You are on page 1of 25

a l i laune

,SA ii, AvL> LCING MM


TiRM M:MORY; 1EVELS OF
ORGUTTING DUF TO DCAY AND PROCESEING; MEASURIS OF
MEMOP
Eoi of Menmory
INTERERENCE

Meniofy S Ciearly a
expeienCSQE Ihepast. WekAoCENS in all aspects oRabohavlour. It deals
with the
ieltaming. an activity, which he had learnt memory all around us. A person or anreproduction
see indications of of eVensa
animal
experien
major part or what is
learnt is forgotten, andpreviously.
Yet forgetting is also a common
past. ror exampie, we when we are required to remember it we tendexperience.
may not remember to reconstruct u
recount t in detal, we do manage to relatethe complete story of a bad film. However, if a friend àSKS
a coherent U
nus memory an active, dynamic, complex
is account, because we fill in the 'gaps in our
philosophers for Over 2000 process. Human memory has been the topic of
meri
years. Both Plato and Aristotle discussed the nåture
and operation
speculation d
owever, the scientific study of memory started only with or menoy
Ebbinghaus
ge0RIS (transiated as Memory in 1913, after an extensive five who wrote volume entitled eEL
a,
proçesses of his own memory. year long investigation of the factors a d

EbDinghaus invented the nonsensesyllables consisting of a vowel


a subject Ebbinghaus would look at between two consonants (Such asAW KU
CEW et.). Using himself as
each syllable in the group for a fraction o o
second, and he wOuld pause fifteen seconds before starting
T ne
cOuld. recite each syllable in the through the group again. He continued this
without making a mistake. At that point he noted in way
exposures (trlals) it took before masterygroup was reached. how may
He also plotted the errors made in each trial,
creating psychology's first learning curve. At various intervals following thus
a group of nonsense syllables. The number of trials for original mastery, Ebbinghaus went back
and
relearned
riais tor
original caming was, calledsavings He plotted savings as relearning subtracted from the number or
a function of time
earning, thus creating psychotbgy's first retention curve. This elapsed since original
ast for the first few
hours following a léarning graph indicated that the rate of forgetting, is very
experience and very slow thereafter. He also found that over
2arning reduces tthe rate of forgetting considerably. Another
12terial, was that the learning rate was faster and retention wasresult, using words from Don Juan as meaningful.
superior for meaningful material.
Today Ebbinghaus is famous for his material (NsS), method
retention curve is one of_the major pioneer attempts_in(experimental)
psychology
and analysis of data (savings). His
to reduce experimental data to a
mathenatical form. Not much of a theorist, Ebbinghaus was essentialy an
novative experimental design _and methodology. Ebbinghaus emancipated experimenter.from His strength wasB
demonstrating that the mental processes of learning and psychology philosophy by
memory could be studied experimentally.
Iminense though his contribution was, Ebbinghaus essentially researched the for nonsense. In contrast
rtlett, a twentieth century British psychologist was interested in people's memory
*search after meaning". Bartlett
32) had his students read stories or study pictures, and reproduce them after 15 minutes, few hours, and
eral weeks. He found evidence for reconstruction in
that, Over a period of time, memory makes
re
coherent, tomplete, and logical, even though they may not have been so at the time of experiences
inghaus, Bartlett devised rescarch methods and theories that influence research in the learning. Like
aningisl materia memory for

Tne researcher's in tre area of meniory cenceptuialize


memory as a static structure whereas otthers think of if as
dyriamic process. Memory is often viewed as a kind of
store-house of knowiedg that is constantly modified b
aperience ThES¢ 3se3rthers conceptuaiize memory as sensory, short terin, ann
long-term memory (Atkinsc
hif, i91). Alternativeiv those who think of memory as a dynamic
proceSS
bf proceSsing iniormatioh, and that the deeper the-levei at which information hold that there are variou
is processed at the ti:716
in, the better vmernory for it. Ttiey aiso believe that tnere are no
'

(raik and kakt i9/2 separate short-term and ionp: te

. Tulving. (1972) . .
episodic and semantic memory

**i******
be Usualy dated as to when the information was acquired.Semanticmemory s the gencal kIjCwledge
in the Hindu religion, etrc. n e cieS H
Word-ho is the president of United States, the impo tance of Ganga Tulving (19/2) Say
Know when _and where the information in semantic memory was originally acquirecd.
Episodic memarylreceives and stores informalion about temporally dated episodes or events, and teno
spatial relations among these events...Semantic memory Is a mental thesaurus, organized knowiedge. P
ara
possesses aoout woras ana otner vervai synoois, tneir neaning and referents, about relations among tem,
about rules, formulas, and algorithms for the manipulation of these symbols, concepts, and relations".

Another major distinction is between expliclt and implicit memory (Schacter, 1987).Explicit memorylis memory
with awareness. The subject is instructed to remember material for which he is to be tested later on. Traditiona
epenments in memory that tested recall, recognition, etc. were all experiments in explicit_learning. Research in
mplcit memoy started in eanest only in the last two decades of the twentieth century. Implicit memory) does
not
invohve explicit or conscious recollection of any previous experience. It is tested by the change on various
kinds of memory tasks due to prior experiences outside the awareness of the subject or due to Something the
subject had not been explicitly instructed to learn for a future test. Though the task instructions do not marRe an
explicit reference to an episode in the subject's personal history, the performance of the subject is
infuenced by such events. nevertheiess
Despite the many contrasting views in the area
of memory, researchers do agree that memory can be described
in terms of three stages: acquisition, storage and rëtrieval.
n hetacQuisition stage information is encoded into memory. It
inthe brain in the form of a memory trace -the engram. However, was felt initially that all experiences are encoded
haveno speciic neurons associated with them. There are changesrigorous research revealed that experiences
in brain activity of specific areas as the
information is
being.progessed, but these changes are not permanent and leave no traces behind. In his baok lIn
search of an engranm
ashley (1950) mentions that removing specific parts of a rat's brain affected his
but did nct totally eliminate
specific memories. He concluded that if at all they are stored in the behaviour,
are distributed
throughout the braini. More recent reveals that the thin outer brain, memories
cortex is known to be essential for recall
of
layer of the brain, the cerebral
change in certain protein molecules (RNA) ofpast experience. Another possibility is that there is a biochemicai
the neurons. Ward Halstead
Ribcnucieic acid and protein molecules (1949) advanced the
might be the engrams that scientists had sought for theory that
Experimenting rats, Hyden (1967) found noticeable changes in the
on many years.
animals and thus recognized that RNA amount of RNA in the brain of trained
ncreasing the production of several chemical presumably influences cell development.. Nerve cell
molecules whenever a wave of electrical responds by
f the nenve cell. activity sweeps the length
Recently researchers Lynch and Baudey (1985) have identified
esponsible for long term memory. An enzyme called a chemical in the brain
which is
1e brain cell. CALPAIN is responsible for breaking down the proteins of

he storade Siage is the warehousing of the


wantity and quality of the information acquired. Since there is often a difference in the
information_that goes into the store and later comes out of the
SSumethat processing of information_continues even store, researchers
st through forgetting, and a
great deal
when_it is in storage. A great deal of information
may be
may be added through consolidation
searchers, however feel that processes
such and reconstruction. Some
e of memory retrieval.
as
forgetting, consolidation, and reconstruction occur in the third
Meval tefers to the process of
en information is present in thebringing information out of memory.There are two basic.methodsof
mation is not present in the environment, people retrieve it from retrieval
environment they must retrieve it through recal memorythrough_recognitian..Butifthe
various definitions of memory reflect the diversity of views
4) define memory
simply as "doing what one has learnt to do". prevailing in the area. Woodworth and Schlosberg
memory. Morgan, King, Weisz, and Schopler Thus they establish the link between
ding, storage, and retrieval of what was learnt (1986), mentioning the stages of learning
earlier". Crider, Goethals, and memory, say that "it is
the
Storage view
I n contrast,
of memory, hold
that Kavanaugh (1989),
"memory is the ability to store information so that it can be adhering to
Morris (1990) holds_that "memory is used at a later
2ptLializes memory as a process.in the process by_which
:rea. AC.ording to line with the
BarR 201 Meriorv is the information processin approa material is ctaFe.ia
i st: ioa*n d
rari7ity Likiiin iit iaie * rarrO?i .
*********

*** ****
SENSORY, SHORT TERM, AND LONG
ORY MEMORY TERM MEMORY
soy memory is the
t
memory for all the stimuli
emory Tor ali tne impinging upon the the
sensations unat an organism at a .Ol devised the patial
report method to study sensory memory.organisIm
He
nas at any
particular time. speru
our letters each for
50
msecS (1/20 of a second). Then he askedshowed the subjects a display of
threeo
report). The subjects averaged the subjects to report as many
4.5 letters out of twelve. iette
T R X D
F V QB
K SN P
t n e n varied the procedure and asked the subiects to report only some of the letters (Partial repor)
heaccomplished by presenting a tone before the display was turned off or at delay periods up to 1 sSecona h
Sibject heard a high pitched tone when they were to remember the top row, a medium pitched toneh
unat
ch
d report condition, it was found that
low pitched tone for the bottom row. In this partial of he
subject remembered an average of 3.3 letters fron the four letters row they had to
recall. Sperling argued

that since people did not know which row they would be asked to recall as they Iookedat the display enHers
nave nad almost ten letters (3.3 x 3 = 9.9 = 10) in their memory. People could report these many eters
of the entire display. his menta
Decause they had a vivid clear, detailed and exhaustive mental image
-
-

tone
for a very short while. If the was not
mage is called an icon (Neisser, 1967). However, the icon is retained recalled no more than the number
or
unti a full second after the display was turned.off, the subject
SOunded
letters remembered from the whole report procedure.
memories as there are the senses. However, only three
icoretIcally, there should be as many types of sensorystudied names them
identified and by the researchers. Neisser (1967)
n d s of sensory memories iaveb en (related to the skin senses). Of these,
(related to audition), and haptic
Conic memory (related to visionichoicbeen the focus of research.
IConic memory and echoic memory have
T
IConic memory may be defined as the rapidly decaying memory component in excess of the visual memory span.
refers to the afterimages that can be seen even after_the
It is an unsettled question whether iconic memory
negative. If you stare at a light square
on
visual stimulus has ceased. Afterimages can be both positive as wellas the negative
a dark background, the positive afterimage wOuld be a light square on a dark background, whereas sensation
There is a persistence of the initial visual
afterimage woLuld be a dark square on a light background.
termination of the which thereafter can change to
stimulus,
the
M (positive afterimage) for a brief period following there is than one cycle of positive and negative
Sometimes more
a negative (reversed contrast) afterimage.

afterimages.
the of activation in the neurons
physiolagicäl basis of the initial persistence of vision perhaps arepersistence to be the result of fatigue of
is
The
afterimages presumed
originally activated by the stimulus. Subsequent negativeothers. Sakitt suggested that these neurons exist
elements previously activated and the dominance of the (1976)
had rods but no cones. Since rods are activated only at low levels
in the retina. by studying a person whose eyes
in seeing ordinary things in daylight (just as we find it difficult too
Th of illumination, this person experienced difficulty but if the
stimuli are presented in ordinary daylight, the subject does not see them,
dy see things in glare). When of the stimulus, she can then see an icon of the presented
subject loses her eyes after the presentation all time! Adelson (1978),
stimulus! In fact she uses this technique in normal viewing, by going
around blinking
influenced by the intensity of the stimulus, but stimulus
however found that the longevity of afterimages is
icons. Therefore he suggests that the icon is probably stored in a
intensity ha_ no effects on the memory for
the retina. Further, iconic memory, as demonstrated by
mere scentral location of the nervous system than non associative in nature as demonstrated by Wickelgren
Sperling (1960) lasts for about a second only. It is also
concentrated on memory for form rather than
and Whitman in. 1970. Studies of iconic memory have typicaly but there could be other
contrast. Thus it might be contended that afterimages contribute to iconic memory,
in iconic memory as well.
Contributory factors
.

**
***
. *-...******** ********
***************** ********** *****
* - ********~************ **********
**** " **"'.********** **** ********
PAPCK D, SCCIIONA,
UiNII
11,

thepersistence of audition. Crowder and Morton :.


Neisser (1967) uses the term echoic memory to refer to concerned only with the sensorY andnot the se
that itis
call it precateqoricalacquetic storage, suggesting
Crowder (1972) presented the subiects.with.three lists of.anuah
featuresof the stimulus Darwin, TurveY, and left,
headphones, One list_sounded as ifit_came from the
and letters on a stereo tape recorderand presented, the listener savy
the ists were
from the right, and the third from_the middle, Immediately_after this.partiai report procedure yielded better
cue to recall one or tne tnree iiSTS, As in Sperling's study
VISuai memory lasts for only_about a seond,
performance than a comparable whole-list procedure. Hawever_iconic
whereas echoic.memory lasts for as long as 4 seconds.
theorists who do not adhere to the storage mddel of
The importance of sensory memory is such that even those to postulate an initial
view of memory), find it necessary and useful
memory (and instead emphasize the process name: it as the
a time limit. Tulving and Schacter (1990)
sensory register, which however is not charactereized by Mark, Quasi
Representation System (PRS), formerly characterized by Tulving (1985)
as Question
Perceptual
Memory or simply QM. PRS is a presemantic perceptual system that operates independently of episodic meinory.
in which the subjctis
It is thus neither semantic nor episodic in nature. PRS is studied through direct priming,
exposed to the target stimuli beforehand, and then it is tested whether he identifies or generates the targets
from reduced cues in a subsequent test. Studies indicate that it functions for ecologically valid stimuli, only.
Tested with pictorial stimuli, no evidence for priming of structurally impossible objects is observed (Schacter,
and
Cooper, and Delany, 1990). Direct priming also does not occur for pseudo-words among amnesics (Diamorid
Rozin, 1984). The PRS thus exists separately but interacts closely with other memory systems. The major
after
evidence supporting the PRS hypothesis is that amnesic patients cannot remember the study episode even
and
a short interval, yet they show priming effects frequently as large as those in normal subject (Warrington
There also
Weiskrantz, 1974; Graf, Squire, and Mandler, 1984; Schacter, 1985; Shimamura, 1986). are,
developmental dissociations between priming and explicit memory. (Light, Singh, and Capps, 1986). Further,
effect
scopolamine reduce performance in explicit recall and explicit recognition, but have little no
'or
alcohol and
function
on priming. (Parker, Schoenberg, Schwartz, and Tulving, 1983). Finally, priming and explicit memory
independently and are affected by different factors in normal subjects. Schacter, (1987) and Richàrd_on
Klavehn and Bjork (1988) have reviewed this literature.

T condude, we may summarize the characteristics of sensory memory as follows:


impinging on the
. It is very vivid dear and detailed. Perhaps it contains information about all the stimuli
-

sense organs.
2. It lasts for a very short_time. Iconiç memory lasts for only about a second whereas echoic memory lasts

about four seconds.


It is non-associativein nature, It is neither episodic.nor.semantic.
Perhaps it processes only ecologically yalid stimuli.
5. It may be moreimplicit than explicit in natuire.

SHORT TERMMEMORY
When we look up a number in the telephone directory, dial it, but get a busy sighal, we know from experience,
that we should either keep repeating it to ourselves or keep on dialing it. Unless we keep the number in our
short term memory with one of these tricks, we are likely to forget it and have to look it up all over again in the
telephone directory. Short term memory is that part of the memorystructure in which information is stored
temporarily.Unlike sensory memory,short-term memory istemporarilyactive and conscious.It is relativaly.more
permanent than sensory memory, but ESS nduring than long term memory.It is also called the working and in
memoy because it contaiñs information that is currentiy in use by the individual, being acted upon,
Consciousness (Anderson,1985; Baddeley, 1983). The dlassic experiment to establish the existence:of a
transientmemory store was done by Peterson.andPeterson in 1956. On each ofa seriesof trials, they required
theirsubiects to remenmber a non-word consisting of three consanants (such as XRZ). The experiment
prevented thesubjects from any type ofrehearsal repeating thetrigramtooneself) by instructing them to count
backwardsby threes from an arbitrarily selected number (such as 657)Thenumber was fashed. immechately
after the subject saw the trigram. Each subject was tested with six retention intervals =36912,l5,and 18
seconds.At the-endof the Tetention interval, the subject was asked to recallthe trigram. he resilts snowea
that substantial and rapid forgetting occurred in each retention interval. The amount forgotten is directly relaed
3 IheJenath_offhe retention interval. TVearly 10% forgetting occurred after three seconds, 50°% after (; seconas,
*ncimora than a0% after 18 seronde. Itis this transient properh which gives TM its namc

*
r. D, d E 1 O N A, UIVI
D

ce for
s!M aiso comes fom
ed the
best (primacy effect) the serial position effect. the fact that the
icd the lcast. The followed by the last
graphical representation few
first
items (recensy effect)few items in a setia Ae
ed the seriai of the ano
posttion curve,
It usually relationship between a recall
ie miadie (Gfeene ana
Samuei,
shows accurate recall in the beginningvWords
and
po nd ow recall in
are better rehearsed
and are thus
1986). The usual explanation for the the end, t fow items
more likely to enter primacy effect is that ho list. This
explanation is supported by the finding that if the the LTM than
rehearsea, the primacy effect is diminished and items are oresented items in the rofore less likely to be
very fast and are thererorc
Smany, the recency effect is explained by the they are not recalled better than other itei 162).
supportea by the fact idea that the items are still in the STM. nis
that if the recall is
da i further
arithmetic, then the delaved for about 30 seconds by asking the subject ontal
ment
uO
recency effect disappears (Postmen and pe
Phillips, 1965).
Atkinson and Shiffrin (1968) hold that the short
nod yerDal or visual information. A buffer has aterm memory, in part, consists of buffers, oen
relatively limited capacity, and can hold informadonoY for
e a ime. Information enters the buffer one item at a time, Once the buffer is filled, new items displace ue
Older tems. While items are in the short term buffer, information about them De rred to the long
t nemory. The longer an item remains in the short term memory, the more may rai about it enters uie
information
germ memory. Enhancing this model, Baddeley (1986, 1992) claims that there are at least three conpone
nemory: the articulatory loop, the visuospatial sketchpad, and the central executive. The articulatory
Kina of internal speech mechanism that functions to rehearse verbal information. The visuospatial
c n a tunctions to rehearse visual images. Both function independently of each other. The centralexecutive
s the skillor process that makes decisions about which other component of working memory must be activated
accomplish a task.

Capacity.of STMI
AS early as.1885, Ebbinghaus reported that the longest list of nonsense syllables that he could recall correctly
after having studied them only once and therefore having taken them into short term memory - was 6 or 7.
Jacobs (1887) reported the immediate memory span to be 6 or 7 numbers. Most students can accurately repeat
strings of 6 or 7 numbers. In 1956, Miller published a classic paper entitled The Magicalnumber 7, plus
minus two: Some limits on our capacity for processing information. According to Miller (1956)STM hasa memory
Span of 7+2 items. The items may be letters, numbers, NSS, words, etc. the capacity of the STM may be
increased by chunkind Acollection of items may be represented together as a single item known as a chunk.
Thust Old he more_arcurate to say that STM has a memory span of 7+2 chunks. For example it may be
difficult to remember all the following letters after readingthem only once:
I BMPU J F KRAW o PEC
However, it is easy to remember the string if we make five chunks: IBM, PU, JFK, RAW, and OPEC. The capacity
ofSTM may also be increased by grouping itemsusing rhythm, movements etc Grouping_is different from
chunking in that the individual items retain their distinct identity in the collection formed due to grouping,
hereas they aYesubsumed in the new node thot represents a chunk
LEncodingiiSTM
As mentioned earlier, information is retained in the STM for a very short while. In fact the material is
completely
lost in about 20- 30seconds. Information may be retained in STM if the subject uses maintenance rehearsal,
i.e., repeats the cognitive activity on the information again and again. E.g., When we dial a telephone, we try to
remember the number by repeating it to ourselves. As soon as these repetitions are stopped, the information is
lostforever, If by chance, we have to dial the same number again, we need to look it up again. However, the
informationin STM can be transferred to the LTM through elaborative rehearsal that involves a deeperand more
thoroughanafysis of the information. If you notice that the first three numbers of a phone number are the same
as your OwWn phone number, and the last four are the same as the number of your car, you Will neyer
forget fhe
numberagain, because deeper processing ensures that the number goes to the LIM (Craik and Lockhart, 1972).
Baddeley and Hitch (1974) proposed the concept of the articulatory loop a mechanism for
based materiat hence allowing the STM trace to be continually refereshed. recycling speech
.

Storage in STM
Information is stored in STM prinarily in acoustic (sound) codes. Conrad (1964) studied the errors made by the
SIhjects when thev were asked to recall a seriec of conscnants that they had seen in visua.
a display at fiE Fat

************** -*
DA , PAPER D, SELI IUIN A, UIVII Melino
of one every .75 seconds.He found that people tend to confuse letters that sounded alike for exantis -

would confuse Nwith i4, and D with T, but not N_with Vor D with 0, He concluded that
seen the everi thotigh peor
consonants, they must have at least partially encoded them as sournds. Nevetheless, sound
only type of codes people use for STM. Coarad (1972presented is no
h
orthat looked alike- KYXZ,_to a graup of deaf people,and to a lists of letters that either souinded alikg-DVEC
nar peapIe wno near niade more mistakes on letters that sounded group of péople with normal hcaring. He foun!
mistakes on letters that lookedalik, This. allke, but people who were deaf macde more
shows.that.other.typesof codes are also uSed by STM.
The type of code used in the short term
As we have seen, short term memory also influences the amount of information that can be stored.
memory can store seven chunks of information if the information is- pronounced and
encoded acoustically. But
if the information
(Schweickert and Boruff, 1986) and if the takes a long time to be pronounced, short-term capacity iS feduced
information
around three chunks (Zhang and Simon, 1985). One _is_encoded semantically, capacity_may be reduced to
explanation for these differences in
information must be placed into the short term is that capacity.
t o encode information in a semantic form. memory within a period of 1.5 to 3 secords, and it takes longer

LRetrieval from STM


Retrieval is the process through which information stored inn memory is
memory search in short term memory was retrieved. The kind of strategy used for
investigated byaSternberg
presented subiects. with sets of randomly ordered numbers 9, 5, 4, (1966) using a probe technigúe. He
seconds but the length of the sets varied from 1to 6 digits. After 2 etc. they saw each numbér.for 1.2
had to determine whether they had just seen it. Sternberg each set they saw one number (the probe) and
found that as the length of the list increases, the timemeasured how long it took the subjects to answer He
Consequently he concluded that search in STM was serial rather taken by the subject to respond also ncreases.
than parallel. He also found that
respond was not affected by the_poStion of the_probe item
time taken to
exhaustive and not self-terminating. People, it seemed, search in the list. TNS Suggsts that memory.searchis
,

a single decision Since infarmation is limited in SIM, through all_the numbers in STM, and then imake
this.strategy is quite efficient.
However, other rsearchers have found that people do
result that would not be expected if the search respond faster when items are repeated in a
a
process were totally exhaustive (Baddeley and
memory set
Thu response times are partially determined by the Ecob, 1973).
Also, whether people have to recall or recognize the strength/familiarity
of the information stored in the
information influences the search process. Sternberg. memory.
found that when people were shown a set of numbers (such as 5 7 (1967)
number came efore the probe item (8), 8 2 3), and then asked to recall which
they responded more
series, than if it had a later position. This was evidence for a quickly if the probe was present earlier in the
2ffect the recognition time as observed in the earlier self-terminating search. Probe position did not
han the recognition times for the experiment also. The recall times were consistently ionger
probe items. This suggested that during recall, people were
eparate decision about each number, whereas during recognition, they scanned making' a
all the numbers and 'made a
ingle decision. In short, the process for searching and deciding about information stored in
kind of task people face. memory varies with
Conaeopacs

orgettingjin STM_ocCurs due to decay as well as interference. The experiment by


Peterson and Peterson
emonstrated how forgetting increased with an increase in the retention interval. Nearly 10%
terthree seconds 50% after 6 seconds, and more than90% after 18 forgetting occurred
second_.
r decay. Nevertheless a more important reason for forgetting is interference.
This is often taken as evidence

ditems in STM. Interferénce refers to the influence of one task on.another task.New items continuously replace
Greater the similarityhetaleen
e two tasks mare is the interference. Retroactive interference occurs _when
new learning_interferes with old
arning. Proactive.interference occurs when old learning interferes with new learning. If dissimilar material is
roduced in an experiment on any trial, the level of retention for that trial rises to the same high level as for the
st trial. The proactive interference built up over the previous trials disappears. This is called
pactive interferenice (Wickens, 1972). release from

sner (1967) tried to


explain forgetting in STM as an interaction between interference and decay. His
theory describes the process of forgetting as analogous to the disintegration of metal when acidin a
it is placed
of corrosive acid. The degree of disintegration of the metal is a function of both the strength of the acid and
amount of the time the metal is submerged in acid. In human memory, thA degree of simia *t u
affects interference much like the streigth
1:5
id affecis corrosion nf mef; n i iatiiy 1iween
s Imore
interference. But time is
Cnory stoi àge. Ihus even also
important. Interference will be
bath immediately. highly interfering
Alternatively, dissimilar (nonitems wll do little harm greater
gtime, ith little forgetting.
the
to a new longer
works on an item
M
interfering) Items may remairn item acid
remov
in the Da
o Conciuae, we may
onlysummarize the characteristics of STM as
1. It lasts for
20-30 seconds.
a S imited storage capaçity. It can follows:
rormaton can be retained in STM store/ process only 7+2 chunks of Information at aiiy time,
.Lonmation is encoded in STM primarily in
indefinitely through malntenance
Retrieval of information acoustic.cades, reheasd
though other codes may aiso, De
6. from STM is u>
Forgetting in STM in due to both through a serial, exhaustive search.pro.
decay and interference.
LONG TERM MEXORY L
i t n i t (2)Je a.. eee l M E MIT
a n n iong-term nmemory exists indefinitely. There is no time limit in
Olduon long-term memory. TheoredcalnY
does not interfere with it, information in LTM is there forever. Even ordinary peope
O R OVer long periods of time. Long term memory is often retai
Studied the tested
memory of subiects between 17 and 74 years old, who had in natural circumstances. 5aniCK
EEKS to S/ years earlier. The 392 subjects were graduated from high scnool
CidsSmdtes. Even after 50 years, presented with_pictures_and names of ther hign schoo
people who classmate's picture mixed with other pictures coua correuy
saw a
ELOglLe, tnE former classmate over 20% of
the time. When a classmate's name was mixed with rour
aC Ogniion of the name was equally good. However, people had more trouble when asked to thinkouner the
name or the person in a picture. Thus, recognition was easier than recal, In a
tested bU0 people more reent study,
who had taken one or more Spanish courses in high school or college. They were EaiCK
reading Comprehension, vocabulary, and so on. Although these former tested the
students forgot a great deal during ror
first three years, after that their memories were relatively stable.
AeasOGTO CAPACLZ(
LM d0es not appear to be limited.in storage capacity either. It has an unlimited capacity. Even very cld people
can iearn new: things and thus add to their memory store. No one has ever become so learned that he
or she
awoke one morning to declare, "that is it. My brain is full. I cannot ever remember another
of neurons in the.brain_is.estimated to-be about 100 billion, and new connections thing". The number
being formed, indicating different and perhaps better ways of organizing memories. between them are constanty
Ordinary people can retain
extraordinarily.large amounts of information in long-term memory. Shepard (1967)
words and found that they could later recognize 88% correctly. Others correctly recognized presented subjects 540
1200 sentences they had studied to the same extent. between 600 and
People were 97%% accurate in recognizing 600 color
advertisements that they saw. Standing (1973) found that people
oictures they had seen earlier. These correctly recognized 73% of the 10,000
findings as well as our own experience sugge_ts that it must be extremely
iifficult if not impossible to reach the full capacity of long term memory,
AcqUIsition in LTM Cat,sossE.
nformation is acquired intó LTM through elaborativerehearsal. Craik and
Jaborative rehearsal extracts more meaning, and is hence a_deeper level of Lockhart_(1972)_proposed that
better memory. (However, they do not accept that there are separate processing, ultimately leading to
of single central processor with limited capacity.) Not
a memory systems, and postulate the idea
encoding of information in LTM as well. Bransford and Johnsonsurprisingly, what people already know affects the
fhosewho were not told what the paragraph was about, remembered (1972) asked their subjects to read a paragraph.
the paragraph was about) In fact the second group remembered almostfewer ideas than those who knew what
twice as many ideas as the first
rhaps this is whý we retain more information if the teacher or a book outlines the group.
preview of the information, before the information itself is plan of the chapter, or gives
er incidental
Jearning may be due to similar reasons. Bartlettpresented. The superiority of intentional learning
ast reactions and experience takes place, and thus we (1932) believed that an active organization_of
acquire a schema at the time of encoding.information.
Amajor factor in éncoding_of information in LTM is activation or
beteris the memory, This is another reason_why arousal. Generally, greater the arousal leyel
reä intentional rarninc is syperior io uncidenial iearnino. For the
Cia3 eperiences rPIated tn the c!t a 'ctdiiied very
wei Flashbulh memOIies i
DA , PArLx D, >ELJUN A, UYII I1

another kind of evidence for the superior_memory of information_encoded in a highly aroused state
a Ameritdns were asked about their memories for what they were doing at thetime ofdramatir hslolicaf
such as the assassination of John. F. Kennedy, most people remembered vividly the circumstances in vbich
had heard the news. Indeed, so detailed was the memory that it seemed that
they had stóred an icon in the
long term nmemory. Both intentional and incidental details were reported (Brown and Kulik, 1977). Nevertheless
iater resezrchers have established that these flashbulb
memories are not perfect and they are subject to tne
same processes of construction and reconstruction as other
memories
Storage in LTM(e n i sd
Information in LTM is coded in at least three ways:
aemantic, visual, and acoustic. Of these, maximum use isS
made ofgemantic codes. Bousfield_(1953) presented_people with the names of15 animal, 15 vegetables, 15
professions, and 15 people in a random order. When the subjects were asked to recall, there was a pattern to
the order in which they recalled the words, They recalled the
tarm animais, groups of domestic animals, groups of Wild words in semantically related clusters-grOups or
codedin a way thapreserved their meaning. Paivio (1975) animals etc. This suggests that words were stored or
LTM according toisual.codes. Bahrick lark, suggests that we also seem to encode information in
and Bahrick
everyday objects for 2 seconds.each. Later they were asked to 71962) showed their subjects, drawings of. 16
drawing of the object was presented with recognize.the drawings they had.seen, when the
was evaluated. When subjects 10otlier.drawings.that.looked much like it: The nature.of. the errors
made mistakes,
Seen berfore. Since. most of us live in a ültüre they tended to recognize objects that looked most
that relies heavily on visual ike they nad
etfective than semantic and visual coding. But we do store some informatiof3 acoustic codingUS less
example, musicians and linguists rely a lot on information in acoustic codes as
subjects
learnt paired associates acoustic coding. In an
experiment by Nelson and Rothbart
well. 1
returned to the laboratory and consisting of a number and words, such as 27
TACKS. One month
(I972),
identical to those they had learnt
tried to learn, other
pairs of numbers and words. Some of these later, they
entirely new. The subjects found itearlier, some sounded like pairs were
pairs in the earlier list (27 TAX), and somé
easiest
-

to learn identical were


and least easy to learn entirely new pairs, next easier to learn acoustically
simijar
OaQni sotion ()Nelauruo Aedet pairs pairs,
In the enormous warehouse that is
ftat ComanLe Hedel
LTM, information has to be
Psychologists have proposed two major types of organized; otherwise we risk
models and feature detection models. n models by which_semantic information is iosing it forever.
emember networks of interrelated concepts. network models _of semantic_ organized: network
hierarchically There are categories within categories.In some memory,people
models (Quillian, 1967), conceptsareareassumed
to

ts subordinate BIRD, and then its For example, people remember the organized
formation about ANIMALS more subordinate CANARYPSuch models suggest that people superordinate ANIMAL,
his has been easily/rapidly than BIRDS, and about BIRDS more ought to remember
supported
as been criticized
by experiments (Collins and
Quillian, 1969; Cooke et al., easily/rapidly than CANARY.
because it fails to explain why
some members of the same 1986). However, this, model
asily than the others (for example if category are evoked
have to think an
we
asily than a vulture or swan). It does not indicate how example of a bird we think of a sparrow or robin earlier/more
more
ow why at times
lower category information is evoked typical instances within a category are. It also fail[. to
ne model has also
been criticized for not earlier/more easily than higher category
w retrieval reflecting differences in semantic and information.
strategies affect performance, and not episodic memory, not explaining
explaining
how statements are disconfirmed.
ch models have been
refined over the years to take into
rious concepts. They postulate that concepts form a network consideration the strength of association
rarchical order alone. This on the basis of between
in turn explains
why certain strength of association, and. not,
in others. For
example, RED is more closely associated exemplars
with FIRE
seem to be more
typical of a concept rather
cepts is activated, the process of ENGINES than SUNRISE. When one of
spreading activation to other concepts makes them easier to
Icepts that are more strongly associated
the
with the original retrieve. Thosee
closely associated will be retrieved later (Mandler, concept will be retrieved first, and those that
1980) are
feature.compaiSon
tion
models hold that the
features sperifying a concept are arrangen iri
of a feature l his
th, Shoben and hierarchy.reilects Lhe degree to iuich. tte.faatue ierarciiy._ne posil
Rips). Defining features are
essential, isnecessary for_the. concept
whereas cliarKieristic 2aiures describe
Onceptthat are not essential. For example: The pruperti o
a an_mimiC VOIces, et. he defining featurES O EARROT arc '
charariii
D , ELIION A UViT u

re abstract and contain fewer


defining characteristics than narrower, subordinate as
OT). Also we have more
characteristics for typical concepts rather categorles
than atypical ones. ne
s ing
jor characteristic features a concept d
has, the more easily it can be
lains how lypical inslances are more easily learnt and recognized recaEd
or The model
recalled.
ecent resedrners.Delieve that humanmemory is probably organized in a way that combines aspects O o
hierarcnical ana reature comparison model (Glass ans Holvoak, 1986). Yet others have suggested divisions w
the memoy system, which
could be related to the way in which memory is
organized. Tulving (19/2)no
some ormation is dated, i.e., we can recall the time and place where it was acqured, WEreas other
information is general knowledge about the world that we acquire without any "
under which it was
acquired..Consequently, he.distinguishes betweerepisodic.and semanemories. Squire ill
(1985) distinguishes betweeprocedural andfdeclarativ memories. Procedural memory the memoy s
and
naoiS remains intact in amnesics. Declarative memorx, which con_ists offacts that we iove
EXplicit memory is
amnesia. chacter (198Z) makes a distinction'betweenexplicit antimplicit memory
areness. Implicit_memory does not involve_explicit_or conscious recollection or any prevou
expenenCe Another evidence for organization of stored material is reconstruction studied extensivery Dy
bdruet 57). Reconstruction refers to the way a subject organizes a lengthy, ambiguous pleCe or o is called
(In English this
g unkS and details that make_the memory more coherent and logical. Researcners dre O Sulc

reminiscenGe, though the term has an altogether different meaning in psychology).


takes place at the time of retrieval.
netner reconstruction occurs during storage with consolldation, or it
Retneval from LTM
LTM is largely with the help of retrieval cues.
Retneval of information from the huge_warehouse that isthe-goal
Rerieva.Cties prompt people remember by specifying
to of memory search, and by directing uneir
but not accesSIDIE (as
may be available in memory
generation and evaluation of possible answers. Informatiorn for the search. Sometimes the cues
happers during blackout in exams) because cues are not specified
memory
and this may result in_retzCLanhibiion
Specitied may not be appropriate because they are not specific enough
and extract information in a circular way.
orblocking (Stern, 1985). Then people use secondary retrieval cues

The importance of retrieval Cues in memory is attested to bya the encoding specificity hvpothesis (Tulving and
encoded, foor
Thompson, 1973). Whenever a person experiences an event, number of aspects of the event are
actions during the event, and sã on. The
exampl, the time,Jocation, features of the setting, people involved,
encodings. Evidence for
hypothesis saysthat retrieval cues willbe effective to the extent that they match these State dependency (Eich, 1980)
this hypothesis comes from studies of state dependency and context dependenCy,physiological state as they were
is the fact that recall improves when people try to recall information in the same
intoxicated is better recalled when
when they first learnt the information. Thus, what people learn while they are remembered under the influence
better
they are intoxicated; what they learn under the influence of marijuana is
of marijuana; information learnt standing or lying down, is bétter recalled standing and Ilying down respectively;
and information learnt in a happy or sad mood is best remembered in the same mood. Context. dependency
efers to the fact that memory is also influenced by contextual cues from the environment. Godden and Baddeley
1975) used 16 membersof a diving club as subjects. They demonstrated that when lists of words were learnt
they learrnt under water, they better recalled
n land, they were better recalled on.land and if were were

inderwater.

ince LTM is unlimited by the boundaries of time and space, researchers contend that the retrieval of information
om LTM has.to be through a parallel, self-terminating search process. It would be the only efficient research
ethod.

t times, there may be a retrieval failure. Everyone forgets things. But is the material really gone from memory
is it just unavailable? Though many theorists from Ebbinghaus (1885) onwards believe that strength of
emory decines with the passage of time, there is hardly any physiological or experimental evidence of decay in
M. In fact findings contradict the idea of decay. For example, Ballard (1913) found that a poem was recalied
tter after two days rather than immediately after learning it, an effect called reminiscence. Payne (1987) found
idence for hypermnes/a when he found that subjects repeatedly tested on the same material, may recall
sterial on later tests that they did not recall earlier. These finding suggest that a consolidation of the trace
2bably occurs with time. ForgettinginLTM.occursmainly dhue to-interference proactive as weil as retroactive
foRGETTINNG
BA 11, PArER D, StLITUIV A, UINII I

Interference increases duetosimilarity ofthe twotasks,.greateramountoflearning, ets, LTM being semaniaAIca


organized, release from proactiveinterference, demonstrated in STM, may have even more relevance in LTM

To conclude, we may summarize the characteristics of LTM as follows:


1. There is no time limitin long-term.memory.
2. It has an unlimited capacity.
Information is acquired into LTM through elaborativerehearsal.
Information in LTM is coded in at least three ways: semantic, visual,
.
and acoustic. Of these, maximum use is
made of semantic codes.
5. The information in LTM is stored in.an.organized manner.
6. The retrievl of information from LTM depends on retrieval cues and has to be_through a parallel, self-
terminatingsearch.process.
7. Forgetting occurs due to interference - proactive as well as retroactive. Interference increases due to
similarity of the two tasks, greater amount of learning, etC.

ARE THERE TWO MEMORY SYSTEMS?

The theories of memory are embedded in the information processing viewpoints, which established their
importancein psychology in the 1960s after the advent of computers. This viewpoint held that the human brain
s structured and functions like a computer. (Two different aspects of information processing have been
emphasizedin recent theories; (structure and process Eysenck, 1978). Those who emphasize structure have
USUally suggested that the nature ofthelnformation processing system imposes limitations on the rate of flow of
information through it, and that there are separate memory stores to deal with the amount of information in
each stage of memory. Those theorists stressing the importance of process have concentrated more on the
richness and variety of encoding and processing strategies, which can be applied to incoming information.yThese
researchers contend that thinking in terms of a single processor is a simpler, more parsimonious explanation.

There are many varieties ofthe modelsthat focus on structure. They are variously called as STM-LTM model or
the duplex theory or the multi store model. It IS assumed that information is initially held in a modality specific
sensory store, and unless attention is paid to it, it is rapidly lost through decay. Attended items are passed, on to
a limited capacity short-term store. The items, if rehearsed, are passed on to a semi-permanent long-term store.
According to Waugh and Norman (1965) every item that is perceived enters primary memory. An item will be lost
from primany memory unless it is rehearsed, which can be overt or covert, intentional or unintentional. If
rehearsed an item may enter secondary memory. This is considered as a more
permanent store, where
rehearsal is not required. The capacity of primary memory is limited (Glanzer and Razel, 1974). They postulate
two memory stores, one for materials to be retained for short periods of time and one for materials.to be
retained for longer periods of time. From the theories of Broacbent (1958, 1971), Waugh and Normari (1965)
and Atkinson and Shiffrin (1968), it is possible to construct of 'Modal Model. The model that is most.often
referredto and therefore sometimes called the modal model"- was one proposed by Richard Atkinsön and
Richard Shiffrin (1968)
.

The Atkinson-Shifrin Buffer model (1968) introduces a new store, called the 'sensory register, the main
characteristic of which is that information stored within it decays in a 'very' brief period of time. Stimuli from the
environment first enter sensory memory. Sensory memory is a large-capacity storage system that records
information from each of the senses with reasonable accuracy. Although touch, smell, and taste can be also
represented in sensory memory, cognitive psychologists are especially likely to study iconic memory (visual
sensory memory) and echoic memory (auditory sensory memory). In any case, information in sensory memory
decays rapidly.
f encoded, the information moves from sensory_register to short-term store. Short-term memory (abbreviated
SM) contains only the small amount of information that we are actively using. Atkinson and Shiffrin proposed
that verbal information in STM is encoded acoustically in terms of its sounds. Memories in STM are fragile -

nough not as fragile as those in sensory memory and they can be lost from memory within about 30 seconds
uhess they are somehow repeated. Thus, unless rehearsed, the information is thought to decay from short-term
store in a matter of sconds. Rehearsal may contribute to its transfer to long-term store, where it is stóred:quite
permanently.
.'
D, CLIUN A, UVI II

1 0 to the model, the


. material finally
(abbreviated LiM)
ories that arrivea
has a passes from
large capacity and short-term memory to long term -tern
antically,tn terms orSeveral minutes ago. contains
Atkinson and Shiffrinmemorles that are decadesmemory
old,
ts
meaning. Memories in LTM are in addu to
oroposed that information in LTM
relatively permanent, and are not IS enco0u
THE STM- LTM MODEL ADAPTED FROM iEDy
'.. ATKJNSON AND SHIFFRIN (1971)

SENSORY STORES SHORT TERM STORE LONG TERM STORE

Iconic
Temporany Permanent store
Environmental working memory Not subject to decayy
input
Echoic CONTROL
PROCESSES
Habtic
Rehearsal
Coding
Decisions
Retrieval Strategies

RESPONSE OUTPUT|

Suppose, that a friend is reading a magazine article on food in China, and she relates to you an unusual fact:

In Eastern China, a featured item on the memo is deep-ftled scorpion.

n-terms of Atkinson and Shiffrin's model, the words in that sentence would first be registered in the auditory
tore of your sensory memory. That information could be lost, or it couldbe transferred. In most cases invoiving
ransfer, the information would pass on to short-term memory, but that information could possibly be also
ansferred from sensory memory to long-term memory. Suppose, hovwever, that the sentence about the deepP
ied scorpion does arrive in short-term memory. Onc cptionis that this information may be lost from short-term
emory. The other option is that it can pass on to long-term memory. Loss from long-term memory is less likely
an in the other. kinds of memory.

formation in long-term memory can pasS back into shórt-term memory when we want to actively work with
at information again. Suppose, for exafmple, that an evening's conversation has drifted to the topic of deep-
ed scorpions and you wish to sháre your knowledge. You can retrieve this useful information from your
latively inactive long-term memory and bring it back to short-term memory.

far, we have examined the model's structural features, which are the stable memory stores used during
ormation processing. Atkinson and Shiffrin also proposed control processes, which are strategies that people
flexibly and voluntarily, depending upon the nature of the material and their personal preference. One
ortant kind of control pracess is rehearsal or the silent repetition of information that encourages it to recycle
ough short-term memory,)For example, you may silently repeat information about deep-fried scorpions in
er to recal it. later. According to the model, information that is rehearsed frequently and kept for a long time
hort-term memory is more likely to be transferred in long-term memory.

trol processes can operate in other ways in memory. For instance, people can decide whether they want to
heir short-term memory with material that needs to be remembered or to leave "work space" to think about
ething else Furthermore, they can decide whether to use a particular memory strategy - such as a mental
Jre- to encode that sentence about
deep-fried scorpions.

*** *********;"*****

.
UiNII I1
SELIIUN A,
PAPE D,
DA ,
follows:
LTM distinction is as
evidence for the STM
-

The and LTM. The


differences are tabulatod.
ed belo
STM
earlier, the characteristic
features dilfter in
1. As detailed LTM
STM
a. There is notime limit inJong-term memory.
a. Itlasts for only 20-30 seconds. It can store /b. Ithas an
unlimited capacity.
b. t has limited storage capacity.
information at any.
process only 7+2 chunks of
time. Information is acquired into
LTM through
Information can be retained in STM indefinitely c. elaborative rehearsal.
through maintenance rehearsal in at least three
Information in LTM is coded
d.
Ihformation is encoded in STM primarily in d. and acoustic. Of these, visual,
acoustic codes, though other codes may also be | ways: semantic,
used. maximum useis made of semantic codes.
stored in
organizing information, becausee. The information in an
LTM I
need for
Lrttie
T it is not a permanent store. organized manner,
f. Retrieval of information from STM is through a The retrieval of information from LTM depends
serial, exhaustive search process. on retrieval cues and has to be through a
parallel, self-terminating search process.
9.
Forgetting inItSTM in due to both decay and g. Forgetting occurs due to interference
interference. lasts for only 20-30 seconds. proactive as well as retroactive. Interference
increases due to similarity of the two tasks,
greater amount of learning, etcC.
2The evidence from_the serialposition.efteatalso supports the STM - LTM distinction. The serial position effect
is the fact that the
first few items in a serial task are recalled the best (primacy effect), followed by the last
few items (recency effect),
and the middle items are recalled the least. The usual explanation for the
T effect is that the first few items are better rehearsed and primacy
the middle or end of the list. This are thus more likely to enter the LTM than items in
explanation is supported by two findings of Murdock (1962): (1) if the
items are presented very fast and are therefore less likely to be rehearsed, the primacy effect is diminished.
(2) longer the list, lesser the primacy effect. Similarly, the
recency
items are still in the STM. This idea is further supported by the fact thateffect_is explained by the idea that the
seconds by asking the subject to perform mental arithmetic then the if the recall is delayed for about 30
Phillips, 1965). recency effect disappears (Postman and
Evidence regarding kinds of codes: In one influential
study, Kintsch, and Buschke (1969) asked people
learn 16 Engish words in order. They proposed that the to
LTM when recall was reguested because so words from the beginning Of the list would be in
hand, the most recent items should still _bemuch time had passed since they were presented. On the other
in STM. Their study focused uipon one
is coded in terms of its acoustic or sound distinction
theorists had proposed: material in SIM that duplex
material in LTM is Coded-in-teFms-ofItS_semantic or meaning characteristics whereas
whether items at the beginning of the list which were
characteristicS The first study examined
semantic factors. The second study examined whether items presumably in LTM
-

would be influenced: by
at the end of the list which were
in STM-would be influenced by acoustic factors. presumably
SYNONYMS HOMONYMS
Angry Tacks (Det make
Pleased 0 menam)
Forest Buy
Sofa
Owe
Ocean Tied
Woods Sew
Carpet Their
Sea
Tax
Happy By
Rug There
Mad Oh
****** Couci . . . ..-.
Tide
st list contained pairs of
synonyms, which are words
wial had been presented, the that are similar to
cipants were requested to
supply
experimenters
the supplied one word from the
each other in
meaning. AeT
pose tnat a person.contuses the next word in the list, for example,
,ord ug às the answer, because word list,
pleased with Its.synonym The correct answer would be forestplease.
nis kind of, semantic confusion rug follows happy. Howev
that occurredhappy. Kintsch ana buschke Then thistneperson might SUppiy u
begiiing o ne ist produced a for items In each measured numper or
result suggests that greater number part of the list. They found instances
items at the of semantic confusiorns than items at the tnate e
meaning beginning of the list, which should be in end of the
list.ns
LTM, are coded in terms or thei
The
secona uist contains pairs of
two words that sound the same,homonyms, which are words
then he or he might see similar to each other in sound. If a person conruse
conrused witn se, which the
before their. Kintsch and word so and respond ther because
were
more tikely at the end appeaied Buschke
should be in STM, are coded in(1969)
of the list, which found that acoustic
terms of their souna. conrusiot
Mcal evidence may also be
pauent H.M; whose hippocampus used to support the STM -
LTM distinction. Milner (1958) cited
the a
in that area. The patient could case or
was removed
remember,old material, and his measured IQ remained because of lesions
an atempt to cure H.M.'s serious the same, but he culd not therearter
acquire. new material. ):
epilepsy,
nippocampus. The operation successfully curedneurosurgeons removed portions of his temporal
H.M, Can the epilepsy, but it left him with a severe lodes and
accurately recalls or kind
surgery and his short-term memory is memory
events that occurred before his ios5.
However, he cannot learn also
retain any new information. For
or normal.
example anyone H.M. meets on a Monday woula
not look tamiliar on a Tuesday. Furthermore, he cannot
his recall more than six numbers in
short-term memory is normal but he lacks the ability to transfer
term memory.
order, suggesting that
material from short-term memory to
(Squire, 1987). The case of a second man,known
long
symptoms. K.F. had been in a as K.F. suggests roughly the opposite
His motorcycle accident,
which damaged portion of the left side of his
long-teri retention is normal, but his
short-term memory is severely limited cerebral cortex.
Similarly, patients with Korsakoff's syndrome
(due to chronic alcoholism) cannot(Shallice and 'Warrington, 1970).
they can remember events and experiences that remember the recent past but
say that LTM is intact but STM does not function inhappened long ago. Duplex theorists interpret these findings to
these patients.
The main criticism of
duplex theory comes from the
1972). They hold that we do not have separate depth / stores. levels of processing viewpoint
by Craik and Lockhart
ingle limited capacity processor. The rate of memory Rather memory is a process. We
forgetting or strength of retention depend on the possesssa
vhich information is processed when it is depth/ level to
acquired.
1eaning when the experience is being encoded. They Deeper processing implies that the person extracts more
1. The duration of STM, argue against the STM- LTM distinctions as
which is its defining feature, can be increased follows:
informationeven at the shallow level of maintenance rehearsal. indefinitely if we continue to process
2. The fact that memory limits
itself to 7t2 items is due to limited
processing capacity rather than limited
storage capacity. This is why chunking helps to process more
be changed, but it is hard to information. A process being
decreased.
conceptualize how a static structure as a store can be dynamic can
3.
increased or
Acoustically encoded information is not retained well
semantically encoded material is processed to a because it processed at a surface level. In contrast,
remembered. deeperlevel and is hence better retained and
4. The STM - LTM distinction
cannot be based the
difference between serial, exhaustive
on
parallel, self terminating search, because these processing differences search and
task rather than differences are dictated
by the
between two stores. The nature of the
search for recall, and a
serial, exhaustive search for person probably uses a parallel, self-terminating
5. Forgetting from both STM as well as LTM is recognition.
feel that the difference dependent
between STM and LTM cannot be interference. Thus the single process theorists
on

serial position effect also does sustained on this basis either.


6. The evidence from the
not support
Lockhartargued that the primacy effect occurs because the STM LTM distinctions. Craik and
the initial items are
whereas the recency effect shows
data, in thís respect, was that the final items receive processed more deeply,
only shallow
reported by Bernbach (1975). He found that if processing. More provocative
the list will be, the the subjert! is not told how
recency effect disappeans iong
completeiv. Further, if the apciiiener
controls the Gder
DA I1, PAPCR D, StLIUiN A, UINII 11

of output, then too the serial position effect


disappears. Thus he argued that when the subieci.
know the length of the list, and/ or the order of output, he cannot subject dkes
adopt different processing
Consequently, ha processes all items in a similar manner, and the serial position effect SLat
is not show
7. The clinical evidence from amnesic
patients interpreted on the basis of laboratory
is
hippocampus is involved in attention (Hendrickson et al., 1969). Thus Single process research hat
amnesic patients do not learn new things because theorists feel th
tney cannot pay adequate attention to tivern. This
supported by the fact that amnesic patients may have lost their declarative memory but their ic
memory is intact (Squire, 1985). Also, they perform procedural
impairment on implicit learning tasks (Schacter, 1985).poorly on explicit memory tasks but show no

The Current Status of the


Duplex Model
Levels of processing, being a more
today. However, all researchers acceptparsimonious explanation, is therefore accepted by many more researchers
the existence of a
between short-term memory and long-term memory revealsseparate sensory. register. Research on the distinction
mixed results. Some studies support the distinction
and many current models
feature separate memories. Research on information
periods, frequently focuses on the limits stored in memory for brief
of memory. In contrast, research on information stored in memory for
longer periods usually examines
research suggests that topics, such as autobiographical memory, encoding, and mnemonics.
memory processes are similar whether the information is stored in Other
seconds or for many memory for several
years. Many current cognitive theories still include a basic
memory and long-term memory (Estes, distinction between short-term
is too simple. For 1991). However, most of the models acknowledge that the duplex model
examples some proposals suggest that short-term
limited capacity. Instead, it is is not a single storehouse with a
memory
probably a collection of temporary storehouses (Squire, 1987).

LEVELS OFPROCESSING
Craik and Lockhart proposed the
influentialin the area of human levels-of-processing approach in 1972. Their paper has been one of the most
prior to 1980. The approach proposes Roediger (1980) pointed out that it had bean quoted at least 700 times
memory.
that deep, meaningful kinds of information
permanent retention than shallow, sensory.kinds of processing (Craik, 1979). For example processing lead to more
words when you judged a word's
meaning (for example, whether the word would at in a you should recall more
its physical appearance (for example whether it is in sentence), rather than
capital letters) or its sound (for example, whether it
with a word). [Because of its emphasis on whether the rhymes
depth-of-processing approach. In general, then, the more processing is deep on shallo_the theory is called the
greater the depth of processing. meaning _a_ person extractsfrom a stimulus, the

The depth / levels of processing viewpoint is rooted


in.the.criticism.af.SIMLIM theory. Craik and Lockhart
argued that acoustically encoded information is not retained well because it processed ata surface level. In
Contrast, semantically encoded material is processed to a_deeper level andis hence better retained and
remembered.|They reinterpret the evidence from the serial position effect alsoTiTéy hold that the primacy effect
OcCurs because the initial items are prgcessed more deeply, whereas the
recency
tems receive only shallow processing.he clinical evidence from amnesic patients..is effect shows that the final
reinterpreted-onthe-basis
aboratory research that the hippocampus is involved in attention (Hendrickson et al., 1969). ThuS single af
process
theorists feel that amnesic_ patients da not.learn_ new things because they cannot _pay adequate attention to
them. This iS Supported by the fact that amnesic patients mayhave lost their declarative memory but their
Procedural memory is intact (squire, 1985). AISO, they.perform_poorly_on explicit memory tasks but show.n0
Mpairment on implicit learning.tasks (Schacter, 1985).
reik and Lockhart (1972) proposed that people cananalvze.stimuli.at.anumber.of differentlevels. The shalow
evelsinvolve anaiysis in terms of physical or sensory characteristics, such asbrightness or pitch. Thedeep levels
nvolve analysis in terms of meaning When you analyze for meaning, you may think of other, related
SSociations, images, and past experiences related to the stimulus. The by-product of all this analysis is a
pemory trace. |If the stimulus is analyzed at a very shallow level (perhaps in terms of whether it had capital
etters or whether it was printed in red), then that memory_trace will be fragile and may be quickly forgotten.
owever, if the stimulus is analyzed at a very.deep.levelCperhaps in terms of its semantic appropriateness in a
ntence or in terms of the meaning category to which it belonn«) thar that meineory tracn will b da
eu.|
cls, D, >CCIION A, UiNIT 11

or
hvpotheSIs
emerging from Craik and
TOduce better recallhis Lockhart's
about three umes as IKely hypothesis has been widely(1972) paper was that
to tested, deeper levels of processi
Craik and Tulving
hey had onginalyY recall a Word if they had
answered
Overed that people who guestions about the word's originally answered questions(1975) fuund that peo
C

ter on a
tier
made semantic judgments abOUE IES Incd oN
physical appearance.JSindtly
surprise recall test about a word's category or its synonym n
number or voweis Contained in a than did people wno
mage nonsemanuc peo
employed to investigate the Levels word or whether it had been printed onlv uggements (ior exarnple, adout
of in capital letters). A bypical
superior recal ror items processed Processing theory is the incidental learning paraaig
(EYsenck, 19/4: Hyde & Jenkins, 1969). deeply compared to those items paradigm. ResUs evea
proauces Reviews of dozens
processed at the more
snauo
Melkman, nigner
recall scores than shallow of studies conclude that
1987). processing (Baddelely, 1990; Horton anddeep processing9e
Mills, 1984, KOriat
ar
Craik and
Lockhait(1986) beljeve that deep levels of
Valsuincuveness andelaborationDistinctiveness means
processing encourage reca fa tors:
i ) . For_instance, people recall words with thata stimuluS is different from all otnermemoy
pKnakL and atghan, better than words with common_ distinctive sequences of short and tall letterS, Sudi s
alrway (Hunt & Eliot, 1980). The second factor that orthographic sequences, such as eakY KEDNe ana
operates with deep levels of processing is
invovesnch processing in terms of meaning (Anderson & Reder, elaborationWnicn
1979; Cohen et al., 1986)
he proSS of cycling information through memory is
rehearsal, Craik and Lockhart
enears aintenance rehearsal merely repeats the "Knd of analysis that has alreadyproposed been
two kinds or
Contraslaborative rehearsal involves a deeper, more meaningful analysis of the carried Out. in
Word bOOK, You could use stimulus. Thus, IT you see the
maintenance rehearsal and simply repeat the sound of that word to
other hand, you couid use elaborative
rehearsal
yourself. Dn thne
to another word on the list. by thinking of an image of a book or by relating the word boOk

What will happen if you spend more time rehearsing? Cr


and Lockhart (1972) predicted
that the answer to this
question depends on the kind of rehearsal you
using. If you are using shallow maintenance rehearsal, then
are
increasingréhearsal time will not influence later recall. Simply repeating the word book five more times will
make it any. möre memorable. However, noot
if_you are using deep elaborative rehearsal, then an increase in
rehearsal time will be helpful. During that time, you can diq out all kinds of
memories to enrich the stimulus, and later recall will be more accurate. For example, extra images associations, and
strong visual, images or many associations with existing knowledge ill be information that involves
information that is being attended to receives more processed at a deeper level. Similarly,
processing than other stimuli/events. The theory also
supports the finding that we remember things that are meaningful
than meaningless stimul. Research on elaboration also emphasizes toa us because thisinrequires more processing
wili recall more if you elaborate on the material. practical point: studying for a test, you

Processing of information
e
at different levels is unconscious and
automatic unless we attend to that level. For
xample, we are normally not aware of the sensory properties of stimuli, or what we have
nless we are asked to specifically identify such information. This in working memory,
ir suggests that the mechanism of attention is an
nterruption in processing rather than a cognitive process in its own
right.
Much of the research on levels of prOcessing has focused
The encoding and. rétrieval, and the explanations on face recognition, the compatibility between
dyri 2viewed belae: for the effectiveness of deep processing. Some such research is

Depth af ProcesÍng and Eace


is2cognize someone with whom Recognition:
Most of us can recall embarrassing incidents where we failed to
we had interacted for many hours. For instance, a student in a class
rMichelle recalled.how she had taken named
dancing
each week for two years. One day Michelle saw the
lessons with another female student,
totaling about three hours
other student in a shopping mall and did not
Michelle had apparently failed to use deep recognize her.
Mow processing to encode her face; with different clothing, that face was
unrecognizable. Research has shown that shallow processing of
ads to p0or récall. For instance research faces, like shallow processing of words
Make judgments aboút whther a person is participants
honest,
recognize a greater number of photoS of faces if they
i SOn S nose (SDorer rather than the gender ot c person or the widii of tie
19aT Bloom and Mudd (1CC1) üvi provideui dlt a{pDeaing evnlanation for
thie**i

. -.
UiViI II
PATX D, SCLIIUY A, whetrer à persor ViEs
to judge
DA 1,
instructed
who had been instn who had ueen
demonstrated that people to people
Theirresearch mIovements as compared
made more eye processing ie2ds to enco
at the faces longer and 1hese authors argue that deeper
was male or
female.
whethe a person and therefore superior recall. Alternately, hen people mare Chardcter judgerme
of features, (Wells and HryCiW, 1984)
in terms of isolated features
number
faces holistically, rather than
encode the
and Craik (176)prODOSEG that retrieval congi
ncoding and retrieval: Moscovitch and C
OIAGODILS xducen in order for deep processing to be highly
effective (MOScovitch

SpOUK dUplicate encOding conditions better


on a riyming tesi
In support, Bransford and his colleagues (1979) found
that people performed
1976). rather than sentence encoding task.
Shaillow acoustc
they had originally performed the rhyming encoding task,
when memory is tested by aSKing wnether any
processing9 may be better than deeper semantic processing, the encoding specifidity nypotnesis which
words on the list rhymed with toy. In support, Tulving (1972) gave
Says nat retrieval
yponesiS
cues will be effective to the extent that they match the encoding cues. Evidernce for his
comes from studies of state dependency and context dependency. State deperGency (Eicn, 1980)is
thefact.that recallimproves when people try to recallinformmation in the same physiological state as tre ¥ere
When they first learnt the information. Thus, what peoplelearn while they are intoxicated is better recaied wiren
e y dre ntoxicated; what they learm under the influence of marijuana is better remembered under h e int+uence
mariuana information learnt standing or lying down, is better recalled standing and lying down respectively;
and intormation learnt in a happy or sad mood is best remembered in the same mood. Context dependercy
rerers to the fact that memory is also influenced by contextual cues from the environment. Godden arnd Bacaeey
(1975) used 16 members of a diving dub as subjects. They demonstrated that when ists
onlandthey were better recalled on land and if they were learnt under of words werE EEnt
underwater water, they were better rEcalle

he SelfReference Efec: We often deal with new information by relating it to


taking
a course in
abnormal psychology often suffer from "medicai students' ourselves. For example, stidents
disorders seem to fit themselves! The professor describes how syndrome" most psychologica
-

future, and suddenly dozens of


a
depressed person feels pessimistic about the
depressed. students are wonderingif their ow pessimism means that they are ciinicaliy
This personal framework for new information
Specifically, the self-reference effect points is an important topic in the levels-of-processing approach.
that information to themselves. In the dassic out that people recall
demonstratíon of the more informatton when they tryto raiate
Kirker (1977) asked participants to process lists of words self-reference effect, Rogers, Kuiper,
according and
leveis-of-processing research, that is, in terms of their
to the kinds of
instructions usuaily studied in
characteristicc, or their semantic (meaning) characteristics. physicaB characteristics, their acoustic (sound)
of self-reference:
People were asked to decide whether a However, other words were to be processed in terms
self-reference task produced the best recall. pa:ticular word could be applied to
ourselves, we develop a particularly memorable Apparently, when we think about a word in themselves. The
connection witn
to decide whether the
word coding for that word. For exampie, suppose that you are trying
:o a friend who had missed generous applies to you yourself. You might remember
how
ciass, and you shared a box of loaned your notes
generous does apply. The mental candy with the other peopleyou in the iounge
processes invoived in the
recalled. The self-reference effect has selif-reference task seem to increase the chances yes,
-

an item will be
nstructions to create mental been demonstrated that
vith imagery (Brown, et al., 1986), with words related to repeatedly, for example with
paragraph-long prose passages (Reeder et al., creativity (Katz,
Oung as 10 years of age (Halpin et al., 1987). The self-reference effect aiso works with 1987), and
1984), as well as elderly
adults children as
(Rogers, 1983).
Vhy should we recall information
af is treated as a
especially weil when it is related to the self?
rich and
organized set of internal cues to Belezza (1984)
hlstrom (1986) focus which information can be suggests that the
ey suggested that when theespecially on
organization as associated. Kiein and
task instructions asked explanation for the self-reference effect. In
an

word was printed inparticular,


pital letters, the
participants inspected the words one people
gether. When the task instructions
to judge whether a
at a time and were Small or
not inclined to
ikely to organize the words asked whether a word fit the organize those words
together. However, meaning of a sentence,
people wêre
ords applied to themselves, it when the instructions asked
people to think about :hether
a'so
ICh other, because might have been tempting to consider
they were all applied to the same how some of these words were
relaled to
- IGCa iiiai iiie seil
iS SomehOw
per sor Furcher, sornei
unini:e in
it ai, N.cust uu: iei itiiV i V
hat the self
worksIKe
reference task
any other
knowledge
structure, but it is simply
encOuragesespecially
greater organization,and the self deep levels of proçessing, a very rich source of
ideas. may
may serve as a rich sourcc of self-reference instruCuo
rent státus of levels of processing: asOCiations.
vchologists appreciate
during learning. But the hoW the
approachlevels-of-processing approach emphasizes the menua
independent assessment of depth also has drawbacks. Firstly,circularity exists We
OCCur
have no
will be better, Ihen we (Nelson, 1977). Specifically, we say that if in the theory otel
say that because the retention processing is deep, nen re
circularlY [eSearGhers heed a.measyre was better, processing
of depth-of-processing in must naVE
product ne_Destretention. Secondlleep processing is not alwaysadvance that can predict wnidan cditions will
the way memorY Is tested. For better, because performance dEe oon
Processing, wnen memory is testedexample, shallow acoustic. processing may be better ntic
by asking whether _und
any words on the list rhymed win Loy DE
Nevertheless, most recent
Bransrord psychologists agree with the
(1979) who suggests that it is not only levels of processing approach.butIt has Deen
how the.information
processed, is nie
accessed. when the
demands for accessing information more how tne inioG
closely match the methods used to eldooa
E rormation, more is remembered. Then the connectionistic
model proposed by Rumelhart ana
MClelland (1986) is one of the dominant forms of current research in cognitive
wIth the most psychology and is consise
recent brain research
IS stored in
(Scientific American, 2000). This model
emphasizes the fact that informauon
multiple locations throughout 'the brain connections. It is consistent witn
in the form of networks of
the levels-of-processing approach in that the more connections to a single idéa or concept, the more likely It is to
be remembered.

MEASURES OF MEMORY
THE SUBJECTS

Primarily because of convenience, subjects in most memory experiments have been college students, usually
enrolledin an introductory course in psychology. Tt i5 debatable how Tar they represent the general human
population. Even those who 'volunteer find the task rather tedious. They would like to finish and escape as soon
as possible.They are also_under stress because they dónot want to appear dull. Thustheir motivation is quite
high. Cognitively also they are rather well prepared for the task. It has been estimated that a student who jõins
college as a fresher has read about 60 to 70million words. He lives in a virtual deluge of words, and would have
read at least 100 million words by the time he graduates. Nevertheless he is certainly more representative of the
hyman p0pulation as compared to the other most popular subjectin psychology-thewhite rat.
E MATERIALS
It is axiomatic that the learning material is carefully selgcted in memory experiments. The material may be
classifiedalengtwo dimensionsterba/-nonverba/ andTheaningless-meaningful.The following figure gives
examples for each kind:

VERBAL NONVERBAL

MEANINGLESS Nonsense syllables Abstract drawings

MEANINGFUL Words, sentences, prose or verse Drawings of real objects

********."*****i***:*******
SCLIIUY A, UiVI I 11
DA 11, PAPrK D,

yverbal máterial fundamentalLuNits of language .

learnt by definition, made n_of the materials mav


The verbal materials_to he
are,
GenerallneaninalessOrna learning
numbers. hese do not necessarily form words. Ebbinghaus (1885)to coñtfol.the fac or
nonsense syllables were invented bY
The srings orletters tnat nave mcaninng.
words. no
nornsense sylables or syiabies are
memory. iNonsense
meaningfulness in his stuaies or nis own and is_theretore.alsa.called a CVC.
EDbingha
consists of a vowel between bwo consonants,
A
SYlable generally nonsense sylable_ consisting of all, possible
used CVCS research, constructing a
in his poolof about2300 be a consönant,
restriction that the first letter
permutations and combinations of consonants and vowels with the
advantages of the nonsense syllables:;
vowel, and the third a consonant again. He saw threenonsense
the second a with syllables were relatively uniformin
Compared selections of real language prose or poetry,
diffculty. to draw similar lists.
"Ihe 00 ttem pool provided a large number of comparablecombinations from.which
"ne sylable being a discrete,_precise unit, the amount of material learnt could be measured and varid
eficiently
it is now quite clear that nonsense syllables are not "meaningless" if we define meaning as the number or
associations a string of letters can evoke. A CVCsuch as BAL is likely to be learnt more easily than TA, because
tne rormer evokes associations more easily. Indeed there have been numerous attempts to measure the
meaningrurness of nonsense syllables by studying tine number and kind of associations they evoke (Glazë, 1928;
Krueger, 1934). The use of nonsense syllables in experimental research gradually decreased over time as it a s
realized that their
most favorable characteristic uniformity -was a myth. As compared to
-

meaningru Words,
they are more difficult to learn, and their use unnecessarily made the task more difficult and demotivating.ror.Une
SUbjects. Moreover, the association value of wordswas also
measured and it,
assoCiation values, and hence similar in meaningfuBness cOuld be used in verbal was felt that words With simlar
learning experimenis.
Meaninafl verbal materials.may be used in the form of words, sentences, or even whole
verse. Bartlett (1932) pionecred the use of meaningful material such as passages of prose or
experiments. When subiects are asked to reproduce the material, passages, pictures and stories in.memory
they make a variety of reliable errors. Three
major.kinds of inaccuracies that creep in their reproductions are:
Leveling: omission of extraneous informaticn
Sharpening: exaggeration of relevant or important information
Assimlation: changing the information to mirror the subject's own experiences, culture,
or vocabulary.
The meaningfulness of words is defined as the number of
associates generated in response to the word. Galton
(1879-1880) had introduced the measurement of free associations to words.
meaningfulness words was more commonly measured on the basis of their However,
of the familiarity and hence
language. Thorndike and Lorge (1944) published such a count in 1944. Sincefrequencymost
of occurrence in
English
used this source. then, experimenters have

With all kinds of verbal materials, there are two


characteristics.that imake them vary in the ease with which they
may be learnt and associated with each other: theiinternal structure and theirelations with ether_items (intra-
ist similarity). The internal structure of individual letters is determined
form them. The combination of the letters BEK forms a by the way in which features combine to
good syllable in English. There is no English word like it,
but it is easy to pronounce and has the feel of an
English word. XEF on the other hand, is a rare combination of
letters and most people would have
difficulty in pronouncing it. Intra-list simlarity
interferencefrom similar material and hence retards memorization of nonsense syllables.increases the possiility of
For meaningful words,
however, it may be an advantage, for it helps the subject to chunk similar words together.
Monbercal material
Nhen the recall of animals or infants is studied in the laboratory, non-verbal tasks are often used.
ubject is shown some interesting object, _which-is-then.hidden while he is Usually, the
e Is permitted to go and get it. If the subject remembers where it is hiddenwatching and after an intervalof_timne
0 to the place immediately, thus and is properly motivated, he will
demonstrating regah for that time interval. The nonverbal material in memory
xperiments on adults consists of figures or drawingsMeaningful
obiecte found in the real wGrid. Bahrick, Clark arcd BarriCK nonverhal material implies the use of diawing9s
a CIUnShairs afinheir xaMeii i i s i wihether
(1967)uSed drawings of GixtCC CVtiyiay viEl,
subier is COuliIecognize i 2t
test, each drawing was shown
toll, When the
e abiects
obje that iookedmistakes made byalong
the
with ten
other drawings
most ike those subjects that looked
looked much like the
they had seen wee analyzed, jt was found that ike
drawing
n nonverbamaterial
ninglessS,
tne reai wOra. consists of abstract
SChacter
before. they tendedto
nsense drawings or
figures. tuales show that (198/) conducted
ecur
in the
real wOrid. iested with sensory memory extensivenonsense figures do not depict any object
experiments on that
abserved (Schacter, Cooper, and pictorial stimuli, nofunctions for ecologically valld
sensory me
memory with such
Delany, 1990). evidence for priming of stimulL only 1.e stimuli
t sthat
HE EXPOSURE TIME structurally p is
Besides the
learning material, another
IS Sow tO uue factor of
Words gave a subject. Ebbinghaus significance in the
quick glance". Some (1885) wrote each nonsense experiment is the time ro s
dran dt not ony controlled the experimenters used syllable on a
separate card, and own
Racent investigators exposure time but alsotachistoscopes, but most preferrea to
use maintained precisely the same order for
computers and an
interactive.program a suuhects.
AHÉ TASKKS
.

for
experimens
Serial learning
Serial learning is the method used originally
De presented to the subject. The by Ebbinghaus in his studies. It involves
ney are presented, When he sees thesubject is to recall not only the items themselves,making
but
a list of tne itenso
second, he responds with the third first item, the subject has also
to respond with the second;the order in wnic
and so on, wnen ne sees u
nis requirement of correct order in the
effect and the obscurity of the functional respCiise of the subject produces two characteristics
Both are little understood despite a large the serial
-

position
research.dhe serial position effect is the stimulus.
nearly amount of relevant
most rapidly, the last few a little les rapidly, andinvariable tendency of the subjects to learn the first few items
ofthe functional stimulus refers to the items just beyond the middle, the least
fact that it rapidly>The obScurity
learning task. It has been identified as the serial is not possible to clearly identify the stimulus in the serial
preceding clump of items, or all the preceding itemsposition of each item the immediately preceding item)somne
it implies that together. If the functional stimulus is not the serial position,
each item is both a stimulus and a response. Therefore it is
response characteristics separately in a serial learning task. For impossible to study stimulus and
of intra-list similarity on the stimulus example, it isimpossible to determine the effect
side, meaningfulnessof the responses, or
selection. response coding versus stimulus

aired associate learning


(he most common method used is paired
Though the sübjects need only to recognize associate learning. It consists of bwo items to be learnt as pairs.
the stimulus item they-MUst_actually be able fo
esponse item In a laboratory experiment, subjects produce the
will usually be required to learn a set of stimulus
bairs, Generally a list consists of 12- 15pairs. Each complete response
fems are presented in a different order in each trial In the presentation of the list constitutes a trial, and the
stimulus item, and then the pair of stimulus and response anticipation method the subject initially sees the
to produce the correct response before the together. From the next trial, the subject is supposed
stimulus and the
method, the entire set of pairs is first presented, and then the response appear together. In the study-test
imuli one at'a time. Also known as the
method of complete
recatl of responses is asked by presenting the
sults until the eñtire series of
items has been studied. presentation, it does not provide knowledge of
rrect response is Another
presented along with the stimLtús before the
variant is the prompting
method, in which the
arning may then be given either before the subject produces the Tests of
which only the stimuli are preséntation of the stimulus response pair or onresponse.
independent trials
presented. Stifl another
0gressively increasing the number of itéms on successivevariant is the cumulative method, which consists of
trials. It is based on the assumption that the
3bably concentrates on mastering a few items subject
AEd o seteucni triais. per trial, and ignores thc
rrmning items teniporariit, 0 DE

****** ******"**
UiVII i
FAPK D, E L I IUN A,
DA 11,

amount of human learning, Such as acquir


Paired associate learning
is representative of a large of FEnglish words, etc. N
Liiring
learning Hindi equUvalents
a specific question, inemorizing state capitals, assóciates method Is theoretical, It is the o es
inmpetus bchind the frequent use of the paired an
greatest stimulus and the response. Both the c
as assoclation between the
of the traditional view of learning looked upon paired associato
iaentinadie. MMany early theorists
the response are precise and Cieariy earnin
a few of these viewed9a
an ideanway of manipulating
stimulus and response varíables independently. Quite
and Golin (1959) considered the sti
special case of conditioning. For example, Goss, Morgan,
learning
term ad aG^,
as the the response term as the US, and the subjects' representational response as the UR or CR af
learning. Besida_ these attempts at analogy, explanations have also been evolved within the area'of Verhal
learning. The most popular of these is the two-stage model introduced by Underwood, Runquist, and Schul
(1959), They made distinction between response learning phase and a hook up or associative phase in paired
associate learning. In the former, the subjects just learnt to give the response with or Without the correct
stimulus. In the latter, the responses were paired with the correct stimulus. Though response learning always
preceded the associative phase, different items within a list might differ in their stage of progression. Other
theorists have emphasized additional processes such as stimulus differentiation, stimulus coding, intra-ist
Similarity, etc. In a research review, Battig (1968) suggested that it may be necessary to consider as many as ten

arate processes or stages in of a thelearning "simple"


paired associate list.
THE SCORE
Thescoring in a memory experiment can be done cn the basis of accuracy of response or speedof response
ACCuracyY of respoAse The measures of memory are based primarily on the amount of
Subject on the test of retention e.g., the number oi items correct/y work. achieved by. the
ACCurate scoring depends on the extent to which the learning materials recalled, recognized, and reconsructed.
Lists of ncnsense can
syllables, digits, and diacrete meaningful words offer nobe hroken down into discrete items.
Connected passage of prose or
poetiy does
not provide equally clear-cut
difficulty. On the other hand, a
number of words correctly recalled
because many of them
units. One cannot simply use:the
frequently recur in the same passage. (articles, adjectives, pronouns, prepositions, etc:)
the learned has mastered the content of Moreover, such a scoring procedure would fail to show to what degree
the. passage. For these
into ideas' or "thought units' estimated to be reasons, a connected passage is usually
of equal difficulty, and the divided
of such thought units
correctly reproduced. It is also possible to gain final score is determined by the rnumber
additional information about degree of
retention by observing the subject's behavior
wellas the results of performance. during the test, i.e. by
examining the nature of the
performance as
:
Speed of response.The teimroral characteristics of the
egree of retention. The speed of response_ may response, in particular, are also sensitive indices of.
the
esponse Response latency is time between the onsetbeofassessed in two ways: responselatency, and rate of
the stimulus and the onset of the response. The
he latency, the greater is the strergth of the smaller
aired associates. The speed with which the response.
This measure is
especially applicable to the method of
subject supplies the missing member of the
vailability of the response, and an index of the strength of retention pair is a measure. of the
mount retained. Another
temporal index of supplementary. to measures of sheer
he speed with_which they follow each other.strength retention is the rate at which _responses are given i.e.,
of
Measurements of rate are easily applicable to verbal learning9,
specially to active recall (method of retained members).
eproduced, the greater is the avallability of the responses. The faster_the rate.at which the correct items are
hich supplements measures of accuracy. hus rate, like latencyis a measure of retention,

HE METHODS

emory is evident in different ways. A student


shows menmory of poetry by
professional wine tester demonstrates of a certain blend by
reciting passage without prompting
tountant shows some memory of his memory selecting it in a taste test. And a retired
er. These schoolboy German by the ease with which he relearns the language
examples illustrate the three basic ways years
Orecall, in which the subiect demonstrates psychologists have established for
measuring remembering:
memory
ecognition, in which the_subiect shawS a familiaritywithout the aid.of sigaificant stimuli
with some stirnisic is presen
Telearring; i wtii the subject sliovws à saving of time
or effot i l i i rEviiusiv a rari y l
thesubject is
iye recall because asKed
to
t is tne reproduce the orlginal
d ie period
o subject's task actively to response in some form. The
c his response in iedming.
n
reproduce the correct
experimentai work the subiect procedure is best aesi d
å
Diank. here is little responses,
often responds vocally ur which ne nds a
oember as many Ttems trom a list as prompting in recall testina, r
1easure of retention. Ihere are possible. As might be Often the student may
is
even D a
simply instrue
several variations of
active
expected, a test of recal generally yierdsu west
1. In tests recall:
oy the method of retained
hecan rememoer. He may
or may not havemembers,) the subiect is simply required to
difticunt One dha dsually requires to do so in
longer period of practice.a definite order. Of coursereproduce as
the latter many
Correctly reproauced. There .The task is uE
In free recall the subject is are two types of instructions that areretention score is simply the numoer o
Sertal recal the suDject mustrequired to reproduce a list usually given whern verbal recd
of stimulus items,
recall both the but he maY do in
way of testing vwhat has items and the order in which they were SO a
pernaps tne simplest oall
ee recal) presentations of a list of presente@.
been learnt. After one (single trial free recall) or
s

witnout regard to order, in some specified items, the subiect is simply asked to
reproduce the items severa l
materla.. Keal ife examples of free recall occurperiod of time. The emphasis is simply on the availabilrey that ne recdi
O
atempt to
name all the countries of Europe, A when we are making up a list of u
variant of
people to invite to
a paryo
given at Ene ume of testing free recall is the cued recall method, in whicn
Deginning or Ene test or only helping the subject to recall some of
after the subject has the items. The cue(s) may be a cue
apparently recalled everything he can by
presented in
means of free recail.
Reçall depends on the organization of thematerial. When free recall and serial recall methods are
(using tasks, too difficult
Tirst rew
for success on the first
trial), serial recall results in compared
greater success, except auring uie
trials. In the early trials, while the subject is still
Words, herever they learning the
appear; therefore, free recall is best. For the test, he usually responds with the easiest
or tirteen
to.twenty items, the subject is more successful when hecomplete task, however, as in learning a list
is required to learn the material
(Waugh, 1961; Earhard, 1967). Miller serially
(1956)
the formation
of larger and larger 'chunks' of emphasizes organization based upon recoding, which he views as
material.. The adult memory span
fixed at approximately seven separate items unless the material is
according to this view appears
learner does; recode many subjectively
individual items into larger chunks of categories, thenorganized into"chunks". When a
such categories promoting the. he is often able to recall seven
investigator
experiments, in which various numbers of itemsspeak
to of the "magical number seven
t 2" (Miller, 1956). Some
were used, have
ather thah the total amount of information, which is demonstrated that it is the number ofchunks"
important in recall.
Recent investigators have made a strong case for the view that successful recall is
subjective organization imposed on the material by the learner. chiefly the result of a
To form groups among the items in, the list on the Subjective organization is the learner's tendency
basis of his own
on
successive, free recall trials of the same set of words, people experience. Tulving (1966) demonstrated that
Mough the list itself is presented in a different order on each settle down to a fixed sequence in recall even
eflects the principles of trial. Each individual's
organization unique to that person. Mandler sequence is different and
onsensus in subjective organization and follow the (1970) argues that most people show
act that memory is hierarchical in organization dictated by the nature of the task,
uch of the current interest in free organization. cued recall is better than despite
It is for this reason the the
recall has been free recall.
ms in an generated by the fact that subjects
orderly way, and this also seems to be related to the reorganize the recalled
al to trial.
Researchers assume that this improvement shown in multi-trial recall from
the very least it indicates the reorganization
trend of encoding and
reflects the
organization of verbal material in the brain.
neral analytical value of free response processing abilities of the brain.
recall! is quite limited because
pecially the stimulus responsible for eliciting the stimuli Otherwise, the
the first item recalled. Most eliciting the responses are nebulous,
hough the specification of these cues is not likely, it is elicited by contextual
possible. cues,
2.Thememoy-span
M, it method is closely related to
the method of retained
was introduces in 1887 by Jacobs. The members. Generally used to
study
digits, letters, or words, and is required subject is presented (visually or orally) with a list of
iRMSIS Successivel Creascd until tiie to reproduce them immediatey after items, such
suojeci faiis in his attempt esentation. he iuilhbe! of
at repiCGuciion. Tine iorngest seriAs.
vhich he
UiVII I1
PArCR 0, tLIIUN A,
r A4,

can reproduce without error, defines hismemory span. The memory span is defined as that lengt
Obtalhed in this manner, the spa o
can be reproduced correctly on 50 percunt of the trials. closcy
It is threshold defining the limit of immediate retention.
psychophysical threshold. a

3. The methodof anticipationis another procedure in which retention is measured by autive


recall
pecuiary cnaracterized oy the ract that each item must be reproduced in reSponse to a specific stim.
item immediately preceding it in the serie. The number of correct anticipations gives the retention.c t h
serial anticipation, each response becomes the stimulus for the next one as well, and it is difficult to manis n
the stimulus and response separately. In the method of paired assóciates one member of a pair becomece
Specific stimulus for the reproduction of the second member. The number of correctly completed pairS isthe
retention score.

4 Thelmethod of reconstruction/puts its main emphasis on the


learning task consists of items arranged in a definite order. Someretention of sequence or order in general. The
time after the end of
given the items he has learnt, but scrambled in a practice, the Subject is
random_arrangement.
order. Ihe retention Score is based on the number of. items put.in His task is to reconstruCtthe _orginal
time taken to reconstruct the original order. their.proper position, or thE DUMber of.tiais.or
5.
Uestimony: In studies of testimony,
only once. The situation is similar subiects are asked to reports on events, which they
have
events of twenty-third of last January? Intestifying in a courtroom. An attorney may ask: "Wl you Witnessedd
to that of
the
recall the
without looking at it pure recall, no additional aids to
again, how can accurately memory are given. For exampie,
you recall the cover of a
particular book?
Eidetic Imagery:
or picture has not beenSometimes
removed.
persons especially children give
ceptionally accurate testimony, if the scene
a
memory images" have been calledIn fact, a child may report that he still "sees" the scene.
hallucinations. However, in ejdetic imagery, a term which suggests that Exceptionaliy accurate
hallucinations, the individuals believe that what he they possess the clearness
present in the outer e+vironment. or
his mind only. The person with gazes at, hears, or
The most common eidetic imagery knows that feels is
and ask the child to test of eidetic he is
describe what he has imagery is to present a responding to an
image in
apart from him and describe what seen. The eidetic child seemsvery detailed picture, then remove it
image dozens of detail, which he he still perceives. A child with to project the
image at a
eported. A child with acoustical eidetic noted earlier when.looking at the real imagery can observe in a istance
has not good eidetic
pjcture
idetic child looked at a
desk imagery can repeat long lists of picture. Acoustic images are
orward and backward while a list of a dozen or so digits after hearing them once. also
hildren. In fact, one
was still
looking the desk. Eidetic
at digits wäs read to him, then he One
ercent in a study of study showed 26.5 imagery appears in
"read" the list off
percent eidetic imagers retarded as
00 por cent of the normal children (Siipola and among retarded children as well as gifted
Hayden, opposed to 7a
ain injury, but more reliable children were brain-injured. 1965; Haber and Haber 1964). Furthermore
retarded
1ggest that one's measurement is needed. These findings may have
culture may be
89 to'
significance for
CrOSS-cultural studies
ar influential factor (Feldman, of diagno_ing
African and American
1968). children
ECOGNITION
the most
general case of
er time he is presented recognition, the subiect is first
h item is from the withlistthese original items mixed exposed to a number of verbal items,
criminate between wonew or
the original_one. his randomly among new items. His then at some
or method
more distinct stimuli. It varies not in requires only that the task is to tell whether
|akes on the subject. subject is
generate them. Response alternatives are presented to the method for
d not the of able t0
presentation,
subject recognition; the
but in the
demand.
e subject himseif
forced choice procedure, the subject must select the
example: Who wrote The correct response from
Vatson B. Skinner Behavior of Organisms? two or more
alternatives
he ves ng procedure, the C. Hebb
xample: Did Hebb write 7hesubject decides whether or not a Single item
es Behavior of passes the threshold of
B. No Organisnms? acceptability.
ion is the type or
retention measured by_a
7nitiorn

09rect alternatiVe
rrect answerfiom among several. In an multiple-choice question. The
tcome individuals are from among many.
more inclined
experimental situation, the subj
Recognitlon
learneris required
ubject would be require
xd in order to render the scores
uced to guess tests
than others,inevitably invite guessing or. the part of the
ne percent or of different and a correction for guessing me
sUDUILUng
iteins subiects comparable. The be
ore is calculated as follows: incorrectly recognized from the scores are adequately djusted
adu
percent Correctly recognized. The recognition
Rccognition score = R W 100
N
where R: right recognitions
W wrong recognitions
N total number of
item 3
given in the
recognition test

With recognition measures, even weak retention


S Prompr fOr the is revealed. The preseed of the rrect response_among the

choices on the basis of length,correct_choice. In addition,subiects (and. students) can eliminate some inco
position, incorrect grammar, etc. This increases the chance o
eeuo,
the
QISCrnihation being much aetire of
easier than reproduction of the response, it is a very _ensitve mea
amaount that has been learnt. It often shows learning where. other methods do not. It
pure, measure ot storage or the extent to which people differentiate among items presentedistoote
tnem.

Oweve, a-reçognition test may or may not be a successful test of retention. If there is little or no similanity
Detween the ojd and the new items, the test will be much too insensitive to measure retention. An extreme
ampie w ilustrate the point. If our subject learns list no-nonsense syllables and we test his retention for
them b mng these syllables with an equal number of words, he will certainly identify all the syllables and
reject al the words. Obviously, retention for nonsense syllables can be tested best by mixing them with other
nonsense sylables, retention for adjectives by mixing them with other adjectives, and so on. The sensitivity of
the recogntion testillLdepend uponthe degree of similarity between theeold and new item.
Responses which were previously associated with stimuli similar to the test stimuli are more likely to be selected
erroneously. The phenomenon of false recognition can be seen in the familiar feeling that one has been in a
certain situation, before. This feeling, called deja vue occurs when enough aspects of the immediate situation
resemble a previous situation.Weincorrectly identify the past with the present. We may feel that we have met a
particular person before because he bears some similarity to ån acquaintance in features, such as hair color, gait,
physique, or manner of dress. Some aspect of the presnt stimulation leads us to think of him as someone
encountered previously.

In daily life the inaccuracy of memory is often emphasized, but we should also stress the tremendous capacity of
the human information-processing system, particularly when recognition is all that is required. Subjects in one
experiment showed great success in recognizing large amounts of visually presented information. They were
xposed to approximately 60 randomly selected visual stimuli, called an inspection series. Then some of these
stimuli were paired with new stimuli, and the subject was asked to indicate which member of the pair he had
been previously. The median correct recognition was 90.0 per cent for the words, 88.2 per cent for sentences,
and 98.5 per cent for pictures.

RELEARNING
Kelearning given by Ebbinghaus (1887) is sometimes called the savings methoc) Using this nmethod, the subject
hst learns Some material and, after yarious lengths of time, relearns that same material to the same criterion
evel. The number Qf trials to relearn the material is usually fewer than the number required to learn it the first
inef a rat learned a maze to the point at which it could run it three times in succession without error, it must
earn to this riterion. We then compare the work required in relearning with the original effort. When a rat
uires twenty trials to learn originally and only five trials some time later, the saving is fifteen trials, or 75 per
t of the
trials originally required. The fornula for relearning expresses the percantage of 'Saveü.
UiVII
DA 11, PAPEK B, StLIIUIV A,

Relearning score L-L2x100


Li
rclcarn it, the rele:.
materlal originally and only four trials to
For example: If it took ten trials to learn the
Score would equal 100 x (10-4)/10 = 60%
thefirst (test)
demonstrates full memory ornthe trial, which
relearning trials are required, that is, if the subject number of original and
t no savng_score is 100%. If
aoes not count as an attempt_at relearning, then_the in relearning. Finally, if the relearning rials are greater
relearning trials are the same, clearly, there is no saving is a negative number and suggests that some
nan the original trials (which is an unlikely outcome), the result other method because
ype or intererence has occurred. Relearning usually reveals more retention than any
more stimulus conditions are present in both the original learning and the relearning. ThuS It Is tne most sensitive
measure of retention savings are shown even if the subject fails to recall or recognize any materna.

Conciusion: There is no one perfect measure of retention, There are several experimental operations for gauging
ne degree of retention, each of them valid in its own right, showing the availability of certain types of response.
s r e s ot retention do, however, vary in sensitivity; and some operations are more suitable than others for
reveallng small differences in degree of retention. Luh (1922) conducted a study to compare the varioUS metnoas
thse a l that relearning is the most sensitive method of assessing memory because it can show even
those small amounts of retention that are not shown by recognition or recall tests. Genera!lly, relearning is the
most sensitive, followed by recognition, followed by recall. Thus we find different retention rates depenaing on
how we test for retention.

FORGETTING DUE TO DECAY AND INTERFERENCE


Forgetting is the obverse of
remembering a failure to retain what has been acquired. Retention
amount of original learning that is still referS to the
typical retention curve: effective, while forgetting refers to the amount lost. The
following is a

100

AALNT I R3OTTEN

AMOLNT RETAINED

HOUAS
1
.

the graph indicates, by far the


arning. Thereafter, the curve levelsgreatest amount_of forgetting takes place in
Durs.
off. The amount retained
after 5 hours is only the first feWmoments.after
slightly greater_than after. 15
Arly behaviourists took a "black-box"
Sponses they produce while ignoring whatapproach to animal learning by focusing on external
owed a similar path. For went on inside the stimuli and the
ie:irer thar even the
severai organism. The pioneers in memory
decades, inost of their studies were báseu on the researcdh
most complex racnGies dre notion of
asenriatin,
uitimately based on asioCiations beiwec!i Sunpie
iciea.
ctive, Ebbinghaus's decision to
smed that once develop a simple and pire
riables, thhe same lawspsychologists
could be understood how associative measure of memory makes pere
applied more complex stimuli bonds were formed between simple
to

imenta Subject wno happened to memorize such as words, images, and


acas
ped Dy. external rorces. For a
particular list was seen as a passive
long tists. he assumption was example, many studies revealed that short
that this cxternal
participant, meemory
lists were learned
th of the 0onds varlable (the length of a list) directly nore
13n mind or whether between
the
stimuli, and there was lIttle inu
feelings and expectations of the interest in how this process occurred wiu t the
memoriser made any difference.
Thus many different views of
forgetting have been suggested; we discuss the four
major vews as
o
THE PSYCHODYNAMICAPPROACH:
MOTIVATED FORGETTING
reud OL noias that some events or experiences are iust too painful to remember,
Gerense mecnaniSm, we forget them. Thus an unconscTOus consequenuy
e process of repression
soy o.d.nan whose memory was blocked by the tip of the tongue phenomenon. or memoy ina a familiar
poem, the man kept getting stuck at the words "with the white sheets". wnie re
His analyst asked him
to tne expression white sheet. An abbreviated version of the man's train of thought went something lIKehic.
uis
e r e m i n d s me of a linen sheet over a corpse (pause) -I think of a friend who recently died of heart
Gseasehe was fat and did not exercise enough - my qrandfather died of heart disease - I am a little chubby
myseir cDuld die of heart disease." Thus the apparently innocent forgetting of a line of a poem involved
repression or unconscious conflicts over a fear of death. Thus memory may not be evident because stumul
appropnate ;for recall are not present. You may be oblivious to some earlier experience of childhood, yet
suddenly recall it because a particular odor, name, or other stimulus led to your recall. The effectiveness Or the
free associa ion method of psychoanalysis maybe attributed to the fact that it arouses verbal stimull, whicn in
turn activate recall of seemingly forgotten events.

THE BIOLOGICAL APPROACH: FROM ENGRAMS TO ALZHEIMER'S DISEASE


Itwas felt initially that all experiences are encoded in the braininthe form of a memory trace-the engram.In
his book n search of an engram Lashley (1950 mentions that removing speciífic parts of a rae's brain affected
the animal's behaviour, but did_not totaly eliminate specific memories. His rigorous researeh revealedthat
experiences. have no specificneurons associated withthemHe was thus never able to identify the physiological
mmory. More recent researchers, using brain-imaging techniques, have found that there are changes
basis fo
in brain activity of specific areas as the information is being processed, but these chanyes are not permanent and
leave no traces behind. If at all memories are stored in the brain, they are distributed throughout the brain.

A lot of curent work focuses on the role of neurotransmitters in memory storage. A fascinating line of research is
exploring therole of the neurotransmittersin the memorylapses associated with Alzheimersdisease. Initallythe
patients just have trouble with everyday tasks - leaving the water running, for forgetting to turn off the stove.

But gradually a profound memory loss occurs such that the patient forgets the names of even close friends,
relatives, and sometimes even their own names. The disease is generally fatal within five years of its appearance.
Identified by the German neurologist Alois Alzheimer in 1860, it has been known_that the disorder leads to
degeneration of brain tissue. In 1983, Coyle, Price, and Delong reported that those who died of Alzheimer's have
60 to 70% less acetylcholine in their brain than comparable victims of other disorders. The disease destroys
neurons in the nuUCieus_basals in the brain, which supplies acetylcholine to the cortex and the hippocampus.
Clinical evidence (the case of H.M.) shows that the hippocampus plays a critical role in memory.

A COGNITIVE APPROACH: CUE DEPENDENT FORGETTING


The cognitive approach explores the structure and organization of the mind and the active processes used by the
learner. It distinguishes between three stages of memory: encoding, storage, and retrieval. Theorists generaly
believe that ltteinformation islost from storage in LTM. What we have lostis the ability tofind or retrieve the
ntormation. Themindislike a giant reference library, where the books neverleave the shelves. The problem is,
eoltEn do not ianow wlere to look for a specific volume. Acording to Tulving's (1974)theory of cue dependenr
ellng, lorgeiny is Calused Lyd alure tOetrreve inforialio omi storuyedue to inadequate memory
cue

You might also like