Memory

•       Outline of the five lectures

Lecture 1 : video

Lecture 2 : two case studies + some principles

Lecture 3 : types of memory

Lecture 4 : memory and the brain

Lecture 5 : forgetting

•       Resources for the five lectures

•       Four episodes of a talk by Eric Kandel and Thomas Jessell, on the VULA website  HIGHLY RECOMMENDED

–      Large files, view them on campus (bring headphones or earphones)

•       Classic papers on H.M. by Scoville & Milner, as well as by Corkin

•       Additional film clips etc., to be advised as we proceed

•       Books to read

•       Case study 1

Clive Wearing

–      Radio broadcaster and musician

–      Contracted encephalitis when 47 yrs old

•       Bilateral damage to hippocampus

•       Left temporal and frontal lobes also damaged

–      Resulted in severe amnesia (both retrograde and anterograde)

•       What happened to Clive Wearing?

•       Case Study 2 : Henry Molaison

•       Most studied patient in medical history

•       Bicycle accident at age 10

•       Epilepsy, at time of operation 10 times per day + one major seizure every day

•       Treated with surgery to bilaterally remove medial temporal lobes, including hippocampus

•       Operation 9/1953, 27 years old

•       H.M.

•       General knowledge intact but “stuck in time”.

–      Did not learn words introduced after 1953: “jacuzzi”, “granola”, “flower-child”

–      Also ‘lost’ the years between 16 and 27

•       Was able to form some memories

–      Initially couldn’t learn how to get to his new home. Took many years to learn his own house

–      Could learn to mirror reverse read and mirror trace

•       Case study 3

Ronald Cotton

–      Accused of raping Jennifer Thompson in 1984

–      Identified in lineup by Thompson

–      Sentenced to life imprisonment + 54 years

–      Conviction and sentence confirmed on appeal

–      After 10.5 years, DNA evidence exonerated him

•       Origins of memory - ‘plasticity’ & brain

•       Where are memories stored?

–      Lashley demonstrated memory to be diffuse, memory better seen as a global property of brain

–      More recent evidence suggests specialised areas e.g. faces, language, but still widely distributed

•       Principle is that entire brain exhibits ‘plasticity’, that is it can learn (change as a function of experience), hence remember

–      Principle is that entire brain exhibits ‘plasticity’, that is it can learn (change as a function of experience), hence remember

–      It changes physically, greatest changes being in childhood

–      by ‘change physically’ we mean among other things, i) neuronal connections strengthening, growing new neural matter e.g. synapses

•       Approaching memory as a topic

•       How do we organise our approach to memory?

–      Usually by distinguishing ‘types’ of memory

•        based on psychological experience (everyday and expert)

•       tested in the usual cognitive neuropsychological ways

–      are they dissociable (functionally and anatomically)?

–      what do scanning studies suggest?

–      Also by presumed mechanism or process/stage

–      Some classifications / typologies

•       Short term (STM) vs Long term memory (LTM)

•       STM vs Working memory (WM)

•       Sensory memory vs STM/WM vs LTM

•       Declarative memory vs procedural memory

•       Implicit vs explicit memory

•       Episodic vs semantic memory

•       Cumbersome….

•       Three stage model of memory

Encoding, Storage, Retrieval

•       Must add consolidation - something happens over time to make memories last.

•       Damage to the brain, is never so severe as to lead to total loss, usually affects more recent memories, leaves older ones less affected

STM is part of the same set of processes

•       Atkinson & Shiffrin (1968) Model of Memory

Short-term memory (STM) is a limited capacity store for information -- place to rehearse new information from sensory buffers

Items need to be rehearsed in short-term memory before entering long-term memory (LTM)

Probability of encoding in LTM directly related to time in STM

•       Short term memory

•       What is in mind, in consciousness

•       Separate stores for sounds (phonological STM), and visual information

•       It does seem to be a separate thing from LTM

–      Clive Wearing and HM have intact short term memory, cannot learn (much) new information ie. Info gets in, and they are aware of it, then lost

•       Patient KF

–      Motorcycle accident, damage to parietal/occipital cortex, has an STM of only 1 or 2 items

•       STM test

7 8 5

9 1 3 4

8 3 9 1 2

9 7 5 8 3 1

8 5 6 9 3 4 2

3 8 2 9 5 1 4 6

4 1 7 3 8 2 5 9 4

•       STM test

7 8 5

9 1 3 4

8 3 9 1 2

9 7 5 8 3 1

8 5 6 9 3 4 2

3 8 2 9 5 1 4 6

4 1 7 3 8 2 5 9 4

•       Is STM needed for long-term learning?

Atkinson & Shiffrin (1968) (modal) model says "yes"; model assumes amount of time in short-term store predicts later learning

•       Evaluating Modal Memory Model

•       Pro

–      provides good quantitative accounts of many findings

•       Contra

–      assumption that all information must go through STM is probably wrong

–      Model proposes one kind of STM but evidence suggests we have multiple kinds of STM stores

•       Baddeley’s working memory model

•       Are there multiple LTM memory systems?

•       How do you learn a new skill?

•       How do you learn a new fact?

•       How about learning about an event?

•       Is there one long-term memory (LTM) system for these types of knowledge or are there multiple LTM systems?

STM test

7 8 5

9 1 3 4

8 3 9 1 2

9 7 5 8 3 1

8 5 6 9 3 4 2

3 8 2 9 5 1 4 6

4 1 7 3 8 2 5 9 4

Is STM needed for long-term learning?

Atkinson & Shiffrin (1968) (modal) model says "yes"; model assumes amount of time in short-term store predicts later learning

Evaluating Modal Memory Model

•       Pro

–      provides good quantitative accounts of many findings

•       Contra

–      assumption that all information must go through STM is probably wrong

–      Model proposes one kind of STM but evidence suggests we have multiple kinds of STM stores

Baddeley’s working memory model

Are there multiple LTM memory systems?

How do you learn a new skill?

How do you learn a new fact?

How about learning about an event?

Is there one long-term memory (LTM) system for these types of knowledge or are there multiple LTM systems?

A taxonomy of memory systems

Memory test

Explicit test of memory:  recall

Write down the words you  remember from the list in the earlier slide

Implicit test of memory: word fragment  completion

On the next slide, you will see some words missing letters, some “word fragments” and some anagrams.  Guess what each word might be.

Implicit Memory Tasks

Word-fragment completion is an implicit memory task.
Fragments are (often) completed with words previously studied in the absence of an explicit instruction to remember the word

Amnesiacs often show spared implicit memory

                dissociation suggest different systems for implicit and explicit memory systems

Taxonomy of memory

Taxonomy of memory

Taxonomy of memory

Taxonomy of memory

Episodic memory

Memory for personally experienced events that occurred in particular place at a specific time (defined by Tulving, 1972)

Contextual, spatiotemporal, autobiographical, “remembering”

“Direct” memory tests:

                                                                                Encoding                                               Retrieval

                                Free recall                                             CAT                                        ?

                                                                                DOG                                       ?

                                Cued recall                                            CAT – DOG                                           EAGLE- ?

                                                                                EAGLE – NEST                    CAT- ?

                                Recognition                          CAT                                        CAT        Ö             

                                                                                DOG                                       SUN        X

                                Source Memory                   CAT                                        CAT        bold

                                                                                DOG                                        DOG       italics

Episodic memory - neuropsychology

Organic amnesia after damage to Medial Temporal Lobes shows

                                semantic memory (e.g, language)                                                                   working memory (e.g, <5 mins)              remain intact                    nondeclarative memory

Episodic memory - neuropsychology

Taxonomy of memory

Semantic memory

Memory for facts, general knowledge, word meanings

Acontextual: Independent of where or when the information was encoded                         

Semantic memory - neuropsychology

Taxonomy of memory

Working Memory - neuropsychology

Taxonomy of memory

Taxonomy of memory

A change in speed, accuracy or bias of processing a stimulus owing to prior exposure to that stimulus

Perceptual vs Conceptual

Example “Indirect” Memory Tests:

                Perceptual Identification – Gollin Figures    ->

Amnesics with Medial Temporal damage show intact Perceptual Priming and intact Conceptual Priming

Huntington’s patients with Basal Ganglia damage show intact priming

Alzheimer’s patients with diffuse Temporal Lobe damage show intact perceptual priming but impaired conceptual priming

Patients with right occipital lesions show no perceptual priming, but intact conceptual priming

Taxonomy of memory

Procedural memory

Skill learning   (e.g. riding a bicycle) 

Requires multiple trials

Indexed by improved accuracy or  RTs

Amnesic patients usually  intact on different kinds of tasks

Alzheimer’s patients usually  intact on different kinds of tasks

Rotary Pursuit (Gabrieli et al 1993)

Mirror Tracing (Heindel et al 1989)

Parkinson’s patients impaired on several kinds of tasks

Huntington’s patients impaired on several kinds of tasks

Cerebellar lesions impair Mirror Tracing (Sanes et al 1990)

Taxonomy of Memory

Classical Conditioning

Changes in response (R) to conditioned stimulus (CS) after repeated conditioned – unconditioned stimulus (US) pairing

Example:

                                Existing:                 e.g. air puff to eye (US) – blink reflex (R)

                                Training:                                e.g. tone in ear (CS) – air puff to eye (US)

                                Result:                          tone in ear (CS) – blink reflex (R)

                  

Taxonomy of Memory

•       Amnesia

•       The abnormal loss of material from LTM or the failure to transfer it from STM/WM to LTM

–      Not ‘normal’ forgetting, BUT gives important clues to how memory normally works

•       E.g. damage to different regions has different consequences, \ different memory systems

–      Amnesia can arise from neurosurgery (e.g. HM), strokes, head injury (KF), certain viruses (CW), or as a consequence of long-term alcoholism

•       RA and AA

•       Amnesia - a deficit in consolidation?

•       Consolidation

–      the process by which recent memories are transferred into long-term memory.

–      see work by Eric Kandel and colleagues on the neurobiology of this

•       Distinction between ‘shorter term’ long term memory and ‘longer term’ long term memory: potentiation  vs synaptic growth

–      at a brain systems level, the medial temporal lobe, and the hippocampal areas, in particular, seem vital for memory consolidation, and for ‘linking’ different parts of the brain to make up a particular memory

•       damage to this area always seems to result in the pattern seen in HM and CW – inability to lay down new declarative memories

•       Amnesia – a failure of context reinstatement?

•       Episodic memories contain rich contextual detail, are specifiable in time & place

–      recall your first day at school

•       Contextual details may be linked together by structures in the medial temporal lobe, incl. the hippocampus, and the memory gradually consolidated over time

•       Newly learned semantic facts may initially be context-dependent (appearance, function, name – to be bound together), but become less so over time, hence more resistant to damage

•       Memory biases (ordinary ‘forgetting’)

•       Memory is better for meaningful significant features than for details of language or perception

à gist is remembered better than detail

•       Reconstructive nature of memory

•       Memory is often side-effect of comprehension

–      details can be filled in or reconstructed at retrieval time

•       Constructive approach to memory:

–      Memory = actual events + knowledge, experiences, expectations

•       Verbal labels can distort visual memories

•       False memory

•       A memory that is either partly or wholly inaccurate,but is accepted as a real memory by the person doing the remembering

•       Hot air balloon ride  Wade, Garry, Read & Lindsay

Over three interviews, subjects thought about a (false) photograph showing them on a hot air balloon ride and tried to recall the event by using guided-imagery exercises.

Fifty percent of the subjects created complete or partial false memories.

•       Role of PFC in evaluating retrieved memories

•       Memories are not stored in some immutable form waiting to be recovered (like video file waiting to be replayed)

•       Brain regions, and collections of neurons retaining information about the past are also used for processing information in the present

•       This means that memories may be distorted and the process of remembering is an act of inference/evaluation (i.e. inferring what may have happened in the past based on what is presently known) – constructive memory

•       PFC (pre-frontal cortex) plays an important role in this process

•       Source monitoring is the process by which retrieved memories are attributed to their original context (e.g. seen v. imagined; told v. read)

•       closely related to the process of recollection, but involves an active evaluation process rather than directly retrieving information

•       Marcia Johnson claims that qualitative characteristics of retrieved memories are used for source monitoring

•       Patients with PFC damage can have poor source memory despite good recognition memory

•       Functional imaging comparing source monitoring with recognition memory shows bilateral DLPFC (dorsolateral prefrontal cortex) activity

•       Confabulation following PFC Damage

•       Confabulations = false and sometimes self contradictory memories that the patient believes to be real (i.e. without an intention to lie)

•       Associated with damage to different regions  (orbitofrontal & ventromedial PFC) than in classical amnesia

•       The neuropsychology of confabulation

•       Some confabulations may have an obvious relationship to events from the person’s past, whereas others are harder to account for.

•       Patients tend to believe in the false memory and even act upon it.

•       Confabulating patients vs. amnesics

–      When recalling a story, amnesic patients may forget most of the information whereas confabulating patients may embellish it or introduce new facts

•       Confabulation following PFC Damage

•       Possible explanations:

–      Failure of strategic retrieval operations (e.g. Moscovitch)

–      Failure of source monitoring (e.g. Johnson)

–      Failure to suppress irrelevant memories so that they intrude into the current retrieval attempt (Schnider)  (temporal context confusion)

•       Is there a purpose to forgetting?

•       Why (should) we have bad memory?

–      Luria (1975): Shereshevskii’s ‘virtually limitless’ memory

–

bad at inductive reasoning (‘filling in the blanks’)

– could not forget irrelevant details

–      It is efficient for our memory system to make recent and frequent memory more readily accessible

•       Algorithmic level explanations of forgetting

•       Decay

–      Memories just fade and disappear (not much evidence for this)

•       Interference

–      Memory is still there but we can’t retrieve it

–      newer memories interfere with older memories à Blocking

–      Suppression & repression

controversial (!)

               

•       Example

•       You call a friend, but realize you need an older phone number that you have not used for a while. With effort, you recall the correct old phone number

•       Blocking

•       One explanation: The old number is blocked (or inhibited) by the new association

•       Retrieval induced forgetting

•       An alternative explanation for the problem of retrieving the old phone number is that the old memory has been suppressed because the new phone number was retrieved à retrieval induced forgetting