Long-Term Memory and Amnesia Lecture Notes PDF

Long-term Memory and the Amnesias Mind, Brain, & Behaviour 1 Learning & Cognition Week 4 Lecture 2 Meredith McKague [email protected] 0 Learning Outcomes Name and describe the two broad divisions of LTM in terms of declarative and non-declarative memory. Name and describe the sub-divisio...

Long-term Memory and the Amnesias Mind, Brain, & Behaviour 1 Learning & Cognition Week 4 Lecture 2 Meredith McKague [email protected] 0 Learning Outcomes Name and describe the two broad divisions of LTM in terms of declarative and non-declarative memory. Name and describe the sub-divisions of declarative and non-declarative memory. Name and describe the two types of amnesia and how each affects the retrieval of memories from different periods of time. Describe and explain the role of the hippocampus in the consolidation of declarative memories, based on the case study of patient H.M., reported by Scoville and Milner (1957). Describe and explain the evidence from cases of amnesia that support the distinction (dissociation) between declarative and non-declarative forms of memory 1 1 2 In this lecture we will be unpacking the very last of the three memory stores proposed in Atkinson and Shiffrin’s (1968) Multi-store model of memory. As in our lecture on STM and WM, what begins as a single box (unitary store) will be unpacked into multiple boxes representing various different aspects of what we call LTM. When we use the term LTM in everyday language, we tend to speak of it as if it is a single entity. What we will come to see in today’s lecture is that LTM is the most multifaceted of all the memory stores. We will begin by breaking it down into two broad categories we will call declarative and nondeclarative forms of LTM. Then we will see how each of these broad categories is further broken down into sub-components. Finally, we will explore another very famous case of severe amnesia to illustrate both the past and future focused aspects of memory, and to provide evidence for the proposed divisions that we make between the different memory systems, particularly between declarative (hippocampal dependent) and non-declarative forms of memory. 2 Divisions of LTM: Declarative and Nondeclarative Memory 1. Declarative memory (explicit): • “Knowing what, why, where, and when” • Facts, events, locations, autobiographical knowledge • Personally experienced events • Hippocampal-dependent 2. Non-declarative memory (implicit): • “knowing how” • Motor skills (e.g., riding a bike) • Cognitive skills (e.g., reading) • Non-hippocampal dependent 3 There are multiple forms of long-term memory that differ in their information processing properties, and in the brain structures that support them. Broadly, the various forms of LTM can be categorised into two classes: declarative and non-declarative memory. Declarative memory is sometimes referred to as explicit memory and non-declarative memory is sometimes referred to as implicit memory. However, the terms explicit and implicit are better used to refer to the form of retrieval mechanism that is employed respectively in declarative and non-declarative memory. This lecture will explore declarative and non-declarative memory and we will examine the evidence gained from studies conducted with amnesic patients that shows that the two forms of memory are dissociable – that is, that they rely on different neurological networks in the brain. In short, declarative memory is dependent on the structures of the medial temporal lobes (MTL), including the hippocampus, whereas non-declarative memories can be encoded, stored and retrieved even when the hippocampus and associated areas in the MTL have been removed. Declarative memory supports the encoding, consolidation, and retrieval of knowledge that can be consciously remembered and described, or “declared,” to other people at the time of retrieval, such as memory for personally experienced events (episodic memory) and for facts and concepts (semantic memory). When you recognise someone, you rely on episodic memory to remember details about your earlier encounter––perhaps their tastes in cuisine, their name, their politics––and you are aware of the contents of your memory and their relation to your past. To launch your new conversation with them you also rely on semantic memory to retrieve knowledge of relevant concepts––say, the views of the relevant political party––and consciously use this knowledge to guide your discussion. Declarative memory depends on the medial temporal lobes. 3 Nondeclarative memory supports forms of long-term knowledge that are implicitly expressed as a change in behavior rather than as conscious remembering. We are often unaware of the operations of nondeclarative memory and how such memories shape our thoughts and actions. Thus, your ability to process the face of someone you recognize is likely facilitated (that is, ‘primed’) by your having previously processed that face—and although you most likely do not notice the change, your second perceptual processing of a face is performed more quickly than was the first. Nondeclarative memory systems support skill learning, conditioning, habit memory, and priming, and all depend on brain structures outside the medial temporal lobes. 3 Sub-divisions of Declarative Memory: Episodic and Semantic Memories • Endel Tulving proposed that declarative memory can be subdivided into the episodic and semantic memory systems. • Episodic memory: • Knowledge of personally experienced events • When/where memories • Contextualised memory • ‘Mental time travel’ • Semantic memory: • General knowledge of facts about the world and yourself • What/Why memories. • Abstract knowledge • Declarative memory is revealed through explicit memory tests 4 Declarative memory (also, explicit memory) refers to forms of long-term memory that can be consciously recollected and “declared,” or described to other people, such as memory for facts, concepts, ideas, and events. Declarative memory encompasses episodic memory, the memory of events in our own personal past, and semantic memory, our general knowledge about things in the world and their meaning. Episodic memory refers to memory for particular events, situated in space and time, as well as the underlying cognitive processes and neural mechanisms involved in remembering those events. A key ingredient of episodic memory that distinguishes it from other forms of memory is the retrieval of information regarding the spatial and/or temporal context in which the remembered event occurred. In contrast, semantic memory refers to the component of long-term memory that contains the representation of our knowledge of objects, facts and concepts as well as words and their meanings. This distinction between episodic and semantic memory was first proposed by Endel Tulving in 1972. Tulving defined episodic memory as the conscious knowledge of temporally dated, spatially located, and personally experienced events or episodes. Tulving defined semantic memory as knowledge about words and concepts, their properties, and interrelations. Episodic memory has a context: According to Tulving, when we recollect details about an event we engage in a kind of “mental time travel” to recover the contextual details associated with the event (when, where, who, etc). In contrast when we retrieve semantic knowledge, the memory is not bound to the specific context in which the knowledge was acquired. This is because we accumulate semantic knowledge across multiple experiences in a variety of contexts. 4 Tests that assess declarative memory are referred to as explicit memory tests because they require the retrieval of an explicit (consciously reportable) description of knowledge from memory. You can go deeper here https://www.sciencedirect.com/science/article/pii/B9780123705099000474 4 Non-declarative Memory • Non-declarative memory is revealed when previous experience facilitates (improves) performance on a task • The improvement in performance does not require conscious recollection of the prior learning experiences. • We get better at things with experience and practice. • We learn associations between recurring stimuli in the environment • Non-declarative memory is revealed through implicit memory tests. 5 Non-declarative memory (also, implicit memory) refers to forms of long-term memory that are expressed as a change in behaviour without conscious recollection. Tests of non-declarative memory are referred to as implicit memory tests. Such tests do not require description of the contents of memory, but rather reveal memory processes indirectly (implicitly) through observed changes in performance. For example, the gradual perfection of a motor skill reflects implicit memory. We will compare and contrast explicit and implicit forms of memory tests later in the following lecture. 5 • Procedural memory: • learning and performance of motor and cognitive skills Subdivisions of Non-declarative memory • Priming: • demonstrated by a change in the ability to identify a stimulus as the result of prior exposure to that stimulus, or a related stimulus. • Repetition priming • For example, prior exposure to a word in a lexical decision task will make that word easier to respond to next time it is encountered. • Associative/semantic priming • e.g., the prior presentation of the word “nurse” facilitates subsequent identification of the word “doctor” 6 The lexical decision task is a commonly used experimental task that records response time and accuracy data for recognising words and discriminating them from ‘nonwords’. Nonwords are madeup word-like strings of letters that are pronounceable, following the rules of a given language (e.g. English). Here’s some examples of nonwords – blump, plirf, quist, lusp, thurch, wustle, forble, sumpkit, frustish, dimble, hingort, prostinanglible. This is fun! The task requires a simple button-press response of either “yes (word)” or “no (nonword)”. A simple repetition priming effect is shown when a word is repeated some number of trials after an initial response to the same word. This priming effect can persist over many trials. The priming effect manifests as a significantly faster response time on the second occasion, controlling for the effect of practice on the task over trials. Semantic priming can be shown in lexical decision when response times are faster for words that are related in meaning to previously presented words. For example, having previously responded to the word ”doctor”, the response time to the word “nurse” is significantly faster compared to response time to nurse when it has not previously been preceded by a word that is strongly related in meaning. 6 Sub-divisions of Non-declarative Memory • Classical conditioning (associative learning): • Learning to attend to a formerly neutral stimulus because it has become associated with a meaningful stimulus. • Operant conditioning (associative learning) • Learning to produce/avoid a behaviour because it has become associated with rewarding/punishing consequences • Non-associative learning: • Habituation: learning to ignore a stimulus because it is trivial (e.g., screening out background noise). • Sensitization: Learning to attend to a potentially threatening stimulus. 7 7 The architecture of LTM 8 Kandel, E. R., Kupferman, I., and Iverson, S. 2000. Learning and Memory. In: E. R. Kandel, J. H. Schwartz, and T. M. Jessell (eds.) Principles of Neural Science, pp. 1227–1246. New York: McGraw-Hill, Fig. 62-4 8 • Deficits in memory caused by brain damage, disease, drug abuse, or psychological trauma. The Amnesias • The selective deficits in memory processes seen in cases of amnesia provide support for the proposed division between the declarative and non-declarative memory systems. 9 Much of what we know about the structure and divisions of LTM comes from studies conducted with amnesic patients. 9 Retrograde and Anterograde amnesia • Retrograde amnesia: • An inability to remember knowledge acquired before the brain injury • Usually temporally graded (see next slide). • Anterograde amnesia: • An inability to recall anything since the time of the brain injury • Inability to learn new information 10 10 The case of H.M. • Henry Gustav Molaison (1926-2008) • Removal of the medial portion of both temporal lobes, including the hippocampi, to treat epilepsy. Excellent podcast about H.M. from ABC radio’s All in the Mind here 11 Scoville, W. B., & Milner, B. (1957). Loss of recent memory after hippocampal bilateral lesions. J. Neurol. Neurosurg. Psychiatry, 20, 11-12. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC497229/ Henry Gustav Molaison (February 26, 1926 – December 2, 2008). H.M. was knocked down by a bicycle at age 9. He began to have minor seizures at age 9-10, and then major seizures from age 16. In 1953, at the age of 27, he was so incapacitated by seizures that he could not work or lead a normal life. William Scoville offered an experimental surgery that was carried out with H.M.’s approval and that of his family. This involved the removal of the medial portion of both temporal lobes, including both hippocampi. Psychologist, Brenda Milner, documented his case over the ensuing decades. The initial description of H.M.’s condition was reported in a paper written by Scoville and Milner in 1957 (now one of the most cited papers in the history of neuroscience). HM's general condition has been described as dense anterograde amnesia, as well as temporally graded retrograde amnesia . HM was unable to form new long-term memories of new events or new semantic knowledge – he lived in the past (Corkin, 2002 – see readings). Since HM did not show any memory impairment before the surgery, the removal of the medial temporal lobes can be held responsible for his memory disorder. Consequently, the medial temporal lobes can be assumed to be a major component involved in the formation of semantic and episodic long-term memories. Further evidence for this assumption has been gained by studies of other patients with lesions of their medial temporal lobe structures. In 1957, four years after H.M.'s treatment, Scoville and Millner published the first written account of his treatment and the extent of his amnesia. The paper, published in the Journal of Neurology, Neurosurgery and Psychiatry, also contained accounts of 8 psychotic patients on whom Scoville had performed surgery similar to that of H.M.'s. The paper can be found in the readings folder. Scoville, W. B. & Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions. J. Neurol. Neurosurg. Psychiat. 20: 11-21. http://www.abc.net.au/radionational/programs/allinthemind/hm---the-man-with-no-memory/5067570 11 Amnesia: The case of H.M. Scoville & Milner (1957) 12 Image from https://scienceblogs.com/neurophilosophy/2007/07/28/rememberinghenry-m https://www.sciencedirect.com/science/article/abs/pii/S1878875015004337 12 13 The quotation and image come from Scoville and Milner (1957). The observations indicate both a short period of retrograde amnesia (temporally graded) and severe antergrade amnesia. Find evidence for each from the quotation given as a means for testing your understanding. 13 Ventral (from below) surface of H.M.’s brain 14 Figure 1 | Ventral surface of H.M.’s brain. The fixed specimen was photographed after removal of the leptomeninges. Evidence of the surgical lesions in the temporal lobes is highlighted by white geometric contours (a, b). Annese, J., Schenker-Ahmed, N. M., Bartsch, H., Maechler, P., Sheh, C., Thomas, N., ... & Klaming, R. (2014). Postmortem examination of patient HM’s brain based on histological sectioning and digital 3D reconstruction. Nature communications, 5(1), https://www.nature.com/articles/ncomms4122 14 After the surgery… • Epileptic seizures controlled But…. • Temporally-graded retrograde amnesia • Memory worst for personally experienced events from years just before the operation • Severe anterograde amnesia • Could not consolidate or retrieve new episodic memories • Severely impaired ability to learn new semantic facts. • For example, H.M. did not acquire new vocabulary introduced since 1953 despite frequent exposure to radio and TV. • Normal sensory and working memory (STM) • For example, normal digit span ability 15 Tests of H.M.’s cognitive abilities revealed that his deficits were highly specific, as his intelligence and some memory functions are relatively preserved. For example, when presented with a short list of numbers and asked to remember them for 30 seconds, H.M. performs as well as those with intact medial temporal lobes. This observation indicates that STM/working memory, information that is maintained over a period of seconds or minutes, does not depend on medial temporal lobe structures. H.M. also had preserved long-term memory for general information acquired before his operation. He remembered his name and former occupation, and he retained a command of language, including vocabulary, indicating a preservation of previously acquired semantic memory. Remote episodic memories were also spared: he could recall childhood events in detail, including riding in a car with his parents when he had his first major seizure on his 16th birthday. However, even though some memory functions were preserved, up until his death in 2008, H.M. suffered from a severe anterograde amnesia, the inability consciously to remember information encountered after brain damage. Thus, although H.M could briefly retain a short list of numbers (because his STM/working memory was intact), he would immediately and completely forget them as soon as the information was no longer attended and so was lost from STM/working memory. This catastrophic forgetting reveals an inability to form, retain, and retrieve new episodic memories. In essence, H.M. was frozen in time from the time of his surgery in 1953––he was unable to update his personal life narrative because of his inability to remember his daily experiences. H.M. himself eloquently described this outcome: “Right now, I’m wondering. Have I done or said anything amiss? You see, at this moment everything looks clear to me, but what happened just before? That’s what worries me. It’s like waking from a 15 dream; I just don’t remember” (Milner, 1966). Thus, when his semantic memory was tested for phrases such as “flower child” that had entered the language after his surgery but to which he had been repeatedly exposed, H.M. did not know their meanings (he guessed that “flower child” meant “a young person who grows flowers”) (Gabrieli, Cohen, & Corkin, 1988). His anterograde amnesia applies to both episodic and semantic Knowledge. H.M. also demonstrated some retrograde amnesia, the forgetting of events that occurred before the damage to the brain. An important aspect of H.M.’s retrograde amnesia was that it was temporally graded: The closer an event had occurred to his surgery, the more likely it was to have been forgotten. In particular, he had greater difficulty remembering experiences that had occurred during the 11 years immediately preceding his surgery than in recalling more remote experiences from his childhood. This pattern of forgetting indicates that episodic memories do not permanently depend on the medial temporal lobes; if this were the case, then even H.M.’s remote memories should have been forgotten. That remote memories were retained suggests that over time some process appears to lodge information in memory so that it remains even after medial temporal lobe damage 9see upcoming slides on the consolidation process with gradual independence from the hippocampus for retrieval). Nonetheless, the pattern of preserved working memory and impaired long-term memory following H.M.’s surgery is a powerful demonstration that the medial temporal lobes are critical for the consolidation of long-term memory to the cortex. 15 16 The temporal gradient on retrograde amnesia. Typically, people suffering retrograde amnesia do not lose the memories formed over a lifetime. It is most often the case that only the memories most recently acquired before the illness/accident/surgery are affected, with more distant memories unaffected. 16 Fergus Craik’s Contention • Memory is the outcome of the normal processes of perceiving and comprehension • Craik (2020) 17 Craik, F.I. (2020). Remembering: An activity of mind and brain. Annual review of Psychology, 71, 1-24. https://www.annualreviews.org/doi/pdf/10.1146/annurev-psych-010419-051027 One other incident in the early 1970s left me pondering the adequacy of the levels account of memory. I had gone to Montreal to visit my friend Morris Moscovitch, who was spending the year in Brenda Milner’s lab. While I was talking to Morris, Brenda popped her head around the door to say hello and to apologize that she could not chat, as she was already late for a meeting. “However,” she said, “my amnesic patients have no trouble perceiving and comprehending events, they are clearly capable of processing to deep semantic levels—yet they don’t remember things. How does that fit with your theory?” “Sorry, can’t stay,” she added, leaving me to worry about the undeniable problem raised by her question. It may therefore be necessary to concede that something else plays a necessary role in memory beyond adequate depth and elaboration of processing. The most obvious candidate for this further ingredient is consolidation, a process with no cognitive correlates as far as I can tell that presumably proceeds automatically after the cognitive processes associated with depth and elaboration. In light of the many investigations carried out on the celebrated amnesic patient HM, and also in light of current studies in neuroscience, 17 it seems likely that such further processing is carried out by interactions between the hippocampus and relevant areas of the cerebral cortex. 17 Brenda Milner’s Interrogation…. • “My amnesic patients have no trouble perceiving and comprehending events, they are clearly capable of processing to deep semantic levels—yet they don’t remember things. • How does that fit with your theory?” 18 https://www.youtube.com/watch?v=JliczINA__Y http://www.psychologywizard.net/scoville--milner-ao1-ao3.html Squire, L. R. (2009). The legacy of patient HM for neuroscience. Neuron, 61(1), 6-9. https://www.sciencedirect.com/science/article/pii/S0896627308010957 18 Role of hippocampus in consolidation of declarative memories • The severe anterograde amnesia that results from removal of hippocampus bilaterally indicates that these structures must be crucial for the consolidation of new declarative information. • Craik (2020) conceded that cases like H.M. forced him to adjust his account of elaborative encoding to include not just the process of interacting meaningfully with information in working memory, but also an additional process of consolidation, mediated by the hippocampus. • Hippocampus also crucial for retrieval of consolidated episodic memories, but not for semantic memories 19 Encoded episodic memories undergo consolidation, a process that modifies these representations such that they become more stable over time and ultimately exist independently of the medial temporal lobes. Evidence for consolidation comes from the observation that H.M.’s, and other amnesic patients’, retrograde amnesia is temporally graded: following removal of the medial temporal lobes, H.M. could still recall childhood memories, but he had difficulty remembering events that happened during the years immediately preceding the surgery. The preservation of his remote episodic memories implies that older memories do not depend on the medial temporal lobes for retrieval––otherwise, access to those memories would have been lost following medial temporal damage. Rather, interactions between the medial temporal lobe and the cortex are thought to store memories outside the medial temporal lobes by slowly forming direct links between the cortical representations of the experience. 19 Learning in Amnesia: Dissociation of declarative and non-declarative memory • Anterograde amnesiacs are capable of new procedural learning (learning a new motor skill). • For example, the mirror-tracing task • Despite stating that they have never performed the task before, they show an improvement over time. • This demonstrates that procedural learning can proceed independently of the brain systems required for declarative memories. 20 Following the understanding that long-term declarative memories depend on the medial temporal lobes, and especially the hippocampi, further tests of H.M.’s memory abilities initiated a second landmark insight into the organisation of memory: the medial temporal lobes are not necessary for all types of long-term memory. Although suffering profound deficits in consolidating new episodic and semantic memories after removal of his medial temporal lobes, H.M. nevertheless was able to form and retain non-declarative long-term memories. The first evidence to this effect came in the 1960s with the observation that H.M. could acquire new motor skills at a normal rate, and that his level of long-term retention of these new skills was comparable to that of healthy controls (Corkin, 1968; Milner,1962). For example, H.M. was able to learn the skill of “mirror tracing.” Given a picture of a star drawn with a double outline, H.M. was to draw a third outline of the star between the two already present—while looking only at the reflection of his hand and the star in a mirror. This task requires the remapping of visual perception onto motor actions because of the mirror-reversed nature of the visual input. Tested over a period of days, his improvement in performance––a measure of implicit learning––was similar to that of participants with no memory deficit. H.M. Became increasingly adept each day, tracing the star more quickly and more accurately, but at the outset of each day he had no conscious recollection of ever having done it before. These observations provided a clear demonstration that declarative and non-declarative (specifically procedural) long-term memory can be distinguished (dissociated) in anterograde amnesia. Corkin, S. (1968). Acquisition of motor skill after bilateral medial temporal-lobe 20 excision. Neuropsychologia, 6(3), 255-265. Milner, B. (1962). Les troubles de la memoire accompagnant des lesions hippocampiques bilaterales. Physiologie de l’hippocampe, 257-272. 20 Learning in Amnesia: Dissociating explicit and implicit memory 21 H.M. shows improvement on tasks involving the learning of skilled movements. The graphs plot, over three days, the number of times during each attempt that H.M. strayed outside the outlines as he drew the star. As with neurologically healthy participants, H.M. improved considerably across the multiple attempts, but he had no conscious recollection that he had ever performed the task before. (Image adapted from Squire & Kandel (2000). Memory: From Mind to Molecules. W.H. Freeman and Company, New York (pp. 13). 21 Preserved nondeclarative memory in anterograde amnesia • Further studies show that patients with anterograde amnesia show other types of preserved non-declarative memories • Intact classical and operant conditioning • Intact priming effects. • Normal habituation and sensitisation • Also, patients with Korsakoff’s syndrome, depressed patients undergoing bilateral ECT, patients with anoxic encephalopathy, who all suffer anterograde amnesia. 22 See, Graf, Squire, & Mandler (1984). The information that amnesic patients do not forget. Journal of Experimental Psychology: Learning, Memory, and Cognition, Vol 10(1), 164-178. https://psycnet.apa.org/record/1984-18111-001 22