Memory Representation in the Temporal Neocortex Flashcards
What does the Hippocampus allow?
recollection of episodes
-> our ability to join together the different elements of memory in order to allow us to recollect memory later
what can recollective experiences be known as?
mental time travel
According to Tulving (1972), there are two types of declarative/explicit long-term memory available to consciousness. What are these?
- event/episodic memory
- semantic memory
episodic memory
dependent on hippocampus and impaired with amnesia
what is the other type of LTM as suggested by Tulving (1972)?
non-declarative / implicit
semantic memory
factual knowledge about the world and preserved in amnesia
episodic memory (according to Tulving, 1972)
- events
- mental time travel (temporarily bound memory about a mode in time)
- self-referential -> you are the centre of information (self moves through time and space, and when you recollect events you see them from your own perspective)
- fragile, easily forgetten
- Affected in amnesia
- Better when young (declines in older adults)
semantic memory (according to Tulving, 1972)
- facts
- time/place NOT coded
- NOT self-referential -> separate from oneself - the concept of who you are doesn’t contribute very much
- more consolidated / durable
- not affected in amnesia (spared)
- better when old (i.e. Ronnlund et al. (2005) -> better time to accumulate knowledge about the world
Patient KC developed amnesia after a motorbike accident, what were their episodic memory like?
- inability to remember recent and new events
- severely impaired retrograde -> inability to remember events before the accident [this includes the period of time where he was working as a skilled machinist] -> he has a rich vocab and concepts about these tools but cannot remember using these tools or requiring them
Patient KC developed amnesia after a motorbike accident, what were his semantic memory like?
- retained some concepts gained as a machinist i.e concepts about tools like keyway shank and feed screw
- knowledgeable about how to change a tire but cannot remember ever changing one (episodic)
KC tasks about his accent, what is this an example of?
personalised semantic memory (learning a new piece of information about themselves but its not a rich detail or recollection of an experience)
what is KC an example of?
single dissociation
-> deficit of one domain (episodic) and preservation of another (semantic)
how can we argue that there are two separate systems supporting different aspects of knowledge?
a double dissociation -> you need other patients who show the reverse impairment (impairment of semantic but not episodic memory)
why is a single dissociation viewed as insufficient to demonstrate separate systems?
episodic memory might be more vulnerable to damage than semantic memory within a single system
- having a single dissociation is thought to be insufficient to show you’ve got two separate systems because some tasks are harder than others
- if you have some degree of brain injury then some of the harder tasks might be more easily impaired than easy tasks -> or perhaps episodic memory is just more difficult (perhaps all memory captured in the hippocampus but episodic impairs first because it’s harder)
-> to prove this is not the case, we need to find individual’s who hippocampus is spared yet has problem with semantic memory in different brain region)
why do people often show issues with episodic and semantic memory?
anterior temporal lobe and hippocampus are exceedingly close together
what types of double dissociations can we look at to test if semantic and episodic memory are from two different systems?
amnesia (episodic impaired, semantic preserved) and semantic detention (semantic impaired, episodic preserved)
what is a key deficit in semantic dementia?
poor understanding of words and objects
-> semantic memories deficits encompasses all aspects of knowledge about the world (i.e. concepts etc)
what is semantic dementia?
- subtype of frontotemporal dementia (FTD) -> bilateral atrophy in the anterior temporal lobe (peak of atrophy is in the ventral of anterior temporal lobe -> right next to/beside the anterior part of the hippocampus)
- progressive loss of conceptual knowledge across modalities
What did Jefferies et al. (2006) find about episodic memory in semantic dementia patients in a double dissociation with amnesia?
relatively intact memory for recent events (able to recall experiences) and impaired semantic memory (cannot name particular concepts)
(Graham et al., 2000)
Patients with SD were asked to recognise objects and later on were asked to pick out which of the three objects they had seen previously
* Condition 1: was exactly the same (had to pick up perceptually identical object)
* Condition 2: had to pick a perceptually non-identical version of the object (same concept of the one you studied earlier)
Looked at which words people could still use: picture naming and picture matching as well -> sort items into two types: semantically known and semantically unknown
-> In SD: As their semantic memory degrades, particularly ideas are affected. And these particular ideas are affected across all tasks and modalities -> reliable fashion, consistent with the idea the concept itself is degrading (i.e. persevered concept of a telephone but not a television)
- SD patients can recognise objects when they are identical
- for non-identical objects meaning is important. Performance is predicted by which items are still understood on semantic tests
What did Graham et al. (2000) find (in detail)
- If item is still known, able to recognise all of the objects -> they could do if they knew the idea in both conditions when it was perceptually identical and perceptually different
- When they didn’t know the idea (semantic impairment), their episodic memory for the perceptual identical items were still good but they do struggle for perceptually non-identical ideas
â—‹ Shows episodic memory in many ways is preserved in semantic dementia (3/4 of these conditions) -> people can have normal visual recognition performance
○ BUT this is not the case for amnesia because they wouldn’t even remember looking at the pictures - Shows semantic memory does contribute to certain types of memory if you cant rely on perceptual features to recognise an object then you have to rely on semantic, and if it’s degrading then performance will be impaired
What do Patient KC and studies of semantic detention provide us with?
a double disocciation between episodic and semantic memory
* difference between hippocampus and anterior temporal lobe
what is the hippocampus responsible for?
encodes and recreates unique multimodal experiences of people/places/objects in events (i.e. episodic memory)
-> allows us to separate experiences from all the other experiences we’ve had (by binding them together, the unique elements allow us to separate memories
what is the ATL responsible for?
extraction of similarities between multimodal experiences to create concepts (i.e. semantic memory)
-> may be critical for seeing similarities between different experiences, extract information which is common across these experiences in order to form conceptual information that can predict things which will be similar in the future
what is the complementary learning system hypothesis?
- building complementation model which can stimulate brain responses and the effect of brain injury -> shows you can build computer models that stimulate the effect of learning in the hippocampus and ATL
-> hypothesis suggests that the hippocampus and ATL work differently
-> idea that hippocampus and ATL have different memory jobs and complementary roles to play in our cognition links to this
what does hypothesis suggest about the hippocampus?
FAST learning in the hippocampius
- hippocampus neurons learn really quickly and bind together elements from unique episodes -> capture episodic memories in one go for unique memory that occurs on time
- uses sparse code (relatively few neurones for each item, allowing us to separate unique memories -> as different neurones are involved)
What does the hypothesis suggest about the ATL?
SLOWER learning in the neocortex to extract similar features which happen across multiple experiences
- similar features shared by multiple experiences are encoded strongly
- useful for semantic category learning
- prevents catastrophic interference -> loss of old memories when new material is learned (way cortex learns to avoid this problem)
- by turning down learning rate, the model overcomes catastrophic interferences allowing memories to be formed from multiple experiences that happen over a long period of time
- This slow learning allows us to then understand why memories are better for more frequent things
- Can also explain why semantic concepts have features which are very strong part of the concept which tend to occur in all the different learning exemplars and then peripheral features which are sometimes true (dogs are brown, but they aren’t all brown)
what is emotional memory?
emotional valence is formed quite slowly as a concept and something largely intact during amnesia
* more semantic than episodic
what is the opposite view to two distinct systems for episodic and semantic memory?
transfer of information from hippocampus to neocortex in memory consolidation during sleep
what does Squire suggest about episodic and semantic memory?
hippocampus has time-limited role
-> system consolidation (information being transferred to the neocortex)
Kim & Fanselow (1992). What are the differences between synaptic consolidation and system consolidation.
Synaptic consolidation:
- looking at inhibiting the synthesis of proteins at the synapse in the hippocampus and at the impact of inhibiting protein system on the duration of memory
i.e. a drug prevents any structural changes happening as a consequence of learning
Inhibition of protein synthesis which allows structural changes at the synapse in long-term potentiation
-> disrupts memory for 1hr
- if you impair protein synthesis an 60 or 90 minutes after learning, it visually has no impact on how much the rat remembers later on
-> suggesting most changes happening at the synapse are down to LTP and happening in the first hour after learning
Kim & Fanselow (1992). What are the differences between synaptic consolidation and system consolidation.
System consolidation:
- if you lesion the hippocampus immediately after learning something, memory will be really poor
- if you lesion / damage hippocampus, two weeks after the memory has been formed, the rat only forgets 50% of what was encoded -> some but not all of the information has been transformed out of the hippocampus
- a month after encoding, a little bit of memory loss but it’s memory will be pretty good -> perhaps the hippocampus is still playing some kind of role which is why a tiny bit of memory is lost