Consolidation & Forgetting Flashcards
Consolidation definition
Integrating new information with old information
Forgetting definition
Information loss
When we forget something, is the information lost?
- Info may be lost - memory trace itself is poor/deteriorating - PASSIVE forgetting - problem with storage
- Info may be inaccessible - lack of retrieval cues / interference - problem with retrieval - PASSIVE forgetting
- Info may be inhibited - ACTIVE forgetting
Forgetting is adaptive
Brain damage resulting in inability to form new memories
Damage to the hippocampus and other MTL structures
e.g. case study of HM
What is the hippocampus critical for?
Critical for the acquisition of declarative memories (opposite to procedural) e.g. episodic (first day of school) or semantic (general knowledge facts; capital of France)
Amnesics - can’t form new memories
Normal controls - when activation in MTL structures on brain scan - that can predict whether info will be later recalled
For long term memory; consolidation process
Hippocampus –> neocortex
First step of consolidation
Hippocampus binds information together to form a higher level representation
Second step of consolidation
This is stored and distributed across networks in other cortical areas and INTEGRATES with old information
What is knowledge?
Structured/organised information
What is the Complementary Learning Systems model?
Memories are first stored via synaptic changes in the hippocampal system - these changes support reinstatement of recent memories in the neocortex
Neocortical synapses change a little on each reinstatement, and that remote memory is based on accumulated neocortical changes. Models that learn via changes to connections help explain this organization.
These models discover the structure in ensembles of items if learning of each item is gradual and interleaved with learning about other items.
This suggests that the neocortex learns SLOWLY to discover the structure in ensembles of experiences.
The hippocampal system permits RAPID learning of new items without disrupting this structure, and reinstatement of new memories interleaves them with others to integrate them into structured neocortical memory systems.
Damage to the hippocampus
Damage to the hippocampal system disrupts recent memory but leaves remote memory intact
Catastrophic interference
New information can interfere and replace old information
First step to complementary systems model
Fast, initial episodic learning mediated by the hippocampus (vulnerable memories)
Second step to complementary systems model
Slow, long-term storage in NEOCORTEX involving integration with existing knowledge (integration reduces interference)
Less vulnerable memories, as they are being embedded
Explanation of consolidation process
Slow neocortical learning process that “stabilises” or “fixes” memories in LTM across hours/days
Retrograde amnesia
Retrograde amnesia is a form of amnesia where someone is unable to recall events that occurred before the development of the amnesia, even though they may be able to encode and memorise new things that occur after the onset - can’t recall recently formed memories as well due to disrupted consolidation process
Retrograde amnesia usually follows damage to areas of the brain other than the hippocampus (the part of the brain involved in encoding new memories), because already existing long-term memories are stored in the neurons and synapses of various different brain regions
Why does consolidation happen in sleep?
Memories are transferred/laid down in neocortex in sleep due to minimal external input at that time
Sleep and learning
Particularly for procedural (non-declarative)
also affects declarative (episodic and semantic)
Early sleep - important for?
Early sleep is slow wave sleep
Important for consolidation of HIPPOCAMPUS-dependent declarative learning
Late/REM sleep - important for?
Procedural learning
Walker et al (2002) findings
Procedural learning - compare performance after x amount of hours being asleep and x hours being awake = - Overnight improvement
- Further improvement over subsequent nights
- Sleep deprivation reduces improvement
Declarative memory
Sleep may “PROTECT” memories from interference
% recall is a lot higher after sleep as opposed to awake condition - study two sets of word pairs (Paired-associate learning)
Clay et al (2007)
Learning of new words tested by assessing the presence of semantic interference with new words
Slower naming with related distractors but only after 1 week
Dumay & Gaskell (2007)
Learning of new words
AM group: study 8am - test 8pm - test 8am
PM group: study 8pm - test 8am - test 8pm
After sleep = better declarative memory
Forgetting curves (Ebbinghaus)
Ebbinghaus examined own ability to learn lists of nonsense syllables - found that forgetting function may be LOGARITHMIC
Forgetting is generally faster shortly after learning and then slower decline
graph = retention (%) vs elapsed time (days)
Savings method
Reduction in number of trials necessary to relearn materials over different retention intervals
Bahrick (1984) - logarithmic forgetting curves
Consistent with Ebbinghaus Retention of Spanish vocab over 50 years Rapid loss in first 3 years Relatively stable until 35 years 35-50 years further decline
Tested % Spanish original vocab retained
Permastore
Memories that appear to be very resistant to loss
Meeter, Murre & Janssen (2005) - logarithmic forgetting curves
14000+ ppts answered questions about publicised world events
Consistent with Ebbinghaus
Rapid loss at short intervals then slower loss
Forgetting rates DON’T depend on degree of learning
Remembering can cause forgetting (RIF)
Retrieval can modify memories, recalling some information changes the record of what you remember
Adaptive mechanism of remembering
Repeated retrieval strengthens a memory trace
Makes related, un-retrieved information less accessible
Retrieval-induced forgetting (Anderson et al, 1994)
To allow successful retrieval, we use inhibition to resolve competition or interference arising during retrieval - we facilitate remembering by facilitating forgetting
Initial RIF study by Anderson
Paired-associate learning - word pairs e.g. fruit-lemon, fruit-banana, metal-silver
Learning task e.g. retrieval practice with fruit-le___
Final test e.g. fruit-L, fruit-B, metal-S
Fruit-L = practiced relevant word (Rp+) - good recall due to benefits of practice
Fruit-B - unpracticed relevant word (Rp-) - poor recall due to inhibited to prevent interference (forgetting is cue-dependent)
Metal-S - unpracticed irrelevant word (NRP) - baseline control pair
Why are some items inhibited?
Similar related words to practice words - inhibited because they compete during retrieval
Another Anderson et al (1994) study - items of low and high taxonomic strength
Study with items of high and low taxonomic strength
fruit-orange = high strength
fruit-kiwi = low strength
Low taxonomic strength items compete loss, these items suffer less from RIF
Storm et al (2007)
Directed forgetting
Participants studied a list then received instruction to forget this list and study a new list
“Forgotten” items should not compete and indeed don’t suffer from RIF
Remembering and forgetting facts study
Practiced topic vs. related topic vs. baseline unrelated topics
Practiced topic: actor “is playing” either guitar // oboe
Guitar was the practiced item
Oboe was the shared topic
Related topic: teacher “is playing” saxophone // drum - shared relation items
Baseline topic with completely different sentence structure
Practiced item recall 60%
Shared topic recall 29%
Shared relation item recall 31%
Baseline topic 27%
Low integrators showed stronger impairment than high integrators
Integration reduces RIF
Integrated bits of information are less likely to interfere and compete with one another
Study cue-target pairs: standard instructions vs. integrative instructions
Integrative instructions are less vulnerable . to RIF
In complex structures or networks; concepts are connected by associative links - network functions as a “unit” - presence of links = ACTIVATION not competition
In complex structures or networks; concepts are connected by associative links
Networks function as a unit
Presence of links = activation, not competition
Scaling up to stories and RIF e.g. eyewitness
Remembering and forgetting in eyewitness testimony
- High motivation to remember
- High likelihood of telling same story over and over
Experiment 1 (RIF and stories)
Participants take policeman’s perspective; study pictures of objects stolen from 2 houses
Experiment 2 (RIF and stories)
Participants study series of slides showing 2 events (2 women making bogus charity collections) .
clear RIF
Time-course effects (Temporary effects)
Inhibits info irrelevant for current task
Weaker effects with delayed retrieval practice
Education - classroom learning - experts
Due to paradox of expertise, experts have more knowledge - retrieval should result in more inhibition (more knowledge to inhibit) - but it doesn’t as INTEGRATION of knowledge PROTECTS against forgetting
Comparing experts vs novices on symptoms of schizophrenia and autism (Anderson & Bell, 2001)
Test immediately or 1 day later
Results: novices = more inhibition (higher RIF effect)
Downside of RIF
Simplification
Can we suppress memories?
Freudian repression
We do have some control over unwanted memories (Anderson & Green, 2001)
Link to amount of RIF and depression - studies show if there is deficit to one’s RIF –> can’t suppress negative memories
least RIF - clinically more depressed than individuals exhibiting the most retrieval-induced forgetting
If a deficit in one's susceptibility to retrieval-induced forgetting can lead to negativity biases in autobiographical memory, and to an increased propensity for experi- encing anxiety and intrusive memories, then perhaps such a deficit stands as a significant risk factor for depression. In support of this hypothesis, correlations between retrieval-induced forgetting and depression have been observed. Groome and Sterkaj (2010), for example, found that 21 individuals diagnosed with clinical depression exhibited significantly less retrieval-induced forgetting than 21 control participants who were not clinically depressed. Similarly, in the study by Storm and Jobe (2012b), when data from participants suffering at least mild levels of depression were analyzed (i.e., scores of 14 or higher on the BDI-II; Beck et al., 1996), a significant correlation was observed such that individuals exhibiting the least retrieval-induced forgetting were significantly more depressed (as measured by severity on the BDI-II) than were individuals exhibiting the most retrieval-induced forgetting
Executive Deficit Hypothesis
People differ in the ability to suppress unwanted memories because they differ in executive control abilities
Most inhibition seen in people with
High working memory capacity High inhibition Highly traumatic experiences Younger (vs older) adults Older children
RIF studies
Memory phenomenon where remembering causes forgetting of other information - context of retrieval inhibition 3-phase experiment 1. study 2. practice of some studied material 3. final test of all studied material
studied/practiced words = Rp+
words related to practice words = Rp-
unrelated words = Nrp
Who does this phenomenon occur?
Process actively inhibiting information
OR
Interference from other information in memory
Which words suffer as a result of prior retrieval?
Rp- (words related to but unstudied)
poorly recalled in comparison to unrelated words
Calculating RIF effect
Rp+ 81%
Rp- 55%
NRp 68%
NRp - Rp- = 68 - 55 = 13% RIF effect
take away related word recall from unrelated words
Boundary conditions of RIF
SEMANTIC INTEGRATION
- Influence of integration of learned material - strength of the semantic relationship between competing targets
- Duration of RIF effect - long retention interval
- Effect of type of test format most susceptible to RIF - retrieval-cue test (paired associative learning) vs. free recall tests (inconsistent results)
- Mood state and stress levels
- some studies show RIF is only seen in cue-dependent tests NOT cue-independent
Retrieval-induced forgetting summary
Memory is modified through the act of retrieval. Although retrieving a target piece of information may strengthen the retrieved information itself, it may also serve to weaken retention of related information. This phenomenon, termed retrieval-induced forgetting, has garnered substantial interest for its implications
Anderson et al (1994) study
?
Stickgold/Diekelman
Stickgold & Walker x 2
Sleep/consolidation studies
Storm et al
RIF - learning/eyewitness/social cognition
Anderson & Green
Suppressing unwanted memories by executive control (executive deficit hypothesis)