingual puzzle that the implicit memory system subjects also showed little priming when was distinct from the explicit memory they studied words in one language but system, while studies on non-patient were tested in another (Durgunoglu & Roediger, 1987). Even changing the 3

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

typecase or font of the words can adversely studies (Buckner et al., 1995; Schacter, affect priming (Jacoby & Hayman, 1987). Alpert, Savage, Rauch & Albert, 1996), as The apparent finicky and inflexible nature well as functional magnetic resonance of implicit memory was very unlike that of imaging (fMRI) studies (Demb et al., 1995; explicit memory. Wagner, Desmond, Demb, Glover, & Gabrieli, 1997), all found decreases in These early visual word priming activity in regions relevant to the task experiments all suggested that the under consideration. For example, stem processes behind implicit memory were completion tasks decreased activity in physiologically and computationally posterior ventral occipital lobes (Squire et different than the ones responsible for al., 1992), whereas semantic priming tasks explicit memory. Contrary to explicit would decrease activity in the left inferior memory tasks, priming was uninfluenced prefrontal cortex (Demb et al., 1995). by semantic processes in depth of These neural activity reductions have been encoding manipulations, has high correlated with priming behaviour, such specificity, and can produce long term that greater reductions in prefrontal effects due to its independence from cortex activity were associated with faster interference. Furthermore, priming performance in word classification tasks occurred in both non-amnesic and (Maccotta & Buckner, 2004). This effect is amnesic subjects. The dominant view also robust, in that it has been shown to which emerged suggested implicit memory last even with intervening stimuli was a function of an entirely different (Henson, Shallice, & Dolan, 2000) and system, one which did not depend on across several days (Turennout, crucial explicit memory structures in the Bielamowicz, & Martin, 2003; Turennout, medial temporal lobe (MTL) like the Ellmore, & Martin, 2000), mimicking the hippocampus (Schacter, 1992; Schacter et behavioural results of the early al., 1993; Squire, 1992; Tulving & behavioural priming studies. Although Schacter, 1990). neural activity reductions have not been shown to last for as long as behavioural Neural Correlates of Priming priming effects, there appears to be an With advances in neuroimaging absence of studies examining such long- technology, researchers are now able to term effects. In any case, any memory look for the brain structures necessary for traces would likely be altered and shifted priming. An early positron emission with time, providing immensely tomography (PET) study by Squire et al. challenging methodological problems for (1992) found significantly decreased blood researchers with current neuroimaging flow in the area of the lingual gyrus of the techniques. posterior occipital lobe during a stem Neuroimaging studies all suggest completion task. There was also an one neural correlate for priming: increase activity in the right hippocampus. anatomically-specific decreases in However, the same authors suggested activation of brain regions responsible for some explicit recognition of familiar word any particular task, with lower-level areas stems was possible, thus ruling out showing smaller reductions than higher- hippocampal activity as a contributor to level areas (Schacter & Buckner, 1998). priming performance. Subsequent PET This effect has been termed neural 4

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

priming. Neural priming converges with a phenomenon called repetition suppression (RS), first discovered concurrently by Brown, Wilson, and Riches (1987) and Balyis and Rolls (1987) in single-cell recordings of monkey neurons. They found reduced responses to familiar stimuli compared to novel stimuli in 15-26% of neurons in the inferior temporal cortex. RS has subsequently been shown in other brain areas like the prefrontal cortex Figure 1. Models of neural priming are depicted. (A) The sharpening model predicts strengthening of initially optimally responding neurons and (Miller, Erickson, & weakening of other neurons via lateral inhibition, resulting in less neurons Desimone, 1996), and responding overall. (B) The fatigue model predicts a global decrease in firing provides a convincing rate, with initially optimally responding neurons having a greater decrease in firing rate. Output from initially weak responding neurons are now relatively mechanism for neural stronger than output from optimally responding neurons providing a priming in human brain mechanism for novelty detection. (C) The facilitation model predicts a global systems. increase in firing rate, presumably through strengthened connections.

Neural Models of Priming Three neural models have been identification in the remaining neurons. proposed to explain neural priming (see How the network is sharpened is thought Figure 1). The first, and perhaps most to be due to competitive interactions and popular, is the sharpening or tuning model lateral inhibition of neurons in the proposed by Desimone (1996) and Wiggs network (Grill-Spector, Henson, & Martin, and Martin (1998). As a stimulus is 2006). A computational neural network repeatedly presented, neurons that code model of the MTL, developed by Norman nonessential features for recognition of and O’Reilly (2003), revealed that neurons the stimulus decrease firing, thereby which respond strongest initially increase weakening and eventually eliminating their activity with subsequent repetitions connections to the rest of the neurons in while also inhibiting weaker neurons. the network representing the stimulus. As Since the weak neurons outnumber strong a result, the network becomes sparser, ones, a global decrease in activation (i.e., with a smaller neuronal population neural priming) is observed despite overall. This sharpening of the network is quicker processing. presumably advantageous because it reduces metabolic cost, and elicits The second model of neural increasingly efficient and rapid stimulus priming is termed the fatigue model (Grill- 5

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

Specotr et al., 2006). In this model, all be dependent on different brain regions, neurons which initially respond to the time, or task conditions (Grill-Spector et stimulus show proportional reductions in al., 2006). While these neural models were responding on subsequent exposures. developed to specifically explain priming, Firing decreases in the highest proportion the principles underlying these models within neurons that respond optimally to likely apply to other manifestations of the repeated stimulus, creating relatively implicit memory as well. Sharpening, enhanced sensitivity to different stimuli. fatigue, and facilitation are theorized This model provides a mechanism for principles which attempt to explain how novel detection since novel stimuli would the implicit memory system works, and create higher levels of activation in the more generally, how the brain can inferior temporal cortex and prefrontal successfully adapt to its environment and cortex (the areas most associated with modify behaviour based on prior neural priming), and via top-down experience. feedback to attentional systems, would drive the organism to automatically seek WHY HAVE IMPLICIT MEMORY? novel items (Li, Miller, & Desimone, Priming is a potentially adaptive 1993). This model however, provides no process because organisms evolved in an suitable explanation as to how reduced environment where stimuli encountered firing can result in increased performance once are likely to be encountered again, so on priming tasks. it would be beneficial for organisms to The last model of neural priming is interact more quickly and more efficiently the facilitation or accumulation model, in familiar environments (Squire, 2009a). which predicts shorter durations of neural The neural priming models discussed firing causing faster processing of above may provide satisfactory repeated stimuli (James & Gauthier, explanations as to how priming works, but 2006). Because the temporal resolution of they do not answer one of the most basic fMRI is over several seconds, shorter questions in implicit memory research: durations of firing would lead to smaller why should such processes be implicit? responses. While this model is consistent Why have such inflexible implicit memory with evidence from fMRI and behavioural systems at all? Is it simply an evolutionary priming data, it cannot account for the spandrel serving little purpose in the decreases in firing rate found in single-cell human species with such developed forms recording studies – individual neurons of explicit memory? The fact that should fire more under this model, not monkeys take several hundred trials to less. learn a simple pattern discrimination task (i.e., □ versus +) using medial temporal While all three models may have lobe (MTL) independent skill learning that their faults, they each offer unique insights humans can learn in a few trials using into how the brain can potentially change MTL dependent memorization reveals the itself due to experience. In other words, power of human explicit memory (Zola- they are potential neuroplastic Morgan & Squire, 1984). In order to mechanisms of priming. It is unlikely that explore how implicit memory contributes these models are mutually exclusive, but to behaviour, the interactions between each model may operate in parallel or may 6

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

implicit and explicit memory systems improved caudate activity increased and must first be discussed. MTL activity decreased. The study also showed a negative correlation between the Competitive Interactions between Implicit activation of the caudate and MTL, Memory and Explicit Memory suggesting a competitive interaction. Interactions between implicit and The competitive interactions explicit memory can be broadly viewed as between caudate-dependent and MTL- either competitive or collaborative. dependent learning are regarded as the Competitive interactions, as described by bases of habit learning, where Poldrack and Packard (2003), stem largely dopaminergic inputs from substantia from habit-learning tasks, which nigra and ventral tegmental area form a constitute a form of implicit memory. In reward circuit with the striatum, which the another study, Packard and McGaugh caudate nucleus is a part of (Packard & (1996) trained rats in a cross maze to Knowlton, 2002; Squire, 2009a). Such approach the west arm from the south arm habit learning mechanisms, which are for a food incentive. On days 8 and 16, the associated with extreme repetitive start location was changed to the opposite behaviours like addiction and obsessive- (north) arm and the rats were evaluated compulsive disorder, should not be seen as on whether they showed place learning maladaptive. Sherry and Shacter (1987) (i.e., approach the same west arm where suggest that multiple memory systems the food was in training), or response may have evolved due to the functional learning (i.e., make the same right turning incompatibility of different demands for a response as during training). Normal rats single memory system. Once one learns showed place learning on day 8 but how to drive a car, for example, the response learning on day 16. When place activity becomes automatic independent learning was destroyed early in training by of whether it is day or night, rain or shine. lidocaine injections to the hippocampus, When an organism encounters a predator, rats showed no preference for either arm. however, it suddenly becomes vital to Interestingly, when lidocaine was injected remember whether it was day or night, into the caudate nucleus to destroy rain or shine, in order to avoid future response learning, rats still demonstrated encounters. In habit learning, place learning. This study in particular remembering the details of a situation yielded two important observations: (1) may detract from task performance. But behaviour initially supported by the in explicit episodic memory, details may hippocampus in a task is later dominated be essential for survival. Competitive by behaviour which did not require the memory systems may be required to hippocampus, and (2) even when arbitrate between the incompatible need behaviour was dominated by the caudate to flexibly access detailed knowledge and nucleus, hippocampal information about the need for fast, automatic responses the place remained available. Support for (Poldrack et al., 2001). competitive interactions between memory systems also comes from human fMRI Collaborative Interactions between Implicit studies. Poldrack et al. (2001) found that Memory and Explicit Memory MTL activity early in a classification learning task, but as performance 7

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

An emerging view suggests implicit The implications of a collaborative and explicit memory may collaborate— interaction between implicit and explicit instead of compete—to produce behaviour. memory processes to produce optimal This view is endorsed by the component learning remain largely unknown. Did process model by Moscovitch (2008), implicit and explicit memory systems which predicts that priming can be evolve to serve fundamentally influenced by explicit memory processes incompatible functions and were later co- dependent on the MTL at an unconscious opted to facilitate and improve each level. At first it would seem that the other’s processing? Is such a collaborative priming evidence discussed previously interaction unique to humans? Do explicit would discount this view as they all memory tasks also employ resources suggested a distinct system independent of traditionally observed for implicit memory the MTL. However, implicit memory tasks processes? Recent fMRI (Turk-Browne, Yi, were designed specifically to preclude any & Chun, 2006) and use of explicit memory strategies, creating electroencephalographic studies suggest an artificial situation in which any natural that they do (Voss & Paller, 2007, 2008). use of the MTL would never be seen. If both implicit and explicit systems share both implicit and explicit processes, the Evidence for MTL activation during lines used to define and separate these implicit tasks has begun to emerge. A systems begin to vanish. Bussey and study by Sheldon and Moscovitch (2010) Saksida (2005) have suggested that found that performance on visual word looking at brain regions in terms of what priming tasks was higher when subjects computations they perform or what were able to remember the details or representations they contain would be context in which the words appeared more useful than modularising areas during the study phase, compared to based on their psychological function. words for which they could not recollect Perhaps in a similar manner, any specific information. In a separate dichotomizing memory systems based on study, Alzheimer’s patients with extensive conscious awareness may also be no MTL degeneration retained priming longer useful. The findings of effects for a perceptually based visual collaborative interactions certainly opens priming task, but exhibited impaired up numerous questions which will no performance compared to healthy controls doubt transform the way we think about (Koenig et al., 2008). Additionally, fMRI memory systems and the human brain in data has shown MTL activation in both general. perceptual (Koenig et al., 2008) and conceptual implicit learning tasks, CONCLUSION AND FUTURE DIRECTIONS particularly in the perirhinal cortex (Wang, Lazzara, Ranaganath, Knight, & Since H.M. and the early priming Yonelinas, 2010). This new evidence tasks used to investigate the newly suggests that the largely disregarded discovered implicit memory system, hippocampal activation discovered by research in implicit memory has expanded Squire et al. (1992) could potentially have rapidly, benefiting immensely from new been due to MTL processes supporting imaging techniques like fMRI. This paper implicit learning behaviour. analyzed evidence focused on the mechanism of implicit memory, and why 8

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

such a system is needed. Several neural REFERENCES models for priming which attempt to explain both historic behavioural priming Baylis, G. C. and E. T. Rolls (1987) and neural priming evidence were Responses of neurons in the inferior discussed, and competitive and temporal cortex in short term and serial collaborative interactions between implicit recognition memory tasks. and explicit memory systems showed how Experimental Brain Research 65(3), both systems contribute to behaviour. 614-622. Competitive interactions form the basis Brown, M. W et al. (1987) Neuronal for habit learning, while collaborative evidence that inferomedial temporal interactions have been found to produce cortex is more important than optimal performance on memory tasks. It hippocampus in certain processes should be noted that this paper was underlying recognition memory. Brain necessarily selective, and major fields in Research 409(1), 158-162. which implicit memory processes have Bowers, J. S. and D. L. Schacter (1990) also had substantial contributions have Implicit memory and test awareness. been passed over. Fear conditioning in Journal of Experimental Psychology: the amygdala (LeDoux, 2003), Learning, Memory, and Cognition conditioned motor responses in the 16(3), 404-416. cerebellum (Thompson & Steinmetz, Buckner, R. L. et al. (1995) Functional 2009), sleep (Stickgold, 2005), grammar anatomical studies of explicit and learning (Reber, 1967), visual search implicit memory retrieval tasks. The (Chun & Jiang, 1998) and selective Journal of Neuroscience 15(1), 12-29. attention (Vuilleumier, Schwartz, Duhoux, Bussey, T. J. and L. M. Saksida (2005) Dolan, & Driver, 2005) are just some of Object memory and perception in the the additional fields in which implicit medial temporal lobe: An alternative memory has been implicated in. approach. Current Opinion in Neurobiology 15(6), 730-737. As technology continues to Chun, M.M. and Y. Jiang (1998) improve, our knowledge of implicit Contextual cueing: Implicit learning memory processes will continue to grow. and memory of visual context guides More powerful neuroimaging and spatial attention. Cognitive Psychology computational modeling techniques will 36, 28–71. allow researchers to refine and distinguish Corkin, S. (1968) Acquisition of motor between neural models of priming, skill after medial temporal-lobe increasing our knowledge of excision. Neuropsychologia 6(3), 255- neuroplasticity and benefiting applications 265. in machine learning and artificial Cohen, N. and L. R. Squire (1980) intelligence. Further study of the Preserved learning and retention of interactions between implicit and explicit pattern-analyzing skill in amnesia: memory systems will certainly force us to Disoociation of knowing how and rethink our theoretical conceptions of knowing that. Science 210(4466), 207- memory and consciousness. The future 209. promises of new and exciting findings in Demb, J. B. et al. (1995) Semantic implicit memory research. encoding and retrieval in the left inferior prefrontal cortex: A functional 9

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

MRI study of task difficult and process Psychology: Learning, Memory, and specificity. The Journal of Neuroscience Cognition 14(4), 749-757. 15(9), 5870-5878. Henson, R. Et al. (2000) Neuroimaging Desimone, R. (1996) Neural mechanisms evidence for dissociable forms of for visual memory and their role in repetition priming. Science 287(5456), attention. Proceedings of the National 1269-1272. Academy of Sciences of the United Jacoby, L. L. and C. A. G. Hayman (1987) States of America 93(24), 13494-13499. Specific visual transfer in word Durgunoglu, A. Y. and H. L. Roediger identification. Journal of Experimental (1987) Test differences in accessing Psychology: Learning, Memory, and bilingual memory. Journal of Memory Cognition 13(3), 456-463. and Language 26(4), 377-391. James, T. W. and I. Gauthier (2005) Graf, P. and G. Mandler (1984) Activation Repetition-induced changes in BOLD makes words more accessible, but not response reflect accumulation of neural necessarily more retrievable. Journal of activity. Human Brain Mapping 27(1), Verbal Learning and Verbal Behavior 37-46. 23(5), 553-568. Koenig, P. et al. (2008) Medial temporal Graf, P. et al. (1982) Simulating amnesic lobe involvement in an implicit memory symptoms in normal subjects. Science task: Evidence of collaborating implicit 218(4578), 1243-1244. and explicit memory systems from Graf, P. and D. L. Schacter (1989) fMRI and Alzehimer’s Disease. Cerebral Unitization and grouping mediate Cortex 18(12), 2831-2843. dissociations in memory for new LeDoux, J. (2003) The emotional brain, associations. Journal of Experimental fear, and the amygdala. Cellular and Psychology: Learning, Memory and, Molecular Neurobiology 23(4-5), 727- Cognition 15(5), 930-940. 738. Graf, P. et al. (1985) Priming across Li, L.et al. (1993) The representation of modalities and priming across category stimulus familiarity in anterior inferior levels: Extending the domain of temporal cortex. Journal of preserved function in amnesia. Journal Neurophysiology 69(6), 1918-1929. of Experimental Psychology: Learning, Maccotta, L. and R. L. Buckner (2004) Memory, and Cognition 11(2), 386-396. Evidence for neural effects of repetition Graf, P. et al. (1984) The information that that directily correlate with behavioural amnesic patients do not forget. Journal priming. Journal of Cognitive of Experimental Psychology: Learning, Neuroscience, 16(9), 1625-1632. Memory, and Cognition 10(1), 164-178. Milner, B. (1962) Les troubles de la Grill-Spector, K. et al. (2006) Repetition memoire accompagnant des lesions and the brain: neural models of hippocampiques bilaterales, In P. stimulus-specific effects. Trends in Passouant, Editor, Phisiologie de Cognitive Sciences, 10(1), 14-23. l’hippocampe (257-272). Paris: Centre Hashtroudi, S. et al. (1988) Data-driven National de la Recherche Scientifique. and conceptually driven processes in Miller, E. K. et al. (1996) Neural partial-word identification and mechanisms of visual working memory recognition. Journal of Experimental in prefrontal cortex of the macaque.

10

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

The Journal of Neuroscience 16(16), Annals of The New York Academy of 5154-5164. Sciences 608, 543-571. Moscovitch, M. (2008) The hippocampus Schacter, D. L. et al. (1996) Functional as a ―stupid‖, domain-specific module: anatomical studies of explicit and Implications for theories of recent and implicit memory retrieval tasks. remote memory, and of imagination. Proceedings of the National Academy of Canada Journal of Experimental Sciences of the United States of Psychology 62(1), 62-79. America 93(1), 321-325. Norman, K. A. and R. C. O’Reilly (2003) Schacter, D. L. and R. L. Buckner (1998) Modeling hippocampal and neocortical Priming and the brain. Neuron 20(2), contributions to recognition memory: A 185-195. complementary-learning-systems Schacter, D. L. et al. (1993) Implicit approach. Psychological Review 110(4), memory: A selective review. Annual 611-646. Review Neuroscience 16, 159-182. Packard, M. G. et al. (1989) Differential Scoville, W.B. and B. Milner (1957) Loss of effects of fornix and caudate nucleus recent memory after bilateral lesions on two radial maze tasks: hippocampal lesions. Journal of Evidence for multiple memory systems. Neurology, Neurosurgery & Psychiatry Neuroscience 9, 1465-1472. 20, 11-21. Packard, M. G. and B. J. Knowlton (2002) Sheldon, S. A. M. and M. Moscovitch Learning and memory functions of the (2010) Recollective performance basal ganglia. Neuroscience 25, 563- advantages for implicit memory tasks. 593. Memory 18(7), 681-697. Packard, M. G. and J. L. McGaugh (1996) Sherry, D. F. and D. L. Schacter (1987) Inactivation of hippocampus or caudate The evolution of multiple memory nucleus with lidocaine differentially systems. Psychological Review 94(4), affects expression of place and response 439-454. learning. Neurobiology of Learning and Sloman, S. A. et al. (1988) Forgetting in Memory 65, 65-72. primed fragment completion. Journal Poldrack, R. A. et al. (2001) Interactive of Experimental Psychology: Learning, memory systems in the human brain. Memory, and Cognition 14(2), 223-239. Nature 414, 546-550. Squire L. R. (1992) Memory and the Poldrack, R. A. and M. G. Packard (2003) hippocampus: A synthesis from Competition among multiple memory findings with rats, monkeys, and systems: Converging evidence from humans. Psychological Review 99(2), animal and human brain studies. 195-231. Neuropsychologia 1497, 1-3. Squire L. R. (2009) Memory and brain Reber, A.S. (1967) Implicit learning of systems: 1969-2009. The Journal of artificial grammars. Journal of Verbal Neuroscience 29(41), 12711-12716. Learning and Verbal Behavior 6(6), Squire L. R. (2009) The legacy of patient 855-863. H.M. for neuroscience. Neuron 61(1), 6- Schacter, D. L. (1992) Perceptual 9. representation systems and implicit Squire L. R. et al. (1992) Activation of memory: Toward a Resolution of the hippocampus in normal humans: A multiple memory systems debate. functional anatomical study of memory.

11

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators

Proceedings of the National Academy of memory and their contamination of Science of the United States of America putative neural correlates of explicit 89(5), 1837-1841. memory. Learning & Memory 14, 259- Squire L. R. et al. (1987) Strength and 267. duration of priming effects in normal Voss, J. L. and K. A. Paller (2008) Brain subjects and amnesic patients. substrates of implicit and explicit Neuropsycholgia 25(1), 195-210. memory: The importance of Stickgold, R. (2005) Sleep-dependent concurrently acquired neural signals of memory consolidation. Nature 437, both memory types. Neuropsychologia 1272-1278. 46(13), 3021-3029. Thompson, R. F. and J. E. Steinmetz Vuilleumier, P. et al. (2005) Selective (2009) The role of the cerebellum in attention modulates neural substrates classical conditioning of discrete of repetition priming and ―implicit‖ behavioural responses. Neuroscience visual memory: Supressions and 162(3), 732-755. enhancements revealed by fMRI. Tulving, E. et al. (1991) Long-lasting Journal of Cognitive Neuroscience perceptual priming and semantic 17(8), 1245-1260. learning in amnesia: A case experiment. Wager, A. D. et al. (1997) Semantic Journal of Experimental Psychology: repetition priming for verbal and Learning, Memory and Cognition 17(4), pictorial knowledge: A functional MRI 595-617. study of left inferior prefrontal cortex. Tulving, E. and D. L. Schacter (1990) Journal of Cognitive Neuroscience 9(6), Priming and human memory systems. 714-726. Science, 247(4940), 301-306. Wang, W. C. et al. (2010) The medial Tulving, E. et al. (1982) Priming effects in temporal lobe supports conceptual word-fragment completion are implicit memory. Neuron 68(5), 835- independent of recognition memory. 842. Journal of Experimental Psychology: Weldon, M. S. and H. L. Roediger (1987) Learning, Memory and Cognition 8(4), Altering retrieval demands reverses the 336-342. picture superiority effect. Memory & Turennout, M. V. et al. (2003) Modulation Cognition, 15(4), 269-280. of neural activity during object naming: Wiggs, C. L. and A. Martin (1998) Effects of time and practice. Cerebral Properties and mechanisms of Cortex 13(4), 381-391. perceptual priming. Current Opinion in Turennout, M. V. et al. (2000) Long- Neurobiology 8, 227-233. lasting cortical plasticity in the object Zola-Morgan, S. and L. R. Squire (1984) naming system. Nature Neuroscience 3, Preserved learning in monkeys with 1329-1334. medial temporal lesions: Sparing of Turke-Browne, N. B. et al. (2006) Linking motor and cognitive skills. The Journal implicit and explicit memory: common of Neuroscience 4(4), 1072-1085. encoding factors and shared representations. Neuron 49(6), 917- 927. Voss, J. L. and K. A. Paller (2007) Neural correlates of conceptual implicit

12

JYI | July 2011 | Vol. 22 Issue 1  2011 The Journal of Young Investigators