 |  Learning: Principles and Applications, 4/e Stephen B Klein,
Mississippi State University
Memory Retrieval and Forgetting
Chapter Outline
Chapter Outline
ATTRIBUTES OF MEMORY
Underwood proposes that memories can be conceptualized as
collections of different types of information called memory
attributes. According to Underwood there are 10 memory
attributes that not only assist in differentiating various
memories from each other but also serve as cues to assist
in the retrieval of particular memories.
The Function of an Attribute
Attributes serve two functions: reducing
interference between memories and enhancing
retrieval. Memories, which consist of different
attributes, are more easy to distinguish and
interference is less likely. When we encounter a
stimulus that is a salient attribute of a
previously established memory, we often retrieve
that memory with little effort.
Types of Attributes
Acoustic Attributes
The acoustical properties of an event serve
as a retrieval cue for that experience. One
acoustical property of language is rhyming.
Orthographic Attributes
The structural characteristics of letters
and words may be stored as part of a memory
for a particular event. Letters differ in
shape, words differ in length, number of
syllables, unusual letter combinations,
etc.
Frequency Attributes
In Underwood's view, a counting mechanism
registers every experience and records the
number of times a specific event has
occurred.
Spatial Attributes
Many students have experienced the spatial
aspect of a memory when they are able to
remember the location on a page where a
particular piece of information may be
found.
Temporal Attributes
Significant temporal information becomes
stored as an attribute of memory. It is
important that the temporal information
have some significance to the person such
as a birthday, a holiday, or Friday the
13th. Insignificant temporal information
does not become associated with the memory.
Modality Attributes
The modality in which an event is
experienced can serve to retrieve the
memory for that event. For example, did you
hear it, see it, feel it, read about it,
etc.
Context Attributes
Reexposure to the background environment
can prompt retrieval of an event associated
with that background. An example might be
returning to your old elementary school or
a house you previously lived in. Odors are
particularly effective contextual cues such
as the odor of your grandmother's house or
the specific cologne/perfume used by a past
significant other. Likewise, music can be a
powerful retrieval agent evidenced by the
popularity of "oldies" radio stations.
Affective Attributes
Many events evoke emotional responses that
then become associated with the memory for
those events.
State-dependent Learning is an
example of the utility of the
affective attribute. When learning
material in one emotional state,
subsequent recall is better when in
that emotional state than when in a
different emotional state.
The Kamin Effect and the role of
ACTH in memory is also an example
of the affective attribute at work.
The Kamin effect refers to a
U-shaped function for an avoidance
response. Rats have excellent
recall of the avoidance response
when tested immediately after
learning or 24-hrs later. However,
retention of the avoidance response
is relatively poor when tested 1-4
hrs after the original learning.
Research indicates that inhibition
of ACTH release is the factor
promoting poor avoidance
performance at intermediate
retention intervals.
Relative Contributions of Internal
versus External Memory Attributes.
Although a change in the internal
state can produce retrieval
failure, forgetting does not occur
because other memory attributes
continue to be present.
Verbal Associate Attributes
When exposed to a particular word, the word
may activate a variety of verbal
associates. The presence of these verbal
associates can assist in retrieval of the
desired word. There are two types of verbal
associate attributes: parallel associatives
and class associatives.
Parallel Associative Attributes
There are three major types of
parallel associative attributes:
antonyms, synonyms, and functional
associates. Functional associatives
are based on functional contiguity
such as "cup-saucer" or "key-lock".
Class Associative Attributes
When a word elicits a category name
to which it belongs, the category
name is a class associative
attribute of the word and may
assist in later retrieval.
Transformational Attributes
Many words or ideas can be transformed in a
number of different ways to enhance later
retrieval. Transformational attributes
represent recoding of the original input.
Three of these transformations are images,
natural language mediators, and order
transformations.
Images
Transforming a word or idea into an
image facilitates retrieval in much
the same way that the method of
loci functions as an effective
mnemonic.
Natural Language Mediators
Natural language mediators assist
in retrieval in paired associate
learning by bridging the gap
between the two words. For example,
if the pair of words to be learned
is dog and cat, using a mediator
such as dog - chase - cat creates a
meaningful link and may assist in
retrieval.
Order Transformations
Meaningless, nonsense syllables can
be transformed into a meaningful
word by rearranging the order of
the letters.
FORGETTING
After memorizing lists of 10 to 12 nonsense syllables,
Ebbinghaus set out to relearn the lists. He memorized over
150 lists using a variety of retention intervals. He found
that after 24 hours he had forgotten almost half of a prior
list and by six days he could recall only about 25% of the
list. Three theories have been proposed to explain
forgetting: decay, interference, and lack of memory
attributes.
Decay of a Memory
The decay theory of memory claims that disuse of
the memory over time leads to its removal. The
decay theory was explored in the classic Jenkins
and Dallenbach (1924) experiment. Modern
researchers have focused on metabolic changes as
the possible mechanism for decay.
Interference
Interference theory proposes that forgetting is the
result of competition between similar kinds of
information at the time of recall. Two types of
interference have been identified, proactive
interference and retroactive interference. In
proactive interference (PI), old information in
memory blocks the recall of newer information
stored in memory. In retroactive interference (RI),
newer information coming into the memory system
disrupts the recall of older information in memory.
Melton and Irwin's Two-Factor Theory of
Interference
According to Melton and Irwin (1940),
response competition is a source of both
proactive and retroactive interference.
However, another factor, unlearning was
assumed to operate in retroactive
interference. In the retroactive
interference paradigm, Melton and Irwin
propose that in order for subjects to learn
the second list, they must unlearn or
extinguish the first list. Thus, trying to
recall the first list responses immediately
after learning the second list is poor.
However, over time, the suppression effect
extinguishes, allowing the first list
responses to recover in strength and
provide competition for the second list
responses.
Underwood's List Differentiation View
Underwood (1983) suggested that
interference is not caused by response
competition but instead by a failure of
task differentiation. According to
Underwood, subjects remember responses to
both tasks but fail to remember which task
a particular response was associated with.
Thus, when asked to recall the response
learned from the first task, subjects can
remember the responses from both tasks, but
will appear to have forgotten them because
the subject cannot remember which response
came from the first list. A considerable
amount of research indicates that the
failure of list differentiation, rather
than competition, is a source of
forgetting.
Postman's Generalized Competition View
As the retention interval increases in the
proactive interference paradigm, the amount
of interference increases. Conversely, as
the retention interval increases in the
retroactive interference paradigm, the
amount of interference decreases. Barnes
and Underwood (1959) initially explained
that the retention interval effects were
due to spontaneous recovery of the first
list responses in the interference
paradigms. Postman has suggested that
generalized competition is a "set" or
disposition to continue to respond in the
same manner most recently learned.
Furthermore, second list learning in the
interference paradigms causes the "set" to
dissipate, leading to a decline in
generalized competition. Therefore, the
degrees of proactive interference and
retroactive interference respectively
increase and decrease over longer retention
intervals.
Forgetting and Short-Term Memory
Peterson and Peterson's (1959) results suggest that
memories, without being rehearsed, fade rapidly in
short-term memory. However, Melton (1963), having
noted that both proactive and retroactive could
have occurred, offered an interference explanation
of Peterson and Peterson's data. Subsequently,
Keppel and Underwood (1962) found compelling
evidence that proactive interference operates to
disrupt short-term memory and may have been a
factor in the Peterson and Peterson experiment.
Later, Waugh and Norman (1965), using their probe
digit technique, found evidence for retroactive
interference operating in short-term memory.
According to some (Baddeley & Cook, 1971; Shiffrin
& Cook, 1978) decay does occur, but is simply not
apparent under most conditions because interference
is a more powerful and common cause of forgetting.
Support for this notion has been found by Baddeley
and Scott (1971).
RECONSTRUCTION OF THE PAST
It appears that retrieval of information from the long-term
store involves memory reconstruction. Therefore, if the
reconstructions are not accurate, then retrieval errors
occur and the memory may not be an accurate representation
of past events. The phenomenon of memory reconstruction was
first assessed by Bartlett in 1932.
Memory Reconstruction Studies
Bartlett (1932) investigated memory reconstruction
in a classic study in which he gave his British
subjects a story about Native American mythology
called The War of the Ghosts. The subjects had to
serially reproduce the story several times later
after the initial reading. Bartlett was interested
in the type of errors the subjects made in their
attempts to recall this unfamiliar story. The error
pattern in the recalls indicated to Bartlett that
the subjects were attempting to reconstruct the
story based on their own personal knowledge system.
In addition to the narrative type of
reconstruction, Bartlett demonstrated additional
memory distortions by presenting a figure to a
subject and asking them to reproduce it for a
second subject. After ten subjects the figure had
undergone an extensive change as each subject
reproduced it based on their own personal knowledge
system. Bartlett's research indicates that we base
our reconstruction on "what must have happened."
The Accuracy of Eyewitness Testimony
Loftus (1980, 1992) has revealed the importance of
memory reconstruction as a factor in the accuracy
of eyewitness testimony. Loftus contends that a
memory can be altered by information received from
others. According to Loftus, witnesses at a trial
recall only what they think they saw or heard.
Lawyers can alter a witnesses recollection by
asking a question in a particular way. For example,
after witnessing a slide presentation of a minor
accident at an intersection Loftus asked subjects
one of two questions: "did you see the broken
headlight" or "did you see a broken headlight." The
key words are "the" versus "a." Subjects asked the
first question were much more likely to report that
they observed a broken headlight than those asked
the second question. The word "the" implies a
certainty that there actually was a broken
headlight, which there was not. Similar misleading
questions used by Loftus involve stop signs versus
yield signs with similar results. Finally, Loftus
and her associates showed that not all people are
susceptible to memory distortion, young (up to 20)
and older subjects (over 65) are most likely to
exhibit memory alteration.
False Memory Syndrome
Many psychologists believe that individuals can
repress memories of extremely distressing childhood
experiences which may later surface during
psychotherapy for an unrelated problem. This
phenomenon has been labeled the repressed memory
syndrome. Several court cases have involved the
apparent recall by adults in therapy of childhood
sexual abuse who have then brought charges against
the alleged perpetrators. However, Loftus and
others question the accuracy of these recovered
memories and claim that the memories are actually
the result of distortions produced by the
patient-therapist alliance. The creation of an
inaccurate record of childhood sexual abuse is now
called the false memory syndrome. Loftus and other
researchers have experimentally demonstrated that
it is relatively easy to induce someone to believe
an entirely false event actually happened some time
in their past simply through suggestion.
Motivated Forgetting
Based on Freudian ideas, repression has been
suggested as a principle of forgetting. The
function of repression is to block the recall of
anxiety provoking memories. Subsequently,
investigators have explored, in the directed
forgetting paradigm, whether animals can be
motivated to forget. Evidence indicates that, when
subjects receive a " remember" cue, performance is
enhanced, but receipt of a "forget" cue weakens
performance, in a delayed matching-to-sample
procedure. Human studies investigating directed
forgetting have yielded mixed results. Some
investigators find directed forgetting both in
recall and recognition tasks while others find
directed forgetting only with a recall task. The
inconsistent data has hampered the development of
an explanation for the directed forgetting
phenomenon. Finally, future studies will have to
discern whether directed forgetting principles
operate in some cases of repression.
ANATOMICAL BASIS OF MEMORY RETRIEVAL
Based upon studies done by Warrington and Weiskrantz (1968,
1970, 1975), interference effects are very prominent in
list learning done by amnesiacs. The interference effects
can be decreased when amnesiacs are given powerful
retrieval cues at the time of list recall. Finally, lesions
of the hippocampus positively correlate with retention
difficulties. Therefore, the hippocampus appears to be
crucial for both the storage and retrieval of declarative
knowledge.
|
|
2002 McGraw-Hill Higher Education