L6 - Memory, Sleep and Consolidation Flashcards
What is studied more and why: anterograde or retrograde amnesia?
Anterograde is better studied because retrograde is more difficult to assess - there are several reasons that might cause people to forget past events.
What are the observations that suggest memories become more stable over time?
Memories formed closer to the time of the trauma are remembered less effectively than memories formed further in the past (Ribot, 1882).
‘Fresh’ memories are more vulnerable to disruption via interference (Muller & Pilzecker, 1900)
What does consolidation assume?
Consolidation assumes that memories, over time, become more stable and less vulnerable to disruption, and also less dependent on the hippocampus.
What are the two main levels of explanation of memory consolidation?
- Cellular/molecular level: how do memories become stabilised at the synapse level?
- System: How are the neural circuits storing a memory reorganised over time?
What is one finding that supports the idea of consolidation processes occurring in the hippocampus, and storage of distant memories belonging to the cortex?
Lesions to the hippocampus causes most impairment to the most recent events (temporally graded amnesia)
Lesions to the neocortex causes most impairment to more distant events.
- suggests that memories are initially dependent on the hippocampus, but are then gradually transferred to the neocortex.
What is the stability-plasticity dilemma?
The brain needs to acquire new information, which requires plasticity, but also needs effective retention of that information without overriding older knowledge, which requires stability.
Having a process which is plastic will automatically lack stability, and a process which is stable will automatically lack plasticity, so how does the brain do both?
What is McClelland et al.’s (1995) suggestion as a solution for the stability-plasticity dilemma?
Suggests a two-stage system:
The hippocampus is a fast learner (good for plasticity) and the cortex is a slow learner (good for stability)
This allows the hippocampus to learn as much as it needs to, and provides the cortex with enough time for the new memories being passed onto it to be coded for with sufficient stability.
What evidence does Bontempi et al., (1999) provide for the occurrence of consolidation in rodents?
Rats learnt a maze task where locations had to be discriminated. Retention was assessed at 5 days after and 25 days after initial learning, with glucose measurements taken to reflect brain activity in a given area.
Found that the hippocampus had greater activation than other areas at 5 days, but the PFC had greater activation than other areas at 25 days.
Concluded that time-dependent re-organisation between the hippocampus and the neocortex mediates cortical long term memory storage.
Is the PFC active or passive in it’s role in memory storage?
Initially thought to be passive, but has been found to actively compare the information already stored within the cortex, with the information being learned by the hippocampus. It has been found to potentially instruct the hippocampus to stop learning information if it is already something that is stored cortically as a long term memory.
What evidence did Takashima et al., (2005) find in support of consolidation in humans?
Tested recognition after increasing time intervals, in response to picture stimuli. Each testing session introduced new stimuli and tested old.
Found that there was less hippocampal activation the longer the delay between encoding and recognition. Also found that the medial PFC became more active with increased delay.
What are the differences in the predictions made for memory performance between the standard model and multiple trace theory (MTT), both in terms of:
- degree of hippocampal damage
- recency of memory
- type of information impaired
- Overall, no difference between the theories in expected memory depending on damage - both agree that those with partial damage will remember slightly more than those with full damage, but not as much as healthy controls.
- Both theories agree that controls will remember more for recent memories compared to remote, but that patients will remember more for remote memories than recent ones.
- Standard model makes no distinction between the type of information that can be recalled. MTT assumes that memory for remote events will not be possible - the same as for recent events, in those with full hippocampal impairments, but only for episodic information (previous trends are predicted, only if semantic information).
Explain the main difference in predicted memory performance between standard model of consolidation and multiple trace theory - what do they predict that differs, and why, according to the theory?
The main difference is that the MTT assumes impaired recall of distant episodic memories, in those with complete hippocampal lesions.
This is because the MTT assumes that access to distant memories is always reliant on the hippocampus, but only for episodic memories.
When does McClelland et al., (1995) propose that memory consolidation needs to take place?
During offline periods when the brain is not learning new information - i.e. when sleeping.
Describe the relationship between type of sleep and the improvement for memory types.
Less forgetting of declarative memories/information after slow-wave than REM sleep
Implicit/motor memories are enhanced after REM and not slow-wave sleep.
What did Gais et al., (2007) find about hippocampal activity after sleep?
Sleep group had increased hippocampal activity at recall after 2 days compared to 30 minutes after stimulus exposure, and compared to the no sleep group.
The mPFC (which also plays a role in consolidation of declarative memories) was more activated after 6 months, but only in the sleep group.
