Exploring the Mind
COGS 100, Summer 2025
Memory
Reading assignment: Ranganath, Libby & Wong, Human Learning and Memory
Take-home messages
• Memory is not a single undifferentiated system.
• Evidence from behaviour.
• Evidence from localization in the brain.
• Remembering is a constructive behaviour.
Memory as information processing
Encoding Storage Retrieval
Human Memory System(s)
Recall: Serial position effects
Presentation rate
Recency effect increased/low
100 frequency words used:
Primacy effect
Proportion recalled
75 attenuated
Primacy effect Recall task delayed :
50
Recency effect
25 attenuated
Conclusion: These two
0
0 3 6 9 12 15 18 21 24 27 30 effects are subserved by
Position of item in list
different parts of the
human memory system.
6
Short- vs. long-term memory
• Interpretation: Early items encoded into a long-term
memory store, recency effect due to active status in STM.
• William James (1890): Primary vs. secondary memory
• Implies a limit on primary memory.
• Miller (1956): Limit on short term memory of 7+/1 2 items.
• Information processed in STM can create an LTM memory
trace.
LTM/STM = Modal model
Questioning the modal model
• Continuous distractor
task
• 12 seconds of
arithmetic after each
pair of words on a
study list.
• Recency effects were
still found!
Bjork & Whitten, 1974, Cognitive Psychology
Working memory
Working memory
“The theoretical concept of working memory assumes that
a limited capacity system, which temporarily maintains and
stores information, supports human thought processes by
providing an interface between perception, long-term
memory and action.”
Baddeley, 2003
Baddeley & Hitch Model
Baddeley and Hitch (1974)
proposed a model of working
memory with 3 components:
A visuospatial sketchpad
A phonological loop
A central executive that controls
processing.
Later versions also included an
episodic buffer.
The phonological loop
• Phonological store: bu er with ~2s of storage.
• Articulatory rehearsal: active process of rehearsing verbal material.
• Can be sub-vocal or out loud.
Phonological store Passive
Phonological Loop
Articulatory
Active
rehearsal
ff
Suppressing the loop
• Articulatory suppression tasks:
Ba ba
• When rehearsal is prevented between ba ba
stimulus presentation and recall,
performance is impaired (e.g.,
Baddeley, Thomson, Buchanan, 1975).
• Rehearsal prevented by repetition of
irrelevant sounds.
• Length e ects:
• List recall (in terms of syllables but also
raw duration).
ff
Articulatory suppression
Suppressing the phonological
loop (syllable repetition) impairs
performance on sentence
comprehension tasks.
Rogalsky et al. (2008)
Section summary
• The modal model of memory may be oversimplistic.
• Evidence for working memory.
• A limited-capacity store of information currently being
used.
• Model by Baddeley & Hitch includes phonological loop,
visuo-spatial sketchpad and central executive.
The medial temporal lobes
The medial temporal lobes
• Medial = toward the
midline
• Includes the
hippocampus, perirhinal,
entorhinal, and
parahippocampal cortex.
Squire et al., 2004, Annual
Review of Neuroscience
H.M.
Henry G. Molaison
• Operation in 1953 removed sections of the Medial
Temporal Lobe bilaterally, including the
hippocampus.
• Surgery was meant to treat uncontrolled
epileptic seizures.
Annese et. al., 2013, Nature Communications
See studies by Scoville, Corkin, Milner and their colleagues.
H.M.
• H.M.’s intellectual ability and
personality were relatively
unchanged.
• Attentional and working memory
capacities normal.
• Memory for long past events was
intact.
• Severe anterograde amnesia: H.M.
could not form new long-term
Annese et. al., 2013, Nature Communications memories (of a certain type).
H.M.
• H.M. could form new procedural
memories:
• Motor skills (e.g. mirror drawing).
• Classical conditioning (e.g., blink
response).
• Perceptual learning (e.g.,
recognition of incomplete pictures).
• Could not describe in words what he
Annese et. al., 2013, Nature Communications had learned to do.
H.M.
Drawing of the
oorpan of a
home HM lived in
from 1958-1974
Second drawing
several years after
moving away.
Corkin, 2002, Nature Reviews Neuroscience
fl
Repetition priming
• Facilitated processing of a stimulus after a prior
presentation.
• For those without neurological disorders, priming
effects can occur days or even months later!
_AR_VA__
(Sloman et al., 1988)
• Participants with amnesia have shown priming effects like
those without amnesia, e.g., by Cave & Squire, 1992.
• Picture naming was facilitated, however recognition
memory was not.
Role of the hippocampus
• While the main de cit in patients with
medial temporal lobe/hippocampus
damage is formation of long-term
memories, the picture may not be so
simple.
• Nichols et al. studied working memory for
faces.
• Amnesic participants (most of whom
had MTL damage): impaired
performance even at short intervals.
• Functional MRI recording: Activation in
the MTL during maintenance in working
memory.
Nichols et al., 2006 MTL involvement in working memory
fi
Section summary
• Studies of patient H.M. elucidated the role of the medial
temporal lobe structures in memory.
• While his ability to form new semantic or declarative
memory was highly impaired, other forms of learning did
lead to long-term retained representations.
• However, the picture may not be so simple (face
memory).
Conclusions and questions
• We know that the medial temporal lobes are critically
involved in memory, but what is their precise role?
• What does this tell us about distinctions in memory
mechanisms in the human mind/brain?
Distinctions in memory types
What major divisions have been proposed?
• Long-term/short term (modal model)
• Working memory (vs. Long-term memory)
• Declarative vs. Procedural
• Semantic vs. Episodic
• Recollection vs. Familiarity
What major divisions have been proposed?
• Long-term/short term (modal model)
• Working memory (vs. Long-term memory)
• Declarative vs. Procedural
• Semantic vs. Episodic
• Recollection vs. Familiarity
Long-term
memory
“Knowing that” “Knowing how”
Declarative Procedural
Semantic Episodic
Declarative memory
Knowledge of facts and events = Explicit memory
Semantic memory Episodic memory
Memory for
Memory for
knowledge and
experiences
facts
Not tied to a
Tied to a certain
certain point in
point in time
time
Includes words Includes the
and their context of an
meanings event
Victoria is the I took the ferry
capital of British to Vancouver
Columbia. Island in 2005.
Declarative memory in the brain
• Associated with the
hippocampus and
adjacent structures in the
medial temporal lobe.
Squire et al., 2004, Annual
Review of Neuroscience
Procedural memory
Memory for skills, habits
• Includes priming, classical
conditioning
• “muscle memory”
• Sensory-motor adaptations
= Implicit memory
Procedural memory in the brain
• Subserved by a complex network of brain areas:
• Premotor/motor cortex
• Subcortical structures:
• Striatum
• Cerebellum
Recall: H.M.
• H.M. could form new procedural
memories:
• Motor skills (e.g. mirror drawing).
• Classical conditioning (e.g., blink
response).
• Perceptual learning (e.g.,
recognition of incomplete pictures).
• Could not describe in words what he
Annese et. al., 2013, Nature Communications had learned to do.
Amnesia and mirror reading
esoidnarg - suoicirpac - elggardeb
• People with amnesia
were able to learn mirror
reading at a comparable
rate to non-amnesiacs.
Cohen & Squire, 1980, Science
Recollection and familiarity
• Recognition: identi cation as
previously known.
• Mandler (1980)’s bus example:
• Imagine you recognize someone
on a bus.
• Do you know their name, which
context you know them from?
• Or just a feeling of familiarity?
See e.g., work by Andrew P. Yonelinas
fi
Recollection and familiarity
• Recollection:
• Recall with associated information.
• Like episodic memory.
• Familiarity:
• Based on strength of the
representation.
• Like semantic memory.
See e.g., work by Andrew P. Yonelinas
Section summary
• Several key divisions exist within long-term memory and have been
supported by behavioural and neuroimaging studies.
• Declararative/procedural
• Semantic/Episodic
• Recollection/Familiarity.
• Distinction between the types of memory that require MTL
involvement vs. those that do not.
• Or ner-grained distinctions, e.g., with familiarity vs. recollection.
fi
Memory encoding and retrieval
• Does the process by which information is
encoded have an in uence on how easily it can
be retrieved?
• If processing is ‘deeper,’ is the representation
of better quality?
• The Levels of Processing Hypothesis
• Evidence: Better recall for words encoded
with participants attention drawn to semantic
content rather than phonological content (in
between) or orthographic content. (Craik &
Tulving 1975)
Memory by SAM Designs from NounProject.com
fl
Remembering and forgetting
Memory encoding and retrieval
• Factors other than depth of processing
may be in uential in determining recall
success, including retrieval cues:
• Encoding speci city: Whether a retrieval
cue is associated with a stimulus.
• Congruence of the task with the
encoding process.
• Context of encoding/retrieval (same
room, emotional state).
Memory by SAM Designs from NounProject.com
fl
fi
Forgetting
• You can forget information that has been encoded
(processed).
• Typically takes place within 48h of the event of learning.
• Forgetting - failure to consolidate?
• Forgetting - failure to retrieve?
• Information learned before or after a critical memory
can interfere.
Reconstructive nature of memory
Deese/Roediger/McDermott paradigm
• Memory list:
nurse sick lawyer medicine health hospital dentist
physician ill patient of ce stethoscope surgeon clinic
cure
• Studied/not studied judgment task:
lawyer clinic weasel health dentist pillow patient lightbulb
of ce doctor stethoscope surgeon pickle desk courage
fi
fi
“All remembering is constructive in nature…The
illusion of remembering events that never happened
can occur quite readily. Therefore, as others have
also pointed out, the fact that people may say they
vividly remember details surrounding an event
cannot, in itself, be taken as convincing evidence
that the event actually occurred.”
Roediger & McDermott (1995, p.812)
“Remembering” a crash
• Loftus & Palmer
(1974):
• Participants
watched a video of
a car accident.
• How fast were the
cars going when
they contacted/
bumped/smashed
each other?
“Remembering” a crash
• Loftus & Palmer (1974):
• Estimations of speed differed based on the verb.
• Contacted < Bumped < Smashed
“Remembering” a hit
• Attempts to replicate the
original result have been
mixed.
• Godschmied et al. (2017)
replicated the experiment in a
hockey setting.
• NHL collision video.
• Manipulated audio included
crowd response.
“Remembering” a hit
• Questionnaire after the video
described contact/bump/
smash.
• The original experiment failed
to nd an effect of verb
condition.
• BUT - a subsequent
experiment with
“sportscaster” commentary
did nd a difference.
fi
fi
Section summary
• Human memory is a constructive process.
• In many ways, our recall of an event may not represent
what actually happened.
• The conditions under which the event was encoded
and retrieved play a role.
General summary
• Today we saw that human memory is not a unitary
phenomenon.
• Experimental studies and studies of memory disorders
show that memory is separable both behaviourally and
in the brain.
• We also saw that recall of events is in uenced by framing
of those events, for example by the language used to
describe them.
fl
