C H A P T E R 6
L E A R N I N G , M E M O R Y ,
A N D F O R G E T T I N G
ARCHITECTURE OF
MEMORY
Throughout most of the history of memory
research, it has been assumed that there is an
important distinction between short-term memory
and long-term memory. It seems reasonable that
the processes involved in briefly remembering
a telephone number are very different from those
involved in long-term memory for theories and
research in psychology. This traditional view
is at the heart of multi-store models, which are
discussed initially. In recent times, however, some
theorists have argued in favour of unitary-store
models in which the distinction between shortterm
and long-term memory is much less clear-cut
than in the traditional approach. We will consider
unitary-store models shortly.
Multi-store model
Several memory theorists (e.g., Atkinson &
Shiffrin, 1968) have described the basic architecture
of the memory system. We can identify
a multi-store approach based on the common
features of their theories. Three types of memory
store were proposed:
Sensory stores, each holding information•
very briefly and being modality specific
(limited to one sensory modality).
Short-term store of very limited capacity.•
Long-term store of essentially unlimited•
capacity holding information over very long
periods of time.
INTRODUCTION
This chapter and the next two are concerned
with human memory. All three chapters deal with
intact human memory, but Chapter 7 also considers
amnesic patients. Traditional laboratorybased
research is the focus of this chapter, with
more naturalistic research being discussed in
Chapter 8. As we will see, there are important
links among these different types of research.
Many theoretical issues are relevant to braindamaged
and healthy individuals whether tested
in the laboratory or in the field.
Theories of memory generally consider both
the architecture of the memory system and the
processes operating within that structure. Architecture
refers to the way in which the memory
system is organised and processes refer to the
activities occurring within the memory system.
Learning and memory involve a series of
stages. Processes occurring during the presentation
of the learning material are known as
“encoding” and involve many of the processes
involved in perception. This is the first stage. As
a result of encoding, some information is stored
within the memory system. Thus, storage is the
second stage. The third (and final) stage is retrieval,
which involves recovering or extracting stored
information from the memory system.
We have emphasised the distinctions between
architecture and process and among encoding,
storage, and retrieval. However, we cannot have
architecture without process, or retrieval without
previous encoding and storage.
206 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
The basic multi-store model is shown in
Figure 6.1. Environmental stimulation is initially
received by the sensory stores. These stores
are modality-specific (e.g., vision, hearing).
Information is held very briefly in the sensory
stores, with some being attended to and processed
further by the short-term store. Some
information processed in the short-term store
is transferred to the long-term store. Longterm
storage of information often depends on
rehearsal. There is a direct relationship between
the amount of rehearsal in the short-term store
and the strength of the stored memory trace.
There is much overlap between the areas of
attention and memory. Broadbent’s (1958) theory
of attention (see Chapter 5) was the main influence
on the multi-store approach to memory. For
example, Broadbent’s buffer store resembles
the notion of a sensory store.
Sensory stores
The visual store is often known as the iconic
store. In Sperling’s (1960) classic work on this
store, he presented a visual array containing three
rows of four letters each for 50 ms. Participants
could usually report only 4–5 letters, but claimed
to have seen many more. Sperling assumed this
happened because visual information had faded
before most of it could be reported. He tested
this by asking participants to recall only part
of the information presented. Sperling’s results
supported his assumption, with part recall being
good provided that the information to be recalled
was cued very soon after the offset of the visual
display.
Sperling’s (1960) findings suggested that
information in iconic memory decays within
about 0.5 seconds, but this may well be an underestimate.
Landman, Spekreijse, and Lamme (2003)
pointed out that the requirement to verbally
identify and recall items in the part-recall condition
may have interfered with performance. They
imposed simpler response demands on participants
(i.e., is a second stimulus the same as the
first one?) and found that iconic memory lasted
for up to about 1600 ms (see Figure 4.12).
Iconic storage is very useful for two reasons.
First, the mechanisms responsible for visual
perception always operate on the icon rather
than directly on the visual environment. Second,
information remains in iconic memory for
upwards of 500 ms, and we can shift our
attention to aspects of the information within
iconic memory in approximately 55 ms (Lachter,
Forster, & Ruthruff, 2004; see Chapter 5).
This helps to ensure we attend to important
information.
The transient auditory store is known
as the echoic store. In everyday life, you may
sometimes have been asked a question while
your mind was on something else. Perhaps you
replied, “What did you say?”, just before realising
that you do know what had been said.
This “playback” facility depends on the echoic
store. Estimates of the duration of information
in the echoic store are typically within the
range of 2–4 seconds (Treisman, 1964).
Sensory
stores
Short-term
store
Long-term
store
RehearsalAttention
Decay Displacement InterferenceFigure 6.1 The multi-store
model of memory.
echoic store: a sensory store in which
auditory information is briefly held.
KEY TERM
6 LEARNING, MEMORY, AND FORGETTING 207
Short- and long-term stores
The capacity of short-term memory is very
limited. Consider digit span: participants listen
to a random series of digits and then repeat
them back immediately in the correct order.
Other span measures are letter span and word
span. The maximum number of units (e.g.,
digits) recalled without error is usually “seven
plus or minus two” (Miller, 1956). However,
there are two qualifications concerning that
finding. First, Miller (1956) argued that the
capacity of short-term memory should be
assessed by the number of chunks (integrated
pieces or units of information). For example,
“IBM” is one chunk for those familiar with
the company name International Business
Machines but three chunks for everyone else.
The capacity of short-term memory is often
seven chunks rather than seven items. However,
Simon (1974) found that the span in chunks
was less with larger chunks (e.g., eight-word
phrases) than with smaller chunks (e.g., onesyllable
words).
Second, Cowan (2000, p. 88) argued that
estimates of short-term memory capacity are
often inflated because participants’ performance
depends in part on rehearsal and on long-term
memory. When these additional factors are
largely eliminated, the capacity of short-term
memory is typically only about four chunks.
For example, Cowan et al. (2005) used the
running memory task – a series of digits ended
at an unpredictable point, with the participants’
task being to recall the items from the end of
the list. The digits were presented very rapidly
to prevent rehearsal, and the mean number of
items recalled was 3.87.
The recency effect in free recall (recalling
the items in any order) refers to the finding
that the last few items in a list are usually much
better remembered in immediate recall than
those from the middle of the list. Counting
backwards for 10 seconds between the end
of list presentation and start of recall mainly
affects the recency effect (Glanzer & Cunitz,
1966; see Figure 6.2). The two or three words
susceptible to the recency effect may be in the
short-term store at the end of list presentation
and so especially vulnerable. However, Bjork
0.80
0.70
0.60
0.50
0.40
0.30
0.20
1 5 10 15
Serial position
Probabilitycorrect
0-sec delay
10-sec delay
30-sec delay
Figure 6.2 Free recall as
a function of serial position
and duration of the
interpolated task.Adapted
from Glanzer and Cunitz
(1966).
chunk: a stored unit formed from integrating
smaller pieces of information.
recency effect: the finding that the last few
items in a list are much better remembered than
other items in immediate free recall.
KEY TERMS
208 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
and Whitten (1974) found that there was still
a recency effect when participants counted
backwards for 12 seconds after each item in
the list was presented. According to Atkinson
and Shiffrin (1968), this should have eliminated
the recency effect.
The above findings can be explained by
analogy to looking along a row of telephone
poles. The closer poles are more distinct than
the ones farther away, just as the most recent
list words are more discriminable than the
others (Glenberg, 1987).
Peterson and Peterson (1959) studied the
duration of short-term memory by using the
task of remembering a three-letter stimulus
while counting backwards by threes followed
by recall in the correct order. Memory performance
reduced to about 50% after 6 seconds
and forgetting was almost complete after 18
seconds (see Figure 6.3), presumably because
unrehearsed information disappears rapidly
from short-term memory through decay (see
Nairne, 2002, for a review). In contrast, it is
often argued that forgetting from long-term
memory involves different mechanisms. In particular,
there is much cue-dependent forgetting,
in which the memory traces are still in the
memory system but are inaccessible (see later
discussion).
Nairne, Whiteman, and Kelley (1999) argued
that the rate of forgetting observed by Peterson
and Peterson (1959) was especially rapid for
two reasons. First, they used all the letters of
the alphabet repeatedly, which may have caused
considerable interference. Second, the memory
task was difficult in that participants had to
remember the items themselves and the presentation
order. Nairne et al. presented different
words on each trial to reduce interference, and
tested memory only for order information and
not for the words themselves. Even though
there was a rehearsal-prevention task (reading
aloud digits presented on a screen) during the
retention interval, there was remarkably little
forgetting even over 96 seconds (see Figure 6.4).
100
90
80
70
60
50
40
30
20
10
0
0 3 6 9 12 15 18
Retention interval (s)
Letterrecall(%)
Figure 6.3 Forgetting over
time in short-term memory.
Data from Peterson and
Peterson (1959).
0.85
0.80
0.75
0.70
0.65
0.60
4 24 48 96
Retention interval (secs)
Proportioncorrectresponses
Figure 6.4 Proportion of correct responses as a
function of retention interval. Data from Nairne
et al. (1999).
6 LEARNING, MEMORY, AND FORGETTING 209
This finding casts doubt on the notion that
decay causes forgetting in short-term memory.
However, reading digits aloud may not have
totally prevented rehearsal.
Finally, we turn to the strongest evidence that
short-term and long-term memory are distinct.
If short-term and long-term memory are separate,
we might expect to find some patients with
impaired long-term memory but intact shortterm
memory and others showing the opposite
pattern. This would produce a double dissociation.
The findings are generally supportive.
Patients with amnesia (discussed in Chapter 7)
have severe impairments of many aspects of
long-term memory, but typically have no problem
with short-term memory (Spiers, Maguire,
& Burgess, 2001). Amnesic patients have damage
to the medial temporal lobe, including the
hippocampus (see Chapter 7), which primarily
disrupts long-term memory (see Chapter 7).
A few brain-damaged patients have severely
impaired short-term memory but intact long-term
memory. For example, KF had no problems
with long-term learning and recall but had a
very small digit span (Shallice & Warrington,
1970). Subsequent research indicated that his
short-term memory problems focused mainly
on recall of letters, words, or digits rather than
meaningful sounds or visual stimuli (e.g., Shallice
& Warrington, 1974). Such patients typically
have damage to the parietal and temporal lobes
(Vallar & Papagno, 2002).
Evaluation
The multi-store approach has various strengths.
The conceptual distinction between three kinds
of memory store (sensory store, short-term store,
and long-term store) makes sense. These memory
stores differ in several ways:
temporal duration•
storage capacity•
forgetting mechanism(s)•
effects of brain damage•
Finally, many subsequent theories of human
memory have built on the foundations of the multistore
model, as we will see later in this chapter.
However, the multi-store model possesses
several serious limitations. First, it is very oversimplified.
It was assumed that the short-term
and long-term stores are both unitary, i.e., each
store always operates in a single, uniform way.
As we will see shortly, Baddeley and Hitch (1974)
proposed replacing the concept of a single shortterm
store with a working memory system
consisting of three different components. That
is a more realistic approach. In similar fashion,
there are several long-term memory systems
(see Chapter 7).
Second, it is assumed that the short-term
store acts as a gateway between the sensory
stores and long-term memory (see Figure 6.1).
However, the information processed in the
short-term store has already made contact
with information stored in long-term memory
(Logie, 1999). For example, consider the phonological
similarity effect: immediate recall of
visually presented words in the correct order
is worse when they are phonologically similar
(sounding similar) (e.g., Larsen, Baddeley, &
Andrade, 2000). Thus, information about the
sounds of words stored in long-term memory
affects processing in short-term memory.
Third, Atkinson and Shiffrin (1968) assumed
that information in short-term memory represents
the “contents of consciousness”. This implies
that only information processed consciously
can be stored in long-term memory. However,
learning without conscious awareness of what
has been learned (implicit learning) appears to
exist (see later in the chapter).
Fourth, multi-store theorists assumed that
most information is transferred to long-term
memory via rehearsal. However, the role of
rehearsal in our everyday lives is very limited.
More generally, multi-store theorists focused
too much on structural aspects of memory rather
than on memory processes.
Unitary-store models
In recent years, various theorists have argued
that the entire multi-store approach is misguided
and should be replaced by a unitary-store model
(see Jonides, Lewis, Nee, Lustig, Berman, &
210 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Moore, 2008, for a review). Unitary-store models
assume that, “STM [short-term memory] consists
of temporary activations of LTM [long-term
memory] representations or of representations
of items that were recently perceived” (Jonides
et al., 2008, p. 198). Such activations will often
occur when certain representations are the focus
of attention.
Unitary-store models would seem to have
great difficulty in explaining the consistent finding
that amnesic patients have essentially intact
short-term memory in spite of having severe
problems with long-term memory. Jonides et al.
(2008) argued that amnesic patients have special
problems in forming novel relations (e.g., between
items and their context) in both short-term and
long-term memory. Amnesic patients apparently
have no problems with short-term memory
because short-term memory tasks typically do
not require relational memory. This leads to a
key prediction: amnesic patients should have
impaired short-term memory performance on
tasks requiring relational memory.
According to Jonides et al. (2008), the
hippocampus and surrounding medial temporal
lobes (typically damaged in amnesic patients)
play a crucial role in forming novel relations
(sometimes called binding) (see Chapter 7).
Multi-store theorists assume that these structures
are much more involved in long-term
memory than in short-term memory. However,
it follows from unitary-store models that the
hippocampus and medial temporal lobes would
be involved if a short-term memory task required
forming novel relations.
Evidence
Evidence supporting the unitary-store approach
was reported by Hannula, Tranel, and Cohen
(2006). They studied patients who had become
amnesic as the result of an anoxic episode
(involving deficient oxygen supply). In one experiment,
scenes were presented for 20 seconds.
Some scenes were repeated exactly, whereas others
were repeated with one object having been
moved spatially. Participants decided whether
each scene had been seen previously. It was
assumed that short-term memory was involved
when a given scene was repeated in its original
or slightly modified form immediately after its
initial presentation (Lag 1) but that long-term
memory was involved at longer lags.
The findings are shown in Figure 6.5. Amnesic
patients performed much worse than healthy
controls in short-term memory (Lag 1) and the
performance difference between the two groups
was even larger in long-term memory. The crucial
issue is whether performance at Lag 1 was only
due to short-term memory. The finding that
amnesics’ performance fell to chance level at
longer lags suggests that they may well have
relied almost exclusively on short-term memory
at Lag 1. However, the finding that controls’ performance
changed little over lags suggests that
they formed strong long-term relational memories,
and these long-term memories may well account
for their superior performance at Lag 1.
Further support for the unitary-store approach
was reported by Hannula and Ranganath (2008).
They presented four objects in various locations
and instructed participants to rotate the
display mentally. Participants were then presented
with a second display, and decided whether the
second display matched or failed to match their
mental representation of the rotated display.
This task involved relational memory. The
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
Comparison group
Anoxic patients
Proportioncorrect
Lag 1 Lag 5 Lag 9
Figure 6.5 Proportion of correct responses for
healthy controls (comparison group) and amnesics
(anoxic patients).The dashed line represents chance
performance. From Hannula et al. (2006) with
permission from Society of Neuroscience.
6 LEARNING, MEMORY, AND FORGETTING 211
key finding was that the amount of activation
in the anterior and posterior regions of the
left hippocampus predicted relational memory
performance.
Shrager, Levy, Hopkins, and Squire (2008)
pointed out that a crucial issue is whether memory
performance at short retention intervals actually
depends on short-term memory rather than
long-term memory. They argued that a distinguishing
feature of short-term memory is that
it involves active maintenance of information
throughout the retention interval. Tasks
that mostly depend on short-term memory are
vulnerable to distraction during the retention
interval because distraction disrupts active maintenance.
Shrager et al. divided their memory
tasks into those susceptible to distraction in
healthy controls and those that were not. Amnesic
patients with medial temporal lobe lesions
had essentially normal levels of performance
on distraction-sensitive memory tasks but were
significantly impaired on distraction-insensitive
memory tasks. Shrager et al. concluded that
short-term memory processes are intact in
amnesic patients. Amnesic patients only show
impaired performance on so-called “short-term
memory tasks” when those tasks actually
depend substantially on long-term memory.
Evaluation
The unitary-store approach has made memory
researchers think deeply about the relationship
between short-term and long-term memory.
There are good reasons for accepting the notion
that activation of part of long-term memory
plays an important role in short-term memory.
According to the unitary-store approach (but
not the multi-store approach), amnesic patients
can exhibit impaired short-term memory under
some circumstances. Some recent evidence (e.g.,
Hannula et al., 2006) supports the prediction
of the unitary-store approach. Functional neuroimaging
evidence (e.g., Hannula & Ranganath,
2008) also provides limited support for the
unitary-store approach.
What are the limitations of the unitarystore
approach? First, it is oversimplified to
argue that short-term memory is only activated
by long-term memory. We can manipulate
activated long-term memory in flexible ways
and such manipulations go well beyond simply
activating some fraction of long-term memory.
Two examples of ways in which we can manipulate
information in short-term memory are
backward digit recall (recalling digits in the
opposite order to the presentation order) and
generating novel visual images (Logie & van
der Meulen, 2009). Second, there is no convincing
evidence that amnesic patients have
impaired performance on relational memory
tasks dependent primarily on short-term memory.
It seems likely that amnesic patients only perform
poorly on “short-term memory” tasks that
depend to a large extent on long-term memory
(Shrager et al., 2008). Third, there is no other
evidence that decisively favours the unitary-store
approach over the multiple-store approach.
However, the search for such evidence only
recently started in earnest.
WORKING MEMORY
Baddeley and Hitch (1974) and Baddeley (1986)
replaced the concept of the short-term store
with that of working memory. Since then, the
conceptualisation of the working memory system
has become increasingly complex. According
to Baddeley (2001) and Repovš and Baddeley
(2006), the working memory system has four
components (see Figure 6.6):
A modality-free• central executive resembling
attention.
A• phonological loop holding information
in a phonological (speech-based) form.
central executive: a modality-free, limited
capacity, component of working memory.
phonological loop: a component of working
memory, in which speech-based information is
held and subvocal articulation occurs.
KEY TERMS
212 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
A• visuo-spatial sketchpad specialised for
spatial and visual coding.
An• episodic buffer, which is a temporary
storage system that can hold and integrate
information from the phonological loop,
the visuo-spatial sketchpad, and long-term
memory. This component (added 25 years
after the others) is discussed later.
The most important component is the
central executive. It has limited capacity, resembles
attention, and deals with any cognitively
demanding task. The phonological loop and
the visuo-spatial sketchpad are slave systems
used by the central executive for specific purposes.
The phonological loop preserves the order
in which words are presented, and the visuospatial
sketchpad stores and manipulates spatial
and visual information. All three components have
limited capacity and are relatively independent
of each other. Two assumptions follow:
If two tasks use the same component, they(1)
cannot be performed successfully together.
If two tasks use different components,(2)
it should be possible to perform them as
well together as separately.
Numerous dual-task studies have been
carried out on the basis of these assumptions.
For example, Robbins et al. (1996) considered
the involvement of the three original components
of working memory in the selection of
chess moves by weaker and stronger players.
The players selected continuation moves from
various chess positions while also performing
one of the following tasks:
Repetitive tapping• : this was the control
condition.
Random number generation• : this involved
the central executive.
Pressing keys on a keypad in a clockwise•
fashion: this used the visuo-spatial sketchpad.
Rapid repetition of the word “see-saw”• :
this is articulatory suppression and uses the
phonological loop.
Robbins et al. (1996) found that selecting
chess moves involved the central executive
and the visuo-spatial sketchpad but not the
phonological loop (see Figure 6.7). The effects
of the various additional tasks were similar on
stronger and weaker players, suggesting that
Rehearsal Rehearsal
Episodic buffer
Holds and integrates
diverse information
Phonological loop
(inner voice)
Holds information in
a speech-based form
Visuo-spatial sketchpad
(inner eye)
Specialised for spatial
and/or visual coding
CENTRAL
EXECUTIVE
Figure 6.6 The major
components of Baddeley’s
working memory system.
Figure adapted from
Baddeley (2001).
visuo-spatial sketchpad: a component of
working memory that is involved in visual and
spatial processing of information.
episodic buffer: a component of working
memory that is used to integrate and to store
briefly information from the phonological
loop, the visuo-spatial sketchpad, and
long-term memory.
articulatory suppression: rapid repetition of
some simple sound (e.g.,“the, the, the”), which
uses the articulatory control process of the
phonological loop.
KEY TERMS
6 LEARNING, MEMORY, AND FORGETTING 213
both groups used the working memory system
in the same way.
Phonological loop
Most early research on the phonological loop
focused on the notion that verbal rehearsal
(i.e., saying words over and over to oneself) is of
central importance. Two phenomena providing
support for this view are the phonological
similarity effect and the word-length effect.
The phonological similarity effect is found
when a short list of visually presented words
is recalled immediately in the correct order.
Recall performance is worse when the words
are phonologically similar (i.e., having similar
sounds) than when they are phonologically dissimilar.
For example, FEE, HE, KNEE, LEE, ME,
and SHE form a list of phonologically similar
words, whereas BAY, HOE, IT, ODD, SHY,
and UP form a list of phonologically dissimilar
words. Larsen, Baddeley, and Andrade (2000)
used those word lists, finding that recall of
the words in order was 25% worse with the
phonologically similar list. This phonological
similarity effect occurred because participants
used speech-based rehearsal processes within
the phonological loop.
The word-length effect is based on memory
span (the number of words or other items recalled
immediately in the correct order). It is defined
by the finding that memory span is lower
for words taking a long time to say than for
30
25
20
15
10
Control Articulatory
suppression
Visuo-spatial
sketchpad
suppression
Central
executive
suppression
Secondary task
Weaker chess players
Stronger chess players
Qualityofchess-moveselection
Figure 6.7 Effects of
secondary tasks on quality
of chess-move selection in
stronger and weaker players.
Adapted from Robbins et al.
(1996).
According to Robbins et al. (1996), selecting
good chess moves requires use of the central
executive and the visuo-spatial sketchpad, but
not of the phonological loop.
phonological similarity effect: the finding
that serial recall of visually presented words is
worse when the words are phonologically
similar rather than phonologically dissimilar.
word-length effect: the finding that word span
is greater for short words than for long words.
KEY TERMS
214 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
those taking less time. Baddeley, Thomson, and
Buchanan (1975) found that participants recalled
as many words presented visually as they could
read out loud in 2 seconds. This suggested that
the capacity of the phonological loop is determined
by temporal duration like a tape loop.
Service (2000) argued that these findings depend
on phonological complexity rather than on
temporal duration. Reassuringly, Mueller, Seymour,
Kieras, and Meyer (2003) found with very carefully
chosen words that memory span depended
on the articulatory duration of words rather
than their phonological complexity.
In another experiment, Baddeley et al.
(1975) obtained more direct evidence that the
word-length effect depends on the phonological
loop. The number of visually presented words
(out of five) that could be recalled was assessed.
Some participants were given the articulatory
suppression task of repeating the digits 1 to 8
while performing the main task. The argument
was that the articulatory suppression task would
involve the phonological loop and so prevent it
being used on the word-span task. As predicted,
articulatory suppression eliminated the wordlength
effect (see Figure 6.8), suggesting it
depends on the phonological loop.
As so often in psychology, reality is more
complex than was originally thought. Note that
the research discussed so far involved the visual
presentation of words. Baddeley et al. (1975)
obtained the usual word-length effect when there
was auditory presentation of word lists. Puzzlingly,
however, there was still a word-length effect with
auditorily presented words even when articulatory
suppression was used (see Figure 6.8). This led
90
80
70
60
50
40
30
Short words Long words
Meanpercentageofwordscorrectlyrecalled
Auditory presentation;
no suppression
Visual presentation;
no suppression
Auditory presentation;
suppression
Visual presentation;
suppression
(%)
Figure 6.8 Immediate word
recall as a function of
modality of presentation
(visual vs. auditory), presence
versus absence of
articulatory suppression, and
word length.Adapted from
Baddeley et al. (1975).
Auditory
word
presentation
Phonological
store
Articulatory
control
process
Visual word
presentation
Figure 6.9 Phonological
loop system as envisaged by
Baddeley (1990).
6 LEARNING, MEMORY, AND FORGETTING 215
Baddeley (1986, 1990; see Figure 6.9) to argue that
the phonological loop has two components:
A passive phonological store directly concerned•
with speech perception.
An articulatory process linked to speech•
production that gives access to the phonological
store.
According to this account, words presented
auditorily are processed differently from those
presented visually. Auditory presentation of
words produces direct access to the phonological
store regardless of whether the articulatory
control process is used. In contrast, visual
presentation of words only permits indirect
access to the phonological store through subvocal
articulation.
The above account makes sense of many
findings. Suppose the word-length effect observed
by Baddeley et al. (1975) depends on the rate
of articulatory rehearsal (see Figure 6.8). Articulatory
suppression eliminates the word-length
effect with visual presentation because access
to the phonological store is prevented. However,
it does not affect the word-length effect with
auditory presentation because information about
the words enters the phonological store directly.
Progress has been made in identifying the
brain areas associated with the two components
of the phonological loop. Some brain-damaged
patients have very poor memory for auditoryverbal
material but essentially normal speech
production, indicating they have a damaged
phonological store but an intact articulatory
control process. These patients typically have
damage to the left inferior parietal cortex (Vallar
& Papagno, 1995). Other brain-damaged patients
have an intact phonological store but a damaged
articulatory control process shown by a lack of
evidence for rehearsal. Such patients generally
have damage to the left inferior frontal cortex.
Similar brain areas have been identified
in functional neuroimaging studies on healthy
volunteers. Henson, Burgess, and Frith (2000)
found that a left inferior parietal area was
associated with the phonological store, whereas
left prefrontal cortex was associated with rehearsal.
Logie, Venneri, Della Sala, Redpath, and Marshall
(2003) gave their participants the task of recalling
letter sequences presented auditorily in the
correct order. All participants were instructed to
use subvocal rehearsal to ensure the involvement
of the rehearsal component of the phonological
loop. The left inferior parietal gyrus and the inferior
and middle frontal gyri were activated.
Evaluation
Baddeley’s theory accounts for the word-length
effects and for the effects of articulatory suppression.
In addition, evidence from brain-damaged
patients and from functional neuroimaging
studies with healthy participants indicates the
existence of a phonological store and an articulatory
control process located in different brain
regions. Our understanding of the phonological
loop is greater than that for the other components
of the working memory system.
What is the value of the phonological loop?
According to Baddeley, Gathercole, and Papagno
(1998, p. 158), “The function of the phonological
loop is not to remember familiar words
but to learn new words.” Supporting evidence
was reported by Papagno, Valentine, and Baddeley
(1991). Native Italian speakers learned pairs
of Italian words and pairs of Italian–Russian
words. Articulatory suppression (which reduces
use of the phonological loop) greatly slowed
the learning of foreign vocabulary but had little
effect on the learning of pairs of Italian words.
Several studies have considered the relationship
between children’s vocabulary development
and their performance on verbal short-term
memory tasks involving the phonological loop.
The capacity of the phonological loop generally
predicts vocabulary size (e.g., Majerus, Poncelet,
Elsen, & van der Linden, 2006). Such evidence
is consistent with the notion that the phonological
loop plays a role in the learning of
vocabulary. However, much of the evidence is
correlational – it is also possible that having a
large vocabulary increases the effective capacity
of the phonological loop.
Trojano and Grossi (1995) studied SC,
a patient with extremely poor phonological
functioning. SC showed reasonable learning
216 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
ability in most situations but was unable to
learn auditorily presented word–nonword pairs.
Presumably SC’s poorly functioning phonological
loop prevented the learning of the phonologically
unfamiliar nonwords.
Visuo-spatial sketchpad
The visuo-spatial sketchpad is used for the
temporary storage and manipulation of visual
patterns and spatial movement. It is used in many
situations in everyday life (e.g., finding the route
when walking; playing computer games). Logie,
Baddeley, Mane, Donchin, and Sheptak (1989)
studied performance on a complex computer game
called Space Fortress, which involves manoeuvring
a space ship around a computer screen. Early
in training, performance on Space Fortress was
severely impaired when participants had to performasecondary
visuo-spatial task. After 25 hours’
training, the adverse effects on the computer
game of carrying out a visuo-spatial task at the
same time were greatly reduced, being limited to
those aspects directly involving perceptuo-motor
control. Thus, the visuo-spatial sketchpad was
used throughout training on Space Fortress, but
its involvement decreased with practice.
The most important issue is whether there
is a single system combining visual and spatial
processing or whether there are partially or
completely separate visual and spatial systems.
According to Logie (1995; see Figure 6.10),
the visuo-spatial sketchpad consists of two
components:
Visual cache• : This stores information about
visual form and colour.
Inner scribe• : This processes spatial and
movement information. It is involved in the
rehearsal of information in the visual cache
and transfers information from the visual
cache to the central executive.
Recent developments in theory and research
on the visuo-spatial sketchpad are discussed
by Logie and van der Meulen (2009).
Klauer and Zhao (2004) explored the issue
of whether there are separate visual and spatial
systems. They used two main tasks – a spatial
task (memory for dot locations) and a visual
task (memory for Chinese ideographs). There
were also three secondary task conditions:
A movement discrimination task (spatial•
interference).
A colour discrimination task (visual•
interference).
A control condition (no secondary task).•
What would we expect if there are somewhat
separate visual and spatial systems? First,
the spatial interference task should disrupt
performance more on the spatial main task
than on the visual main task. Second, the visual
interference task should disrupt performance
more on the visual main task than on the
spatial main task. Both predictions were supported
(see Figure 6.11).
Additional evidence supporting the notion
of separate visual and spatial systems was
reported by Smith and Jonides (1997) in an
ingenious study. Two visual stimuli were presented
together, followed by a probe stimulus.
Inner scribe
(active
rehearsal)
Visual cache
(stores visual
information)
Central
executive
Figure 6.10 The visuo-spatial sketchpad or working
memory as envisaged by Logie.Adapted from Logie
(1995), Baddeley, Mane, Donchin, and Sheptak.
visual cache: according to Logie, the part of
the visuo-spatial sketchpad that stores
information about visual form and colour.
inner scribe: according to Logie, the part of
the visuo-spatial sketchpad that deals with
spatial and movement information.
KEY TERMS
6 LEARNING, MEMORY, AND FORGETTING 217
Participants decided whether the probe was
in the same location as one of the initial stimuli
(spatial task) or had the same form (visual
task). Even though the stimuli were identical
in the two tasks, there were clear differences
in patterns of brain activation. There was more
activity in the right hemisphere during the spatial
task than the visual task, but more activity
in the left hemisphere during the visual task
than the spatial task.
Several other studies have indicated that
different brain regions are activated during
visual and spatial working-memory tasks (see
Sala, Rämä, & Courtney, 2003, for a review).
The ventral prefrontal cortex (e.g., the inferior
and middle frontal gyri) is generally activated
more during visual working-memory tasks than
spatial ones. In contrast, more dorsal prefrontal
cortex (especially an area of the superior
prefrontal sulcus) tends to be more activated
during spatial working-memory tasks than
visual ones. This separation between visual and
spatial processing is consistent with evidence
that rather separate pathways are involved in
visual and spatial perceptual processing (see
Chapter 2).
Evaluation
Various kinds of evidence support the view that
the visuo-spatial sketchpad consists of somewhat
separate visual (visual cache) and spatial
(inner scribe) components. First, there is often
little interference between visual and spatial
tasks performed at the same time (e.g., Klauer
& Zhao, 2004). Second, functional neuroimaging
data suggest that the two components of the
visuo-spatial sketchpad are located in different
brain regions (e.g., Sala et al., 2003; Smith &
Jonides, 1997). Third, some brain-damaged
patients have damage to the visual component but
not to the spatial component. For example, NL
found it very hard to describe details from the
left side of scenes in visual imagery even though
his visual perceptual system was essentially intact
(Beschin, Cocchini, Della Sala, & Logie, 1997).
Many tasks require both components of
the visuo-spatial sketchpad to be used in combination.
It remains for the future to understand
more fully how processing and information from
the two components are combined and integrated
onsuchtasks.Inaddition,muchremainsunknown
about interactions between the workings of the
visuo-spatial sketchpad and the episodic buffer
(Baddeley, 2007).
Central executive
The central executive (which resembles an
attentional system) is the most important and
versatile component of the working memory
system. Every time we engage in any complex
cognitive activity (e.g., reading a text; solving
a problem; carrying out two tasks at the same
time), we make considerable use of the central
executive. It is generally assumed that the prefrontal
cortex is the part of the brain most
involved in the functions of the central executive.
Mottaghy (2006) reviewed studies using
repetitive transcranial magnetic stimulation
(rTMS; see Glossary) to disrupt activity within
the dorsolateral prefrontal cortex. Performance
on many complex cognitive tasks was impaired
by this manipulation, indicating that dorsolateral
prefrontal cortex is of importance in central
executive functions. However, we need to be
20
15
10
5
0
–5
Movement
Colour
Dots Ideographs
Interferencescores(%)
Figure 6.11 Amount of interference on a spatial
task (dots) and a visual task (ideographs) as a
function of secondary task (spatial: movement vs.
visual: colour discrimination). From Klauer and
Zhao (2004), Copyright © 2000 American
Psychological Association. Reproduced with
permission.
218 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
careful about associating the central executive
too directly with prefrontal cortex. As Andrés
(2003) pointed out, patients with damage to
prefrontal cortex do not always show executive
deficits, and some patients with no damage to
prefrontal cortex nevertheless have executive
deficits.
One way of trying to understand the importance
of the central executive in our everyday
functioning is to study brain-damaged individuals
whose central executive is impaired. Such
individuals suffer from dysexecutive syndrome
(Baddeley, 1996), which involves problems with
planning, organising, monitoring behaviour,
and initiating behaviour. Patients with dysexecutive
syndrome typically have damage within the
frontal lobes at the front of the brain (adverse
effects of damage to the prefrontal cortex on
problem solving are discussed in Chapter 12).
However, some patients seem to have damage to
posterior (mainly parietal) rather than to frontal
regions (e.g., Andrés, 2003). Brain-damaged
patients are often tested with the Behavioural
Assessment of the Dysexecutive Syndrome (BADS;
Wilson, Alderman, Burgess, Emslie, & Evans,
1996). This consists of various tests assessing the
ability to shift rules, to devise and implement a
solution to a practical problem, to divide time
effectively among various tasks, and so on.
Individuals with dysexecutive syndrome as assessed
by the BADS typically have great problems in
holding down a job and functioning adequately
in everyday life (Chamberlain, 2003).
The conceptualisation of the central executive
has changed over time. As Repovš and
Baddeley (2006, p. 12) admitted, it was originally
“a convenient ragbag for unanswered questions
related to the control of working memory
and its two slave subsystems.” In the original
model, the central executive was unitary, meaning
that it functioned as a single unit. In recent
years, theorists have increasingly argued that
the central executive is more complex. Baddeley
(1996) suggested that four of the functions of
the central executive were as follows: switching
of retrieval plans; timesharing in dual-task
studies; selective attention to certain stimuli
while ignoring others; and temporary activation
of long-term memory. These are examples of
executive processes, which are processes that serve
to organise and co-ordinate the functioning of
the cognitive system to achieve current goals.
Miyake et al. (2000) identified three executive
processes or functions overlapping partially
with those of Baddeley (1996). They assumed
these functions were related but separable:
Inhibition function• : This refers to “one’s
ability to deliberately inhibit dominant,
automatic, or prepotent responses when
necessary” (p. 55). Friedman and Miyake
(2004) extended the inhibition function
to include resisting distractor interference.
For example, consider the Stroop task, on
which participants have to name the colours
in which words are printed. In the most
difficult condition, the words are conflicting
colour words (e.g., the word BLUE printed
in red). In this condition, performance is
slowed down and there are often many
errors. The inhibition function is needed to
minimise the distraction effect created by
the conflicting colour word. It is useful in
preventing us from thinking and behaving
in habitual ways when such ways are
inappropriate.
Shifting function• : This refers to “shifting back
and forth between multiple tasks, operations,
or mental sets” (p. 55). It is used
when you switch attention from one task
to another. Suppose, for example, you are
presented with a series of trials, on each of
which two numbers are presented. In one
dysexecutive syndrome: a condition in which
damage to the frontal lobes causes impairments
to the central executive component of
working memory.
executive processes: processes that organise
and co-ordinate the functioning of the cognitive
system to achieve current goals.
Stroop task: a task in which the participant has
to name the colours in which words are printed.
KEY TERMS
6 LEARNING, MEMORY, AND FORGETTING 219
condition, there is task switching: on some
trials you have to multiply the two numbers
and on other trials you have to divide one
by the other. In the other condition, there
are long blocks of trials on which you
always multiply the two numbers and there
are other long blocks of trials on which you
always divide one number by the other.
Performance is slower in the task-switching
condition, because attention has to be
switched backwards and forwards between
the two tasks. Task switching involves the
shifting function, which allows us to shift
attention rapidly from one task to another.
This is a very useful ability in today’s 24/7
world.
Updating function• : This refers to “updating
and monitoring of working memory representations”
(p. 55). It is used when you
update the information you need to remember.
For example, the updating function is
required when participants are presented
with members of various categories and have
to keep track of the most recently presented
member of each category. Updating is useful
if you are preparing a meal consisting of
several dishes or, more generally, if you are
trying to cope with changing circumstances.
Evidence
Various kinds of evidence support Miyake
et al.’s (2000) identification of three executive
functions. First, there are the findings from
their own research. They argued that most
cognitive tasks involve various processes, which
makes it difficult to obtain clear evidence for
any single process. Miyake et al. administered
several tasks to their participants and then used
latent-variable analysis. This form of analysis
focuses on positive correlations among tasks
as the basis for identifying the common process
or function involved. Thus, for example, three
tasks might all involve a common process
(e.g., the shifting function) but each task might
also involve additional specific processes. Latentvariable
analysis provides a useful way of
identifying the common process. Miyake et al.
found evidence for three separable executive
functions of inhibition, shifting, and monitoring,
but also discovered that these functions
were positively correlated with each other.
Second, Collette et al. (2005) administered
several tasks designed to assess the same three
executive processes, and used positron emission
tomography (PET; see Glossary) to compare
brain activation associated with each process.
There were two main findings. First, each executive
process or function was associated with
activation in a different region within the prefrontal
cortex. Second, all the tasks produced
activation in the right intraparietal sulcus, the
left superior parietal sulcus, and the left lateral
prefrontal cortex. Collette et al. suggested that
the right intraparietal sulcus is involved in
selective attention to relevant stimuli plus the
suppression of irrelevant information; the left
superior parietal sulcus is involved in switching
and integration processes; and the lateral prefrontal
cortex is involved in monitoring and
temporal organisation.
Are there executive processes or functions
not included within Miyake et al.’s (2000)
theory? According to Baddeley (1996), one
strong contender relates to the dual-task situation,
in which people have to perform two
different tasks at the same time. Executive processes
are often needed to co-ordinate processing
on the two tasks. Functional neuroimaging
studies focusing on dual-task situations have
produced somewhat variable findings (see
Chapter 5). However, there is sometimes much
activation in prefrontal areas (e.g., dorsolateral
prefrontal cortex) when people perform two
tasks at the same time but not when they perform
only one of the tasks on its own (e.g.,
Collette et al., 2005; Johnson & Zatorre, 2006).
Such findings suggest that co-ordination of two
tasks can involve an executive process based
mainly in the prefrontal cortex.
Further support for the notion that there
is an executive process involved specifically in
dual-task processing was reported by Logie,
Cocchini, Della Sala, and Baddeley (2004).
Patients with Alzheimer’s disease were compared
with healthy younger and older people
220 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
on digit recall and tracking tasks, the latter of
which involved keeping a pen on a red oval
that moved randomly. The Alzheimer’s patients
were much more sensitive than the healthy
groups to dual-task demands, but did not differ
in their ability to cope with single-task demands.
These findings suggest that Alzheimer’s patients
have damage to a part of the brain involved
in dual-task co-ordination. MacPherson, Della
Sala, Logie, and Wilcock (2007) reported very
similar findings using verbal memory and visuospatial
memory tasks.
Dysexecutive syndrome
Stuss and Alexander (2007) argued that the
notion of a dysexecutive syndrome is flawed
because it implies that brain damage to the
frontal lobes typically damages all central
executive functions of the central executive.
They accepted that patients with widespread
damage to the frontal lobes have a global dysexecutive
syndrome. However, they claimed
there are three executive processes based in
different parts of the frontal lobes:
Task setting• : This involves planning and
was defined as “the ability to set a stimulus–
response relationship . . . necessary in the early
stages of learning to drive a car or planning
a wedding” (p. 906).
Monitoring• : This was defined as “the process
of checking the task over time for ‘quality
control’ and the adjustment of behaviour”
(p. 909).
Energisation• : This involves sustained attention
or concentration and was defined as
“the process of initiation and sustaining of
any response. . . . Without energisation . . .
maintaining performance over prolonged
periods will waver” (pp. 903–904).
All three executive processes are very general
in that they are used across an enormous range
of tasks. They are not really independent,
because they are typically all used when you
deal with a complex task. For example, if
you have to give a speech in public, you would
first plan roughly what you are going to say
(task setting), concentrate through the delivery
of the speech (energisation), and check that
what you are saying is what you intended
(monitoring).
Stuss and Alexander (2007) tested their
theory of executive functions on patients having
fairly specific lesions within the frontal lobes.
In view of the possibility that there may be
reorganisation of cognitive structures and processes
following brain damage, the patients were
tested within a few months of suffering brain
damage. A wide range of cognitive tasks was
administered to different patient groups to try
to ensure that the findings would generalise.
Public speaking involves all three of Stuss and
Alexander’s (2007) executive functions: planning
what you are going to say (task setting);
concentrating on delivery (energisation); and
checking that what you say is as intended
(monitoring).
6 LEARNING, MEMORY, AND FORGETTING 221
Stuss and Alexander found evidence for
the three hypothesised processes of energisation,
task setting, and monitoring. They also
discovered that each process was associated
with a different region within the frontal cortex.
Energisation involves the superior medial region
of the frontal cortex, task setting involves the
left lateral frontal region, and monitoring involves
the right lateral frontal region. Thus, for example,
patients with damage to the right lateral frontal
region generally fail to detect the errors they
make while performing a task and so do not
adjust their performance.
Why do the processes identified by Stuss
and Alexander (2007) differ from those identified
by Miyake et al. (2000)? The starting point
in trying to answer that question is to remember
that Stuss and Alexander based their conclusions
on studies with brain-damaged patients,
whereas Miyake et al. studied only healthy
individuals. Nearly all executive tasks involve
common processes (e.g., energisation, task setting,
monitoring). These common processes are
positively correlated in healthy individuals and
so do not emerge clearly as separate processes.
However, the differences among energisation,
task setting, and monitoring become much
clearer when we consider patients with very
specific frontal lesions. It remains for future
research to show in more detail how the views
of Stuss and Alexander and of Miyake et al.
can be reconciled.
Evaluation
There has been real progress in understanding
the workings of the central executive. The
central executive consists of various related but
separable executive processes. There is accumulating
evidence that inhibition, updating,
shifting, and dual-task co-ordination may be
four major executive processes. It has become
clear that the notion of a dysexecutive syndrome
is misleading in that it suggests there is a single
pattern of impairment. Various executive processes
associated with different parts of frontal
cortex are involved.
Two issues require more research. First, the
executive processes suggested by behavioural
and functional neuroimaging studies on healthy
individuals do not correspond precisely with
those suggested by studies on patients with
damage to the frontal cortex. We have speculated
on the reasons for this, but solid evidence
is needed. Second, while we have emphasised
the differences among the major executive processes
or functions, there is plentiful evidence
suggesting that these processes are fairly closely
related to each other. The reasons for this
remain somewhat unclear.
Episodic buffer
Baddeley (2000) added a fourth component
to the working memory model. This is the
episodic buffer, in which information from
various sources (the phonological loop, the visuospatial
sketchpad, and long-term memory) can
be integrated and stored briefly. According to
Repovš and Baddeley (2006, p. 15), the episodic
buffer, “is episodic by virtue of holding
information that is integrated from a range of
systems including other working memory components
and long-term memory into coherent
complex structures: scenes or episodes. It is a
buffer in that it serves as an intermediary between
subsystems with different codes, which it combines
into multi-dimensional representations.”
In view of the likely processing demands
involved in integrating information from different
modalities, Baddeley (2000, 2007) suggested
that there would be close links between
the episodic buffer and the central executive.
If so, we would expect to find prefrontal activation
on tasks involving the episodic buffer, because
there are associations between use of the central
executive and prefrontal cortex.
episodic buffer: a component of working
memory that is used to integrate and to store
briefly information from the phonological
loop, the visuo-spatial sketchpad, and
long-term memory.
KEY TERM
222 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Why did Baddeley add the episodic buffer
to the working memory model? The original
version of the model was limited because its
various components were too separate in their
functioning.Forexample,Chincotta,Underwood,
Abd Ghani, Papadopoulou, and Wresinki (1999)
studied memory span for Arabic numerals and
digit words, finding that participants used both
verbal and visual encoding while performing
the task. This suggests that participants combined
information from the phonological loop
and the visuo-spatial sketchpad. Since these
two stores are separate, this combination and
integration process must take place elsewhere,
and the episodic buffer fits the bill.
Another finding hard to explain within
the original working memory model is that, in
immediate recall, people can recall about five
unrelated words but up to 16 words presented
in sentences (Baddeley, Vallar, & Wilson, 1987).
The notion of an episodic buffer is useful,
because this is where information from longterm
memory could be integrated with information
from the phonological loop and the
visuo-spatial sketchpad.
Evidence
Zhang et al. (2004) obtained evidence consistent
with the notion that the episodic buffer is often
used in conjunction with the central executive.
Their participants had to recall a mixture of
digits and visual locations, a task assumed to
require the episodic buffer. As predicted, there
was greater right prefrontal activation in this
condition than one in which digits and visual
locations were not mixed during presentation.
Baddeley and Wilson (2002) provided support
for the notion of an episodic buffer. They
pointed out that it had generally been assumed
that good immediate prose recall involves the
ability to store some of the relevant information
in long-term memory. According to this view,
amnesic patients with very impaired long-term
memory should have very poor immediate prose
recall. In contrast, Baddeley and Wilson argued
that the ability to exhibit good immediate prose
recall depends on two factors: (1) the capacity
of the episodic buffer; and (2) an efficiently
functioning central executive creating and maintaining
information in the buffer. According to
this argument, even severely amnesic patients with
practically no delayed recall of prose should have
good immediate prose recall provided they have
an efficient central executive. As predicted, immediate
prose recall was much better in amnesics
having little deficit in executive functioning than
in those with a severe executive deficit.
Other studies suggest that the episodic buffer
can operate independently of the central executive.
Gooding, Isaac, and Mayes (2005) failed
to replicate Baddeley and Wilson’s (2002) findings
in a similar study. Among their amnesic
patients (who were less intelligent than those
studied by Baddeley and Wilson), there was a
non-significant correlation between immediate
prose recall and measures of executive functioning.
It is possible that using the central executive
to maintain reasonable immediate prose recall
requires high levels of intelligence. Berlingeri
et al. (2008) found in patients with Alzheimer’s
disease that 60% of those having almost intact
performance on tasks requiring the central
executive nevertheless had no immediate prose
recall. This finding also casts doubt on the
importance of the central executive on tasks
involving the episodic buffer.
Rudner, Fransson, Ingvar, Nyberg, and
Ronnberg (2007) used a task involving combining
representations based on sign language and
on speech. This episodic buffer task was not
associated with prefrontal activation, but was
associated with activation in the left hippocampus.
This is potentially important because the
hippocampus plays a key role in binding together
different kinds of information in memory (see
Chapter 7). An association between use of the
episodic buffer and the hippocampus was also
reported by Berlingeri et al. (2008). They found
among patients with Alzheimer’s disease that
those with most atrophy of the anterior part of
the hippocampus did worst on immediate prose
recall.
Evaluation
The addition of the episodic buffer to the working
memory model has proved of value. The
6 LEARNING, MEMORY, AND FORGETTING 223
original three components of the model were
too separate from each other and from longterm
memory to account for our ability to
combine different kinds of information (e.g.,
visual, verbal) on short-term memory tasks.
The episodic buffer helps to provide the
“glue” to integrate information within working
memory.
Some progress has been made in tracking
down the brain areas associated with the
episodic buffer. The hippocampus is of central
importance in binding and integrating information
during learning, and so it is unsurprising
that it is associated with use of the episodic
buffer. The evidence suggests that use of the
episodic buffer is sometimes associated with the
central executive, but we do not know as yet
what determines whether there is an association.
It is harder to carry out research on the
episodic buffer than on the phonological loop
or the visuo-spatial sketchpad. We have to use
complex tasks to study the episodic buffer because
it involves the complicated integration of information.
In contrast, it is possible to devise relatively
simple tasks to study the phonological loop
or the visuo-spatial sketchpad. In addition, there
are often close connections between the episodic
buffer and the other components of the working
memory system. That often makes it difficult
to distinguish clearly between the episodic buffer
and the other components.
Overall evaluation
The working memory model has several advantages
over the short-term memory store proposed
by Atkinson and Shiffrin (1968). First, the working
memory system is concerned with both active
processing and transient storage of information,
and so is involved in all complex cognitive
tasks, such as language comprehension (see
Chapter 10) and reasoning (see Chapter 14).
Second, the working memory model explains
the partial deficits of short-term memory observed
in brain-damaged patients. If brain damage
affects only one of the three components of working
memory, then selective deficits on short-term
memory tasks would be expected.
Third, the working memory model incorporates
verbal rehearsal as an optional process
within the phonological loop. This is more
realistic than the enormous significance of
rehearsal within the multi-store model of
Atkinson and Shiffrin (1968).
What are the limitations of the working
memory model? First, it has proved difficult
to identify the number and nature of the main
executive processes associated with the central
executive. For example, disagreements on the
nature of executive functions have emerged from
approaches based on latent-variable analyses
of executive tasks (Miyake et al., 2000) and
on data from brain-damaged patients (Stuss &
Alexander, 2007). One reason for the lack of
clarity is that most complex tasks involve the
use of more than one executive process, making
it hard to establish the contribution that each
has made.
Second, we need more research on the
relationship between the episodic buffer and
the other components of the working memory
system. As yet, we lack a detailed account of
how the episodic buffer integrates information
from the other components and from long-term
memory.
LEVELS OF PROCESSING
What determines how well we remember information
over the long term? According to Craik
and Lockhart (1972), what is crucial is how
we process that information during learning.
They argued in their levels-of-processing approach
that attentional and perceptual processes at
learning determine what information is stored
in long-term memory. There are various levels
of processing, ranging from shallow or physical
analysis of a stimulus (e.g., detecting specific
letters in words) to deep or semantic analysis;
the greater the extent to which meaning is
processed, the deeper the level of processing.
They implied that processing nearly always
proceeds in a serial fashion from shallow
sensory levels to deeper semantic ones. However,
they subsequently (Lockhart & Craik, 1990)
224 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
admitted that that was an oversimplification
and that processing is often parallel.
Craik and Lockhart’s (1972) main theoretical
assumptions were as follows:
The level or depth of processing of a stimulus•
has a large effect on its memorability.
Deeper levels of analysis produce more•
elaborate, longer lasting and stronger memory
traces than do shallow levels of analysis.
Craik and Lockhart (1972) disagreed with
Atkinson and Shiffrin’s (1968) assumption that
rehearsal always improves long-term memory.
They argued that rehearsal involving simply
repeating previous analyses (maintenance
rehearsal) does not enhance long-term memory.
In fact, however, maintenance rehearsal typically
has a rather small (but beneficial) effect
on long-term memory (Glenberg, Smith, &
Green, 1977).
Evidence
Numerous studies support the main assumptions
of the levels-of-processing approach. For
example, Craik and Tulving (1975) compared
recognition performance as a function of the
task performed at learning:
Shallow graphemic task• : decide whether each
word is in uppercase or lowercase letters.
Intermediate phonemic task• : decide whether
each word rhymes with a target word.
Deep semantic task• : decide whether each
word fits a sentence containing a blank.
Depth of processing had impressive effects on
memory performance, with performance more
than three times higher with deep than with
shallow processing. In addition, performance
was generally much better for words associated
with “Yes” responses on the processing task
than those associated with “No” responses.
Craik and Tulving used incidental learning –
the participants did not realise at the time of
learning that there would be a memory test.
They argued that the nature of task processing
rather than the intention to learn is crucial.
Craik and Tulving (1975) assumed that the
semantic task involved deep processing and
the uppercase/lowercase task involved shallow
processing. However, it would be preferable to
assess depth. One approach is to use brainimaging
to identify the brain regions involved
in different kinds of processing. For example,
Wagner, Maril, Bjork, and Schacter (2001) found
there was more activation in the left inferior
frontal lobe and the left lateral and medial
temporal lobe during semantic than perceptual
processing. However, the findings have been
somewhat inconsistent. Park and Rugg (2008b)
presented word pairs and asked participants to
rate the extent to which they shared a semantic
theme (deep processing) or sounded similar
(shallow processing). Memory was better following
semantic processing than phonological pro-
cessing.However,successfulmemoryperformance
was associated with activation in the left ventrolateral
prefrontal cortex regardless of the encoding
task. This finding suggests that there is no simple
relationship between processing task and patterns
of brain activation.
Craik and Tulving (1975) argued that elaboration
of processing (i.e., the amount of processing
of a particular kind) is important as well as depth
of processing. Participants were presented on
each trial with a word and a sentence containing
a blank, and decided whether the word fitted
into the blank space. Elaboration was manipulated
by using simple (e.g., “She cooked the
____”) and complex “The great bird swooped
down and carried off the struggling ____”)
sentence frames. Cued recall was twice as high
for words accompanying complex sentences.
Long-term memory depends on the kind
of elaboration as well as the amount. Bransford,
Franks, Morris, and Stein (1979) presented either
minimally elaborated similes (e.g., “A mosquito
maintenance rehearsal: processing that
involves simply repeating analyses which have
already been carried out.
KEY TERM
6 LEARNING, MEMORY, AND FORGETTING 225
is like a doctor because they both draw blood”)
or multiply elaborated similes (e.g., “A mosquito
is like a raccoon because they both have heads,
legs, jaws”). Recall was much better for the
minimally elaborated similes than the multiply
elaborated ones, indicating that the nature of
semantic elaborations needs to be considered.
Eysenck (1979) argued that distinctive or
unique memory traces are easier to retrieve than
those resembling other memory traces. Eysenck
and Eysenck (1980) tested this notion using
nouns having irregular grapheme–phoneme
correspondence (i.e., words not pronounced
in line with pronunciation rules, such as “comb”
with its silent “b”). In one condition, participants
pronounced these nouns as if they
had regular grapheme–phoneme correspondence,
thus producing distinctive memory traces.
Other nouns were simply pronounced normally,
thus producing non-distinctive memory traces.
Recognition memory was much better in the
former condition, indicating the importance of
distinctiveness.
Morris, Bransford, and Franks (1977)
argued that stored information is remembered
only if it is of relevance to the memory test.
Participants answered semantic or shallow
(rhyme) questions for lists of words. Memory
was tested by a standard recognition test, in
which list and non-list words were presented,
or by a rhyming recognition test. On this latter
test, participants selected words that rhymed
with list words: the words themselves were not
presented. With the standard recognition test,
the predicted superiority of deep over shallow
processing was obtained (see Figure 6.12). However,
the opposite result was reported with the
rhyme test, which disproves the notion that deep
processing always enhances long-term memory.
Morris et al. (1977) argued that their findings
supported transfer-appropriate processing
theory. According to this theory, different kinds
of learning lead learners to acquire different
kinds of information about a stimulus. Whether
the stored information leads to subsequent
retention depends on the relevance of that
information to the memory test. For example,
storing semantic information is essentially
irrelevant when the memory test requires the
identification of words rhyming with list words.
What is required for this kind of test is shallow
rhyme information. Further evidence supporting
transfer-appropriate theory is discussed later
in the chapter.
Nearly all the early research on levels-ofprocessing
theory used standard memory tests
(e.g., recall, recognition) involving explicit memory
(conscious recollection). It is also important
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
Rhyme Semantic
Orienting task
Proportionrecognised
Standard test
Rhyming test
Figure 6.12 Mean
proportion of words
recognised as a function of
orienting task (semantic or
rhyme) and of the type of
recognition task (standard
or rhyming). Data are from
Morris et al. (1977), and are
from positive trials only.
226 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
to consider the effects of level of processing
on implicit memory (memory not involving
conscious recollection; see Chapter 7). Challis,
Velichkovsky, and Craik (1996) asked participants
to learn word lists under various conditions:
judging whether the word was related to
them (self-judgement); simple intentional learning;
judging whether it referred to a living thing
(living judgement); counting the number of
syllables (syllable task); or counting the number
of letters of a certain type (letter type). The
order of these tasks reflects decreasing depth
of processing. There were four explicit memory
tests (recognition, free recall, semantic cued
recall involving a word related in meaning to
a list word, and graphemic cued recall involving
a word with similar spelling to a list word), and
two implicit memory tests. One of these tests
involved answering general knowledge questions
in which the answers corresponded to
list words, and the other involved completing
word fragments (e.g., c _ pp _ _).
For the four explicit memory tests, there
was an overall tendency for performance to
increase with increasing depth of processing,
but there are some hard-to-interpret differences
as well (see Figure 6.13). We turn now to the
implicit memory tests. The word-fragment test
failed to show any levels-of-processing effect,
whereas level of processing had a significant
6
5
4
3
2
1
0
LSDs
Self
Intentional
Living
Syllable
Letter
Recognition Free
recall Semantic
cued recall Graphemic
cued recall General
knowledge Word
fragment
Figure 6.13 Memory performance as a function of encoding conditions and retrieval conditions.The findings
are presented in units of least significant differences (LSDs) relative to baseline performance, meaning that
columns of differing heights are significantly different. Reprinted from Roediger (2008), based on data in Challis
et al. (1996), Copyright © 1996, with permission from Elsevier.
6 LEARNING, MEMORY, AND FORGETTING 227
effect on the general knowledge memory
test. The general knowledge memory test is a
conceptual implicit memory test based on meaning.
As a result, it seems reasonable that it would
be affected by level of processing, even though
the effects were much smaller than with explicit
memory tests. In contrast, the word-fragment test
is a perceptual implicit memory test not based on
meaning, which helps to explain why there was
no levels-of-processing effect with this test.
In sum, levels-of-processing effects were generally
greater in explicit memory than implicit
memory. In addition, there is some support for
the predictions of levels-of-processing theory
with all memory tests other than the wordfragment
test. Overall, the findings are too
complex to be explained readily by levels-ofprocessing
theory.
Evaluation
Craik and Lockhart (1972) argued correctly
that processes during learning have a major
impact on subsequent long-term memory. This
may sound obvious, but surprisingly little
research before 1972 focused on learning processes
and their effects on memory. Another
strength is the central assumption that perception,
attention, and memory are all closely
interconnected, and that learning and remembering
are by-products of perception, attention,
and comprehension. In addition, the approach
led to the identification of elaboration and distinctiveness
of processing as important factors
in learning and memory.
The levels-of-processing approach possesses
several limitations. First, it is generally
difficult to assess processing depth. Second,
Craik and Lockhart (1972) greatly underestimated
the importance of the retrieval environment
in determining memory performance.
As Morris et al. (1977) showed, the typical
levels effect can be reversed if stored semantic
information is irrelevant to the requirements
of the memory test. Third, long-term memory
is influenced by depth of processing, elaboration
of processing, and distinctiveness of processing.
However, the relative importance of
these factors (and how they are inter-related)
remains unclear. Fourth, findings from amnesic
patients (see Chapter 7) cannot be explained
by the levels-of-processing approach. Most
amnesic patients have good semantic or deep
processing skills, but their long-term memory
is extremely poor, probably because they have
major problems with consolidation (fixing of
newly learned information in long-term memory)
(Craik, 2002; see Chapter 7). Fifth, Craik and
Lockhart (1972) did not explain precisely why
deep processing is so effective, and it is not clear
why there is a much smaller levels-of-processing
effect in implicit than in explicit memory.
IMPLICIT LEARNING
Do you think you could learn something without
being aware of what you have learned?
It sounds improbable. Even if we do acquire
information without any conscious awareness,
it might seem somewhat pointless and wasteful
– if we do not realise we have learned something,
it seems unlikely that we are going to
make much use of it. What we are considering
here is implicit learning, which is, “learning
without conscious awareness of having learned”
(French & Cleeremans, 2002, p. xvii). Implicit
learning has been contrasted with explicit
learning, which involves conscious awareness
of what has been learned.
Cleeremans and Jiménez (2002, p. 20) provided
a fuller definition of implicit learning:
“Implicit learning is the process through which
we become sensitive to certain regularities in
the environment (1) in the absence of intention
to learn about these regularities, (2) in the
absence of awareness that one is learning, and
(3) in such a way that the resulting knowledge
implicit learning: learning complex
information without the ability to provide
conscious recollection of what has been learned.
KEY TERM
228 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
is difficult to express.” You probably possess
skills that are hard to express in words. For
example, it is notoriously difficult to express
what we know about riding a bicycle.
There are clear similarities between implicit
learning and implicit memory, which is memory
not depending on conscious recollection (see
Chapter 7). You may wonder why implicit learning
and implicit memory are not discussed
together. There are three reasons. First, there
are some differences between implicit learning
and implicit memory. As Buchner and Wippich
(1998) pointed out, implicit learning refers to
“the [incidental] acquisition of knowledge about
the structural properties of the relations between
[usually more than two] objects or events.” In
contrast, implicit memory refers to “situations in
which effects of prior experiences can be observed
despite the fact that the participants are not
instructed to relate their current performance to
a learning episode” (Buchner & Wippich, 1998).
Second, studies of implicit learning have typically
used relatively complex, novel stimulus materials,
whereas most studies of implicit memory have
used simple, familiar stimulus materials. Third,
relatively few researchers have considered the
relations between implicit learning and implicit
memory.
How do the systems involved in implicit
learning differ from those involved in explicit
learning and memory? Reber (1993) proposed
five such characteristics (none has been established
definitively):
Robustness• : Implicit systems are relatively
unaffected by disorders (e.g., amnesia) affecting
explicit systems.
Age independence• : Implicit learning is
little influenced by age or developmental
level.
Low variability• : There are smaller individual
differences in implicit learning and memory
than in explicit learning and memory.
IQ independence• : Performance on implicit
tasks is relatively unaffected by IQ.
Commonality of process• : Implicit systems
are common to most species.
We can identify three main types of research
on implicit learning. First, there are studies to
see whether healthy participants can learn fairly
complex material in the absence of conscious
awareness of what they have learned. According
to Reber (1993), individual differences in such
learning should depend relatively little on IQ.
It is often assumed that implicit learning makes
minimal demands on attentional resources. If
so, the requirement to perform an additional
attentionally-demanding task at the same time
should not impair implicit learning.
Second, there are brain-imaging studies. If
implicit learning depends on different cognitive
processes to explicit learning, the brain areas
associated with implicit learning should differ
from those associated with explicit learning.
More specifically, brain areas associated with
conscious experience and attentional control
(e.g., parts of the prefrontal cortex) should be
much less activated during implicit learning
than explicit learning.
Third, there are studies on brain-damaged
patients, mostly involving amnesic patients
having severe problems with long-term memory.
Amnesic patients typically have relatively intact
implicit memory even though their explicit
memory is greatly impaired (see Chapter 7). If
amnesic patients have intact implicit learning but
impaired explicit learning, this would provide
Implicit learning is “learning without conscious
awareness of having learned”. Bike riding is an
example of implicit learning in which there is no
clear conscious awareness of what has been
learned.
6 LEARNING, MEMORY, AND FORGETTING 229
evidence that the two types of learning are very
different.
You might imagine it would be relatively
easy to decide whether implicit learning has
occurred – we simply ask participants to perform
a complex task without instructing them to
engage in deliberate learning. Afterwards, they
indicate their conscious awareness of what
they have learned. Implicit learning has been
demonstrated if learning occurs in the absence
of conscious awareness of the nature of that
learning. Alas, there are several reasons why
participants fail to report conscious awareness
of what they have learned. For example,
there is the “retrospective problem” (Shanks &
St. John, 1994): participants may be consciously
aware of what they are learning at the time,
but have forgotten it when questioned at the
end of the experiment. Shanks and St. John
proposed two criteria for implicit learning to
be demonstrated:
Information criterion• : The information participants
are asked to provide on the awareness
test must be the information responsible for
the improved level of performance.
Sensitivity criterion• : “We must be able to
show that our test of awareness is sensitive
to all of the relevant knowledge” (p. 374).
People may be consciously aware of more
task-relevant knowledge than appears on
an insensitive awareness test, leading us to
underestimate their consciously accessible
knowledge.
Complex learning
Much early research on implicit learning involved
artificial grammar learning. On this task, participants
initially memorise meaningless letter
strings (e.g., PVPXVPS; TSXXTVV). After that,
they are told that the memorised letter strings
all follow the rules of an artificial grammar,
but are not told the nature of these rules. Next,
the participants classify novel strings as grammatical
or ungrammatical. Finally, they describe
the rules of the artificial grammar. Participants
typically perform significantly above chance level
on the classification task, but cannot describe
the grammatical rules (e.g., Reber, 1967). Such
findings are less impressive than they appear. As
several researchers have found (e.g., Channon,
Shanks, Johnstone, Vakili, Chin, & Sinclair, 2002),
participants’ decisions on the grammaticality
of letter strings do not depend on knowledge
of grammatical rules. Instead, participants classify
letter strings as grammatical when they
share letter pairs with the letter strings memorised
initially and as ungrammatical when they
do not. Thus, above-chance performance depends
on conscious awareness of two-letter fragments,
and provides little or no evidence of implicit
learning.
The most commonly used implicit learning
task involves serial reaction time. On each trial,
a stimulus appears at one out of several locations
on a computer screen, and participants respond
rapidly with the response key corresponding to
its location. There is typically a complex, repeating
sequence over trials in the various stimulus
locations,butparticipantsarenottoldthis.Towards
the end of the experiment, there is typically a
block of trials conforming to a novel sequence,
but this information is not given to participants.
Participants speed up during the course of the
experiment but respond much slower during the
novel sequence (see Shanks, 2005, for a review).
When questioned at the end of the experiment,
participants usually show no conscious awareness
that there was a repeating sequence or pattern
in the stimuli presented to them.
One strength of the serial reaction time task
is that the repeating sequence (which is crucial
to the demonstration of implicit learning) is
incidental to the explicit task of responding to
the stimuli as rapidly as possible. However, we
need to satisfy the information and sensitivity
criteria (described above) with this task. It seems
reasonable to make the awareness test very
similar to the learning task, as was done by
Howard and Howard (1992). An asterisk
appeared in one of four locations on a screen,
under each of which was a key. The task was
to press the key corresponding to the position
of the asterisk as rapidly as possible. Participants
showed clear evidence of learning the underlying
230 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
sequence by responding faster and faster to the
asterisk. However, when given the awareness
test of predicting where the asterisk would
appear next, their performance was at chance
level. These findings suggest there was implicit
learning – learning occurred in the absence of
conscious awareness of what had been learned.
Contrary evidence that participants have
some conscious awareness of what they have
learned on a serial reaction time task was
reported by Wilkinson and Shanks (2004).
Participants were given either 1500 trials (15
blocks) or 4500 trials (45 blocks) on the task
and showed strong evidence of sequence learning.
Then they were told there was a repeated
sequence in the stimuli, following which they
were presented on each of 12 trials with part
of the sequence under one of two conditions.
In the inclusion condition, they guessed the
next location in the sequence. In the exclusion
condition, they were told they should avoid
guessing the next location in the sequence.
If sequence knowledge is wholly implicit, then
performance should not differ between the
inclusion and exclusion conditions because
participants would be unable to control how
they used their sequence knowledge. In contrast,
if it is partly explicit, then participants
should be able to exert intentional control over
their sequence knowledge. If so, the guesses
generated in the inclusion condition should be
more likely to conform to the repeated sequence
than those in the exclusion condition. The findings
indicated that explicit knowledge was
acquired on the serial reaction time task (see
Figure 6.14).
Similar findings were reported by Destrebecqz
et al. (2005) in another study using the serial
reaction time task. The interval of time between
the participant’s response to one stimulus and
the presentation of the next one was either 0 ms
or 250 ms, it being assumed that explicit learning
would be more likely with the longer interval.
Participants responded progressively faster over
trials with both response-to-stimulus intervals.
As Wilkinson and Shanks (2004) had done,
they used inclusion and exclusion conditions.
Participants’ responses were significantly closer
to the training sequence in the inclusion condition
than in the exclusion condition, suggesting
that some explicit learning occurred, especially
when the response-to-stimulus interval was long.
In addition, as discussed below, brain-imaging
findings from this study suggested that explicit
learning occurred.
If the serial reaction time task genuinely
involves implicit learning, performance on that
task might well be unaffected by the requirement
to perform a second, attentionally-demanding
task at the same time. This prediction was tested
by Shanks, Rowland, and Ranger (2005). Four
different target stimuli were presented across
trials, and the main task was to respond rapidly
to the location at which a target was presented.
Half the participants performed only this
task, and the remainder also carried out
the attentionally-demanding task of counting
targets. Participants with the additional task performed
much more slowly than those with no
additional task, and also showed significantly
inferior sequence learning. Thus, attentional
resources were needed for effective learning of
the sequence on the serial reaction time task,
which casts doubt on the notion that such
learning is implicit. In addition, both groups
8
6
4
2
0
15 block 45 block
Group
Meannumberofcompletions
Inclusion own
Inclusion other
Exclusion own
Exclusion other
Figure 6.14 Mean number of completions (guessed
locations) corresponding to the trained sequence
(own) or the untrained sequence (other) in inclusion
and exclusion conditions as a function of number of
trials (15 vs. 45 blocks). From Wilkinson and Shanks
(2004). Copyright © 2004 American Psychological
Association. Reproduced with permission.
6 LEARNING, MEMORY, AND FORGETTING 231
of participants had significantly more accurate
performance under inclusion than exclusion
instructions, further suggesting the presence of
explicit learning.
As mentioned above, Reber (1993) assumed
that individual differences in intelligence have
less effect on implicit learning than on explicit
learning. Gebauer and Mackintosh (2007) carried
out a thorough study using various implicit
learning tasks (e.g., artificial grammar learning;
serial reaction time). These tasks were given
under standard implicit instructions or with
explicit rule discovery instructions (i.e., indicating
explicitly that there were rules to be discovered).
The mean correlation between implicit task
performance and intelligence was only +0.03,
whereas it was +0.16 between explicit task
performance and intelligence. This supports
the hypothesis. It is especially important that
intelligence (which is positively associated with
performance on the great majority of cognitive
tasks) failed to predict implicit learning
performance.
Brain-imaging studies
Different areas of the brain should be activated
during implicit and explicit learning if they
are genuinely different. Conscious awareness
is associated with activation in many brain
regions, but the main ones are the anterior
cingulate and the dorsolateral prefrontal cortex
(Dehaene & Naccache, 2001; see Chapter 16).
Accordingly, these areas should be more active
during explicit than implicit learning. In contrast,
it has often been assumed that the striatum is
associated with implicit learning (Destrebecqz
et al., 2005). The striatum is part of the basal
ganglia; it is located in the interior areas of the
cerebral hemispheres and the upper region of
the brainstem.
Functional neuroimaging studies have provided
limited support for the above predictions.
Grafton, Hazeltine, and Ivry (1995) found that
explicit learning was associated with activation
in the anterior cingulate, regions in the parietal
cortex involved in working memory, and areas
in the parietal cortex concerned with voluntary
attention. Aizenstein et al. (2004) found that
there was greater activation in the prefrontal
cortex and anterior cingulate during explicit
rather than implicit learning. However, they
did not find any clear evidence that the striatum
was more activated during implicit than explicit
learning.
Destrebecqz et al. (2005) pointed out that
most so-called explicit or implicit learning
tasks probably involve a mixture of explicit
and implicit learning. As mentioned before,
they used inclusion and exclusion conditions
with the serial reaction time task to distinguish
clearly between the explicit and implicit components
of learning. Activation in the striatum
was associated with the implicit component
of learning, and the mesial prefrontal cortex
and anterior cingulate were associated with the
explicit component.
In sum, failure to discover clear differences
in patterns of brain activation between explicit
and implicit learning can occur because the
tasks used are not pure measures of these two
forms of learning. It is no coincidence that
the study distinguishing most clearly between
explicit and implicit learning (Destrebecqz et al.,
2005) is also the one producing the greatest
support for the hypothesised associations of
prefrontal cortex with explicit learning and the
striatum with implicit learning.
Brain-damaged patients
As discussed in Chapter 7, amnesic patients
typically perform very poorly on tests of explicit
memory (involving conscious recollection) but
often perform as well as healthy individuals
on tests of implicit memory (on which conscious
recollection is not needed). The notion that
separate learning systems underlie implicit learning
and explicit learning would be supported
striatum: it forms part of the basal ganglia of
the brain and is located in the upper part of the
brainstem and the inferior part of the cerebral
hemispheres.
KEY TERM
232 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
if amnesic patients showed intact levels of implicit
learning combined with impaired explicit
learning. Explicit learning in amnesics is often
severely impaired, but amnesics’ performance
on tasks allegedly involving implicit learning
is variable (see Vandenberghe, Schmidt, Fery, &
Cleeremans, 2006, for a review). For example,
Knowlton, Ramus, and Squire (1992) found that
amnesics performed as well as healthy controls
on an implicit test on which participants distinguished
between grammatical and ungrammatical
letter strings (63% versus 67% correct,
respectively). However, they performed significantly
worse than the controls on an explicit test
(62% versus 72%, respectively).
Meulemans and Van der Linden (2003)
pointed out that amnesics’ performance on
Knowlton et al.’s (1992) implicit test may have
depended on explicit fragment knowledge (e.g.,
pairs of letters found together). Accordingly, they
used an artificial grammar learning task in which
fragment knowledge could not influence performance
on the test of implicit learning. They also
used a test of explicit learning in which participants
wrote down ten letter strings they regarded
as grammatical. The amnesic patients performed
as well as the healthy controls on implicit learning.
However, their performance was much worse
than that of the controls on explicit learning.
There is evidence of implicit learning in
amnesic patients in studies on the serial reaction
time task. The most thorough such study
was carried out by Vandenberghe et al. (2006).
Amnesic patients and healthy controls were
given two versions of the task: (1) deterministic
sequence (fixed repeating sequence); and (2)
probabilistic sequence (repeating sequence with
some deviations). The healthy controls showed
clear evidence of learning with both sequences.
The use of inclusion and exclusion instructions
indicated that healthy controls showed explicit
learning with the deterministic sequence but
not with the probabilistic one. The amnesic
patients showed limited learning of the deterministic
sequence but not of the probabilistic
sequence. Their performance was comparable
with inclusion and exclusion instructions, indicating
that this learning was implicit.
Earlier we discussed the hypothesis that the
striatum is of major importance in implicit
learning. Patients with Parkinson’s disease (the
symptoms of which include limb tremor and
muscle rigidity) have damage to the striatum,
and so we could predict that they would have
impaired implicit learning. The evidence generally
supports that prediction (see Chapter 7 for a
fuller discussion). Siegert, Taylor, Weatherall, and
Abernethy (2006) carried out a meta-analysis
of six studies investigating the performance of
patients with Parkinson’s disease on the serial
reaction time task. Skill learning on this task
was consistently impaired in the patients relative
to healthy controls. Wilkinson and Jahanshahi
(2007) obtained similar findings with patients
having Parkinson’s disease using a different version
of the serial reaction time task. In addition, they
reported convincing evidence that patients’ learning
was implicit (i.e., lacked conscious awareness).
The patients performed at chance level when
trying to recognise old sequences. In addition,
their knowledge was not under intentional
control, as was shown by their inability to suppress
the expression of what they had learned
when instructed to do so.
We have seen that there is some evidence
that amnesic patients have poor explicit learning
combined with reasonably intact implicit
learning. We would have evidence of a double
dissociation (see Glossary) if patients with
Parkinson’s disease had poor implicit learning
combined with intact explicit learning. This
pattern has occasionally been reported with
patients in the early stages of the disease (e.g.,
Saint-Cyr, Taylor, & Lang, 1988). However,
Parkinson’s patients generally have impaired
explicit learning, especially when the learning
task is fairly complex and involves organisation
of the to-be-learned information (see Vingerhoets,
Vermeule, & Santens, 2005, for a review).
Evaluation
There has been a considerable amount of recent
research on implicit learning involving three
different approaches: behavioural studies on
healthy participants; functional neuroimaging
6 LEARNING, MEMORY, AND FORGETTING 233
studies on healthy participants; and studies on
amnesic patients. Much of that research suggests
that implicit learning should be distinguished
from explicit learning. Some of the most convincing
evidence has come from studies on braindamaged
patients. For example, Vanderberghe
et al. (2006) found, using the serial reaction time
task, that amnesic patients’ learning seemed to be
almost entirely at the implicit level. Other convincing
evidence has come from functional neuroimaging
studies. There is accumulating evidence
that explicit learning is associated with the prefrontal
cortex and the anterior cingulate, whereas
implicit learning is associated with the striatum.
What are the limitations of research on implicit
learning? First, it has proved hard to devise tests
of awareness that can detect all the task-relevant
knowledge of which people have conscious awareness.
Second, some explicit learning is typically
involved on the artificial grammar learning task
and the serial reaction time task (e.g., Destrebecqz
et al., 2005; Shanks et al., 2005; Wilkinson &
Shanks, 2004). Third, the brain areas underlying
what are claimed to be explicit and implicit learning
are not always clearly different (e.g., Schendan,
Searl, Melrose, & Stern, 2003).
What conclusions can we draw about implicit
learning? It is too often assumed that finding
that explicit learning plays some part in explaining
performance on a given task means that
no implicit learning occurred. It is very likely
that the extent to which learners are consciously
aware of what they are learning varies from
individual to individual and from task to task.
One possibility is that we have greatest conscious
awareness when the representations of what
we have learned are stable, distinctive, and strong,
and least when those representations are unstable,
non-distinctive, and weak (Kelly, 2003). All kinds
of intermediate position are also possible.
Sun, Zhang, and Mathews (2009) argued that
learning nearly always involves implicit and
explicit aspects, and that the balance between
these two types of learning changes over time.
On some tasks, there is initial implicit learning
based on the performance of successful actions
followed by explicit learning of the rules apparently
explaining why those actions are successful.
On other tasks, learners start with explicit rules
and then engage in implicit learning based on
observing their actions directed by those rules.
THEORIES OF
FORGETTING
Forgetting was first studied in detail by
Hermann Ebbinghaus (1885/1913). He carried
out numerous studies with himself as the only
participant (not a recommended approach!).
Ebbinghaus initially learned a list of nonsense
syllables lacking meaning. At various intervals
of time, he recalled the nonsense syllables.
He then re-learned the list. His basic measure
of forgetting was the savings method, which
involved seeing the reduction in the number of
trials during re-learning compared to original
learning. Forgetting was very rapid over the
first hour after learning but slowed down
considerably after that (see Figure 6.15). These
findings suggest that the forgetting function is
approximately logarithmic.
Rubin and Wenzel (1996) analysed the
forgetting functions taken from 210 data sets
involving numerous memory tests. They found
(in line with Ebbinghaus (1885/1913) that a
logarithmic function most consistently described
the rate of forgetting (for alternative possibilities,
see Wixted, 2004). The major exception was
autobiographical memory, which showed slower
forgetting. One of the possible consequences of
a logarithmic forgetting function is Jost’s (1897)
law: if two memory traces differ in age but are
of equal strength, the older one will decay more
slowly over any given time period.
Most studies of forgetting have focused on
declarative or explicit memory (see Chapter 7),
which involves conscious recollection of
savings method: a measure of forgetting
introduced by Ebbinghaus, in which the number
of trials for re-learning is compared against the
number for original learning.
KEY TERM
234 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
previously learned information. Comparisons
of forgetting rates in explicit and implicit
memory (in which conscious recollection is not
required) suggest that forgetting is slower in
implicit memory. Tulving, Schacter, and Stark
(1982) carried out a study in which participants
initially learned a list of relatively rare words
(e.g., “toboggan”). One hour or one week later,
they received a test of explicit memory (recognition
memory) or a word-fragment completion
test of implicit memory. Word fragments (e.g.,
_ O _ O _ GA _) were presented and participants
filled in the blanks to form a word without
being told that any of the words came from
the list studied previously. Recognition memory
was much worse after one week than one hour,
whereas word-fragment completion performance
was unchanged.
Dramatic evidence of long-lasting implicit
memories was reported by Mitchell (2006).
His participants tried to identify pictures from
fragments having seen some of them before in
a laboratory experiment 17 years previously.
They did significantly better with the pictures
seen before; thus providing strong evidence for
implicit memory after all those years! In contrast,
there was rather little explicit memory for the
experiment 17 years earlier. A 36-year-old male
participant confessed, “I’m sorry – I don’t really
remember this experiment at all.”
In what follows, we will be discussing the
major theories of forgetting in turn. As you
read about these theories, bear in mind that
they are not mutually exclusive. Thus, it is entirely
possible that all the theories discussed identify
some of the factors responsible for forgetting.
Interference theory
The dominant approach to forgetting during
much of the twentieth century was interference
theory. According to this theory, our ability to
remember what we are currently learning can
be disrupted (interfered with) by previous learning
(proactive interference) or by future learning
(retroactive interference) (see Figure 6.16).
Interference theory dates back to Hugo
Munsterberg in the nineteenth century. For
many years, he kept his pocket-watch in one
particular pocket. When he moved it to a
different pocket, he often fumbled about in
confusion when asked for the time. He had
learned an association between the stimulus,
“What time is it, Hugo?”, and the response of
removing the watch from his pocket. Later on,
the stimulus remained the same. However, a
different response was now associated with it,
thus causing proactive interference.
Research using methods such as those
shown in Figure 6.16 revealed that proactive
100
80
60
40
20
0
0 1 8 24 48 120 744
Length of retention interval (hours)
Savings(%)
Figure 6.15 Forgetting
over time as indexed by
reduced savings. Data from
Ebbinghaus (1885/1913).
6 LEARNING, MEMORY, AND FORGETTING 235
and retroactive interference are both maximal
when two different responses are associated
with the same stimulus and minimal when two
different stimuli are involved (Underwood &
Postman, 1960). Strong evidence of retroactive
interference has been obtained in studies of
eyewitness testimony in which memory of an
event is interfered with by post-event information
(see Chapter 8).
Proactive interference
Proactive interference can be very useful when
circumstances change. For example, if you have
re-arranged everything in your room, it is a
real advantage to forget where your belongings
used to be.
Most research on proactive interference
has involved declarative or explicit memory.
An exception was a study by Lustig and Hasher
(2001). They used a word-fragment completion
task (e.g., A _ L _ _ GY), on which participants
wrote down the first appropriate word coming
to mind. Participants previously exposed to words
almost fitting the fragments (e.g., ANALOGY)
showed evidence of proactive interference.
Jacoby, Debner, and Hay (2001) argued
that proactive interference might occur for two
reasons. First, it might be due to problems in
retrieving the correct response (discriminability).
Second, it might be due to the great strength of
the incorrect response learned initially (bias or
habit). Thus, we might show proactive interference
because the correct response is very weak or
because the incorrect response is very strong.
Jacoby et al. found consistently that proactive
interference was due more to strength of the
incorrect first response than to discriminability.
At one time, it was assumed that individuals
passively allow themselves to suffer from
interference. Suppose you learn something but
find your ability to remember it is impaired by
proactive interference from something learned
previously. It would make sense to adopt active
strategies to minimise any interference effect.
Kane and Engle (2000) argued that individuals
with high working-memory capacity (correlated
with intelligence) would be better able to resist
proactive interference than those with low
capacity. However, even they would be unable to
resist proactive interference if performing an
attentionally demanding task at the same time as
the learning task. As predicted, the high-capacity
participants with no additional task showed the
least proactive interference (see Figure 6.17).
The notion that people use active control
processes to reduce proactive interference has
Proactive interference
Group
Experimental
Control
Learn
–
Learn
A–C
(e.g. Cat−Dirt)
Test
Retroactive interference
Group
Experimental
Control
Learn Learn
–
Test
Note: for both proactive and retroactive interference, the experimental group
exhibits interference. On the test, only the first word is supplied, and the
participants must provide the second word.
A–C
(e.g. Cat−Dirt)
A–C
(e.g. Cat−Dirt)
A–B
(e.g. Cat−Tree)
A–C
(e.g. Cat−Dirt)
A–C
(e.g. Cat−Dirt)
A–B
(e.g. Cat−Tree)
A–B
(e.g. Cat−Tree)
A–B
(e.g. Cat−Tree)
A–B
(e.g. Cat−Tree)
Figure 6.16 Methods of
testing for proactive and
retroactive interference.
236 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
been tested in several studies using the Recent
Probes task. A small set of items (target set) is
presented, followed by a recognition probe.
The task is to decide whether the probe is a
member of the target set. On critical trials, the
probe is not a member of the current target
set but was a member of the target set used on
the previous trial. There is clear evidence of
proactive interference on these trials in the
form of lengthened reaction times and increased
error rates.
Which brain areas are of most importance
on proactive interference trials with the Recent
Probes task? Nee, Jonides, and Berman (2007)
found that the left ventrolateral prefrontal cortex
was activated on such trials. The same brain
area was also activated on a directed forgetting
version of the Recent Probes task (i.e., participants
were told to forget some of the target
set items). This suggests that left ventrolateral
prefrontal cortex may play an important role
in suppressing unwanted information.
Nee et al.’s (2007) study could not show
that left ventrolateral prefrontal cortex actually
controls the effects of proactive interference.
More direct evidence was reported by Feredoes,
Tononi, and Postle (2006). They administered
transcranial magnetic stimulation (TMS; see
Glossary) to left ventrolateral prefrontal cortex.
This produced a significant increase in the error
rate on proactive interference trials, suggesting
that this brain area is directly involved in attempts
to control proactive interference.
Retroactive interference
Numerous laboratory studies using artificial
tasks such as paired-associate learning (see
Figure 6.16) have produced large retroactive
interference effects. Such findings do not necessarily
mean that retroactive interference is
important in everyday life. However, Isurin
and McDonald (2001) argued that retroactive
interference explains why people forget some
of their first language when acquiring a second
one. Bilingual participants fluent in two languages
were first presented with various pictures
and the corresponding words in Russian
or Hebrew. Some were then presented with the
same pictures and the corresponding words in
the other language. Finally, they were tested
for recall of the words in the first language.
There was substantial retroactive interference
– recall of the first-language words became
progressively worse the more learning trials
there were with the second-language words.
Retroactive interference is generally greatest
when the new learning resembles previous
learning. However, Dewar, Cowan, and Della
0.45
0.40
0.35
0.30
0.25
0.20
No load Additional
load
Concurrent task
Proportionalproactiveinterferenceeffect
High attentional capacity
Low attentional capacity
Figure 6.17 Amount of
proactive interference as
a function of attentional
capacity (low vs. high) and
concurrent task (no vs.
additional load). Data from
Kane and Engle (2000).
6 LEARNING, MEMORY, AND FORGETTING 237
Sala (2007) found retroactive interference even
when no new learning occurred during the
retention interval. In their experiment, participants
learned a list of words and were then
exposed to various tasks during the retention
interval before list memory was assessed. There
was significant retroactive interference even
when the intervening task involved detecting
differences between pictures or detecting tones.
Dewar et al. concluded that retroactive interference
can occur in two ways: (1) expenditure
of mental effort during the retention interval;
or (2) learning of material similar to the original
learning material. The first cause of retroactive
interference probably occurs more often than
the second in everyday life.
Lustig, Konkel, and Jacoby (2004) identified
two possible explanations for retroactive
interference in paired-associate learning. First,
there may be problems with controlled processes
(active searching for the correct response).
Second, there may be problems with automatic
processes (high accessibility of the incorrect
response). They identified the roles of these
two kinds of processes by assessing retroactive
interference in two different ways. One way
involved direct instructions (i.e., deliberately
retrieve the correct responses) and the other
way involved indirect instructions (i.e., rapidly
produce the first response coming to mind
when presented with the cue). Lustig et al.
assumed that direct instructions would lead to
the use of controlled and automatic processes,
whereas indirect instructions would primarily
lead to the use of automatic processes.
What did Lustig et al. (2004) find? First,
use of direct instructions was associated with
significant retroactive interference on an immediate
memory test (cued recall) but not one
day later. Second, the interference effect found
on the immediate test depended mainly on
relatively automatic processes (i.e., accessibility
of the incorrect response). Third, the disappearance
of retroactive interference on the
test after one day was mostly due to reduced
accessibility of the incorrect responses. Thus,
relatively automatic processes are of major
importance in retroactive interference.
Evaluation
There is strong evidence for both proactive
and retroactive interference. There has been
substantial progress in understanding interference
effects in recent years, mostly involving
an increased focus on underlying processes.
For example, automatic processes make incorrect
responses accessible, and people use
active control processes to minimise interference
effects.
What are the limitations of interference
theory? First, the emphasis has been on interference
effects in declarative or explicit memory,
and detailed information about interference
effects in implicit memory is lacking. Second,
interference theory explains why forgetting
occurs but not directly why the rate of forgetting
decreases over time. Third, more needs to
be done to understand the brain mechanisms
involved in interference and attempts to reduce
interference.
Repression
One of the best-known theories of forgetting
owes its origins to the bearded Austrian
psychologist Sigmund Freud (1856–1939). He
claimed that very threatening or traumatic
memories are often unable to gain access to
conscious awareness, using the term repression
to refer to this phenomenon. According
to Freud (1915/1963, p. 86), “The essence
of repression lies simply in the function of
rejecting and keeping something out of consciousness.”
However, Freud sometimes used the
concept to refer merely to the inhibition of the
capacity for emotional experience (Madison,
1956). Even though it is often believed that
Freud regarded repression as unconscious,
Erdelyi (2001) showed convincingly that Freud
accepted that repression is sometimes an active
repression: motivated forgetting of traumatic
or other threatening events.
KEY TERM
238 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
and intentional process. It is harder to test the
notion of repression if it can be either unconscious
or conscious.
Most evidence relating to repression is
based on adult patients who have apparently
recovered repressed memories of childhood
sexual and/or physical abuse in adulthood. As
we will see, there has been fierce controversy
as to whether these recovered memories are
genuine or false. Note that the controversy
centres on recovered memories – most experts
accept that continuous memories (i.e., ones
constantly accessible over the years) are very
likely to be genuine.
Evidence
Clancy, Schacter, McNally, and Pitman (2000)
used the Deese–Roediger–McDermott paradigm,
which is known to produce false memories.
Participants are given lists of semantically
related words and are then found to falsely
“recognise” other semantically related words
not actually presented. Clancy et al. compared
women with recovered memories of childhood
sexual abuse with women who believed they
had been sexually abused but could not recall
the abuse, women who had always remembered
being abused, and female controls.
Women reporting recovered memories showed
higher levels of false recognition than any other
group (see Figure 6.18), suggesting that these
women might be susceptible to developing false
memories.
Lief and Fetkewicz (1995) found that 80%
of adult patients who admitted reporting false
recovered memories had therapists who made
direct suggestions that they had been the
victims of childhood sexual abuse. This suggests
that recovered memories recalled inside
therapy may be more likely to be false than
those recalled outside therapy (see box).
Motivated forgetting
Freud, in his repression theory, focused on
some aspects of motivated forgetting. However,
his approach was rather narrow, with its
emphasis on repression of traumatic and other
distressing memories and his failure to consider
the cognitive processes involved. In recent years,
a broader approach to motivated forgetting
has been adopted.
Motivated forgetting of traumatic or other
upsetting memories could clearly fulfil a useful
0.8
0.7
0.6
0.5
0.4
0.3
Controls Always
remembered
abuse
Abused
no recall
Abused
recovered
memories
Falserecognitionperformance
Figure 6.18 False
recognition of words not
presented in four groups of
women with lists containing
eight associates. Data from
Clancy et al. (2000).
6 LEARNING, MEMORY, AND FORGETTING 239
Memories of abuse recovered inside and outside therapy
Geraerts, Schooler, Merckelbach, Jelicic, Haner,
and Ambadar (2007) carried out an important
study to test whether the genuineness of recovered
memories depends on the context in which they
were recovered. They divided adults who had
suffered childhood sexual abuse into three groups:
(1) those whose recovered memories had been
recalled inside therapy;(2) those whose recovered
memories had been recalled outside therapy;and
(3) those who had continuous memories.Geraerts
et al. discovered how many of these memories
had corroborating evidence (e.g., someone else
had also reported being abused by the same
person;the perpetrator had confessed) to provide
an approximate assessment of validity.
What did Geraerts et al.(2007) find?There
was corroborating evidence for 45% of the
individuals in the continuous memory group,for
37% of those who had recalled memories outside
therapy, and for 0% of those who had recalled
memories inside therapy.These findings suggest
that recovered memories recalled outside therapy
are much more likely to be genuine than those
recalled inside therapy. In addition, those individuals
whose memories were recalled outside
therapy reported being much more surprised at
the existence of these memories than did those
whose memories were recalled inside therapy.
Presumably those whose recovered memories
emerged inside therapy were unsurprised at
these memories because they had previously
been led to expect them by their therapist.
Geraerts et al. (2008) asked various groups
of adults who claimed memories of childhood
sexual abuse to recall the most positive and the
most anxiety-provoking event they had experienced
during the past two years.The participants
were then told to try to suppress thoughts
relating to these events, and to keep a diary
record of any such thoughts over the following
week. Adults who had recovered memories
outside therapy were much better at this than
control participants, those who had recovered
memories inside therapy, and those who had
continuous memories.
In sum,it appears that many of the traumatic
memories recovered by women outside therapy
are genuine.The finding that such women are
especially good at suppressing emotional memories
under laboratory conditions helps to explain why
they were unaware of their traumatic memories
for long periods of time prior to recovery.
6.0
5.0
4.0
3.0
2.0
1.0
0
Meanfrequencyofintrusions
Spontaneously
recovered
Recovered
in therapy
Continuous Controls
Target event
Anxious event
Positive event
Figure 6.19 Mean
numbers of intrusions
of anxious and positive
events over seven days
for patients who had
recovered traumatic
memories outside therapy
(spontaneously recovered),
inside therapy (recovered
in therapy), or who had
had continuous traumatic
memories (continuous), and
non-traumatised controls.
Based on data in Geraerts
et al. (2008).
240 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
function. In addition, much of the information
we have stored in long-term memory is outdated
or irrelevant, making it useless for present
purposes. For example, if you are looking
for your car in a car park, there is no point in
remembering where you have parked the car
previously. Thus, motivated or intentional forgetting
can be adaptive (e.g., by reducing proactive
interference).
Directed forgetting
Directed forgetting is a phenomenon involving
impaired long-term memory caused by an
instruction to forget some information presented
for learning (see Geraerts & McNally,
2008, for a review). Directed forgetting has
been studied in two ways. First, there is the
item method. Several words are presented, each
followed immediately by an instruction to
remember or to forget it. After all the words
have been presented, participants are tested for
their recall or recognition of all the words.
Memory performance on recall and recognition
tests is typically worse for the to-be-forgotten
words than for the to-be-remembered words.
Second, there is the list method. Here, participants
receive two lists of words. After the
first list has been presented, participants are
told to remember or forget the words. Then
the second list is presented. After that, memory
is tested for the words from both lists. Recall
of the words from the first list is typically
impaired when participants have been told to
forget those words compared to when they
have been told to remember them. However,
there is typically no effect when a recognition
memory test is used.
Why does directed forgetting occur?
Directed forgetting with the item method is
found with both recall and recognition, suggesting
that the forget instruction has its effects
during learning. For example, it has often been
suggested that participants may selectively rehearse
remember items at the expense of forget
items (Geraerts & McNally, 2008). This explanation
is less applicable to the list method,
because participants have had a substantial
opportunity to rehearse the to-be-forgotten list
items before being instructed to forget them.
The finding that directed forgetting with the
list method is not found in recognition memory
suggests that directed forgetting in recall involves
retrieval inhibition or interference (Geraerts &
McNally, 2008).
Inhibition: executive deficit hypothesis
A limitation with much of the research is that
the precise reasons why directed forgetting has
occurred are unclear. For example, consider
directed forgetting in the item-method paradigm.
This could occur because to-be-forgotten
items receive much less rehearsal than tobe-remembered
items. However, it could also
occur because of an active process designed
to inhibit the storage of words in long-term
memory. Wylie, Foxe, and Taylor (2007) used
fMRI with the item-method paradigm to test
these rival hypotheses. In crude terms, we might
expect less brain activity for to-be-forgotten
items than to-be-remembered ones if the former
simply attract less processing. In contrast, we
might expect more brain activity for to-beforgotten
items if active processes are involved.
In fact, intentional forgetting when compared
with intentional remembering was associated
with increased activity in several areas (e.g.,
medial frontal gyrus (BA10) and cingulated
gyrus (BA31)) known to be involved in executive
control.
Anderson and Green (2001) developed a
variant of the item method known as the think/
no-think paradigm. Participants first learn a
list of cue-target word pairs (e.g., Ordeal–
Roach). Then they are presented with cues
studied earlier (e.g., Ordeal) and instructed to
think of the associated word (Roach) (respond
condition) or to prevent it coming to mind
(suppress condition). Some of the cues were
not presented at this stage (baseline condition).
directed forgetting: impaired long-term
memory resulting from the instruction to
forget information presented for learning.
KEY TERM
6 LEARNING, MEMORY, AND FORGETTING 241
Finally, all the cues are presented and participants
provide the correct target words. Levy
and Anderson (2008) carried out a meta-analysis
of studies using the think/no-think paradigm.
There was clear evidence of directed forgetting
(see Figure 6.20). The additional finding that
recall was worse in the suppress condition than
in the baseline condition indicates that inhibitory
processes were involved in producing
directed forgetting in this paradigm.
What strategies do participants use in the
suppress condition? They report using numerous
strategies, including forming mental images,
thinking of an alternative word or thought, or
repeating the cue word (Levy & Anderson,
2008). Bergstrom, de Fockert, and RichardsonKlavehn
(2009) manipulated the strategy used.
Direct suppression of the to-be-forgotten words
was more effective than producing alternative
thoughts.
Anderson et al. (2004) focused on individual
differences in memory performance
using the think/no-think paradigm. Their study
was designed to test the executive deficit
hypothesis, according to which the ability to
suppress memories depends on individual differences
in executive control abilities. Recall
for word pairs was worse in the suppress condition
than in the respond and baseline conditions.
Of special importance, those individuals
having the greatest activation in bilateral dorsolateral
and ventrolateral prefrontal cortex were
most successful at memory inhibition. Memory
inhibition was also associated with reduced
hippocampal activation – this is revealing
because the hippocampus plays a key role in
episodic memory (see Chapter 7). These findings
suggest that successful intentional forgetting
involves an executive control process in the
prefrontal cortex that disengages hippocampal
processing.
Additional support for the executive deficit
hypothesis was reported by Bell and Anderson
(in preparation). They compared individuals
high and low in working memory capacity (see
Chapter 10), a dimension of individual differences
strongly related to executive control and
intelligence. As predicted, memory suppression
in the think/no-think paradigm was significantly
greater in the high capacity group.
Is research using the think/no-think paradigm
relevant to repression? There are encouraging
signs that it is. First, Depue, Banich, and
Curran (2006, 2007) had participants learn to
pair unfamiliar faces with unpleasant photographs
(e.g., a badly deformed infant; a car
accident) using the paradigm. The findings
were very similar to those of Anderson et al.
(2004). There was clear evidence for suppression
of unwanted memories and suppression was
associated with increased activation of the
lateral prefrontal cortex and reduced hippocampal
activity. Second, Anderson and Kuhl (in preparation)
found that individuals who had
experienced several traumatic events showed
superior memory inhibition abilities than those
who had experienced few or none. This suggests
that the ability to inhibit or suppress memories
improves with practice.
Evaluation
Directed forgetting is an important phenomenon.
The hypothesis that it involves executive
100
95
90
85
80
75
70
Percentrecalled
Respond Baseline Suppress
Figure 6.20 Meta-analysis of final recall
performance in the think/no-think procedure
as a function of whether participants had earlier
tried to recall the item (respond), suppress the
item (suppress), or had had no previous reminder
(baseline). Reprinted from Levy and Anderson (2008),
Copyright © 2008, with permission from Elsevier.
242 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
control processes within the frontal lobes has
received much empirical support. The extension
of this hypothesis to account for individual
differences in directed forgetting has also been
well supported. In addition, the notion that
research on directed forgetting may be of
genuine relevance to an understanding of repression
is important. A major implication of
directed forgetting research is that suppression
or repression occurs because of deliberate
attempts to control awareness rather than
occurring unconsciously and automatically, as
suggested by Freud.
Directed forgetting is clearly one way in
which forgetting occurs. However, most forgetting
occurs in spite of our best efforts to remember,
and so the directed forgetting approach is
not of general applicability. The suppression
effect in the think/no-think paradigm (baseline–
suppression conditions) averages out at only
6% (see Figure 6.20), suggesting it is rather
weak. However, participants spent an average
of only 64 seconds trying to suppress each item,
which is presumably massively less than the
amount of time many individuals devote to
suppressing traumatic memories. Most research
on directed forgetting has used neutral and
artificial learning materials, and this limits our
ability to relate the findings to Freud’s ideas
about repression.
Cue-dependent forgetting
Forgetting often occurs because we lack the
appropriate cues (cue-dependent forgetting).
For example, suppose you are struggling to
think of the name of the street on which a friend
of yours lives. If someone gave you a short list
of possible street names, you might have no
difficulty in recognising the correct one.
Tulving and Psotka (1971) showed the
importance of cues. They presented between
one and six word lists, with four words in six
different categories in each list. After each list,
participants free recalled as many words as
possible (original learning). After all the lists
had been presented, participants free recalled
the words from all the lists (total free recall).
Finally, all the category names were presented
and the participants tried again to recall all the
words from all the lists (free cued recall).
There was strong evidence for retroactive
interference in total free recall, since word
recall from any given list decreased as the number
of other lists intervening between learning
and recall increased. However, there was essentially
no retroactive interference or forgetting
when the category names were available to the
participants. Thus, the forgetting observed in
total free recall was basically cue-dependent
forgetting (due to a lack of appropriate cues).
Tulving (1979) developed the notion of
cue-dependent forgetting in his encoding specificity
principle: “The probability of successful
retrieval of the target item is a monotonically
increasing function of informational overlap
between the information present at retrieval
and the information stored in memory” (p. 408;
emphasis added). If you are bewildered by that
sentence, note that “monotonically increasing
function” refers to a generally rising function
that does not decrease at any point. Tulving
also assumed that the memory trace for an
item generally consists of the item itself plus
information about context (e.g., the setting;
current mood state). It follows that memory
performance should be best when the context
at test is the same as that at the time of
learning.
The encoding specificity principle resembles
the notion of transfer-appropriate processing
(Morris et al., 1977; see earlier in chapter).
The central idea behind transfer-appropriate
processing is that long-term memory is best
when the processing performed at the time of
test closely resembles that at the time of learning.
The main difference between these two notions
is that transfer-appropriate processing focuses
more directly on the processes involved.
encoding specificity principle: the notion
that retrieval depends on the overlap between
the information available at retrieval and the
information in the memory trace.
KEY TERM
6 LEARNING, MEMORY, AND FORGETTING 243
Evidence
Many attempts to test the encoding specificity
principle involve two learning conditions
and two retrieval conditions. This allows the
researcher to show that memory depends on
the information in the memory trace and the
information available in the retrieval environment.
Thomson and Tulving (1970) presented
pairs of words in which the first was the cue
and the second was the to-be-remembered word.
The cues were weakly associated with the list
words (e.g., “Train–BLACK”) or strongly
associated (e.g., “White–BLACK”). Some of
the to-be-remembered items were tested by weak
cues (e.g., “Train–?”), and others were tested
by strong cues (e.g., “White–?”).
Thomson and Tulving’s (1970) findings are
shown in Figure 6.21. As predicted, recall performance
was best when the cues provided at
recall matched those provided at learning. Any
change in the cues reduced recall, even when
the shift was from weak cues at input to strong
cues at recall. Why were strong cues associated
with relatively poor memory performance
when learning had involved weak cues? Tulving
assumed that participants found it easy to generate
the to-be-remembered words to strong
cues, but failed to recognise them as appropriate.
However, that is not the whole story.
Higham and Tam (2006) found that participants
given strong cues at test after weak
cues at learning found it harder to generate the
target words than other participants given
strong cues at test who had not previously
engaged in any learning! This happened because
participants given weak cues at learning had
formed a mental set to generate mainly weak
associates to cues.
Context is important in determining forgetting.
For example, information about current
mood state is often stored in the memory
trace, and there is more forgetting if the mood
state at the time of retrieval is different. The
notion that there should be less forgetting
when the mood state at learning and retrieval
is the same is known as mood-state-dependent
memory. There is reasonable evidence for moodstate-dependent
memory (see Chapter 15).
However, the effect is stronger when participants
are in a positive rather than negative
mood because they are motivated to alter
negative moods.
Input cues
90
80
70
60
50
40
30
Weak Strong
Percentageofwordsrecalled
Strong output cues
Weak output cues
Figure 6.21 Mean word
recall as a function of input
cues (strong or weak) and
output cues (strong or
weak). Data from Thomson
and Tulving (1970).
244 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Other kinds of context are also important.
Marian and Neisser (2000) studied the
effects of linguistic context. Russian–English
bilinguals recalled personal memories when
prompted with cues presented in the Russian
or English language. The participants generated
Russian memories (based on experiences in a
Russian-speaking context) to 64% of the cues
in Russian compared to only 35% when the
cues were in English.
The effects of context are often stronger
in recall than recognition memory. Godden and
Baddeley (1975) asked participants to learn a
list of words on land or 20 feet underwater,
followed by a test of free recall on land or
under water. Those who had learned on land
recalled more on land and those who learned
underwater did better when tested underwater.
Overall, recall was about 50% higher when
learning and recall took place in the same
environment. However, there was no effect of
context when Godden and Baddeley (1980)
repeated the experiment using recognition
memory rather than recall.
We all know that recognition is generally
better than recall. For example, we may be
unable to recall the name of an acquaintance
but if someone mentions their name we instantly
recognise it. One of the most dramatic
predictions from the encoding specificity principle
is that recall should sometimes be better
than recognition. This should happen when
the information in the recall cue overlaps more
than the information in the recognition cue
with the information stored in the memory
trace. Muter (1978) presented participants with
people’s names (e.g., DOYLE, THOMAS) and
asked them to circle those they “recognised as
a person who was famous before 1950”. They
were then given recall cues in the form of brief
descriptions plus first names of the famous
people whose surnames had appeared on the
recognition test (e.g., author of the Sherlock
Holmes stories: Sir Arthur Conan _____; Welsh
poet: Dylan ______). Participants recognised
only 29% of the names but recalled 42%
of them.
Brain-imaging evidence supporting the
encoding specificity principle and transferappropriate
processing was reported by Park
and Rugg (2008a). Participants were presented
with pictures and words and then on a subsequent
recognition test each item was tested with a
congruent cue (word–word and picture–picture
conditions) or an incongruent cue (word–picture
and picture–word conditions). As predicted by
the encoding specificity principle, memory performance
was better in the congruent than in
the incongruent conditions.
Park and Rugg (2008) carried out a further
analysis based on brain activity at learning
for items subsequently recognised. According
to transfer-appropriate processing, it is more
important for successful recognition for words
to be processed at learning in a “word-like” way
if they are tested by picture cues than by word
cues. In similar fashion, successful recognition
of pictures should depend more on “picturelike”
processing at study if they are tested by
pictures cues than by word cues. Both predictions
were supported, suggesting that longterm
memory is best when the processing at
the time of learning is similar to that at the
time of retrieval.
Rugg, Johnson, Park, and Uncapher (2008)
reported similar findings supporting transferMood-state
dependent memory refers to the
enhanced ease in recalling events that have an
emotional tone similar to our current mood.
If we’re feeling happy and content, we are
more likely to recall pleasant memories; when
depressed we are likely to retrieve unpleasant
ones.
6 LEARNING, MEMORY, AND FORGETTING 245
appropriate processing. However, they pointed
out that the similarity in patterns of brain
activation at learning and retrieval was never
very great. This probably happened because
only some of the processing at the time of
learning directly influenced what information
was stored. In addition, only some of the processing
at retrieval directly determined what
was retrieved.
Evaluation
The overlap between the information stored
in the memory trace and that available at the
time of retrieval often plays an important role
in determining whether retrieval occurs. Recent
neuroimaging evidence supports both the encoding
specificity principle and transfer-appropriate
processing. The emphasis placed on the role
of contextual information in retrieval is also
valuable. As we have seen, several different kinds
of context (e.g., external cues; internal mood
states; linguistic context) influence memory
performance.
What are the limitations of Tulving’s
approach? First, it is most directly applicable
to relatively simple memory tasks. Tulving
assumed that the information at the time of
test is compared in a simple and direct way
with the information stored in memory to
assess informational overlap. That is probably
often the case, as when we effortlessly recall
autobiographical memories when in the same
place as the original event (Berntsen & Hall,
2004). However, if you tried to answer the
question, “What did you do six days ago?, you
would probably use complex problem-solving
strategies not included within the encoding
specificity principle.
Second, the encoding specificity principle
is based on the assumption that retrieval
occurs fairly automatically. However, that is
not always the case. Herron and Wilding (2006)
found that active processes can be involved
in retrieval. People found it easier to recollect
episodic memories relating to when and where
an event occurred when they adopted the
appropriate mental set or frame of mind beforehand.
Adopting this mental set was associated
with increased brain activity in the right frontal
cortex.
Third, there is a danger of circularity
(Eysenck, 1978). Memory is said to depend on
“informational overlap”, but this is rarely
measured. It is tempting to infer the amount of
informational overlap from the level of memory
performance, which is circular reasoning.
Fourth, as Eysenck (1979) pointed out,
what matters is not only the informational
overlap between retrieval information and
stored information but also the extent to which
retrieval information allows us to discriminate
the correct responses from the incorrect ones.
Consider the following thought experiment
(Nairne, 2002b). Participants read aloud the
following list of words: write, right, rite, rite,
write, right. They are then asked to recall the
word in the third serial position. We increase
the informational overlap for some participants
by providing them with the sound of the item in
the third position. This increased informational
overlap is totally unhelpful because it does not
allow participants to discriminate the correct
spelling of the sound from the wrong ones.
Fifth, Tulving assumed that context influences
recall and recognition in the same way.
However, the effects of context are often greater
on recall than on recognition memory (e.g.,
Godden & Baddeley, 1975, 1980).
Consolidation
None of the theories considered so far provides
a wholly convincing account of forgetting over
time. They identify factors causing forgetting,
but do not indicate clearly why forgetting is
greater shortly after learning than later on.
Wixted (2004a, 2005) argued that the secret
of forgetting may lie in consolidation theory.
Consolidation is a process lasting for a long
consolidation: a process lasting several hours
or more which fixes information in long-term
memory.
KEY TERM
246 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
time (possibly years) that fixes information
in long-term memory. More specifically, it is
assumed that the hippocampus plays a vital
role in the consolidation of memories (especially
episodic memories for specific events and
episodes), with many memories being stored
ultimately in various parts of the neocortex,
including the temporal lobes. A key assumption
is that recently formed memories still being
consolidated are especially vulnerable to interference
and forgetting. Thus, “New memories
are clear but fragile and old ones are faded but
robust” (Wixted, 2004a, p. 265).
According to some versions of consolidation
theory (e.g., Eichenbaum, 2001), the process
of consolidation involves two major phases.
The first phase occurs over a period of hours
and centres on the hippocampus. The second
phase takes place over a period of time ranging
from days to years and involves interactions
between the hippocampal region, adjacent
entorhinal cortex and the neocortex. This
second phase only applies to episodic memories
and semantic memories (stored knowledge about
the world). It is assumed that such memories
are stored in the lateral neocortex of the temporal
and other lobes.
Consolidation theory is relevant to two of
the oldest laws of forgetting (Wixted, 2004b).
First, there is Jost’s (1897) law (mentioned
earlier), according to which the older of two
memories of the same strength will decay
slower. According to the theory, the explanation
is that the older memory has undergone
more consolidation and so is less vulnerable.
Second, there is Ribot’s (1882) law, according
to which the adverse effects of brain injury on
memory are greater on newly formed memories
than older ones. This is temporally graded
retrograde amnesia. It can be explained on the
basis that newly formed memories are most
vulnerable to disruption because they are at
an early stage of consolidation.
Evidence
Several lines of evidence support consolidation
theory. First, consider the form of the forgetting
curve. A decreasing rate of forgetting over time
since learning follows from the notion that recent
memories are vulnerable due to an ongoing
process of consolidation. Consolidation theory
also provides an explanation of Jost’s law.
Second, there is research on Ribot’s law,
which claims that brain damage adversely
affects recently-formed memories more than
older ones. Such research focuses on patients
with retrograde amnesia, which involves
impaired memory for events occurring before
the onset of the amnesia. Many of these patients
have suffered damage to the hippocampus as
the result of an accident, and this may have
a permanently adverse effect on consolidation
processes. As predicted by consolidation theory,
numerous patients with retrograde amnesia
show greatest forgetting for those memories
formed very shortly before the onset of amnesia
(Manns, Hopkins, & Squire, 2003). However,
retrograde amnesia can in extreme cases extend
for periods of up to 40 years (Cipolotti et al.,
2001).
Third, consolidation theory predicts that
newly-formed memories are more susceptible
to retroactive interference than are older
memories. On the face of it, the evidence is
inconsistent. The amount of retroactive interference
generally does not depend on whether
the interfering material is presented early or
late in the retention interval (see Wixted, 2005,
for a review). However, the great majority of
studies have only considered specific retroactive
interference (i.e., two responses associated
with the same stimulus). Consolidation theory
actually claims that newly-formed memories
are more susceptible to interference from
any subsequent learning. When the interfering
material is dissimilar, there is often more retroactive
interference when it is presented early
in the retention interval (Wixted, 2004a).
retrograde amnesia: impaired memory for
events occurring before the onset of amnesia.
KEY TERM
6 LEARNING, MEMORY, AND FORGETTING 247
Fourth, consider the effects of alcohol on
memory. People who drink excessive amounts
of alcohol sometimes suffer from “blackout”,
an almost total loss of memory for all events
occurring while they were conscious but very
drunk. These blackouts probably indicate a
failure to consolidate memories formed while
intoxicated. An interesting (and somewhat
surprising) finding is that memories formed
shortly before alcohol consumption are often
better remembered than those formed by individuals
who do not subsequently drink alcohol
(Bruce & Pihl, 1997). Alcohol probably prevents
the formation of new memories that would
interfere with the consolidation process of the
memories formed just before alcohol consumption.
Thus, alcohol protects previously formed
memories from disruption.
Fifth, Haist, Gore, and Mao (2001) obtained
support for the assumption that consolidation
consists of two phases. Participants identified
faces of people famous in the 1980s or 1990s.
Selective activation of the hippocampus for
famous faces relative to non-famous ones was
only found for those famous in the 1990s. In
contrast (and also as predicted), there was
greater activation in the entorhinal cortex connected
to widespread cortical areas for famous
faces from the 1980s than from the 1990s.
Evaluation
Consolidation theory has various successes to
its credit. First, it explains why the rate of
forgetting decreases over time. Second, consolidation
theory successfully predicts that retrograde
amnesia is greater for recently formed
memories and that retroactive interference
effects are greatest shortly after learning. Third,
consolidation theory identifies the brain
areas most associated with the two phases of
consolidation.
What are the limitations of consolidation
theory? First, we lack strong evidence that
consolidation processes are responsible for all
the effects attributed to them. For example,
there are various possible reasons why newly
formed memories are more easily disrupted
than older ones. Second, consolidation theory
indicates in a general way why newly formed
memory traces are especially susceptible to
interference effects, but not the more specific
finding that retroactive interference is greatest
when two different responses are associated
with the same stimulus. Third, forgetting can
involve several factors other than consolidation.
For example, forgetting is greater when
there is little informational overlap between
the memory trace and the retrieval environment
(i.e., encoding specificity principle), but this
finding cannot be explained within consolidation
theory. Fourth, consolidation theory ignores
cognitive processes influencing forgetting. For
example, as we have seen, the extent to which
forgetting due to proactive interference occurs
depends on individual differences in the ability to
inhibit or suppress the interfering information.
Architecture of memory•
According to the multi-store model, there are separate sensory, short-term, and long-term
stores. Much evidence (e.g., from amnesic patients) provides general support for the model,
but it is clearly oversimplified. According to the unitary-store model, short-term memory
is the temporarily activated part of long-term memory. There is support for this model
in the finding that amnesics’ performance on some “short-term memory” tasks is impaired.
However, it is likely that long-term memory plays an important role in determining performance
on such tasks.
CHAPTER SUMMARY
248 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Working memory•
Baddeley replaced the unitary short-term store with a working memory system consisting
of an attention-like central executive, a phonological loop holding speech-based information,
and a visuo-spatial sketchpad specialised for spatial and visual coding. More recently,
Baddeley has added a fourth component (episodic buffer) that integrates and holds information
from various sources. The phonological loop and visuo-spatial sketchpad are both
two-component systems, one for storage and one for processing. The central executive
has various functions, including inhibition, shifting, updating, and dual-task co-ordination.
Some brain-damaged patients are said to suffer from dysexecutive syndrome, but detailed
analysis indicates that different brain regions are associated with the functions of task
setting, monitoring, and energisation.
Levels of processing•
Craik and Lockhart (1972) focused on learning processes in their levels-of-processing
theory. They identified depth of processing (the extent to which meaning is processed),
elaboration of processing, and distinctiveness of processing as key determinants of longterm
memory. Insufficient attention was paid to the relationship between processes at
learning and those at retrieval. In addition, the theory isn’t explanatory, it is hard to assess
processing depth, and shallow processing can lead to very good long-term memory.
Implicit learning•
Much evidence supports the distinction between implicit and explicit learning, and amnesic
patients often show intact implicit learning but impaired explicit learning. In addition,
the brain areas activated during explicit learning (e.g., prefrontal cortex) differ from those
activated during implicit learning (e.g., striatum). However, it has proved hard to show
that claimed demonstrations of implicit learning satisfy the information and sensitivity
criteria. It is likely that the distinction between implicit and explicit learning is oversimplified,
and that more complex theoretical formulations are required.
Theories of forgetting•
Strong proactive and retroactive interference effects have been found inside and outside
the laboratory. People use active control processes to minimise proactive interference.
Much retroactive interference depends on automatic processes making the incorrect
responses accessible. Most evidence on Freud’s repression theory is based on adults claiming
recovered memories of childhood abuse. Such memories when recalled outside therapy
are more likely to be genuine than those recalled inside therapy. There is convincing
evidence for directed forgetting, with executive control processes within the prefrontal
cortex playing a major role. Forgetting is often cue-dependent, and the cues can be external
or internal. However, decreased forgetting over time is hard to explain in cue-dependent
terms. Consolidation theory provides an explanation for the form of the forgetting curve,
and for reduced forgetting rates when learning is followed by alcohol.
6 LEARNING, MEMORY, AND FORGETTING 249
Baddeley, A.D. (2007).• Working memory: Thought and action. Oxford: Oxford University
Press. Alan Baddeley, who has made massive contributions to our understanding of working
memory, has written an excellent overview of current knowledge in the area.
Baddeley, A.D., Eysenck, M.W., & Anderson, M.C. (2009).• Memory. Hove, UK: Psychology
Press. Several chapters in this book provide additional coverage of the topics discussed
in this chapter (especially forgetting).
Jonides, J., Lewis, R.L., Nee, D.E., Lustig, C.A., Berman, M.G., & Moore, K.S. (2008).•
The mind and brain of short-term memory. Annual Review of Psychology, 59, 193–224.
This chapter discusses short-term memory at length, and includes a discussion of the
multi-store and unitary-store models.
Repovš, G., & Baddeley, A. (2006). The multi-component model of working memory:•
Explorations in experimental cognitive psychology. Neuroscience, 139, 5–21. This article
provides a very useful overview of the working memory model, including a discussion of
some of the most important experiment findings.
Roediger, H.L. (2008). Relativity of remembering: Why the laws of memory vanished.•
Annual Review of Psychology, 59, 225–254. This chapter shows very clearly that learning
and memory are more complex and involve more factors than is generally assumed to be
the case.
Shanks, D.R. (2005). Implicit learning. In K. Lamberts & R. Goldstone (eds.),• Handbook
of cognition. London: Sage. David Shanks puts forward a strong case for being critical
of most of the evidence allegedly demonstrating the existence of implicit learning.
Wixted, J.T. (2004). The psychology and neuroscience of forgetting.• Annual Review of
Psychology, 55, 235–269. A convincing case is made that neuroscience has much to
contribute to our understanding of forgetting.
FURTHER READING
C H A P T E R 7
L O N G - T E R M M E M O R Y S Y S T E M S
information within a given class or domain.
For example, semantic memory is concerned
with general knowledge of different
kinds.
Properties and relations(2) : The properties
of a memory system, “include types of
information that fall within its domain,
rules by which the system operates, neural
substrates, and functions of the system
(what the system is ‘for’)” (Schacter et
al., 2000, p. 629).
Convergent dissociations(3) : Any given
memory system should differ clearly in
various ways from other memory systems.
Amnesia
Convincing evidence that there are several longterm
memory systems comes from the study of
brain-damaged patients with amnesia. Such
patients have problems with long-term memory,
but if you are a movie fan you may have mistaken
ideas about the nature of amnesia (Baxendale,
2004). In the movies, serious head injuries
typically cause characters to forget the past while
still being fully able to engage in new learning.
In the real world, however, new learning is
generally greatly impaired. In the movies, amnesic
individuals often suffer a profound loss of identity
or their personality changes completely. For
example, consider the film Overboard (1987).
In that film, Goldie Hawn falls from her yacht,
and immediately switches from being a rich,
spoilt socialite into a loving mother. Such personality
shifts are extremely rare. Most bizarrely,
INTRODUCTION
We have an amazing variety of information
stored in long-term memory. For example,
long-term memory can contain details of our
last summer holiday, the fact that Paris is the
capital of France, information about how to
ride a bicycle or play the piano, and so on.
Much of this information is stored in the form
of schemas or organised packets of knowledge,
and is used extensively during language comprehension.
The relationship between schematic
knowledge and language comprehension is
discussed in Chapter 10.
In view of the variety of information in longterm
memory, Atkinson and Shiffrin’s (1968)
notion that there is a single long-term memory
store seems improbable (see Chapter 6). As we
will see, it is generally accepted that there
are several major long-term memory systems.
For example, Schacter and Tulving (1994)
argued that there are four major long-term
memory systems (episodic memory, semantic
memory, the perceptual representation system,
and procedural memory), and their approach
will be discussed. However, there has been
some controversy about the precise number
and nature of long-term memory systems.
What do we mean by a memory system?
According to Schacter and Tulving (1994) and
Schacter, Wagner, and Buckner (2000), we can
use three criteria to identify a memory system:
Class inclusion operations(1) : Any given
memory system handles various kinds of
252 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Only slightly impaired short-term memory•
on measures such as digit span (the ability
to repeat back a random string of digits).
Some remaining learning ability after the•
onset of amnesia.
The reasons why patients have become
amnesic are very varied. Bilateral stroke is one
the rule of thumb in the movies is that the best
cure for amnesia caused by severe head injury is
to suffer another massive blow to the head!
We turn now to the real world. Amnesic
patients are sometimes said to suffer from the
“amnesic syndrome” consisting of the following
features:
Anterograde amnesia• : a marked impairment
in the ability to remember new information
learned after the onset of amnesia. HM is
a famous example of anterograde amnesia
(see box).
Retrograde amnesia• : problems in remembering
events occurring prior to the onset
of amnesia (see Chapter 6).
The famous case HM
HM was the most-studied amnesic patient of
all time. He suffered from very severe epilepsy
starting at the age of ten.This eventually led to
surgery by William Beecher Scoville, involving
removal of the medial temporal lobes including
the hippocampus. HM had his operation on
23 August 1953, and since then he “forgets the
events of his daily life as fast as they occur”
(Scoville & Milner, 1957). More dramatically,
Corkin (1984,p.255) reported many years after
the operation that HM,“does not know where
he lives, who cares for him, or where he ate
his last meal. . . . In 1982 he did not recognise a
picture of himself that had been taken on his
fortieth birthday in 1966.” When shown faces
of individuals who had become famous after the
onset of his amnesia, HM could only identify
John Kennedy and Ronald Reagan. In spite of
everything, HM still had a sense of humour.
When Suzanne Corkin asked him how he tried
to remember things, he replied, “Well, that
I don’t know ’cause I don’t remember [laugh]
what I tried” (Corkin, 2002, p. 158).
It would be easy to imagine that all HM’s
memory capacities were destroyed by surgery.
In fact, what was most striking (and of greatest
theoretical importance) was that he retained
the ability to form many kinds of long-term
memory as well as having good short-term memory
(e.g., on immediate span tasks;Wickelgren,
1968). For example, HM showed reasonable
learning on a mirror-tracing task (drawing objects
seen only in reflection), and he retained some
of this learning for one year (Corkin, 1968). He
also showed learning on the pursuit rotor,which
involves manual tracking of a moving target.HM
showed normal performance on a perceptual
identification task in which he had to identify
words presented very briefly. He identified
more words previously studied than words not
previously studied, thus showing evidence for
long-term memory.
Some reports indicated that his language
skills were reasonably well preserved.However,
Mackay,James,Taylor,and Marian (2007) reported
that he was dramatically worse than healthy
controls at language tasks such as detecting
grammatical errors or answering questions about
who did what to whom in sentences.
HM died on 2 December 2008 at the age
of 82. He was known only as HM to protect
his privacy, but after his death it was revealed
that his real name was Henry Gustav Molaison.
Researchers have focused on the patterns
of intact and impaired memory performance
shown by HM and other amnesic patients.The
theoretical insights they have produced will be
considered in detail in this chapter.
anterograde amnesia: reduced ability to
remember information acquired after the onset
of amnesia.
KEY TERM
7 LONG-TERM MEMORY SYSTEMS 253
amnesic patients. Furthermore, such patients
have provided some of the strongest evidence
supporting these distinctions.
Declarative vs. non-declarative
memory
The most important distinction between different
types of long-term memory is that between
declarative memory and non-declarative memory.
Declarative memory involves conscious recollection
of events and facts – it refers to memories
that can be “declared” or described. Declarative
memory is sometimes referred to as explicit
memory, defined as memory that “requires
conscious recollection of previous experiences”
(Graf & Schacter, 1985, p. 501).
In contrast, non-declarative memory does
not involve conscious recollection. Typically,
we obtain evidence of non-declarative memory
by observing changes in behaviour. For example,
consider someone learning how to ride a bicycle.
We would expect their cycling performance
(a form of behaviour) to improve over time even
though they could not consciously recollect
what they had learned about cycling. Nondeclarative
memory is also known as implicit
memory, which involves enhanced performance
in the absence of conscious recollection.
factor causing amnesia, but closed head injury
is the most common cause. However, patients
with closed head injury often have several
cognitive impairments, which makes interpreting
their memory deficit hard. As a result, most
experimental work has focused on patients who
became amnesic because of chronic alcohol abuse
(Korsakoff’s syndrome; see Glossary). There are
two problems with using Korsakoff patients to
study amnesia. First, the amnesia usually has
a gradual onset, being caused by an increasing
deficiency of the vitamin thiamine associated
with chronic alcoholism. That makes it hard to
know whether certain past events occurred before
or after the onset of amnesia. Second, brain damage
in Korsakoff patients is often rather widespread.
Structures within the diencephalon (e.g.,
the hippocampus and the amygdala) are usually
damaged. There is often damage to the frontal
lobes, and this can produce various cognitive
deficits not specific to the memory system. It would
be easier to interpret findings from Korsakoff
patients if the brain damage were more limited.
Other cases of amnesia typically have damage
to the hippocampus and adjacent areas in the
medial temporal lobes. The brain areas associated
with amnesia are discussed more fully towards
the end of the chapter.
Why have amnesic patients contributed
substantially to our understanding of human
memory? The study of amnesia provides a
good test-bed for existing theories of healthy
memory. For example, strong evidence for the
distinction between short- and long-term memory
comes from studies on amnesic patients (see
Chapter 6). Some patients have severely impaired
long-term memory but intact short-term memory,
whereas a few patients show the opposite pattern.
The existence of these opposite patterns
forms a double dissociation (see Glossary) and
is good evidence for separate short- and longterm
stores.
The study of amnesic patients has also proved
very valuable in leading to various theoretical
developments. For example, distinctions such
as the one between declarative or explicit
memory and non-declarative or implicit memory
(discussed in the next section) were originally
proposed in part because of data collected from
declarative memory: a form of long-term
memory that involves knowing that something
is the case and generally involves conscious
recollection; it includes memory for facts
(semantic memory) and memory for events
(episodic memory).
explicit memory: memory that involves
conscious recollection of information; see
implicit memory.
non-declarative memory: forms of long-term
memory that influence behaviour but do not
involve conscious recollection; priming and
procedural memory are examples of
non-declarative memory.
implicit memory: memory that does not
depend on conscious recollection; see explicit
memory.
KEY TERMS
254 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
extrastriate cortex, the left fusiform gyrus, and
bilateral inferior prefrontal cortex, areas that
are involved in stimulus identification.
Schott et al. (2005) found that different brain
areas were associated with memory retrieval
on declarative memory and non-declarative
tasks. Declarative retrieval was associated with
bilateral parietal and temporal and left frontal
increases in activation, whereas non-declarative
retrieval was associated with decreases in activation
in the left fusiform gyrus and bilateral
frontal and occipital regions. Thus, the brain
areas associated with declarative memory and
non-declarative memory are different both at
the time of encoding or learning and at the
time of retrieval. In addition, retrieval from
declarative memory is generally associated with
increased brain activation, whereas retrieval
from non-declarative memory is associated
with decreased brain activation.
For the rest of the chapter, we will discuss the
various forms of declarative and non-declarative
memory. Figure 7.1 provides a sketch map of
the ground we are going to be covering.
Declarative memory
We all have declarative or explicit memory for
many different kinds of memories. For example,
Declarative memory and non-declarative
memory seem to be very different. Evidence
for the distinction comes from amnesic patients.
They seem to have great difficulties in forming
declarative memories but their ability to form
non-declarative memories is intact or nearly
so. In the case of HM, he had extremely poor
declarative memory for personal events occurring
after the onset of amnesia and for faces
of those who had become famous in recent
decades (see Box on p. 252). However, he had
reasonable learning ability on tasks such as
mirror tracing, the pursuit rotor, and perceptual
identification. What these otherwise different
tasks have in common is that they all
involve non-declarative memory. As we will see
later in the chapter, the overwhelming majority
of amnesic patients have very similar patterns
of memory performance to HM.
Functional imaging evidence also supports
the distinction between declarative and nondeclarative
memory. Schott, Richardson-Klavehn,
Henson, Becker, Heinze, and Duzel (2006) found
that brain activation during learning that predicted
subsequent declarative memory performance
occurred in the bilateral medial temporal lobe
and the left prefrontal cortex. In contrast, brain
activation predicting subsequent non-declarative
memory performance occurred in the bilateral
Long-term memory
Declarative
(explicit)
Nondeclarative
(implicit)
Facts Events
Medial temporal lobe
Priming Precedural
(skills and
habits)
Associative learning:
classical and
operant conditioning
Nonassociative learning:
habituation and
sensitisation
Emotional
responses
Skeletal
musculature
Cortex Striatum Amygdala Cerebellum Reflex pathways
Figure 7.1 The main forms of long-term memory, all of which can be categorised as declarative (explicit) or
nondeclarative (implicit).The brain regions associated with each form of long-term memory are also indicated.
From Kandel, Kupferman, and Iverson (2000) with permission from McGraw Hill.
7 LONG-TERM MEMORY SYSTEMS 255
There are similarities between episodic
and semantic memory. Suppose you remember
meeting your friend yesterday afternoon at Starbuck’s.
That clearly involves episodic memory,
because you are remembering an event at a given
time in a given place. However, semantic memory
is also involved – some of what you remember
depends on your general knowledge about coffee
shops, what coffee tastes like, and so on.
Tulving (2002, p. 5) clarified the relationship
between episodic and semantic memory:
“Episodic memory...shares many features with
semantic memory, out of which it grew,...but
also possesses features that semantic memory
does not. . . . Episodic memory is a recently
evolved, late-developing, and early-deteriorating
past-oriented memory system, more vulnerable
than other memory systems to neuronal
dysfunction.”
What is the relationship between episodic
memory and autobiographical memory (discussed
in Chapter 8)? They are similar in that
both forms of memory are concerned with
personal experiences from the past, and there is
no clear-cut distinction between them. However,
there are some differences. Much information
in episodic memory is relatively trivial and is
remembered for only a short period of time.
In contrast, autobiographical memory stores
information for long periods of time about
events and experiences of some importance to
the individual concerned.
Non-declarative memory
A defining characteristic of non-declarative
memory is that it is expressed by behaviour
we remember what we had for breakfast this
morning or that “le petit déjeuner” is a French
expression meaning “breakfast”. Tulving (1972)
argued that these kinds of memories are very
different, and he used the terms “episodic
memory” and “semantic memory” to refer to
the difference. Episodic memory involves storage
(and retrieval) of specific events or episodes
occurring in a given place at a given time.
According to Wheeler, Stuss, and Tulving (1997,
p. 333), the main distinguishing characteristic
of episodic memory is, “its dependence on
a special kind of awareness that all healthy
human adults can identify. It is the type of
awareness experienced when one thinks back
to a specific moment in one’s personal past and
consciously recollects some prior episode or
state as it was previously experienced.”
In contrast, semantic memory “is the aspect
of human memory that corresponds to general
knowledge of objects, word meanings, facts and
people, without connection to any particular
time or place” (Patterson, Nestor, & Rogers,
2007, p. 976). Wheeler et al. (1997) shed further
light on the distinction between semantic and
episodic memory. They pointed out that semantic
memory involves “knowing awareness” rather
than the “self-knowing” associated with episodic
memory.
Semantic memory goes beyond the meaning of
words and extends to sensory attributes such as
taste and colour; and to general knowledge of
how society works, such as how to behave in a
supermarket.
episodic memory: a form of long-term
memory concerned with personal experiences
or episodes that occurred in a given place at a
specific time; see semantic memory.
semantic memory: a form of long-term
memory consisting of general knowledge about
the world, concepts, language, and so on;
see episodic memory.
KEY TERMS
256 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
hugely influential and accounts for numerous
findings on long-term memory. As you read
through this chapter, you will see that some
doubts have been raised about the distinction.
Towards the end of this chapter, an alternative
approach is discussed under the heading, “Beyond
declarative and non-declarative memory:
amnesia”. Much of that section focuses on
research suggesting that the notion that amnesic
patients have deficient declarative memory but
intact non-declarative memory is oversimplified.
EPISODIC VS. SEMANTIC
MEMORY
If episodic and semantic memory form separate
memory systems, there should be several important
differences between them. We will consider
three major areas of research here.
The first major area of research involves
testing the ability of amnesic patients to acquire
episodic and semantic memories after the onset
of amnesia. In other words, the focus was
on the extent of anterograde amnesia. Spiers,
Maguire, and Burgess (2001) reviewed 147
cases of amnesia involving damage to the
hippocampus or fornix. There was impairment
of episodic memory in all cases, whereas many
of the patients had only modest problems with
semantic memory. Thus, the impact of brain
damage was much greater on episodic than on
semantic memory, suggesting that the two types
of memory are distinctly different. Note that
and does not involve conscious recollection.
Schacter et al. (2000) identified two nondeclarative
memory systems: the perceptual
representation system and procedural memory:
the perceptual representation system “can be
viewed as a collection of domain-specific
modules that operate on perceptual information
about the form and structure of words and
objects” (p. 635). Of central importance within
this system is repetition priming (often just called
priming): stimulus processing occurs faster and/
or more easily on the second and successive
presentations of a stimulus. For example, we
may identify a stimulus more rapidly the second
time it is presented than the first time. What
we have here is learning related to the specific
stimuli used during learning. Schacter, Wig, and
Stevens (2007, p. 171) provided a more technical
definition: “Priming refers to an improvement
or change in the identification, production, or
classification of a stimulus as a result of a prior
encounter with the same or a related stimulus.”
The fact that repetition priming has been obtained
in the visual, auditory, and touch modalities
supports the notion that there is a perceptual
representation system.
In contrast, procedural memory “refers to
the learning of motor and cognitive skills, and
is manifest across a wide range of situations.
Learning to ride a bike and acquiring reading
skills are examples of procedural memory”
(Schacter et al., 2000, p. 636). The term “skill
learning” has often been used to refer to what
Schacter et al. defined as procedural memory.
It is shown by learning that generalises to
several stimuli other than those used during
training. On the face of it, this seems quite
different from the very specific learning associated
with priming.
Reference back to Figure 7.1 will indicate
that there are other forms of non-declarative
memory: classical conditioning, operant conditioning,
habituation, and sensitisation. We
will refer to some of these types of memory
later in the chapter as and when appropriate.
There is one final point. The distinction
between declarative or explicit memory and
non-declarative or implicit memory has been
perceptual representation system: an
implicit memory system thought to be involved
in the faster processing of previously presented
stimuli (e.g., repetition priming).
repetition priming: the finding that stimulus
processing is faster and easier on the second
and successive presentations.
procedural memory/knowledge: this is
concerned with knowing how, and includes the
ability to perform skilled actions; see
declarative memory.
KEY TERMS
7 LONG-TERM MEMORY SYSTEMS 257
great problems with both episodic and semantic
memory? The answer may be that they have
damage to the hippocampus and to the underlying
cortices. This makes sense given that the
two areas are adjacent.
Some support for the above hypothesis was
reported by Verfaellie, Koseff, and Alexander
(2000). They studied a 40-year-old woman
(PS), who, as an adult, suffered brain damage
to the hippocampus but not the underlying
cortices. In spite of her severe amnesia and
greatly impaired episodic memory, she managed
to acquire new semantic memories (e.g., identifying
people who only became famous after
the onset of her amnesia).
We have seen that some amnesic patients
perform relatively better on tasks involving semantic
memory than on those involving episodic
memory. However, there is a potential problem
of interpretation, because the opportunities for
learning are generally greater with semantic
memory (e.g., acquiring new vocabulary). Thus,
one reason why these patients do especially
poorly on episodic memory tasks may be because
of the limited time available for learning.
The second main area of research involves
amnesic patients suffering from retrograde
amnesia (i.e., impaired memory for learning
occurring before the onset of amnesia; see also
Chapter 6). If episodic and semantic memory
form different systems, we would expect to find
some patients showing retrograde amnesia only
for episodic or semantic memory. For example,
consider KC, who suffered damage to several
cortical and subcortical brain regions, including
the medial temporal lobes. According to Tulving
(2002, p. 13), “[KC’s] retrograde amnesia is
highly asymmetrical: He cannot recollect any
personally experienced events..., whereas his
semantic knowledge acquired before the critical
accident is still reasonably intact. His knowledge
of mathematics, history, geography, and
other ‘school subjects’, as well as his general
knowledge of the world is not greatly different
from others’ at his educational level.”
The opposite pattern was reported by
Yasuda, Watanabe, and Ono (1997), who studied
an amnesic patient with bilateral lesions to the
the memory problems of amnesic patients are
limited to long-term memory. According to
Spiers et al. (p. 359), “None of the cases was
reported to have impaired short-term memory
(typically tested using digit span – the immediate
recall of verbally presented digits).”
We would have stronger evidence if we could
find amnesic patients with very poor episodic
memory but intact semantic memory. Such
evidence was reported by Vargha-Khadem,
Gadian, Watkins, Connelly, Van Paesschen, and
Mishkin (1997). They studied three patients, two
of whom had suffered bilateral hippocampal
damage at an early age before they had had
the opportunity to develop semantic memories.
Beth suffered brain damage at birth, and Jon
did so at the age of four. Jon suffered breathing
problems which led to anoxia and caused his
hippocampus to be less than half the normal size.
Both of these patients had very poor episodic
memory for the day’s activities, television programmes,
and telephone conversations. In spite
of this, Beth and Jon both attended ordinary
schools, and their levels of speech and language
development, literacy, and factual knowledge
(e.g., vocabulary) were within the normal range.
Vargha-Khadem, Gadian, and Mishkin (2002)
carried out a follow-up study on Jon at the age
of 20. As a young adult, he had a high level
of intelligence (IQ = 120), and his semantic
memory continued to be markedly better than
his episodic memory. Brandt, Gardiner, VarghaKhadem,
Baddeley, and Mishkin (2006) obtained
evidence suggesting that Jon’s apparent recall
of information from episodic memory actually
involved the use of semantic memory. Thus,
Jon’s episodic memory may be even worse than
was previously assumed.
How can we explain the ability of Beth
and Jon to develop fairly normal semantic
memory in spite of their grossly deficient
episodic memory? Vargha-Khadem et al. (1997)
argued that episodic memory depends on the
hippocampus, whereas semantic memory depends
on the underlying entorhinal, perihinal, and
parahippocampal cortices. The brain damage
suffered by Beth and Jon was centred on the
hippocampus. Why do so many amnesics have
258 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
effects being limited to a period of about ten
years. Third, damage to the neocortex impairs
semantic memory. Westmacott, Black, Freedman,
and Moscovitch (2004) studied retrograde
amnesia in patients suffering from Alzheimer’s
disease (a progressive disease in which cognitive
abilities including memory are gradually
lost). The severity of retrograde amnesia for
vocabulary and famous names in these patients
increased with the progress of the disease.
This suggests that the impairment in semantic
memory was related to the extent of degeneration
of neocortex.
The third main area of research involves
functional neuroimaging. Studies in this area
indicate that episodic and semantic memory
involve activation of somewhat different parts
of the brain. In a review, Wheeler et al. (1997)
reported that the left prefrontal cortex was more
active during episodic than semantic encoding.
What about brain activation during retrieval?
Wheeler et al. reported that the right prefrontal
cortex was more active during episodic memory
retrieval than during semantic memory retrieval
in 25 out of 26 neuroimaging studies.
Further neuroimaging evidence was reported
by Prince, Tsukiura, and Cabeza (2007). The
left hippocampus was associated with episodic
encoding but not with semantic memory retrieval,
whereas the lateral temporal cortex was associated
with semantic memory retrieval but not
with episodic encoding. The greater involvement
of the hippocampus with episodic than with
semantic memory is consistent with the research
on brain-damaged patients discussed above
(Moscovitch et al., 2006). In addition, Prince
et al. (2007) found within the left inferior
prefrontal cortex that a posterior region was
involved in semantic retrieval, a mid-region was
associated with both semantic retrieval and
episodic encoding, and a more anterior region
was associated with episodic encoding only
temporal lobe. She had very poor ability to
remember public events, cultural items, historical
figures, and some items of vocabulary from the
time prior to the onset of amnesia. However, she
was reasonably good at remembering personal
experiences from episodic memory dating back
to the pre-amnesia period.
Kapur (1999) reviewed studies on retrograde
amnesia. There was clear evidence for a
double dissociation: some patients showed more
loss of episodic than semantic memory, whereas
others showed the opposite pattern.
Which brain regions are involved in retrograde
amnesia? The hippocampal complex of
the medial temporal lobe (including the hippocampus
proper, dentate gyrus, the perirhinal,
enterorhinal, and parahippocampal cortices)
is of special importance. According to multiple
trace theory (e.g., Moscovitch, Nadel, Winocur,
Gilboa, & Rosenbaum, 2006), every time an
episodic memory is retrieved, it is re-encoded.
This leads to multiple episodic traces of events
distributed widely throughout the hippocampal
complex. Of key importance, it is assumed
theoretically that detailed episodic or autobiographical
memories of the past always depend
on the hippocampus. Semantic memories initially
depend heavily on the hippocampus, but
increasingly depend on neocortex.
Multiple trace theory has received support
from studies on healthy individuals as well as
patients with retrograde amnesia. For example,
Gilboa, Ramirez, Kohler, Westmacott, Black,
and Moscovitch (2005) studied people’s personal
recollections of recent and very old events
going back several decades. Activation of the
hippocampus was associated with the vividness
of their recollections rather than the age of those
recollections.
There is reasonable support for predictions
following from multiple trace theory. First,
the severity of retrograde amnesia in episodic
memory is fairly strongly related to the amount
of damage to the hippocampal complex, although
frontal areas are also often damaged (Moscovitch
et al., 2006). Second, damage to the hippocampal
complex generally has less effect on semantic
memory than on episodic memory, with any
Alzheimer’s disease: a condition involving
progressive loss of memory and mental abilities.
KEY TERM
7 LONG-TERM MEMORY SYSTEMS 259
during various working-memory tasks, which
raises the possibility that these regions of
prefrontal cortex are involved in executive
processing or cognitive control.
EPISODIC MEMORY
As we saw in Chapter 6, most episodic memories
exhibit substantial and progressive forgetting
over time. However, there are some exceptions.
For example, Bahrick, Bahrick, and Wittlinger
(1975) made use of photographs from highschool
yearbooks dating back many years.
Ex-students showed remarkably little forgetting
of information about their former classmates
at retention intervals up to 25 years. Performance
was 90% for recognising a name as being that
of a classmate, for recognising a classmate’s
photograph, and for matching a classmate’s
name to his/her school photograph. Performance
remained very high on the last two tests even
at a retention interval of almost 50 years, but
performance on the name recognition task
declined.
Bahrick, Hall, and Da Costa (2008) asked
American ex-college students to recall their
academic grades. Distortions in recall occurred
shortly after graduation but thereafter remained
fairly constant over retention intervals up to
54 years. Perhaps not surprisingly, the great
when semantic retrieval was also involved. These
various findings suggested that, “episodic and
semantic memory depend on different but closely
interacting memory systems” (Prince et al.,
2007, p. 150).
Evaluation
There is convincing evidence for separate episodic
and semantic memory systems. The relevant
evidence is of various kinds, and includes studies
of anterograde and retrograde amnesia as well
as numerous neuroimaging studies.
It should be emphasised that the episodic
and semantic memory systems typically combine
in their functioning. For example, suppose you
retrieve an episodic memory of having an enjoyable
picnic in the countryside. To do this, you
need to retrieve semantic information about the
concepts (e.g., picnic; grass) contained in your
episodic memory. We have just seen that Prince
et al. (2007) found evidence that some of the
same brain regions are associated with episodic
and semantic memory. In similar fashion, Nyberg
et al. (2003) found that four regions of prefrontal
cortex were activated during episodic
and semantic memory tasks: left fronto-polar
cortex, left mid-ventrolateral prefrontal cortex,
left mid-dorsolateral prefrontal cortex, and
dorsal anterior cingulate cortex. Nyberg et al.
also found that the same areas were activated
Bahrick et al. (1975) found
that adults were remarkably
good at recognising the
photographs of those with
whom they had been at
school almost so years later.
260 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
of thing academic psychologists think about!),
he realised the man was a ticket-office clerk
at Wimbledon railway station. Thus, initial
recognition based on familiarity was replaced
by recognition based on recollection.
There are various ways of distinguishing
between these two forms of recognition memory.
Perhaps the simplest is the remember/know
task, in which participants indicate subjectively
whether their positive recognition decisions were
based on recollection of contextual information
(remember responses) or solely on familiarity
(know responses). The crucial issue here is
deciding whether recollection and familiarity
involve different processes – sceptics might argue
that the only real difference is that strong memory
traces give rise to recollection judgements and
weak memory traces give rise to familiarity
judgements. Dunn (2008) is one such sceptic.
He carried out a meta-analysis of 37 studies
using the remember–know task, and found
that the findings could be explained in terms
of a single process based on memory strength.
However, as we will see, there is much support
for dual-process models.
We saw earlier that the medial temporal lobe
and adjacent areas are of crucial importance
in episodic memory. There is now reasonable
support for a more precise account of the brain
areas involved in recognition memory provided
by the binding-of-item-and-context model (Diana
et al., 2007) (see Figure 7.2):
Perirhinal cortex receives information(1)
about specific items (“what” information
needed for familiarity judgements).
Parahippocampal cortex receives informa-(2)
tion about context (“where” information
useful for recollection judgements).
The hippocampus receives what and where(3)
information (both of great importance to
episodic memory), and binds them together
to form item-context associations that
permit recollection.
Functional neuroimaging studies provide
support for the binding-of-item-and-context
model. Diana et al. (2007) combined findings
majority of distortions involved inflating the
actual grade.
Bahrick (1984) used the term permastore
to refer to very long-term stable memories.
This term was based on permafrost, which is
the permanently frozen subsoil found in polar
regions. It seems probable that the contents of
the permastore consist mainly of information
that was very well-learned in the first place.
We turn now to a detailed consideration
of how we can assess someone’s episodic memory.
Recognition and recall are the two main types
of episodic memory test. The basic recognitionmemory
test involves presenting a series of
items, with participants deciding whether each
one was presented previously. As we will see,
however, more complex forms of recognitionmemory
test have also been used. There are
three basic forms of recall test: free recall, serial
recall, and cued recall. Free recall involves
producing to-be-remembered items in any order
in the absence of any specific cues. Serial recall
involves producing to-be-remembered items in
the order in which they were presented originally.
Cued recall involves producing to-be-remembered
items in the presence of cues. For example,
‘cat–table’ might be presented at learning and
the cue, ‘cat–?’ might be given at test.
Recognition memory
Recognition memory can involve recollection
or familiarity (e.g., Mandler, 1980). According
to Diana, Yonelinas, and Ranganath (2007,
p. 379), “Recollection is the process of recognising
an item on the basis of the retrieval of
specific contextual details, whereas familiarity
is the process of recognising an item on the
basis of its perceived memory strength but
without retrieval of any specific details about
the study episode.”
We can clarify the distinction with the
following anecdote. Several years ago, the first
author walked past a man in Wimbledon, and
was immediately confident that he recognised
him. However, he simply could not think of
the situation in which he had seen the man
previously. After some thought (this is the kind
7 LONG-TERM MEMORY SYSTEMS 261
out a meta-analysis of recognition-memory
studies involving amnesic patients with and
without lesions in the medial temporal lobes
(including the hippocampus). Of central interest
was the memory performance of these two
groups on measures of recollection and familiarity
(see Figure 7.3). Both groups performed
consistently worse than healthy controls. Most
importantly, however, the patient group with
medial temporal lobe lesions only had significantly
worse performance than the other patient
group with recollection and not with familiarity.
This suggests that the hippocampus and adjacent
regions are especially important in supporting
recollection.
from several studies of recognition memory
that considered patterns of brain activation
during encoding and retrieval (see Figure 7.2).
As predicted, recollection was associated with
more activation in parahippocampal cortex
and the hippocampus than in the perirhinal
cortex. In contrast, familiarity was associated
with more activation in the perirhinal cortex than
the parahippocampal cortex or hippocampus.
It is a reasonable prediction from the above
model that amnesic patients (who nearly always
have extensive hippocampal damage) should
have greater problems with recognition based
on recollection than recognition based on familiarity.
Skinner and Fernandes (2007) carried
Figure 7.2 (a) locations of
the hippocampus (red), the
perirhinal cortex (blue), and
the parahippocampal cortex
(green); (b) the binding-ofitem-and-context
model.
Reprinted from Diana et al.
(2007), Copyright © 2007,
with permission from
Elsevier.
(b)
Hippocampus
“Binding of
items & contexts”
Perirhinal cortex
‘Items’
Parahippocampal
cortex
‘Context’
“What” “Where”
Entorhinal
cortex
Parahippocampa
gyrus
Neocortical
input
(a)
A B
BA
262 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
successful free recall is associated with higher
levels of brain activity in several areas at encoding
and at retrieval than successful recognition
memory. This suggests that free recall is in
some sense more “difficult” than recognition
memory. Third, Staresina and Davachi’s (2006)
finding that some brain areas are associated
with successful free recall but not recognition
memory suggests that free recall involves processes
additional to those involved in recognition
memory. As indicated above, inter-item processing
is the most obvious requirement specific
to free recall.
Is episodic memory constructive?
We use episodic memory to remember past
events that have happened to us. You might
imagine that our episodic memory system would
work like a video recorder, providing us with
accurate and detailed information about past
events. That is not the case. As Schacter and
Addis (2007, p. 773) pointed out, “Episodic
memory is . . . a fundamentally constructive,
rather than reproductive process that is prone
to various kinds of errors and illusions.” Plentiful
evidence for this constructive view of episodic
memory is discussed in other chapters. In
Chapter 8, we discuss research showing how
the constructive nature of episodic memory leads
eyewitnesses to produce distorted memories of
what they have seen. In Chapter 10, we discuss
the influential views of Bartlett (1932). His
central assumption was that the knowledge we
possess can produce systematic distortions and
errors in our episodic memories, an assumption
that has been supported by much subsequent
research.
Why are we saddled with an episodic memory
system that is so prone to error? Schacter and
Addis (2007) identified three reasons. First, it
would require an incredible amount of processing
to produce a semi-permanent record of all our
experiences. Second, we generally want to access
the gist or essence of our past experiences; thus,
we want our memories to be discriminating by
omitting the trivial details. Third, imagining
possible future events and scenarios is important
Recall memory
Some research on recall is discussed in Chapter 6.
Here, we will focus on whether the processes
involved in free recall are the same as those
involved in recognition memory. In an important
study, Staresina and Davachi (2006) used
three memory tests: free recall, item recognition
(familiarity), and associative recognition (recollection).
Successful memory performance on
all three tests was associated with increased
activation in the left hippocampus and left
ventrolateral prefrontal cortex at the time of
encoding. This was most strongly the case with
free recall and least strongly the case with item
recognition. In addition, only successful subsequent
free recall was associated with increased
activation in the dorsolateral prefrontal cortex
and posterior parietal cortex. The most likely
explanation of this finding is that successful
free recall involves forming associations (in this
case between items and the colours in which they
were studied), something that is not required
for successful recognition memory.
What conclusions can we draw? First, the
finding that similar brain areas are associated
with successful free recall and recognition
suggests that there are important similarities
between the two types of memory test. Second,
0.6
0.5
0.4
0.3
0.2
0.1
0
Control MTL lesion Non-MTL lesion
Memoryprocessestimate
Recollection Familiarity
Memory process
Figure 7.3 Mean recollection and familiarity
estimates for healthy controls, patients with medial
temporal lobe (MTL) lesions, and patients with
non-MTL lesions. Reprinted from Skinner and
Fernandes (2007), Copyright © 2007, with permission
from Elsevier.
7 LONG-TERM MEMORY SYSTEMS 263
during the generation phase as well. However,
there were higher levels of activity in several areas
(e.g., the right frontopolar cortex; the left inferior
frontal gyrus) during the generation of future than
of past events. This suggests that more intensive
constructive processes are required to imagine
future events than to retrieve past events.
Evaluation
It has been assumed by many theorists, starting
with Bartlett (1932), that episodic memory
relies heavily on constructive processes, and
there is convincing evidence to support that
assumption (see Chapters 8 and 10). The further
assumption by Schacter and Addis (2007) that
the same constructive processes involved in
episodic memory for past events are also involved
in imaging the future is an exciting development.
The initial findings from amnesic patients
and functional neuroimaging studies are supportive.
However, further research is needed
to clarify the reasons why there are higher levels
of brain activation when individuals imagine
future events than when they recall past events.
SEMANTIC MEMORY
Our organised general knowledge about the
world is stored in semantic memory. The
content of such knowledge can be extremely
varied, including information about the French
language, the rules of hockey, the names of
capital cities, and the authors of famous books.
How is information organised within semantic
memory? Most is known about the organisation
of concepts, which are mental representations
of categories of objects or items. We will start
by considering influential models focusing on
the ways in which concepts are interconnected.
After that, we will consider the storage of information
about concepts within the brain.
to us for various reasons (e.g., forming plans
for the future). Perhaps the constructive processes
involved in episodic memory are also
used to imagine the future.
Evidence
We typically remember the gist of what we
have experienced previously, and our tendency
to remember gist increases with age. Consider
a study by Brainerd and Mojardin (1998).
Children aged 6, 8, and 11 listened to sets of
three sentences (e.g., “The coffee is hotter than
the tea”; “The tea is hotter than the cocoa”;
“The cocoa is hotter than the soup”). On the
subsequent recognition test, participants decided
whether the test sentences had been presented
initially in precisely that form. The key condition
was one in which sentences having the same
meaning as original sentences were presented
(e.g., “The cocoa is cooler than the tea”). False
recognition on these sentences increased steadily
with age.
We turn now to the hypothesis that imagining
future events involves the same processes
as those involved in remembering past events.
On that hypothesis, individuals with very poor
episodic memory (e.g., amnesic patients) should
also have impaired ability to imagine future
events. Hassabis, Kumaran, Vann, and Maguire
(2007) asked amnesic patients and healthy
controls to imagine future events (e.g., “Imagine
you are lying on a white sandy beach in a
beautiful tropical bay”). The amnesic patients
produced imaginary experiences consisting of
isolated fragments of information lacking
the richness and spatial coherence of the
experiences imagined by the controls.
Addis, Wong, and Schacter (2007) compared
brain activity when individuals generated
past and future events and then elaborated on
them. There was considerable overlap in patterns
of brain activity during the elaboration phase.
The areas activated during elaboration of past
and future events included the left anterior
temporal cortex (associated with conceptual
and semantic information about one’s life) and
the left frontopolar cortex (associated with selfreferential
processing). There was some overlap
concepts: mental representations of categories
of objects or items.
KEY TERM
264 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
hierarchy. In contrast, the sentence, “A canary
can fly”, should take longer because the concept
and property are separated by one level
in the hierarchy. The sentence, “A canary has
skin”, should take even longer because two
levels separate the concept and the property.
As predicted, the time taken to respond to true
sentences became progressively slower as the
separation between the subject of the sentence
and the property became greater.
The model is right in its claim that we often
use semantic memory successfully by inferring
the right answer. For example, the information
that Leonardo da Vinci had knees is not stored
directly in semantic memory. However, we
know Leonardo da Vinci was a human being,
and that human beings have knees, and so we
confidently infer that Leonardo da Vinci had
knees. This is the kind of inferential process
proposed by Collins and Quillian (1969).
In spite of its successes, the model suffers
from various problems. A sentence such as, “A
canary is yellow”, differs from, “A canary has
skin”, not only in the hierarchical distance
between the concept and its property, but also
in familiarity. Indeed, you have probably never
encountered the sentence, “A canary has skin”,
in your life before! Conrad (1972) found that
hierarchical distance between the subject and
the property had little effect on verification
time when familiarity was controlled.
Network models
We can answer numerous simple questions about
semantic memory very rapidly. For example,
it takes about one second to decide a sparrow
is a bird, or to think of a fruit starting with p.
This great efficiency suggests that semantic
memory is highly organised or structured.
The first systematic model of semantic
memory was put forward by Collins and Quillian
(1969). Their key assumption was that semantic
memory is organised into hierarchical networks
(see Figure 7.4). The major concepts (e.g.,
animal, bird, canary) are represented as nodes,
and properties or features (e.g., has wings; is
yellow) are associated with each concept. You
may wonder why the property “can fly” is stored
with the bird concept rather than with the
canary concept. According to Collins and
Quillian, those properties possessed by nearly
all birds (e.g., can fly; has wings) are stored only
at the bird node or concept. The underlying
principle is one of cognitive economy: property
information is stored as high up the hierarchy
as possible to minimise the amount of information
stored.
According to the model of Collins and
Quillian (1969), it should be possible to decide
very rapidly that the sentence, “A canary is
yellow”, is true because the concept (i.e.,
“canary”) and the property (i.e., “is yellow”)
are stored together at the same level of the
Animal
Bird Fish
Canary Ostrich Shark Salmon
has skin
can move around
eats
breathes
has wings
can fly
has feathers
has fins
can swim
has gills
can sing
is yellow
has thin,
long legs
is tall
can’t fly
can
bite
is
dangerous
is pink
is edible
swims
upriver
to lay
eggs
Figure 7.4 Collins and
Quillian’s (1969) hierarchical
network.
7 LONG-TERM MEMORY SYSTEMS 265
ostriches are less typical birds than eagles,
which in turn are less typical than robins.
What does this tell us about the structure
of semantic memory? It strongly implies that
Collins and Quillian (1969) were mistaken in
assuming that the concepts we use belong to
rigidly defined categories. Convincing evidence
that many concepts in semantic memory are
fuzzy rather than neat and tidy was reported
by McCloskey and Glucksberg (1978). They
gave 30 people tricky questions such as, “Is a
stroke a disease?” and “Is a pumpkin a fruit?”
They found that 16 said a stroke is a disease,
but 14 said it was not. A pumpkin was regarded
as a fruit by 16 participants but not as a fruit
by the remainder. More surprisingly, when
McCloskey and Glucksberg tested the same
participants a month later, 11 of them had
changed their minds about “stroke” being a
disease, and eight had altered their opinion
about “pumpkin” being a fruit!
Collins and Loftus (1975) put forward a
spreading activation theory. They argued that
There is another limitation. Consider the
following statements: “A canary is a bird” and
“A penguin is a bird”. On their theory, both
statements should take the same length of time
to verify, because they both involve moving
one level in the hierarchy. In fact, however, it
takes longer to decide that a penguin is a bird.
Why is that so? The members of most categories
vary considerably in terms of how typical or
representative they are of the category to which
they belong. For example, Rosch and Mervis
(1975) found that oranges, apples, bananas,
and peaches were rated as much more typical
fruits than olives, tomatoes, coconuts, and dates.
Rips, Shoben, and Smith (1973) found that
verification times were faster for more typical
or representative members of a category than
for relatively atypical members (the typicality
effect).
More typical members of a category possess
more of the characteristics associated with that
category than less typical ones. Rosch (1973)
produced a series of sentences containing the
word “bird”. Sample sentences were as follows:
“Birds eat worms”; “I hear a bird singing”;
“I watched a bird fly over the house”; and “The
bird was perching on the twig”. Try replacing
the word bird in each sentence in turn with
robin, eagle, ostrich, and penguin. Robin fits all
the sentences, but eagle, ostrich, and penguin
fit progressively less well. Thus, penguins and
The typicality effect determines that it will take longer to decide that a penguin is a bird than that a canary
is a bird. A penguin is an example of a relatively atypical member of the category to which it belongs, whereas
the canary – being a more representative bird – can be verified more quickly.
typicality effect: the finding that objects can
be identified faster as category members when
they are typical or representative members of
the category in question.
KEY TERM
266 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
According to spreading activation theory,
whenever a person sees, hears, or thinks about
a concept, the appropriate node in semantic
memory is activated. This activation then spreads
most strongly to other concepts closely related
semantically, and more weakly to those more
distant semantically. For example, activation
would pass strongly and rapidly from “robin”
to “bird” in the sentence, “A robin is a bird”,
because “robin” and “bird” are closely related
semantically. However, it would pass more
weakly and slowly from “penguin” to “bird”
in the sentence, “A penguin is a bird”. As a
result, the model predicts the typicality effect.
Other predictions of the spreading activation
model have been tested experimentally.
For example, Meyer and Schvaneveldt (1976)
the notion of logically organised hierarchies
was too inflexible. They assumed instead that
semantic memory is organised on the basis
of semantic relatedness or semantic distance.
Semantic relatedness can be measured by asking
people to decide how closely related pairs of
words are. Alternatively, people can list as many
members as they can of a particular category.
Those members produced most often are regarded
as most closely related to the category.
You can see part of the organisation of
semantic memory assumed by Collins and Loftus
in Figure 7.5, with the length of the links
between two concepts indicating their degree
of semantic relatedness. Thus, for example,
red is more closely related to orange than to
sunsets.
Street
Vehicle
Car
Truck
Bus
Ambulance
Fire
engine
House
Fire
Red
Orange
Yellow
Green
Apples
Cherries Pears
Violets Roses
Flowers
Sunsets
Sunrises Clouds
Figure 7.5 Example of a
spreading activation semantic
network. From Collins and
Loftus (1975). Copyright ©
1975 American Psychological
Association. Reproduced
with permission.
7 LONG-TERM MEMORY SYSTEMS 267
network model. An important reason is that
it is a much more flexible approach. However,
flexibility means that the model typically does
not make very precise predictions. This makes
it difficult to assess its overall adequacy.
Organisation of concepts
in the brain
It is often assumed (e.g., Bartlett, 1932; Bransford,
1979) that we have schemas (organised packets
of knowledge) stored in semantic memory. For
example, our schematic knowledge leads us to
expect that most kitchens will have an oven,
a refrigerator, a sink, cupboards, and so on.
What is known about the organisation of
schematic knowledge in the brain is discussed
in Chapter 10.
In this section, we focus on our semantic
knowledge of concepts and objects. How is
that knowledge organised in the brain? One
obvious possibility is that all information we
possess about any given object or concept is
stored in one location in the brain. Another
possibility is that different kinds of information
(features) about a given object are stored in
different locations in the brain. This notion is
incorporated in feature-based theories. According
to such theories, “Object concepts may be
represented in the brain as distributed networks
of activity in the areas involved in the processing
of perceptual or functional knowledge” (Canessa
et al., 2008, p. 740). As we will see, both of
these possibilities capture part of what is actually
the case.
Perceptual–functional theories
An influential feature-based approach was put
forward by Warrington and Shallice (1984) and
Farah and McClelland (1991). According to
this approach, there is an important distinction
between visual or perceptual features (e.g.,
what does the object look like?) and functional
features (e.g., what is the object used for?).
Our semantic knowledge of living things is
mostly based on perceptual information. In
contrast, our knowledge of non-living things (e.g.,
tools) mainly involves functional information.
had participants decide as rapidly as possible
whether a string of letters formed a word. In
the key condition, a given word (e.g., “butter”)
was immediately preceded by a semantically
related word (e.g., “bread”) or by an unrelated
word (e.g., “nurse”). According to the model,
activation should have spread from the first word
to the second only when they were semantically
related and this activation should have made
it easier to identify the second word. Thus,
“butter” should have been identified as a word
faster when preceded by “bread” than by “nurse”.
Indeed, there was a facilitation (or semantic
priming) effect for semantically related words.
McNamara (1992) used the same basic
approach as Meyer and Schvaneveldt (1976).
Suppose the first word was “red”. This was
sometimes followed by a word one link away
(e.g., “roses”), and sometimes by a word two
links away (e.g., “flowers”). More activation
should spread from the activated word to
words one link away than those two links
away, and so the facilitation effect should have
been greater in the former case. That is what
McNamara (1992) found.
Schacter, Alpert, Savage, Rauch, and Albert
(1996) used the Deese–Roediger–McDermott
paradigm described in Chapter 6. Participants
received word lists constructed in a particular
way. An initial word (e.g., “doctor”) was selected,
and then several words closely associated with
it (e.g., “nurse”, “sick”, “hospital”, “patient”)
were selected. All these words (excluding the
initial word) were presented for learning,
followed by a test of recognition memory. When
the initial word was presented on the recognition
test, it should theoretically have been highly
activated because it was so closely related to
all the list words. Schacter et al. compared
brain activation on the recognition test when
participants falsely recognised the initial word
and when they correctly recognised list words.
The pattern and intensity of brain activation
were very similar in both cases, indicating that
there was substantial activation of the initial
word, as predicted by the model.
The spreading activation model has generally
proved more successful than the hierarchical
268 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
from 44 patients. Of the 38 patients having a
selective impairment for knowledge of living
things, nearly all had damage to the anterior,
medial, and inferior parts of the temporal lobes.
In contrast, the six patients having a selective
impairment for knowledge of man-made objects
had damage in fronto-parietal areas extending
further back in the brain than the areas damaged
in the other group.
Support for perceptual–functional theories
has also come from neuroimaging studies. Lee,
Graham, Simons, Hodges, Owen, and Patterson
(2002) asked healthy participants to retrieve
perceptual or non-perceptual information about
living or non-living objects or concepts when
presented with their names. Processing of perceptual
information from both living and nonliving
objects was associated with activation
of left posterior temporal lobe regions. In contrast,
processing of non-perceptual information
(e.g., functional attributes) was associated with
activation of left posterior inferior temporal
lobe regions. Comparisons between living and
non-living objects indicated that the same brain
regions were activated for both types of concept.
Thus, what determined which brain areas
were activated was whether perceptual or nonperceptual
information was being processed.
Similar findings were reported by Marques,
Canessa, Siri, Catricala, and Cappa (2008). Participants
were presented with statements about
the features (e.g., form, colour, size, motion)
of living and non-living objects, and patterns
of brain activity were assessed while they decided
whether the statements were true or false. Their
findings largely agreed with those of Lee et al.
(2002): “The results...highlighted that feature
type rather than concept domain [living versus
non-living] is the main organisational factor
of the brain representation of conceptual knowledge”
(Marques et al., 2008, p. 95).
An additional assumption of the perceptual–
functional approach is that semantic memory
contains far more information about perceptual
properties of objects than of functional properties.
Farah and McClelland (1991) examined
the descriptors of living and non-living objects
given in the dictionary. Three times more of
the descriptors were classified as visual than
as functional. As predicted, the ratio of visual
to functional descriptors was 7.7:1 for living
objects but only 1.4:1 for non-living objects.
Two major predictions follow from the
perceptual–functional approach. First, brain
damage should generally impair knowledge of
living things more than non-living things. Brain
damage is likely to destroy more information
about perceptual features than functional features
because more such information is stored in the
first place. Second, neuroimaging should reveal
that different brain areas are activated when
perceptual features of an object are processed
than functional features.
We turn now to a consideration of the
relevant evidence. Some research has focused
on brain-damaged patients who have problems
with semantic memory and other research
has used neuroimaging while healthy participants
engage in tasks that involve semantic
memory.
Evidence
Many brain-damaged patients exhibit categoryspecific
deficits, meaning they have problems
with specific categories of object. For example,
Warrington and Shallice (1984) studied a patient
(JBR). He had much greater difficulty in identifying
pictures of living than of non-living
things (success rates of 6% and 90%, respectively).
This pattern is common. Martin and
Caramazza (2003) reviewed the evidence. More
than 100 patients with a category-specific deficit
for living but not for non-living things have
been studied compared to approximately 25 with
the opposite pattern. These findings are as
predicted by perceptual–functional theories.
Why do some patients show greater impairment
in recognising non-living than living
things? Gainotti (2000) reviewed the evidence
category-specific deficits: disorders caused
by brain damage in which semantic memory
is disrupted for certain semantic categories.
KEY TERM
7 LONG-TERM MEMORY SYSTEMS 269
taste, and tactile. For example, there are similarities
among fruits, vegetables, and foods
because sensory features associated with taste
are important to all three categories.
Cree and McRae (2003) identified seven
different patterns of category-specific deficits
occurring following brain damage (see Table 7.1).
They pointed out that no previous theory could
account for all these patterns. However, their
multiple-feature approach can do so. When
brain damage reduces stored knowledge for
one or more properties of objects, semantic
memory for all categories relying strongly on
those properties is impaired.
The multiple-property approach is promising
for various reasons. First, it is based on a
recognition that most concepts consist of several
properties and that these properties determine
similarities and differences among them. Second,
the approach provides a reasonable account of
several different patterns of deficit in conceptual
knowledge observed in brain-damaged patients.
Third, it is consistent with brain-imaging findings
suggesting that different object properties are
stored in different parts of the brain (e.g., Martin
& Chao, 2001).
Distributed-plus-hub theory vs.
grounded cognition
As we have seen, there is general agreement
that much of our knowledge of objects and
concepts is widely distributed in the brain. Such
knowledge is modality-specific (e.g., visual or
auditory) and relates to perception, language,
and action. This knowledge is probably stored
in brain regions overlapping with those involved
in perceiving, using language, and acting.
Does semantic memory also contain relatively
abstract amodal representations not
associated directly with any of the sensory
Multiple-property approach
The findings discussed so far are mostly consistent
with perceptual–functional theories.
However, there is increasing evidence that such
theories are oversimplified. For example, many
properties of living things (e.g., carnivore; lives
in the desert) do not seem to be sensory or
functional. In addition, the definition of functional
feature has often been very broad and
included an object’s uses as well as how it is
manipulated. Buxbaum and Saffran (2002) have
shown the importance of distinguishing between
these two kinds of knowledge. Some of the
patients they studied suffered from apraxia,
a disorder involving the inability to make
voluntary bodily movements. Apraxic patients
with frontoparietal damage had preserved
knowledge of the uses of objects but loss of
knowledge about how to manipulate objects.
In contrast, non-apraxic patients with damage
to the temporal lobe showed the opposite pattern.
Functional knowledge should probably be
divided into “what for” and “how” knowledge
(Canessa et al., 2008).
Canessa et al. (2008) reported functional
magnetic resonance imaging (fMRI; see Glossary)
findings supporting the above distinction. Healthy
participants were presented with pictures of pairs
of objects on each trial. They decided whether
the objects were used in the same context (functional
or “what for” knowledge) or involved the
same manipulation pattern (action or “how”
knowledge). Processing action knowledge led
to activation in a left frontoparietal network,
whereas processing functional knowledge activated
areas within the lateral anterior inferotemporal
cortex. The areas associated with these
two kinds of knowledge were generally consistent
with those identified by Buxbaum and
Saffran (2002) in brain-damaged patients.
Cree and McRae (2003) showed that the
distinction between perceptual and functional
properties of objects is oversimplified. They
argued that functional features should be
divided into entity behaviours (what a thing
does) and functional information (what humans
use it for). Perceptual properties should be
divided into visual (including colour), auditory,
apraxia: a neurological condition in which
patients are unable to perform voluntary bodily
movements.
KEY TERM
270 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Evidence
As predicted by theories of grounded cognition,
modality-specific information is very important
in our processing of concepts. Consider a study
by Hauk, Johnsrude, and Pulvermüller (2004).
Tongue, finger, and foot movements produced
different patterns of activation along the motor
strip. When they presented participants with
words such as “lick”, “pick”, and “kick”, these
verbs activated parts of the motor strip overlapping
with (or very close to) the corresponding
part of the motor strip. Thus, for example,
the word “lick” activated areas associated with
tongue movements.
The findings of Hauk et al. (2004) show
that the motor system is associated with the
processing of action words. However, these
findings do not necessarily mean that the motor
and premotor cortex influence the processing
of action words. More convincing evidence
was reported by Pulvermüller, Hauk, Nikulin,
and Ilmoniemi (2005). Participants performed
a lexical decision task in which they decided
whether strings of letters formed words. Different
parts of the motor system were stimulated with
transcranial magnetic stimulation (TMS; see
Glossary) while this task was performed. The
key conditions were those in which arm-related
or leg-related words were presented while TMS
was applied to parts of the left-hemisphere
modalities? There has been much recent controversy
on this issue. Barsalou (2008) argued
that the answer is, “No”. He argued in favour
of theories of grounded cognition which, “reject
the standard view that amodal symbols represent
knowledge in semantic memory...[they] focus
ontherolesofsimulationincognition....Simulation
is the re-enactment of perceptual, motor, and
introspective states acquired during experience
(p. 618).
According to the distributed-plus-hub theory
(Patterson et al., 2007; Rogers et al., 2004),
the answer is, “Yes”. There is a hub for each
concept or object in addition to distributed
modality-specific information. Each hub is a
unified conceptual representation that “supports
the interactive activation of [distributed] representations
in all modalities” (Patterson et al.,
2007, p. 977). According to Patterson et al.,
concept hubs are stored in the anterior temporal
lobes. Why do we have hubs? First, they provide
an efficient way of integrating our knowledge
of any given concept. Second, they make it easier
for us to detect semantic similarities across
concepts differing greatly in their modalityspecific
attributes. As Patterson et al. pointed
out, scallops and prawns are conceptually related
even though they have different shapes, colours,
shell structures, forms of movement, names,
and so on.
TABLE 7.1: Cree and McRae’s (2003) explanation of why brain-damaged patients show various patterns
of deficit in their knowledge of different categories. From Smith and Kosslyn (2007). Copyright © Pearson
Education, Inc. Reproduced with permission.
Deficit pattern Shared properties
1. Multiple categories consisting of living creatures Visual motion, visual parts, colour
2. Multiple categories of non-living things Function, visual parts
3. Fruits and vegetables Colour, function, taste, smell
4. Fruits and vegetables with living creatures Colour
5. Fruits and vegetables with non-living things Sound, colour
6. Inanimate foods with living things
(especially fruits and vegetables)
Function, taste, smell
7. Musical instruments with living things Function
7 LONG-TERM MEMORY SYSTEMS 271
concrete concepts or objects that we can see
and interact with. On the face of it, the
approach seems less useful when applied to
abstract concepts such as “truth”, “freedom”,
and “invention”. However, Barsalou and WiemerHastings
(2005) argued that abstract concepts
can potentially be understood within the grounded
cognition approach. Participants indicated the
characteristic properties of various abstract
concepts. Many properties referred to settings
or events associated with the concept (e.g.,
scientists working in a laboratory for “invention”),
and others referred to relevant mental
states. Thus, much of the knowledge we have
of abstract concepts is relatively concrete.
According to the distributed-plus-hub theory,
hubs or amodal conceptual representations
are stored in the anterior temporal lobes. What
would happen if someone suffered brain damage
to these lobes? Theoretically, this should
lead to impaired performance on all tasks
requiring semantic memory. Thus, performance
motor strip associated with arm or leg movements.
There was a facilitation effect: arm-related
words were processed faster when TMS was
applied to the arm site than to the leg site, and
the opposite was the case with leg-related words
(see Figure 7.6).
Evidence that perceptual information is
involved in our use of concepts was reported
by Solomon and Barsalou (2001). Participants
decided whether concepts possessed certain
properties. The key issue was whether verification
times would be speeded up when the same
property was linked to two different concepts.
There was a facilitation effect only when the
shape of the property was similar in both cases,
indicating that perceptual information influenced
task performance. For example, verifying that
“mane” is a property of “pony” was facilitated
by previously verifying “mane” for “horse” but
not by verifying “mane” for “lion”.
The grounded cognition approach is clearly
useful in understanding our knowledge of
620
600
580
560
540
520
500
480
Arm site Leg site Arm siteLeg site
TMS to left
hemisphere
TMS to right
hemisphere
Sham
situation
Responsetimes(ms)
Arm Arm
Leg words
Arm words
Leg Leg
Figure 7.6 Top: sites to
which TMS was applied.
Bottom left: response times
to make lexical (word vs.
non-word) decisions on
arm- and leg-related words
when TMS was applied to
the left language-dominant
hemisphere. Bottom middle
and right: findings from
control experiments with
TMS to the right hemisphere
and during sham stimulation.
From Pulvermüller et al.
(2005). © 2005 Federation
of European Neuroscience
Societies. Reprinted
with permission of
Wiley-Blackwell.
272 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Evaluation
Much progress has been made in understanding
the organisation of semantic memory (see also
Chapter 10). The distributed-plus-hub theory
provides a more comprehensive account of
semantic memory than previous theories. The
evidence from brain-damaged patients with
category-specific deficits indicates that different
object properties are stored in different brain
areas. In addition, patients with semantic dementia
provide evidence for the existence of concept
hubs stored in the anterior temporal lobes.
What are the limitations of distributed-plushub
theory? First, more remains to be discovered
about the information contained within concept
hubs. For example, is more information stored
in the hubs of very familiar concepts than of
less familiar ones? Second, how do we combine
or integrate concept hub information with
distributed modality-specific information? It
would seem that complex processes are probably
involved, but we do not as yet have a clear sense
of how these processes operate.
NON-DECLARATIVE
MEMORY
The essence of non-declarative memory is that
it does not involve conscious recollection but
instead reveals itself through behaviour. As
discussed earlier, repetition priming (facilitated
processing of repeated stimuli) and procedural
memory (mainly skill learning) are two of the
major types of non-declarative memory. There
are several differences between repetition priming
and procedural memory. First, priming often
occurs rapidly, whereas procedural memory
or skill learning is typically slow and gradual
would be poor regardless of the modality of
input (e.g., objects; words; sounds) and the
modality of output (e.g., object naming; object
drawing).
The above predictions have been tested
using patients with semantic dementia. Semantic
dementia involves loss of concept knowledge
even though most cognitive functions are reasonably
intact early in the disease. It always
involves degeneration of the anterior temporal
lobes. As predicted by the distributed-plus-hub
theory, patients with semantic dementia perform
very poorly on tests of semantic memory across
all semantic categories regardless of the modalities
of input and output (see Patterson et al.,
2007, for a review). Patients with semantic
dementia are unable to name objects when
relevant pictures are presented or when they
are given a description of the object (e.g., “What
do we call the African animal with black and
white stripes?”). They are also unable to identify
objects when listening to their characteristic
sounds (e.g., a phone ringing; a dog barking).
Theoretically, we would expect functional
neuroimaging studies to indicate strong activation
in the anterior temporal lobes when healthy
participants perform semantic memory tasks.
In fact, most studies have found no evidence
for such activation! Rogers et al. (2006) identified
two likely reasons. First, most studies used
fMRI, which is poor at detecting activation in
the anterior frontal lobes. Second, the semantic
memory tasks used in most fMRI studies have
not required objects to be classified with much
precision or specificity, but patients with semantic
dementia have greater problems with more precise
categories. Rogers et al. carried out a study
on healthy participants using PET rather than
fMRI. Their task involved deciding whether
an object belonged to the category specified
by a previous word. The category was specific
(e.g., BMW; labrador) or more general (e.g.,
car; dog). There was activation in the anterior
temporal lobes when the task involved specific
categories. Thus, we finally have solid evidence
of the involvement of the anterior temporal
lobes in semantic memory from a functional
neuroimaging study.
semantic dementia: a condition in which
there is widespread loss of information about
the meanings of words and concepts but
executive functioning is reasonably intact
in the early stages.
KEY TERM
7 LONG-TERM MEMORY SYSTEMS 273
non-words presented in a mirror were read as
fast as possible. Activity in different areas of
the brain was assessed by fMRI. The findings
were reasonably clear-cut:
[Skill] learning...was associated with
increased activation in left inferior temporal,
striatal, left inferior prefrontal and right
cerebellar regions and with decreased
activity in the left hippocampus and left
cerebellum. Short-term repetition priming
was associated with reduced activity in
many of the regions active during mirror
reading and...long-term repetition priming
resulted in a virtual elimination of
activity in those regions. (p. 67)
The finding that very similar areas were
involved in skill learning and priming is consistent
with the hypothesis that they involve
the same underlying memory system. However,
evidence less supportive of that hypothesis is
discussed later.
Repetition priming
We can draw a distinction between perceptual
priming and conceptual priming. Perceptual
priming occurs when repeated presentation of
a stimulus leads to facilitated processing of its
perceptual features. For example, it is easier to
identify a word presented in a degraded fashion
if it has recently been encountered. In contrast,
conceptual priming occurs when repeated presentation
of a stimulus leads to facilitated processing
of its meaning. For example, people can
decide faster whether an object is living or
nonliving if they have seen it recently.
(Knowlton & Foerde, 2008). Second, there is
stimulus specificity. Priming is tied to specific
stimuli whereas skill learning typically generalises
to numerous stimuli. For example, it would
not be much use if you learned how to hit
backhands at tennis very well, but could only
do so provided that the ball came towards you
from a given direction at a given speed! Third,
there is increasing evidence that different brain
areas are involved in repetition priming and
skill learning (Knowlton & Foerde, 2008).
If repetition priming and skill learning
involve different memory systems, then there
is no particular reason why individuals who
are good at skill learning should be good at
priming. There is often practically no correlation
between performance on these two types of
task. Schwartz and Hashtroudi (1991) used
a word-identification task to assess priming
and an inverted-text reading task to assess skill
learning. There was no correlation between
priming and skill learning. However, the interpretation
of such findings is open to dispute.
Gupta and Cohen (2002) developed a computational
model based on the assumption that
skill learning and priming depend on a single
mechanism. This model accounted for zero
correlations between skill learning and priming.
It is probable that priming and skill learning
involve separate memory systems. However,
most of the evidence is not clear-cut because
the tasks assessing skill learning and repetition
priming have been very different. This led
Poldrack, Selco, Field, and Cohen (1999) to
compare skill learning and priming within a
single task. Participants entered five-digit
numbers as rapidly as possible into a computer
keypad. Priming was assessed by performance
on repeated digit strings, whereas skill learning
was assessed by performance on non-repeated
strings. Skill learning and the increase in speed
with repetition priming were both well described
by a power function, leading Poldrack et al.
to conclude that they both involve the same
learning mechanism.
Poldrack and Gabrieli (2001) studied skill
learning and repetition priming using a mirrorreading
task in which words and pronounceable
perceptual priming: a form of repetition
priming in which repeated presentation of a
stimulus facilitates perceptual processing of it.
conceptual priming: a form of repetition
priming in which there is facilitated processing
of stimulus meaning.
KEY TERMS
274 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
patients were presented with a list of words
followed by a priming task. This task was perceptual
identification, and involved presenting
the words at the minimal exposure time needed
to identify them. The performance of the amnesic
patients resembled that of control participants,
with identification times being faster for the
primed list words than for the unprimed ones.
Thus, the amnesic patients showed as great a
perceptual priming effect as the controls. Cermak
et al. also used a conventional test of recognition
memory (involving episodic memory) for the
list words. The amnesic patients did significantly
worse than the controls on this task.
Graf, Squire, and Mandler (1984) studied
a different perceptual priming effect. Word lists
were presented, with the participants deciding
how much they liked each word. The lists were
followed by one of four memory tests. Three
tests involved declarative memory (free recall,
recognition memory, and cued recall), but the
fourth test (word completion) involved priming.
On this last test, participants were given threeletter
word fragments (e.g., STR ____) and
simply wrote down the first word they thought
of starting with those letters (e.g., STRAP;
STRIP). Priming was assessed by the extent
to which the word completion corresponded
to words from the list previously presented.
Amnesic patients did much worse than controls
on all the declarative memory tests, but the
groups did not differ on the word-completion
test.
Levy, Stark, and Squire (2004) studied
conceptual priming and recognition memory
(involving declarative memory) in amnesic
patients with large lesions in the medial
temporal lobe, amnesic patients with lesions
limited to the hippocampus, and healthy controls.
The conceptual priming task involved
deciding whether words previously studied or
not studied belonged to given categories. The
findings were striking. All three groups showed
very similar amounts of conceptual priming.
However, both amnesic groups performed poorly
on recognition memory (see Figure 7.7). Indeed,
the amnesic patients with large lesions showed
no evidence of any declarative memory at all.
Much evidence supports the distinction
between perceptual and conceptual priming.
Keane, Gabrieli, Mapstone, Johnson, and Corkin
(1995) studied perceptual and conceptual priming
in LH, a patient with bilateral brain damage
within the occipital lobes. LH had an absence
of perceptual priming but intact conceptual
priming. In contrast, patients with Alzheimer’s
disease have the opposite pattern of intact perceptual
priming but impaired conceptual priming
(see Keane et al., 1995, for a review). According
to Keane et al., the impaired conceptual priming
shown by Alzheimer’s patients is due to
damage within the temporal and parietal lobes.
The findings suggest the existence of a double
dissociation (see Glossary), which provides
reasonable support that different processes
underlie the two types of priming.
Evidence
If repetition priming involves non-declarative
memory, then amnesic patients should show
intact repetition priming. This prediction has
been supported many times. Cermak, Talbot,
Chandler, and Wolbarst (1985) compared the
performance of amnesic patients and nonamnesic
alcoholics on perceptual priming. The
Perceptual priming occurs when repeated
presentation of a stimulus leads to facilitated
processing of its perceptual features. For example,
it would be easier to identify words that had
been eroded and had faded in the sand, if they
had previously been seen when freshly etched.
7 LONG-TERM MEMORY SYSTEMS 275
words spoken in the same voice. After that, they
tried to identify the same words passed through
an auditory filter; the words were spoken in
the same voice or an unfamiliar voice. Amnesic
patients and healthy controls both showed
perceptual priming, with word-identification
performance being better when the words were
spoken in the same voice (see Figure 7.8a).
The findings discussed so far seem neat and
tidy. However, complications arose in research
by Schacter, Church, and Bolton (1995). Their
study resembled that of Schacter and Church
(1995) in that perceptual priming based on
auditory word identification was investigated.
However, it differed in that the words were
The notion that priming depends on memory
systems different from those involved in
declarative memory would be strengthened if
we could find patients having intact declarative
memory but impaired priming. This would be
a double dissociation, and was achieved by
Gabrieli, Fleischman, Keane, Reminger, and
Morell (1995). They studied a patient, MS, who
had right occipital lobe lesion. MS had normal
levels of performance on the declarative memory
tests of recognition and cued recall but impaired
performance on perceptual priming.
Further evidence that amnesics have intact
perceptual priming was reported by Schacter
and Church (1995). Participants initially heard
300
200
100
0
Msec
CON MTL H
(a) Priming
100
80
60
40
Percentagecorrect
CON MTL H
(c) Recognition
15
10
5
0
Percent
CON MTL H
(b) Priming/baseline
Figure 7.7 Performance of healthy controls (CON), patients with large medial temporal lobe lesions (MTL),
and patients with hippocampal damage only (H) on: (a) priming in terms of reaction times; (b) priming in terms
of percentage priming effect; and (c) recognition performance. From Levy et al. (2004). Reprinted with
permission of Wiley-Blackwell.
0.6
0.5
0.4
0.3
0.2
Same
voice
(b)
Meancorrectidentification(proportion)
Controls
Amnesic patients
Re-paired
voice
0.8
0.7
0.6
0.5
0.4
0.3
0.2
Unfamiliar
voice
Same
voice
(a)
Meancorrectidentification(proportion)
Figure 7.8 Auditory word identification for previously presented words in amnesics and controls. (a) All words
originally presented in the same voice; data from Schacter and Church (1995). (b) Words originally presented in
six different voices; data from Schacter et al. (1995).
276 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
frontal gyrus in conceptual priming by delivering
transcranial magnetic stimulation to that area.
The subsequent classification of objects that
had been accompanied by TMS showed an
absence of both conceptual and neural priming.
These findings suggest that the left inferior
temporal cortex plays a causal role in producing
conceptual priming.
Evaluation
There are important similarities and differences
between perceptual and conceptual priming.
They are similar in that most amnesic patients
typically show essentially intact perceptual and
conceptual priming, suggesting that both types
of priming involve non-declarative memory.
However, the finding of a double dissociation
in which some patients are much better at
perceptual than at conceptual priming, whereas
others show the opposite pattern, suggests there
are some important differences between them.
The consistent finding that repetition priming
is associated with reduced brain activation
suggests that people become more efficient at
processing repeated stimuli. Recent research
has supported the hypothesis that there is a
causal link between patterns of brain activation
and priming performance.
Future research needs to establish more
clearly that reduced brain activation during
repetition priming is causally related to enhanced
priming. There is also a need to identify more
precisely the different processes involved in
perceptual and conceptual priming.
Procedural memory or
skill learning
What exactly is skill learning? According to
Poldrack et al. (1999, p. 208), “Skill learning
refers to the gradual improvement of performance
with practice that generalises to a range
of stimuli within a domain of processing.”
Motor skills are important in everyday life.
For example, they are needed in word processing,
writing, and playing a musical instrument.
Foerde and Poldrack (2009) identified
numerous types of skill learning or procedural
initially presented in six different voices. On
the word-identification test, half the words were
presented in the same voice and half were spoken
by one of the other voices (re-paired condition).
The healthy controls showed more priming for
words presented in the same voice, but the
amnesic patients did not (see Figure 7.8b).
How can we explain the above findings?
In both the same voice and re-paired voice
conditions, the participants were exposed to
words and voices they had heard before. The
only advantage in the same voice condition was
that the pairing of word and voice was the same
as before. However, only those participants
who had linked or associated words and voices
at the original presentation would benefit from
that fact. The implication is that amnesics
are poor at binding together different kinds of
information even on priming tasks apparently
involving non-declarative memory (see discussion
later in the chapter).
What processes are involved in priming?
One popular view is based on perceptual fluency:
repeated presentation of a stimulus means it
can be processed more efficiently using fewer
resources. It follows from this view that priming
should be associated with reduced levels of
brain activity (known as neural priming). There
is considerable evidence for this prediction
(e.g., Poldrack & Gabrieli, 2001). The precise
brain regions showing reduced activation vary
somewhat depending on the task and whether
perceptual or conceptual priming is being studied.
Early visual areas in the occipital lobe often
show reduced activity with perceptual priming,
whereas the inferior frontal gyrus and left inferior
temporal cortex show reduced activity
with conceptual priming (see Schacter et al.,
2007, for a review).
The finding that repetition of a stimulus
causes priming and reduced brain activity does
not show there is a causal link between patterns
of brain activation and priming. More direct
evidence was reported by Wig, Grafton, Demos,
and Kelley (2005). They studied conceptual
priming using a task in which participants
classified objects as living or nonliving. Wig
et al. tested the involvement of the left inferior
7 LONG-TERM MEMORY SYSTEMS 277
declarative memory. Thus, the involvement of
procedural and declarative memory on the
probabilistic classification task seemed to depend
on the precise conditions under which the task
was performed.
Evidence
Amnesics often have normal (or nearly normal)
rates of skill learning across numerous tasks.
Spiers et al. (2001), in a review discussed earlier,
considered the memory performance of numerous
amnesic patients. They concluded as follows:
“None of the cases was reported to...be impaired
on tasks which involved learning skills or habits,
priming, simple classical conditioning and simple
category learning” (p. 359).
Corkin (1968) reported that the amnesic
patient HM (see p. 252) was able to learn mirror
drawing, in which the pen used in drawing a
figure is observed in a mirror rather than directly.
He also showed learning on the pursuit rotor,
which involves manual tracking of a moving
target. HM’s rate of learning was slower than that
of healthy individuals on the pursuit rotor. In
contrast, Cermak, Lewis, Butters, and Goodglass
(1973) found that amnesic patients learned the
pursuit rotor as rapidly as healthy participants.
However, the amnesic patients were slower than
healthy individuals at learning a finger maze.
Tranel, Damasio, Damasio, and Brandt
(1994) found in a study on 28 amnesic patients
that all showed comparable learning on the
pursuit rotor to healthy controls. Of particular
note was a patient, Boswell, who had unusually
extensive brain damage to areas (e.g., medial
and lateral temporal lobes) strongly associated
with declarative memory. In spite of this, his
learning on the pursuit rotor and retention over
a two-year period were both at the same level
as healthy controls.
The typical form of the serial reaction time
task involves presenting visual targets in one of
four horizontal locations, with the participants
pressing the closest key as rapidly as possible
(see Chapter 6). The sequence of targets is
sometimes repeated over 10 or 12 trials, and skill
learning is shown by improved performance
on these repeated sequences. Nissen, Willingham,
memory, including the following: motor skill
learning; sequence learning, mirror tracing;
perceptual skill learning; mirror reading; probabilistic
classification learning; and artificial
grammar learning. Some of these forms of skill
learning are discussed at length in Chapter 6.
Here, we will address the issue of whether
the above tasks involve non-declarative or procedural
memory, and thus involve different
memory systems from those underlying episodic
and semantic memory. This issue has been
addressed in various ways. However, we will
mostly consider research on skill learning in
amnesic patients. The rationale for doing this
is simple: if amnesic patients have essentially
intact skill learning but severely impaired
declarative memory that would provide evidence
that different memory systems are involved.
We will shortly turn to the relevant evidence.
Before doing so, however, we need to consider
an important issue. It is easy to imagine that
some tasks involve only non-declarative or
procedural memory, whereas others involve
declarative memory. In fact, matters are rarely
that simple (see Chapter 6). For example,
consider the probabilistic classification task.
Participants predict whether the weather will
be sunny or rainy on the basis of various cues.
Reber, Knowlton, and Squire (1996) found that
amnesics learned this task as rapidly as healthy
controls, suggesting that the task involves
procedural memory.
Foerde, Knowlton, and Poldrack (2006)
obtained evidence suggesting that learning on
the probabilistic classification task can depend
on either procedural or declarative memory.
Participants performed the task on its own or
with a demanding secondary task. Performance
was similar in the two conditions. However,
important differences emerged between the
conditions when the fMRI data were considered.
Task performance in the dual-task condition
correlated with activity in the striatum (part of
the basal ganglia), a part of the brain associated
with procedural learning and memory. In
contrast, task performance in the single-task
performance correlated with activity in the
medial temporal lobe, an area associated with
278 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
in spite of very poor declarative memory. That
provides reasonable evidence that there are
major differences between the two forms of
memory. Shortly, we will consider evidence
indicating that the brain areas associated with
procedural memory differ from those associated
with declarative memory. However, we must not
think of declarative and procedural memory
as being entirely separate. Brown and Robertson
(2007) gave participants a procedural learning
task (the serial reaction time task) and a declarative
learning task (free recall of a word list).
Procedural memory was disrupted when declarative
learning occurred during the retention
interval. In a second experiment, declarative
memory was disrupted when procedural learning
occurred during the retention interval. Thus,
there can be interactions between the two
memory systems.
BEYOND DECLARATIVE
AND NON-DECLARATIVE
MEMORY:AMNESIA
Most memory researchers have argued that
there is a very important distinction between
declarative/explicit memory and non-declarative/
implicit memory. As we have seen, this distinction
has proved very useful in accounting for most
of the findings (especially those from amnesic
patients). However, there are good grounds for
arguing that we need to move beyond that
distinction. We will focus our discussion on
amnesia, but research on healthy individuals also
suggests that the distinction between declarative
and non-declarative memory is limited (see Reder,
Park, & Kieffaber, 2009, for a review).
According to the traditional viewpoint,
amnesic patients should have intact performance
on declarative memory tasks and impaired
performance on non-declarative tasks. There
is an alternative viewpoint that has attracted
increasing interest (e.g., Reder et al., 2009; Ryan,
Althoff, Whitlow, & Cohen, 2000; Schacter
et al., 1995). According to Reder et al. (2009,
p. 24), “The critical feature that distinguishes
and Hartman (1989) found that amnesic patients
and healthy controls showed comparable performance
on the serial reaction time task during
learning and also on a second test one week
later. Vandenberghe et al. (2006) obtained more
complex findings. They had a deterministic
condition in which there was a repeating sequence
and a probabilistic condition in which there was
a repeating sequence but with some deviations.
Amnesic patients failed to show skill learning
in the probabilistic condition, but exhibited
some implicit learning in the deterministic
condition. Thus, amnesic patients do not always
show reasonable levels of skill learning.
Mirror tracing involves tracing a figure
with a stylus, with the figure to be traced being
seen reflected in a mirror. Performance on this
task improves with practice in healthy participants,
and the same is true of amnesic patients
(e.g., Milner, 1962). The rate of learning is
often similar in both groups.
In mirror reading we can distinguish between
general improvement in speed of reading produced
by practice and more specific improvement
produced by re-reading the same groups of words
or sentences. Cohen and Squire (1980) reported
general and specific improvement in reading
mirror-reversed script in amnesics, and there
was evidence of improvement even after a delay
of three months. Martone, Butters, Payne, Becker,
and Sax (1984) also obtained evidence of general
and specific improvement in amnesics.
Cavaco, Anderson, Allen, Castro-Caldas,
and Damasio (2004) pointed out that most
tasks used to assess skill learning in amnesics
require learning far removed from that occurring
in everyday life. Accordingly, Cavaco et al. used
five skill-learning tasks requiring skills similar
to those needed in the real world. For example,
there was a weaving task and a control stick task
requiring movements similar to those involved
in operating machinery. Amnesic patients showed
comparable rates of learning to those of healthy
individuals on all five tasks, in spite of having
significantly impaired declarative memory for the
tasks assessed by recall and recognition tests.
In sum, amnesic patients show reasonably
good skill or procedural learning and memory
7 LONG-TERM MEMORY SYSTEMS 279
more with practice on the old displays than on
the new ones. This involved implicit learning,
because they had no ability to discriminate old
displays from new ones on a recognition test. The
amnesic patients showed general improvement
with practice, and thus some implicit learning.
However, there was no difference between their
performance on new and old displays. This failure
of implicit learning probably occurred because
the amnesic patients could not bind the arrangement
of the distractors to the location of the
target in old displays.
There have been some failures to replicate
the above findings (see Reder et al., 2009, for
a review), perhaps because amnesic patients
differ so much in their precise brain damage and
memory impairments. Park, Quinlan, Thornton,
and Reder (2004) argued that a useful approach
is to use drugs that mimic the effects of amnesia.
They administered midazolam, a benzodiazepine
that impairs performance on explicit memory
tasks but not implicit tasks (e.g., repetition
priming). They carried out a study very similar
to that of Chun and Phelps (1999), and obtained
similar findings. Their key result was that healthy
individuals given midazolam failed to perform
better on old displays than new ones, in contrast
to individuals given a placebo (saline) (see
Figure 7.9). Thus, midazolam-induced amnesia
impairs implicit learning because it disrupts
binding with old displays.
A study by Huppert and Piercy (1976) on
declarative memory supports the binding hypothesis.
They presented large numbers of pictures
on day 1 and on day 2. Some of those presented
on day 2 had been presented on day 1 and
others had not. Ten minutes after the day-2
presentation, there was a recognition-memory
test, on which participants decided which pictures
had been presented on day 2. Successful performance
on this test required binding of picture
and temporal context at the time of learning.
Healthy controls performed much better than
amnesic patients in correctly identifying day-2
pictures and rejecting pictures presented only
on day 1 (see Figure 7.10a).Thus, amnesic patients
were at a great disadvantage when binding
was necessary for memory.
tasks that are impaired from those that are
spared under amnesia hinges on whether the
task requires the formation of an association
(or binding) between the two concepts.” We
will briefly consider research relevant to adjudicating
between these two viewpoints. Before
we do so, note that the binding-of-item-andcontext
model (Diana et al., 2007; discussed
earlier in the chapter) identifies the hippocampus
as of central importance in the binding process.
The relevance of that model here is that amnesic
patients typically have extensive damage to the
hippocampus.
Evidence
Earlier in the chapter we discussed a study by
Schacter et al. (1995) on perceptual priming.
Amnesic patients and healthy controls identified
words passed through an auditory filter
having previously heard them spoken by the
same voice or one out of five different voices.
The measure of perceptual priming was the
extent to which participants were better at
identifying words spoken in the same voice
than those spoken in a different voice. Since six
different voices were used altogether, successful
perceptual priming required binding or associating
the voices with the words when the
words were presented initially. In spite of the
fact that Schacter et al. used a non-declarative
memory task, amnesic patients showed no
better performance for words presented in the
same voice than in a different voice (see Figure
7.8b). This finding is inconsistent with the
traditional viewpoint but is as predicted by the
binding hypothesis.
More evidence that amnesic patients sometimes
have deficient implicit memory was reported
by Chun and Phelps (1999). Amnesic patients
and healthy controls carried out a visual search
task in which the target was a rotated T and the
distractors were rotated Ls. Half the displays
were new and the remainder were old or repeated.
There were two main findings with the healthy
controls. First, their performance improved
progressively throughout the experiment (skill
learning). Second, they improved significantly
280 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
declarative memory tasks successfully provided
that binding is not required.
Evaluation
Since declarative memory tasks generally require
the formation of associations and non-declarative
memory tasks do not, it is often hard to decide
which viewpoint is preferable. However, there
Huppert and Piercy (1976) also used
a familiarity-based recognition memory test.
Participants decided whether they had ever
seen the pictures before. Here, no prior binding
of picture and temporal context was necessary.
On this test, the amnesic patients and healthy
controls performed the task extremely well
(see Figure 7.10b). Thus, as predicted by the
binding hypothesis, amnesic patients can perform
100
80
60
40
20
0
–20
–40
–60
–80
1 2 3 4 5 6
Epoch
Contextualcuing(ms)
Saline
Midazolam
Figure 7.9 The difference
between visual search
performance with old and
new displays (i.e., contextual
cueing effect) as a function
of condition (Midazolam vs.
placebo/saline) and stage of
practice (epochs). From Park
et al. (2004), Copyright ©
2004 National Academy of
Sciences, USA. Reprinted
with permission.
100
90
80
70
60
50
40
30
20
10
0
Recognitionpercentage
Day 2 recognition
Day 1 only
pictures
(i.e. incorrect
recognition)
Day 2 only
pictures
Normal
controls
Korsakoff
patients
(a)
Ever seen recognition
Pictures not
seen before
(i.e. lack of
recognition
= correct)
Pictures
seen before
(b)
100
90
80
70
60
50
40
30
20
10
0
Recognitionpercentage
Normal
controls
Korsakoff
patients
Figure 7.10 Recognition memory for pictures in Korsakoff patients and normal controls. Data from Huppert
and Piercy (1976).
7 LONG-TERM MEMORY SYSTEMS 281
different brain regions contribute to long-term
memory, with an emphasis on the major brain
areas associated with each memory system. As
we will see, each memory system is associated
with different brain areas. This strengthens the
argument that the various memory systems are
indeed somewhat separate. In what follows,
we will discuss some of the evidence. The role
of the anterior temporal lobes in semantic
memory (e.g., Patterson et al., 2007), early visual
areas in the occipital lobe in perceptual priming
(Schacter et al., 2007), and left inferior temporal
cortex in conceptual priming (e.g., Wig et al.,
2005) were discussed earlier in the chapter.
Medial temporal lobe and medial
diencephalon
The medial temporal lobe including the hippocampal
formation is of crucial importance
in anterograde amnesia and in declarative
memory generally. However, we have a problem
because chronic alcoholics who develop
Korsakoff’s syndrome have brain damage to
the diencephalon including the mamillary bodies
and various thalamic nuclei (see Figure 7.11).
Aggleton (2008) argued persuasively that temporal
lobe amnesia and diencephalic amnesia
both reflect damage to the same integrated
brain system involving the temporal lobes and
the medial diencephalon. Aggleton pointed out
is increasing support for the binding hypothesis.
More specifically, we now have studies showing
that amnesic patients sometimes fail to show
non-declarative/implicit memory when binding
of information (e.g., stimulus + context) is
required (e.g., Chun & Phelps, 1999; Schacter
et al., 1995). In addition, amnesic patients
sometimes show essentially intact declarative/
explicit memory when binding of information
is not required (e.g., Huppert & Piercy, 1976).
What is needed for the future? First, we
need more research in which the predictions
based on the traditional viewpoint differ from
those based on the binding hypothesis. Second,
we should look for tasks that differ more clearly
in their requirements for binding than most of
those used hitherto. Third, it is important to
specify more precisely what is involved in the
binding process.
LONG-TERM MEMORY AND
THE BRAIN
Our understanding of long-term memory has
been greatly enhanced by functional imaging
studies and research on brain-damaged patients.
It is clear that encoding and retrieval in longterm
memory involve several processes and are
more complex than was previously thought.
In this section, we will briefly consider how
Body of
fornix
Crus of
fornix
Corpus
callosum
Columns
of fornix
Precommissural fornix
Thalamus
Postcommisural
fornix
Entorhinal
Hippocampus
subiculum
Midbrain
nuclei (e.g. LC)
PREFRONTAL
CORTEX
N.
ACC
DIAGONAL
BAND HYPOTH
ATN LD
SUM
MB
ACSEPTUM
RE
Figure 7.11 The main
interconnected brain areas
involved in amnesia:AC =
anterior commissure;ATN =
anterior thalamic nuclei;
HYPOTH = hypothalamus;
LC = locus coeruleus; LD =
thalamic nucleus lateralis
dorsalis; MB = mammillary
bodies; RE = nucleus
reuniens; SUM =
supramammillary nucleus.
From Aggleton (2008).
282 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
striatum. Parkinson’s disease is a progressive
disorder characterised by tremor of the limbs,
muscle rigidity, and mask-like facial expression.
Siegert, Taylor, Weatherall, and Abernethy (2006)
reported a meta-analysis of learning on the
serial reaction time task (discussed above) by
patients with Parkinson’s disease (see Chapter 6).
Skill learning by Parkinson’s patients was consistently
slower than that by healthy controls.
Strong evidence that the basal ganglia are
important in skill learning was reported by
Brown, Jahanshahi, Limousin-Dowsey, Thomas,
Quinn, and Rothwell (2003). They studied
patients with Parkinson’s disease who had had
posteroventral pallidotomy, a surgical form of
treatment that disrupts the output of the basal
ganglia to the frontal cortex. These patients
showed no implicit learning at all on the serial
reaction time task.
Not all the evidence indicates that Parkinson’s
patients show deficient procedural learning and
memory. Osman, Wilkinson, Beigi, Castaneda,
and Jahanshahi (2008) reviewed several studies
in which Parkinson’s patients performed well
on procedural learning tasks. In their own
experiment, participants had to learn about
and control a complex system (e.g., water-tank
system). Patients with Parkinson’s disease showed
the same level of procedural learning as healthy
controls on this task, which suggests that the
striatum is not needed for all forms of procedural
learning and memory.
Neuroimaging studies have produced somewhat
variable findings (see Kelly & Garavan,
2005, for a review). However, practice in skill
learning is often associated with decreased
activation in the prefrontal cortex but increased
activation in the basal ganglia. It is likely that
the decreased activation in the prefrontal cortex
occurs because attentional and control processes
that the anterior thalamic nuclei and the mammillary
bodies differ from the rest of the medial
diencephalon in that they both receive direct
inputs from the hippocampal formation via the
fornix (see Figure 7.11). Thus, these areas are
likely to be of major importance within the
hypothesised integrated system. Aggleton and
Brown (1999) proposed that an “extended hippocampal
system” consisting of the hippocampus,
fornix, mammillary bodies, and the anterior
thalamic nuclei is crucial for episodic memory.
There is much support for the notion of an
extended hippocampal system. Harding, Halliday,
Caine, and Kril (2000) studied the brains of
alcoholics with Korsakoff’s syndrome and those
of alcoholics without amnesia. The only consistent
difference between the two groups was that the
Korsakoff patients had degeneration of the anterior
thalamic nuclei. There is also evidence for the
importance of the fornix. Patients with benign
brain tumours who suffer atrophy of the fornix
as a consequence consistently exhibit clear signs
of anterograde amnesia (Gilboa et al., 2006).
We have focused on anterograde amnesia
in this section. However, the hippocampal formation
and medial temporal lobe are also very
important in retrograde amnesia (Moscovitch
et al., 2006). In addition, the hippocampus (and
the prefrontal cortex) are of central importance in
autobiographical memory (Cabeza & St. Jacques,
2007; see Chapter 8).
Striatum and cerebellum
Which brain areas are involved in skill learning
or procedural memory? Different types of skill
learning involve different brain areas depending
on characteristics of the task (e.g., auditory
versus visual input). However, two brain areas
are most closely associated with procedural
memory: the striatum (part of the basal ganglia)
in particular but also the cerebellum. The evidence
implicating those brain areas comes from
studies on brain-damaged patients and from
neuroimaging research.
Much research has made use of braindamaged
patients suffering from Parkinson’s
disease, which is associated with damage to the
Parkinson’s disease: it is a progressive
disorder involving damage to the basal ganglia;
the symptoms include rigidity of the muscles,
limb tremor, and mask-like facial expression.
KEY TERM
7 LONG-TERM MEMORY SYSTEMS 283
are important early in learning but become less
so with extensive practice. Debaere et al. (2004)
found, during acquisition of a skill requiring coordination
of hand movements, that there were
decreases in activation within the right dorsolateral
prefrontal cortex, the right premotor cortex,
and the bilateral superior parietal cortex. At the
same time, there were increases in activation
within the cerebellum and basal ganglia.
In sum, the striatum (and to a lesser extent
the cerebellum) are important in procedural
learning and memory. However, we must avoid
oversimplifying a complex reality. The neuroimaging
findings indicate clearly that several
other areas (e.g., the prefrontal cortex; the
posterior parietal cortex) are also involved.
Prefrontal cortex
As discussed in Chapter 5, the prefrontal cortex
is extremely important in most (or all) executive
processes involving attentional control. As we
have seen in this chapter, it is also of significance
in long-term memory. Two relatively
small regions on the lateral or outer surface of
the frontal lobes are of special importance: the
dorsolateral prefrontal cortex (roughly BA9 and
B46) and the ventrolateral prefrontal cortex
(roughly BA45 and BA47) (see Figure 1.4).
Dorsolateral prefrontal cortex
What is the role of dorsolateral prefrontal cortex
in declarative memory? One idea is that this
area is involved in relational encoding (forming
links between items or between an item and its
context). Murray and Ranganath (2007) carried
out a study in which unrelated word pairs were
presented. In one condition, the task involved
a comparison between the two words (relational
encoding) and in the other it did not (itemspecific
encoding). Activation of the dorsolateral
prefrontal cortex was greater during relational
than item-specific encoding. More importantly,
the amount of dorsolateral activity at encoding
predicted successful performance on a recognition
test of relational memory.
Another possible role of dorsolateral prefrontal
cortex in memory is to evaluate the
relevance of retrieved information to current task
requirements (known as post-retrieval monitoring).
The more information that is retrieved,
the more likely the individual will engage in
monitoring. Achim and Lepage (2005) manipulated
the amount of information likely to be
retrieved in two recognition-memory tests. As
predicted, activity within the dorsolateral prefrontal
cortex was greater when there was more
demand for post-retrieval monitoring.
In sum, dorsolateral prefrontal cortex plays
a role at encoding and at retrieval. First, it is
involved in relational encoding at the time of
learning. Second, it is involved in post-retrieval
monitoring at the time of retrieval. In general
terms, dorsolateral prefrontal cortex is often
activated when encoding and/or retrieval is
relatively complex.
Ventrolateral prefrontal cortex
Badre and Wagner (2007) discussed a twoprocess
account of the involvement of the
ventrolateral prefrontal cortex in declarative
memory. There is a controlled retrieval process
used to activate goal-relevant knowledge. There
is also a post-retrieval selection process that
deals with competition between memory representations
active at the same time.
Evidence that both of the above processes
involve the ventrolateral prefrontal cortex was
reported by Badre, Poldrack, Pare-Blagoev,
Insler, and Wagner (2005). A cue word and
two or four target words were presented on
each trial, and the task was to decide which
target word was semantically related to the cue
word. It was assumed that the controlled
retrieval process would be involved when the
target word was only weakly associated with
the cue (e.g., cue = candle; target word = halo).
It was also assumed that the post-retrieval
selection process would be needed when one of
the incorrect target words was non-semantically
associated with the cue word (e.g., cue = ivy;
incorrect target word = league). As predicted,
there was increased activation within the
ventrolateral prefrontal cortex when the task
required the use of controlled retrieval or postretrieval
selection.
284 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
lives. The memories recalled were less vivid
and contained less detail than those of healthy
controls. However, the same patients performed
normally when they were probed for specific
details of their memories.
Cabeza (2008) explained this and other
findings in his dual attentional processes hypothesis.
According to this hypothesis, ventral
parietal cortex is associated with bottom-up
attentional processes captured by the retrieval
output. These attentional processes were damaged
in the patients studied by Berryhill et al.
(2007). In contrast, dorsal parietal cortex is
associated with top-down attentional processes
influenced by retrieval goals. The hypothesis
is supported by two findings (see Cabeza, 2008,
for a review):
There is greater ventral parietal activation(1)
when memory performance is high due
to greater capture of bottom-up attention
by relevant stimuli.
There is greater dorsal parietal activation(2)
when memory performance is low due to
greater demands on top-down attention.
Evaluation
Considerable progress has been made in understanding
the involvement of different brain areas
in the major memory systems. The findings
Kuhl, Kahn, Dudukovic, and Wagner (2008)
studied the post-retrieval selection process.
There was activation of the right ventrolateral
prefrontal cortex and the anterior cingulate
when memories that had previously been selected
against were successfully retrieved. It was
assumed that an effective post-retrieval selection
process was needed to permit previously selectedagainst
memories to be retrieved.
Parietal lobes
What is the involvement of the parietal lobes
in long-term memory? Simons et al. (2008)
carried out a meta-analysis of functional neuroimaging
studies on episodic memory in which
brain activation was assessed during successful
recollection of the context in which events had
occurred. Lateral and medial areas within the
parietal lobes were more consistently activated
than any other areas in the entire brain (see
Figure 7.12).
The picture seems to be very different when
we consider patients with damage to the parietal
lobes. For the most part, these patients do not
seem to have severe episodic memory deficits
(see Cabeza, 2008, for a review). However,
some deficits have been found in such patients.
In one study (Berryhill, Phuong, Picasso, Cabeza,
& Olson, 2007), patients with ventral parietal
damage freely recalled events from their own
100
90
80
70
60
50
40
30
20
10
0
PercentageoffMRIstudiesshowingactivation
APFC VLPFC DLPFC Thalamus MTL Lateral
parietal
Medial
parietal
Figure 7.12 Percentages of
fMRI studies of episodic
memory showing activation
in various brain regions.
APFC = anterior prefrontal
cortex;VLPFC = ventrolateral
prefrontal cortex; DLPFC =
dorsolateral prefrontal cortex;
MTL = medial temporal lobe.
Reprinted from Simons et al.
(2008), Copyright © 2008,
with permission from
Elsevier.
7 LONG-TERM MEMORY SYSTEMS 285
clear. A brain area might be important because
it is needed for initial encoding, for subsequent
storage of information, for control of memoryrelevant
processes, or for retrieval of stored
information. Finding that a given brain area is
activated during a particular memory task does
not immediately indicate why it is activated.
Third, a major task for the future is to
understand how different brain areas interact
and combine during learning and memory.
Learning and memory undoubtedly depend upon
networks consisting of several brain regions,
but as yet we know relatively little about the
structure or functioning of such networks.
from cognitive neuroscience are generally consistent
with those from cognitive psychology.
As a result, we have an increasingly clear overall
picture of how memory works.
What are the limitations of research in this
area? First, the findings from brain-damaged
patients and from functional neuroimaging
sometimes seem inconsistent. Thus, for example,
the importance of the parietal cortex in human
memory seems greater in neuroimaging studies
than in studies on brain-damaged patients.
Second, even when we have established that
a given brain area is important with respect to
some memory system, its role is not always very
Introduction•
There are several long-term memory systems. However, the crucial distinction is between
declarative and non-declarative memory. Strong evidence for that distinction comes from
amnesic patients having severely impaired declarative memory but almost intact nondeclarative
memory and from functional neuroimaging. Declarative memory can be divided
into episodic and semantic memory. Non-declarative memory can be divided into repetition
priming and procedural memory or skill learning.
Episodic vs. semantic memory•
Virtually all amnesic patients have severe problems with forming new episodic memories
but many have only modest problems in forming new semantic memories. Some amnesic
patients have retrograde amnesia mainly for episodic memory, whereas others have
retrograde amnesia mainly for semantic memory. Damage to the hippocampal complex
has less effect on semantic memory than on episodic memory, whereas damage to the
neocortex impairs semantic memory. Functional neuroimaging also indicates that different
brain areas are associated with episodic and semantic memory.
Episodic memory•
There is an important distinction between familiarity and recollection in recognition
memory. According to the binding-of-item-and-context model, familiarity judgements
depend on perirhinal cortex, whereas recollection depends on binding what and where
information in the hippocampus. Free recall involves similar brain areas to recognition
memory. However, it is associated with higher levels of brain activity, and it also involves
some brain areas not needed for recognition memory. Episodic memory is basically constructive
rather than reproductive, and so we remember the gist or essence of our past
experiences. We use the constructive processes associated with episodic memory to imagine
future events.
CHAPTER SUMMARY
286 COGNITIVE PSYCHOLOGY: A STUDENT’S HANDBOOK
Semantic memory•
Collins and Quillian (1969) argued that semantic memory is organised into hierarchical
networks with concept properties stored as high up the hierarchy as possible. This inflexible
approach was superseded by spreading activation theory, in which activation of one
concept causes activation to spread to semantically related concepts. Perceptual–functional
theories assume that the visual or perceptual features of an object are stored in different
locations from its functional features. Such theories are oversimplified. The distributedplus-hub
theory provides the most comprehensive approach to semantic memory. There
are hubs (unified abstract conceptual representations) for concepts as well as distributed
modality-specific information. Evidence from patients with semantic dementia indicates
that these hubs are stored in the anterior temporal lobes.
Non-declarative memory•
Amnesic patients typically have intact repetition priming but impaired declarative memory,
whereas a few patients with other disorders show the opposite pattern. Priming is associated
with perceptual fluency and increased neural efficiency. Amnesic patients generally
(but not always) have high levels of procedural learning and memory. This is the case
whether standard motor-skill tasks are used or tasks requiring skills similar to those
needed in the real world.
Beyond declarative and non-declarative memory: amnesia•
Several theorists have argued that the distinction between declarative and non-declarative
memory is oversimplified and is inadequate to explain the memory deficits of amnesic
patients. According to an alternative viewpoint, amnesic patients are deficient at binding
or forming associations of all kinds. The evidence mostly supports this binding hypothesis
over the traditional viewpoint that amnesic patients are deficient at declarative or explicit
memory.
Long-term memory and the brain•
Research on amnesic patients has shown that an extended hippocampal system is crucial
for episodic memory. Skill learning or procedural memory involves the striatum and the
cerebellum. Patients with Parkinson’s disease have damage to the striatum and are generally
impaired at procedural learning. Neuroimaging studies suggest that the prefrontal
cortex is often involved in the early stages of procedural learning and the striatum at later
stages. The dorsolateral prefrontal cortex is involved in relational encoding and postretrieval
monitoring. The ventrolateral prefrontal cortex is involved in controlled retrieval
and a process dealing with competing memory representations. The parietal cortex is
involved in various attentional processes of relevance to learning and memory.
Baddeley, A.D., Eysenck, M.W., & Anderson, M.C. (2009).• Memory. Hove, UK: Psychology
Press. Several chapters (especially 5, 6, and 11) are of direct relevance to the topics covered
in this chapter.
FURTHER READING
7 LONG-TERM MEMORY SYSTEMS 287
Foerde, K., & Poldrack, R.A. (2009). Procedural learning in humans. In• Encyclopedia of
neuroscience. New York: Elsevier. This chapter gives an excellent overview of theory and
research on procedural learning and procedural memory.
Patterson, K., Nestor, P.J., & Rogers, T.T. (2007). Where do you know what you know?•
The representation of semantic knowledge in the human brain. Nature Reviews Neuroscience,
8, 976–987. The authors provide a succinct overview of our current understanding of
how semantic memory is organised within the brain.
Reder, L.M., Park, H., & Kieffaber, P.D. (2009). Memory systems do not divide on conscious-•
ness: Re-interpreting memory in terms of activation and binding. Psychological Bulletin,
135, 23–49. The distinction between explicit/declarative and implicit/non-declarative
memory systems is evaluated in the light of the evidence and an alternative theoretical
perspective is proposed.
Schacter, D.L., & Addis, D.R. (2007). The cognitive neuroscience of constructive memory:•
Remembering the past and imagining the future. Philosophical Transactions of the Royal
Society B: Biological Sciences, 362, 773–786. Interesting new perspectives on episodic
memory are offered in this article by Schacter and Addis.
Schacter, D.L., Wig, G.S., & Stevens, W.D. (2007). Reductions in cortical activity during•
priming. Current Opinion in Neurobiology, 17, 171–176. Schacter and his co-authors
discuss the main mechanisms underlying priming.