Our memory is not just a single mental capacity.

Our brain has a number of memory systems

that do different things with different types of info and guide our behaviours in different ways.
This is the so called taxonomy of memory. Every memory type belongs to a different brain area.

The first distinction is between short-term and long-term memory

Long-term memory is divided in declarative- explicit (conscious) memory and

nondeclarative- implicit (unconscious) memory.

Often patients have problems with declarative memory and especially episodic memory.
The famous ‘Pin’ anecdote

This is a classical example from the literature with the pin anecdote in which the neurologist
hid a pin in his hand and pricked the patient and at a later occasion the patient refused to
shake hands again but when she interviewed the patient didn't recall anything of the actual
incident being pricked. This suggests that your explicit (conscious memory might be
impaired but the implicit (unconscious) memory is not.

This anecdote is an early illustration of the dissociation between explicit and Implicit
memory, but also as being relevant for theories of conscious recognition

Case of HM

Bilateral removal of hippocampus

at 27 HM had a portion of his brain surgically removed in the hope that it would stop/lessen
his epileptic seizures. The seizures vanished but so did his ability to maintain memories. The
researchers found out that the hippocampus was responsible for declarative memory.
He was able to learn new skills and recall general aspects of his life but was incapable
of remembering anything he just experienced.

Impaired memories in HM

1. New Episodic memories + new semantic memories = declarative memory

2. Some forms of priming that involve new semantic knowledge + conceptual priming
3. Anterograde amnesia

Global (anterograde) amnesia= inability to learn new things, long-term memories before the
brain trauma remain intact. Impaired episodic memory and semantic memory (declarative
memory). The semantic memory deficit concerned inability to learn new semantic facts,
contents, concepts

Anterograde amnesia (AA) refers to an impaired capacity for new learning. People with
anterograde amnesia have trouble making new memories after the onset of amnesia

Retrograde amnesia (RA) refers to the loss of information that was acquired before the onset
of amnesia. People with retrograde amnesia have trouble accessing memories from before
the onset of amnesia.

Episodic memory= personal events (something happened and we were involved)

Semantic memory= general knowledge (perceptual knowledge) of the world, facts, concepts,
explicit meaning
Do we need episodic memory in order to learn new semantic memories?

Findings are mixed:

-Tulving(2001): Learning/encoding proceeds from the perceptual to the semantic to the

episodic system. Hence new semantic learning is possible even when episodic memory is
impaired

-Baddeley(1988): semantic learning is based on the accumulated residu of multiple previous

episodic memories = Η σημασιολογική μάθηση βασίζεται στο συσσωρευμένο υπόλειμμα
πολλαπλών προηγούμενων επεισοδιακών αναμνήσεων

Greenberg & Verfaille(2010): Episodic Memory is a binding of semantic information with

context information. Episodic Memory facilitates new semantic learning. When episodic
memory is impaired, semantic learning is still possible but has to run on the basis of slow
neocortica llearning.

➡️we can say that they are related, you don't need to remember the event in order to use the
semantic knowledge. So you can say that semantic is not contextual and episodic is
contextual

Spared memories in HM

1. Procedural learning (implicit) = he was becoming better and better in mirror

tracing tasks. He performed this task 10 times each day for three days. They
counted the errors every time he went outside the lines. HM drawings improved even
 though he couldn't recall that he had done this test before. So he couldn't form new
explicit long-term memories due to damage to hippocampus and surrounding
areas but he could learn new skills

2. Different sorts of priming (implicit) = faster identification of fragmented pictures &

words when having seen the complete item earlier. Another type of implicit non
impaired memory is priming= an exposure to a stimulus can lead to a faster
response when stimulus is presented again. They don't ask you to recognize this
stimulus and say anything whether is part of a previous learning event

we recognize it but don't recall a previous event/exposure (when we learned about

this= this is priming (implicit memory) = να ξερω κατι αλλα να μην θυμαμαι που και
πως το εμαθα). If you recognize it and you remember where you first learned
about it you are using your explicit memory. It is completely new to him when
they ask him if he remembers doing this task
3. Working memory = the small amount of information that can be held in mind and
used in the execution of cognitive tasks

4. Well-consolidated semantic memory= old semantic knowledge

5. Old episodic memories = at least 11 year before

Kopelman (1989): he asked participants to describe or recall events from different

periods of their life. He had a group of Korsakoff patients, Alzheimer patients and a
control group.
The controls do quite well in recalling things and they best recall the more recent
things.
In patients the old things are relatively spared but they are amnestic for the recent
things this is called the Ribot gradients of amnesia and this is what we see in HM
as well he cannot recall recent episodic memories but his old episodic memories are
spared
 How long does it take for episodic memories to become semantic?

Consolidation for events from 11 years ago is better than current events. One
possibility is that they turn from recollection (remembering) to familiarity (knowing).
Non-contextual remembering might lead to a familiarity based, semanticized explicit
memory

6. Metamemory = meta-thinking about your own memories. A lot of memory failure is a

result of failure in our metamemory. Metamemory might partly depend on episodic
memory. HM realises that his memory is poor (he has this metamemory). His
sense of self is intact = he recognizes himself in the mirror even though his face
doesn’t match the premorbid image

7. Old spatial memories are spared; he remembers his house as a child because he
learned this information as a kid.

But!! He was able to memorise his new house after the surgery -> why is this
surprising? Because it is an explicit-conscious-declarative memory that allowed him
to do that and also it’s new and we said that he wasn’t able to do that so how can we
explain this -> he was there for years and he was exposed to that place multiple
times using multiple sensory inputs so it looks like the procedural learning that
we’ve seen before and he is spared there.
Also he is spared in some areas that are responsible for spatial learning -> and
specifically egocentric

Egocentric = where things are in respect to our own body = we navigate by keeping
track of our own body e.g. i need to turn left. Parietal and somatosensory areas
support egocentric space coding

Allocentric based on the environment = landmarks etc

Allocentric spatial navigation allows us to have perspective free map of the space,
plan, detours only by thinking about the environment. Hippocampal areas support
this.

we could perhaps say that HM still has some egocentric learning and no longer
allocentric learning (after surgery)
Sum of spared and impaired functions in HM

Can he give his consent?

Three criteria for valid consent:

1. Full information
2. Voluntary participation
3. Capacity to make decisions

Sense of self
1. Collection of current and long term goals = working memory + future thinking
(projecting yourself in the future and imagining the future scenes. Patients with
memory impairments have difficulty imagining future processes
 2. Narrative of one’s life = semantic + episodic memory

“Your sense of sense is the results of your memories”

Informed consent
1. Full information = declarative (episodic and semantic) + working memory

2. Consideration of possible outcomes = future thinking

➡️this could have been a problem in HM

Case of Jon

Suffered perinatal anoxia resulting in 50 bilateral hippocampal reduction in volume while

surrounding cortical areas seemed preserved
He has high IQ

Three Long-term memory impairments

1. Episodic forgetting (yesterday’s events)
2. Temporal (appointments = prospective memory)
3. Spatial (where things are)

He is good in recognition = it is argued that it is because of familiarity and not recollection

He is poor in recall

In recall and recognition test you measure episodic memory

On the figure we can see that he is good in recognition but not in recall

Familiarity= you know something but you cannot recall from where you know it
recollection= you remember something and you have a sense of place,time,how, with whom
etc. Space is critical for recollection -> if you don’t have spatial memory you could also have
problems in recollection

Episodic memory is often described as what where and when

Jon only has the what = familiarity

Experiment 2 King et. al.

They asked participants to move around in a VR environment if they are able to do that they
can build a good allocentric map of that environment. Next they presented items from either
view point 1 or 2 or 3 and then they showed the items again either from the same view point
as they initially saw it or from a different view point (shifter perspectives). They also placed
distractors in the scenes.

What they found out is that when you place the distractors further away it’s easier to
make a decision

Jon and controls do well in the same view but Jon performs very badly in the shifted
view
Here we see that Jon’s difference in same vs shifted view is 40% that’s a lot so he really
struggles

Joh is unable to use allocentric spatial memory. ALsM depends on the hippocampus.
His egocentric memory is preserved (same view)

Experiment 3 King et al.

So you have the same view coding again

But now you have to reason from a completely city environment (the background changed) -
you cannot use the snapshot information
Jon is again impaired compared to the controls when the view point changes but he is
also very much impaired when the background has changed. This means that he
doesn't have a good sense of computing a motor factor so he requires this visual snapshot
and when it's not allowed he can't do the task. In this graph the controls have very large
error comparing the shifted view and same view whereas in experiment 2 that was minimum