Chapter 12 The Biology of Learning and Memory Flashcards
Localized Representations of Memory [placeholder]
.
Pavlov’s classical conditioning
- unconditioned stimulus (meat) automatically elicits an unconditioned response (dog salivates)
- conditioned stimulus (tone) initially elicits no response
- conditioned response (dog salivates to tone) is learned with repeated pairings of tone with meat
Localized Representations of Memory cont. [placeholder]
.
Operant Conditioning
- an individual’s response is followed by a reinforcement or punishment
Reinforcement
is an event that increases the future probability of the response
—> e.g.: given fruit loops for a correct response, a rat learns to return to the same arm of maze
Punishment
- is an event that suppresses the frequency of the response
—> e.g.: given a shock for the wrong response, a rat learns to avoid arm of maze
Engram
- the physical representation of what has been learned (e.g., a connection between two brain areas)
Karl Lashley searched for the ….
Engram
- he trained rats on mazes and brightness discrimination tasks (memory tasks) and made cuts between two brain areas or removed part of the brain
—> none of the deep cuts disrupted maze performance
—> maze performance was only decreased when very large amounts of brain were removed (not small areas)
- So learning and memory does not depend on a single cortical connection or area
So from this, Lashley proposed two principles
- equipotentiality
- mass action
equipotentiality
- all parts of the cortex contribute equally to complex behaviors like learning
mass action
- the cortex works as a whole, and the more cortex the better [main difference between this and equipotentiality]
- incorrectly assumed that all kinds of memory were physiologically the same and that the cortical area was the best place to search for an engram
- But pretty good for being the first! (he did this in the 1950s)
lateral interpositus nucleus (LIP) of cerebellum
- critical for classical conditioning of the eye-blink response
- plays a crucial role in motor learning and coordination.
Modern Search for the Engram [placeholder] {info dump}
- 1986 - Thompson et al: in a classical conditioning task, they paired tone with puff of air in eye until rabbit blinked at tone
—> found changes in cells in the lateral interpositus nucleus (LIP) of cerebellum
—> when LIP was cooled or drugged the rabbit did not learn and later when effects wore off rabbit learned at same rate as new rabbits
- Conclusion: LIP has to be active to learn
Red Nucleus
(mid-brain motor area that receives input from LIP)
Modern Search for the Engram cont. [placeholder]
- Thompson et al then suppressed activity in red nucleus (mid-brain motor area that receives input from LIP)
—> rabbits showed no response during training
—> when effects wore off rabbits showed strong learned response to CS
—> thus, red nucleus necessary for performance, not learning, of response
—> And the LIP really strongly looks like it’s involved in learning! (at least in rabbits)
Modern Search for the Engram cont. even more
- Probably same in humans
—> classical conditioning of eye blink produced activity in cerebellum, red nucleus and other areas
—> people with damage to cerebellum are impaired at eye-blink task
—> But people with cerebellum damage aren’t impaired in other types of learning tasks.
- So, just like with vision, learning in humans is most likely done in several different locations, depending on what type of memory is needed.
Hebb: Short and Long-Term Memory [place holder]
.
Short-term
- events that have just occurred
- limited capacity, around 7 unrelated items
- Why phone numbers are 7 digits (seriously!)
- If forgotten is lost forever – can only get it into LTM through rehearsal. Hints don’t help
Long-term
- events from previous times
- not lost, are recoverable, and hints help
- HUGE capacity – never appears to get full – when you learn something new, you don’t have to forget something old to “make room for it”
Consolidation
- is the process of converting short term memory to long term memory
- memories that stay in short-term memory long enough are formed into long-term memory
Consolidation of Long-Term Memory [info dump]
- To study consolidation, researchers used shocks to disrupt memory, looking for short-and long term memory threshold
—> but, shock to head disrupted both short-term and long-term memories
—> also, some disrupted memories could be recovered after reminder of event
- Human studies weaken the distinction between short- and long term memory
—> we can recall events more than a few hours, or days, old but nevertheless will be forgotten in near future
—> So obviously, at least in humans, it’s not quite as simple as STM and LTM
Consolidation of Long-Term Memory cont. part 1 [info dump]
- When we learn, the learning activates several genes that promote learning and memory. It also activates a gene that produces PP1
- Protein phosphatase 1 (PP1) interferes with memory retention by inactivating genes that promote learning
—> This is actually adaptive – it allows us to easily forget single experiences, as we don’t need to remember every single thing that ever happened to you
—> PP1 accumulates during massed practice and declines during distributed practice (why cramming is worse for memory retention than spread out learning)
—> So if you really attend to what you’re trying to learn and learn it in an appropriately spaced manner, the learning genes will be stronger than the PP1 and you’ll remember it. If not, the PP1 “wins” and the memory is lost
Consolidation of Long-Term Memory cont part 2 [info dump]
- Consolidation is gradual and we recall older memories more easily
—> It appears that the earlier a memory is, the more firmly it’s implanted in our memory
—> older people showed most MRI activity to celebrities in the 1990s and least to celebrities in the 1940s
—> people showed more temporal lobe activity to places visited in last 2 years than those visited in last 7 years
—> Interpretation is that we have to work harder to get the newer memories, but over time, they will get more firmly ingrained
- Meaningful and emotional experiences enhance memory consolidation, e.g, receiving award or being in car wreck
—> Emotions increase secretion of cortisol and epinephrine
—> epinephrine stimulates vagus nerve which excites the amygdala; cortisol stimulates the hippocampus – big in memory!
Consolidation of Long-Term Memory cont. part 3
- Direct injections of cortisol or epinephrine enhance storage and consolidation of memories
—> but prolonged stress results in too much cortisol and memory is impaired
- After damage to amygdala, emotional arousal does not enhance memory storage
—> If I gave you a list of 30 words to memorize and some were emotionally charged (taboo), like bitch or penis, you’d remember those more than the non emotional words like boot or apple.
—> But patients with amygdala damage didn’t remember “taboo” words better
—> So the amygdala is involved in emotional memory storage
A Modified Theory: Working Memory
- Needed a more complex alternative to just STM and LTM
Working memory
- we actually store information in LTM by working with it or attending to it