Cognitive Psych Discussion Posts

Discussion posts – at least 2 paragraphs each

(Discussion post 1) Chapters 5 and 6 in your text discuss several iconic experiments in memory that have helped us learn about our memory system. After reading about the experiments, select one of the optional articles featured in the module resources and provide a brief critique of the experiment and its application to the real world. You should be sure to discuss any problems with the methodology of the experiment or any problems with the interpretations of the results. Additionally, consider how the findings of the study can be applied to real-world issues. How could findings from the experiment be applied to develop policies or procedures to improve human memory in different professional settings? What strengths or weaknesses in human memory do the experiments highlight that would be important for professional disciplines? (The three optional articles to choose from are attached to this post.)

(Discussion post 2) Mental imagery is an exciting area of research because it has been used in many real-world applications, including improving sports performance, counseling, promoting well-being, and improving memory. Locate an article on mental imagery published within the past five years. Provide a brief summary and critique of the article. Next, discuss the importance and applications of the article’s findings in terms of contemporary issues. How could you use information from the article to improve cognition in different scenarios and professional settings?

(Discussion post 3) Over the years, psychologists have proposed a number of different theories to describe the interactions between language and other types of cognition. This week’s readings expose you to a number of these theories. Some theories suggest that language shapes the way we think about things. Others suggest that individuals are able to change their thought patterns by learning different languages. Consider the influence of language on human cognition in professional settings by discussing how language promotes or limits communication between people. Finally, discuss how cognition could be improved or problems could be solved by examining language.

(Discussion post 4) Throughout the course, you have been exposed to theories in cognitive psychology and have been asked to use those theories to solve problems and evaluate human cognition. Based on what you have learned, discuss the role you think cognitive psychology plays in understanding human behavior and interactions. How has this field advanced our understanding of human cognition and behavior? How has cognitive psychology helped us to create socially responsible strategies for improving cognition? What directions do you think the field will take next, and how will that research be important in solving contemporary problems or real-world issues?

Journal ol Experimental Psychology: General
1975, Vol. 104, No. 3,

268

-294

Depth of Processing and the Retention of Words
in Episodic Memory

Fergus I. M. Craik and Endel Tulving
University of Toronto, Toronto, Ontario, Canada

SUMMARY

Ten experiments were designed to explore the levels of processing framework
for human memory research proposed by Craik and Lockhart (1972). The basic
notions are that the episodic memory trace may be thought of as a rather auto

–

matic by-product of operations carried out by the cognitive system and that the
durability of the trace is a positive function of “depth” of processing, where depth
refers to greater degrees of semantic involvement. Subjects were induced to
process words to different depths by answering various questions about the words

.

For example, shallow encodings were achieved by asking questions about type-
script; intermediate levels of encoding were accomplished by asking questions
about rhymes; deep levels were induced by asking whether the word would fit into
a given category or sentence frame. After the encoding phase was completed,
subjects were unexpectedly given a recall or recognition test for the words. In
general, deeper encodings took longer to accomplish and were associated with
higher levels of performance on the subsequent memory test. Also, questions lead-
ing to positive responses were associated with higher retention levels than questions
leading to negative responses, at least at deeper levels of encoding.

Further experiments examined this pattern of effects in greater analytic detail.
It was established that the original results did not simply reflect differential encod-
ing times; an experiment was designed in which a complex but shallow task took
longer to carry out but yielded lower levels of recognition than an easy, deeper
task. Other studies explored reasons for the superior retention of words associated
with positive responses on the initial task. Negative responses were remembered
as well as positive responses when the questions led to an equally elaborate encoding
in the two cases. The idea that elaboration or “spread” of encoding provides a
better description of the results was given a further boost by the finding of the
typical pattern of results under intentional learning conditions, and where each
word was exposed for 6 sec in the initial phase. While spread and elaboration
may indeed be better descriptive terms for the present findings, retention depends
critically on the qualitative nature of the encoding operations performed; a
minimal semantic analysis is more beneficial than an extensive structural analysis.

Finally, Schulman’s (1974) principle of congruity appears necessary for a
complete description of the effects obtained. Memory performance is enhanced
to the extent that the context, or encoding question, forms an integrated unit with
the word presented. A congruous encoding yields superior memory performance
because a more elaborate trace is laid down and because in such cases the struc-
ture of semantic memory can be utilized more effectively to facilitate retrieval.
The article concludes with a discussion of the broader implications of these data
and ideas for the study of human learning and memory,

268

DEPTH OF PROCESSING AND WORD RETENTION 269

While information-processing models of
human memory have been concerned largely
with structural aspects of the system, there
is a growing tendency for theorists to focus,
rather, on the processes involved in learning
and remembering. Thus the theorist’s task,
until recently, has been to provide an
adequate description of the characteristics
and interrelations of the successive stages
through which information flows. An al-
ternative approach is to study more directly
those processes involved in remembering—
processes such as attention, encoding, re-
hearsal, and retrieval—and to formulate a
description of the memory system in terms
of these constituent operations. This alter-
native viewpoint has been advocated by
Cermak (1972), Craik and Lockhart
(1972), Hyde and Jenkins (1969, 1973),
Kolers (1973a), Neisser (1967), and Paivio
(1971), among others, and it represents a
sufficiently different set of fundamental
assumptions to justify its description as a
new paradigm, or at least a miniparadigm,
in memory research. How should we con-
ceptualize learning and retrieval operations
in these terms? What changes in the sys-
tem underlie remembering? Is the “mem-
ory trace” best regarded as some copy of
the item in a memory store (Waugh & Nor-
man, 1965), as a bundle of features (Bower,
1967), as the record resulting from the
perceptual and cognitive analyses carried
out on the stimulus (Craik & Lockhart,
1972), or do we remember in terms of the
encoding operations themselves (Neisser,
1967; Kolers, 1973a) ? Although we are
still some way from answering these crucial
questions satisfactorily, several recent stud-
ies have provided important clues.

The incidental learning situation, in which
subjects perform different orienting tasks,

The research reported in this article was sup-
ported by National Research Council of Canada
Grants A8261 and A8632 to the first and second
authors, respectively. The authors gratefully
acknowledge the assistance of Michael Anderson,
Ed Darte, Gregory Mazuryk, Marsha Carnat,
Marilyn Tiller, and Margaret Barr.

Requests for reprints should be sent to F. I. M.
Craik, Erindale College, University of Toronto,
Mississauga, Ontario, LSL 1C6, Canada.

provides an experimental setting for the
study of mental operations and their effects
on learning. It has been shown that when
subjects perform orienting tasks requiring
analysis of the meaning of words in a list,
subsequent recall is as extensive and as
highly structured as the recall observed
under intentional conditions in the absence
of any specific orienting task; further re-
search has indicated that a “process”
explanation is most compatible with the
results (Hyde, 1973; Hyde & Jenkins,
1969, 1973; Walsh & Jenkins, 1973).
Schulman (1971) has also shown that a
semantic orienting task is followed by
higher retention of words than a “struc-
tural” task in which the nonsemantic aspects
of the words are attended to. Similar find-
ings have been reported for the retention of
sentences (Bobrow & Bower, 1969; Rosen-
berg & Schiller, 1971; Treisman & Tux-
worth, 1974) and in memory for faces
(Bower & Karlin, 1974). In all these
experiments, an orienting task requiring
semantic or affective judgments led to
better memory performance than tasks
involving structural or syntactic judgments.
However, the involvement of semantic
analyses is not the whole story: Schulman
(1974) has shown that congruous queries
about words (e.g., “Is • a SOPRANO a
singer?”) yield better memory for the
words than incongruous queries (e.g., “Is
MUSTARD concave?”). Instruction to form
images from the words also leads to excel-
lent retention (e.g., Paivio, 1971; Sheehan,
1971).

The results of these studies have impor-
tant theoretical implications. First, they
demonstrate a continuity between incidental
and intentional learning—the operations
carried out on the material, not the intention
to learn, as such, determine retention. The
results thus corroborate Postman’s (1964)
position on the essential similarity of inci-
dental and intentional learning, although the
recent work is more usually described in
terms of similar processes rather than sim-
ilar responses (Hyde & Jenkins, 1973).
Second, it seems clear that attention to the
word’s meaning is a necessary prerequisite
of good retention. Third, since retrieval

270 FERGUS I. M. CRAIK AND ENDEL TULVING

conditions are typically held constant in
the experiments described above, the dif-
ferences in retention reflect the effects of
different encoding operations, although the
picture is complicated by the finding that
different encoding operations are optimal
for different retrieval conditions (e.g.,
Eagle & Leiter, 1964; Jacoby, 1973).
Fourth, large differences in recall under
different encoding operations have been
observed under conditions where the sub-
jects’ task does not entail organization or
establishment of interitem associations;
thus the results seem to take us beyond
associative and organization processes, as
important determinants of learning and
retention. It may be, of course, that the
orienting tasks actually do lead to organiz-
ation as suggested by the results of Hyde
and Jenkins (1973). Yet, it now becomes
possible to entertain the hypothesis that
optimal processing of individual words, qua
individual words, is sufficient to support
good recall. Finally, the experiments may
yield some insights into the nature of learn-
ing operations themselves. Classical verbal
learning theory has not been much con-
cerned with processes and changes within
the system but has concentrated largely on
manipulations of the material or the experi-
mental situation and the resulting effects
on learning. Thus at the moment, we know
a lot about the effects of meaningfulness,
word frequency, rate of presentation, var-
ious learning instructions, and the like, but
rather little about the nature and character-
istics of underlying or accompanying
mental events. Experimental and theo-
retical analysis of the effects of various
encoding operations holds out the promise
that intentional learning can be reduced
to, and understood in terms of, some com-
bination of more basic operations.

The experiments reported in the present
paper were carried out to gain further in-
sights into the processes involved in good
memory performance. The initial experi-
ments were designed to gather evidence
for the depth of processing view of mem-
ory outlined by Craik and Lockhart (1972).
These authors proposed that the memory
trace could usefully be regarded as the by-

product of perceptual processing; just as
perception may be thought to be composed
of a series of analyses, proceeding from
early sensory processing to later semantic-
associative operations, so the resultant
memory trace may be more or less elab-
orate depending on the number and qualita-
tive nature of the perceptual analyses car-
ried out on the stimulus. It was further
suggested that the durability of the memory
trace is a function of depth of processing.
That is, stimuli which do not receive full
attention, and are analyzed only to a shal-
low sensory level, give rise to very transient
memory traces. On the other hand, stimuli
that are attended to, fully analyzed, and
enriched by associations or images yield a
deeper encoding of the event, and a long-
lasting trace.

The Craik and Lockhart formulation
provides one possible framework to accom-
modate the findings from the incidental
learning studies cited above. It has the
advantage of focusing attention on the pro-
cesses underlying trace formation and on
the importance of encoding operations;
also, since memory traces are not seen as
residing in one of several stores, the depth
of processing approach eliminates the neces-
sity to document the capacity of postulated
stores, to define the coding characteristic of
each store, or to characterize the mechanism
by which an item is transferred from one
store to another. Despite these advantages,
there are several obvious shortcomings of
the Craik and Lockhart viewpoint. Does
the levels of processing framework say any
more than “meaningful events are well
remembered” ? If not, it is simply a collec-
tion of old ideas in a somewhat different
setting. Further, the position may actually
represent a backward step in the study of
human memory since the notions are much
vaguer than any of the mathematical models
proposed, for example, in Norman’s (1970)
collection. If we already know that the
memory trace can be precisely represented
as

I =

(Wickelgren, 1973), then such woolly
statements as “deeper processing yields a

DEPTH OF PROCESSING AND WORD RETENTION 271

more durable trace” are surely far behind
us. Third, and most serious perhaps, the
very least the levels position requires is
some independent index of depth—there are
obvious dangers of circularity present in
that any well-remembered event can too
easily be labeled deeply processed.

Such criticisms can be partially countered.
First, cogent arguments can be marshaled
(e.g., Broadbent, 1961) for the advantages
of working with a rather general theory-—

•

provided the theory is still capable of gen-
erating predictions which are distinguish-
able from the predictions of other theories.
From this general and undoubtedly true
starting point, the concepts can be refined in
the light of experimental results suggested
by the theoretical framework. In this
sense the levels of processing viewpoint will
encourage rather different types of question
and may yield new insights. A further
point on the issue of general versus specific
theories is that while strength theories of
memory are commendably specific and so-
phisticated mathematically, the sophistica-
tion may be out of place if the basic premises
are of limited generality or even wrong. It
is now established, for example, that the
trace of an event can be readily retrieved in
one environment of retrieval cues, while it
is retrieved with difficulty in another (e.g.,
Tulving & Thomson, 1973) ; it is hard to
reconcile such a finding with the view that
the probability of retrieval depends only on
some unidimensional strength.

With regard to an independent index of
processing depth, Craik and Lockhart
(1972) suggested that, when other things
are held constant, deeper levels of process-
ing would require longer processing times.
Processing time cannot always be taken as
an absolute indicator of depth, however,
since highly familiar stimuli (e.g., simple
phrases or pictures) can be rapidly analyzed
to a complex meaningful level. But within
one class of materials, or better, with one
specific stimulus, deeper processing is
assumed to require more time. Thus, in
the present studies, the time to make deci-
sions at different levels of analysis was
taken as an initial index of processing
depth.

The purpose of this article is to describe
10 experiments carried out within the levels
of processing framework. The first experi-
ments examined the-plausibility of the basic
notions and attempted to rule out alterna-
tive explanations of the results. Further
experiments were carried out in an attempt
to achieve a better characterization of depth
of processing and how it is that deeper
semantic analysis yields superior memory
performance. Finally, the implications of
the results for an understanding of learning
operations are examined, and the adequacy
of the depth of processing metaphor ques-
tioned.

EXPERIMENTAL INVESTIGATIONS

Since one basic paradigm is used through-
out the series of studies, the method will be
described in detail at this point. Variations
in the general method will be indicated as
each study is described.

General Method
Typically, subjects were tested individually.

They were informed that the experiment con-
cerned perception and speed of reaction. On each
trial a different word (usually a common noun)
was exposed in a tachistoscope for 200 msec.
Before the word was exposed, the subject was
asked a question about the word. The purpose
of the question was to induce the subject to pro-
cess the word to one of several levels of analysis,
thus the questions were chosen to necessitate
processing either to a relatively shallow level
(e.g., questions about the word’s physical appear-
ance) or to a relatively deep level (e.g., questions
about the word’s meaning). In some experiments,
the subject read the question on a card; in others,
the question was read to him. After reading or
hearing the question, the subject looked in the
tachistoscope with one hand resting on a yes
response key and the other on a no response key.
One second after a warning “ready” signal the
word appeared and the subject recorded his (or
her) decision by pressing the appropriate key
(e.g., if the question was “Is the word an animal
name?” and the word presented was TIGER, the
subject would respond yes). After a series of
such question and answer trials, the subject was
unexpectedly given a retention test for the words.
The expectation was that memory performance
would vary systematically with the depth of
processing.

Three types of question were asked in the
initial encoding phase, (a) An analysis of the
physical structure of the word was effected by
asking about the physical structure of the word

272 FERGUS I. M. CRAIK AND ENDEL TULVING

TABLE 1
TYPICAL QUESTIONS AND RESPONSES USED IN THE EXPERIMENTS

Level of processing
Answer

Question

Yes No

Structural
Phonemic

Category

Sentence

Is the word in capital letters?
Does the word rhyme with WEIGHT?
Is the word a type of fish?
Would the word fit the sentence:
“He met a in the street”?

TABLE
crate
SHARK

FRIEND

table
MARKET
heaven

cloud

(e.g., “Is the word printed in capital letters?”).
(b) A phonemic level of analysis was induced by
asking about the word’s rhyming characteristics
(e.g., “Does the word rhyme with TRAIN?”).
(c) A semantic analysis was activated by asking
either categorical questions (e.g., “Is the word
an animal name?”) or “sentence” questions (e.g.,
“Would the word fit the following sentence:
The girl placed the on the table’?”).
Further examples are shown in Table 1. At each
of the three levels of analysis, half of the ques-
tions yielded yes responses and half no responses.

The general procedure thus consisted of
explaining the perceptual-reaction time task to a
single subject, giving him a long series of trials
in which both the type of question and yes-no
decisions were randomized, and finally giving him
an unexpected retention test. This test was either
free recall (“Recall all the words you have seen
in the perceptual task, in any order”) ; cued recall,
in which some aspect of each word event was re-
presented as a cue; or recognition, where copies
of the original words were re-presented along
with a number of distractors. In the initial en-
coding phase, response latencies were in fact
recorded: A millisecond stop clock was started by
the timing mechanism which activated the tachisto-
scope, and the clock was stopped by the subject’s
key response. Typically, over a group of sub-
jects, the same pool of words was used, but each
word was rotated through the various level and
response combinations (CAPITALS ?-yes; SEN-
TENCE ?-no, and so on). The general prediction
was that deeper level questions would take longer
to answer but would yield a more elaborate mem-
ory trace which in turn would support higher
recognition and recall performance.

Experiment 1
Method. In the first experiment, single subjects

were given the perceptual-reaction time test; this
encoding phase was followed by a recognition test.
Five types of question were used. First, “Is there
a word present?” Second, “Is the word in cap-
ital letters?” Third, “Does the word rhyme with

?” Fourth, “Is the word in the cat-
egory ?” Fifth, “Would the word fit
in the sentence ?” When the first type
of question was asked (“Is there a word pres-
ent?”), on half of the trials a word was present

and on half of the trials no word was present on
the tachistoscope card; thus, the subject could
respond yes when he detected any wordlike pat-
tern on the card. (This task may be rather dif-
ferent from the others and was not used in
further experiments; also, of course, it yields
difficulties of analysis since no word is presented
on the negative trials, these trials cannot be
included in the measurement of retention.)

The stimuli used were common two-syllable
nouns of 5, 6, or 7 letters. Forty trials were
given; 4 words represented each of the 10 condi-
tions (5 levels X yes-no). The same pool of 4

0

words was used for all 20 subjects, but each word
was rotated through the 10 conditions so that, for
different subjects, a word was presented as a
rhyme-jie.? stimulus, a category-wo stimulus and
so on. This procedure yielded 10 combinations
of questions and words; 2 subjects received each
combination. On each trial, the question was
read to the subject who was already looking
in the tachistoscope. After 2 sec, the word was
exposed and the subject responded by saying yes
or no—his vocal response activated a voice key
which stopped a millisecond timer. The experi-
menter recorded the response latency, changed the
word in the tachistoscope, and read the next
question; trials thus occurred approximately
every 10 sec.

After a brief rest, the subject was given a sheet
with the 40 original words plus 40 similar dis-
tractors typed on it. Any one subject had
actually only seen 36 words as no word was
presented on negative “Word present?” trials. He
was asked to check all words he had seen on the
tachistoscope. No time limit was imposed for
this task. Two different randomizations of the
80 recognition words were typed; one random-
ization was given to each member of the pair of
subjects who received identical study lists. Thus
each subject received a unique presentation-
recognition combination. The 20 subjects were
college students of both sexes paid for their
services.

Results and discussion. The results are
shown in Table 2. The upper portion
shows response latencies for the different
questions. Only correct answers were in-

DEPTH OF PROCESSING AND WORD RETENTION 273

eluded in the analysis. The median latency
was calculated for each subject; Table 2
shows mean medians. Although the five
question levels were selected intuitively, the
table shows that in fact response latency
rises systematically as the questions neces-
sitated deeper processing. Apart from the
sentence level, yes and no responses took
equivalent times. The median latency
scores were subjected to an analysis of
variance (after log transformation). The
analysis showed a significant effect of level,
F(4, 171) = 35.4, p < .001, but no effect of response type (yes-no) and no inter- action. Thus, intuitively deeper questions —semantic as opposed to structural deci- sions about the word—required slightly longer processing times (150-200 msec).

Table 2 also shows the recognition re-
sults. Performance (the hit rate) increased
substantially from below 20% recognized
for questions concerning structural charac-
teristics, to 96% correct for sentence-yes
decisions. The other prominent feature of
the recognition results is that the yes re-
sponses to words in the initial perceptual
phase were accompanied by higher sub-
sequent recognition than the no responses.
Further, the superiority of recognition of
yes words increased with depth (until the
trend was apparently halted by a ceiling
effect). These observations were confirmed
by analysis of variance on recognition pro-
portions (after arc sine transformation).
Since the first level (word present?) had
only yes responses, words from this level
were not included in the analysis. Type of
question was a significant factor, F(3, 133)
= 52.8, p < .001, as was response type (yes- no), F(\, 133) =40.2, /X.001. The Question X Response Type interaction was also significant, F(3, 133) = 6.77, p < .001.

The results have thus shown that differ-
ent encoding questions led to different re-
sponse latencies; questions about the sur-
face form of the word were answered com-
paratively rapidly, while more abstract
questions about the word’s meaning took
longer to answer. If processing time is an
index of depth, then words presented after
a semantic question were indeed processed
more deeply. Further, the different encod-

TABLE 2
INITIAL DECISION LATENCY AND RECOGNITION

PERFORMANCE FOR WORDS AS A FUNCTION OF
INITIAL TASK (EXPERIMENT 1)

Response
type

Level of processing
1 2 3 4 S

Response latency (msec)

Yes
No

591
590

614
625

689
678

711
716

746
832

Proportion recognized

Yes
No

.22 .18
.14

.78

.36

.93

.63

.96

.83

ing questions were associated with marked
differences in recognition performance:
Semantic questions were followed by higher
recognition of the word. In fact, Table 2
shows that initial response latency is sys-
tematically related to subsequent recogni-
tion. Thus, within the limits of the present
assumptions, it may be concluded that
deeper processing yields superior retention.

It is of course possible to argue that the
higher recognition levels are more simply
attributable to longer study times. This
point will be dealt with later in the paper,
but for the present it may be noted that in
these terms, 200 msec of extra study time
led to a 400% improvement in retention.
It seems more reasonable to attribute the
enhanced performance to qualitative differ-
ences in processing and to conclude that
manipulation of levels of processing at the
time of input is an extremely powerful
determinant of retention of word events.
The reason for the superior recognition of
yes responses is not immediately apparent—
it cannot be greater depth of processing in
the simple sense, since yes and no responses
took the same time for each encoding ques-
tion. Further discussion of this point is
deferred until more experiments are de-
scribed.

Experiment 2 is basically a replication of
Experiment 1 but with a somewhat tidier
design and with more recognition distrac-
tors to remove ceiling effects.

Experiment 2
Method. Only three levels of encoding were

used in this study: questions concerning type-

274 FERGUS I. M. CRAIK AND ENDEL TULVING

CASE RHYME SENTENCE CASE RHYME SENTENCE

LEVEL OF PROCESSING

FIGURE 1. Initial decision latency and recognition performance for words
as a function of the initial task (Experiment 2).

script (uppercase or lowercase), rhyme questions,
and sentence questions (in which subjects were
given a sentence frame with one word missing).
During the initial perceptual phase 60 questions
were presented: 10 yes and 10 no questions at
each of the three levels. Question type was ran-
domized within the block of 60 trials. The ques-
tion was presented auditorily to the subject; 2
sec later the word appeared in the tachistoscope
for 200 msec. The subject responded as rapidly
as possible by pressing one of two response keys.
After completing the 60 initial trials, the subject
was given a typed list of 180 words comprising
the 60 original words plus 120 distractors. He
was told to check all words he had seen in the
first phase.

All words used were five-letter common con-
crete nouns. From the pool Of 60 words, two
question formats were constructed by randomly
allocating each word to a question type until all
10 words for each question type were filled. In
addition, two orders of question presentation and
two random orderings of the 180-word recogni-
tion list were used. Three subjects were tested
on each of the eight combinations thus generated.
The 24 subjects were students of both sexes paid
for their services and tested individually.

Results and discussion. The left-hand
panel of Figure 1 shows that response
latency rose systematically for both- response
types, from case questions to rhyme ques-
tions to sentence questions. These data
again are interpreted as showing that deeper
processing took longer to accomplish. At

each level, positive and negative responses
took the same time. An analysis of variance
on mean medians yielded an effect of ques-
tion type, F(2, 46) = 46.5, p < .001, but yielded no effect of response type and no interaction.

Figure 1 also shows the recognition
results. For yes words, performance in-
creased from 15% for case decisions to 81%
for sentence decisions—more than a five-
fold increase in hit rate for memory per-
formance for the same subjects in the same
experiment. Recognition of no words also
increased, but less sharply from 19% (case)
to 49% (sentence). An analysis of vari-
ance showed a question type (level of pro-
cessing) effect, F(2, 46) = 118, p < .001, a response type (yes-no} effect, F(\, 23) = 47.9, p < .001, and a Question Type X Response Type interaction, F(2, 46) = 22.5, p < .001.

Experiment 2 thus replicated the results
of Experiment 1 and showed clearly (a)
Different encoding questions are associated
with different response latencies—this find-
ing is interpreted to mean that semantic
questions induce a deeper level of analysis
of the presented word, (b) positive and
negative responses are equally fast, (c)

DEPTH OF PROCESSING AND WORD RETENTION 275

recognition increases to the extent that the
encoding question deals with more abstract,
semantic features of the word, and (d)
words given a positive response are asso-
ciated with higher recognition performance,
but only after rhyme and category ques-
tions.

The data from Figure 1 are replotted in
Figure 2, in which recognition performance
is shown as a function of initial categoriza-
tion time. Both yes and no functions are
strikingly linear, with a steeper slope for
yes responses. This pattern of data sug-
gests that memory performance may simply
be a function of processing time as such
(regardless of “level of analysis”). This
suggestion is examined (and rejected) in
this article, where we argue that level of
analysis, not processing time, is the critical
determinant of recognition performance.

Experiments 3 and 4 extended the gen-
erality of these findings by showing that
the same pattern of results holds in recall
and under intentional learning conditions.

Experiment 3
Method. Three levels of encoding were again

included in the study by asking questions about
typescript (case), rhyme, and sentences. On each
trial the question was read to the subject; after
2 sec the word was exposed for 200 msec on the
tachistoscope. The subject responded by press-
ing the relevant response key. At the end of
the encoding trials, the subject was allowed to
rest for 1 min and was then asked to recall as
many words as he could. In Experiment 3, this
final recall task was unexpected—thus the initial
encoding phase may be considered an incidental
learning task—while in Experiment 4, subjects
were informed at the beginning of the session
that they would be required to recall the words.

Pilot studies had shown that the recall level
in this situation tends to be low. Thus, to boost
recall, and to examine the effects of encoding
level on recall more clearly, half of the words in
the present study were presented twice. In all,
48 different words were used, but 24 were pre-
sented twice, making a total of 72 trials. Of the
24 words presented once only, 4 were presented
under each of the six conditions (three types of
question X yes-no). Similarly, of the 24 words
presented twice, 4 were presented under each of
the six conditions. When a word was repeated,
it always occurred as the 20th item after its first
presentation; that is, the lag between first and
second presentations was held constant. On its
second appearance, the same type of question was
asked as on the word’s first appearance but, for

‘500 600 700 800 900 1000

INITIAL DECISION TIME (msec)

FIGURE 2. Proportion of words recognized as
a function of initial decision time (Experiment
2).

rhyme and sentence questions, a different specific
question was asked. Thus, when the word TRAIN
fell into the rhyme-yes category, the question
asked on its first presentation might have been
“Does the word rhyme with BRAIN?” while on
the second presentation the question might have
been “Does the word rhyme with CRANE?” For
case questions the same question was asked on the
two occurrences since each subject was given the
same question throughout the experiment ‘(e.g.,
“Is the word in lowercase?”). This procedure
was adopted as early work had shown that sub-
jects’ response latencies were greatly slowed if
they had to associate yes responses to both upper-
case and lowercase words.

A constant pool of 48 words was used for all
subjects. The words were common concrete
nouns. Five presentation formats were constructed
in which the words were randomly allocated to
the various encoding conditions. Four subjects
were tested on each format: Two made yes
responses with their right hand on the right
response key while two used the left-hand key
for yes responses. The 20 student subjects were
paid for their services. They were told that the
experiment concerned perception and reaction time;
they were warned that some words would occur
twice, but they were not informed of the final
recall test.

Results and discussion. Response laten-
cies are shown in Table 3. For each sub-
ject and each experimental condition (e.g.,
case-yes) the median response latency was
calculated for the eight words presented on
their first occurrence (i.e., the four words
presented only once, and the first occurrence
of the four repeated words). The median

276 FERGUS I. M. CRAIK AND ENDEL TULVING

TABLE 3
RESPONSE LATENCIES FOR EXPERIMENTS

3 AND 4

Condition Case Rhyme Sentence

1st presentation

Incidental
(Exp. 3)

Yes
No

Intentional
(Exp. 4)

689
70S

816
725

870
872

Yes
No

687
685

796
768

897
911

2nd presentation

Incidental
(Exp. 3)
Yes
No

Intentional
(Exp. 4)
Yes
No

616
634

609
599

689
725

684
716

771
856

793
866

Note. Mean medians of response latencies are presented.

latency was also calculated for the four
repeated words on their second presentation.
Only correct responses were included in the
calculation of the medians. Table 3 shows
the mean medians for the various experi-
mental conditions. There was a systematic
increase in response latency from case ques-
tion to sentence questions. Also, response
latencies were more rapid on the word’s
second presentation—this was especially
true for yes responses. These observations
were confirmed by an analysis of variance.
The effect of question type was significant,
F(2, 38) = 14.4, p < .01, but the effect of response type was not (F < 1.0). Repeated words were responded to reliably faster, F(l, 19) = 10.3, p < .01 and the Number of Presentations X Response Type (yes-no} interaction was significant, F(l, 19) = 5.33, p < .05.

Thus, again, deeper level questions took
longer to process, but yes responses took
no longer than no responses. The extra
facilitation shown by positive responses on
the second presentation may be attributable
to the greater predictive value of yes ques-

tions. For example, the second presenta-
tion of a rhyme question may remind the
subject of the first presentation and thus
facilitate the decision.

Figure 3 shows the recall probabilities
for words presented once or twice. There
is a marked effect of question type (sen-
tence > rhymes > case); retention is again
superior for words given an initial yes
response and recall of twice-presented words
is higher than once-presented words. An
analysis of variance confirmed these obser-
vations. Semantic questions yielded higher
recall, F(2, 38) = 36.9, p < .01; more yes responses than no responses were recalled, F(l, 19) = 21.4, p < .01; two presenta- tions increased performance, F(l, 19) = 33.0, p < .01. In addition, semantically encoded words benefited more from the sec- ond presentation, as shown by the signifi- cant Question Level X Number of Presen- tations interaction, F(2, 38) = 10.8, p < .01.

Experiment 3 thus confirmed that deeper
levels of encoding take longer to accomplish
and that yes and no responses take equal
encoding times. More important, semantic
questions led to higher recall performance
and more yes response words were recalled
than no response words. These basic re-
sults thus apply as well to recall as they do
to recognition. Experiments 1-3 have used
an incidental learning paradigm; there are
good reasons to believe that the incidental
nature of the task is not critical for the ob-
tained pattern of results to appear (Hyde
& Jenkins, 1973). Nevertheless, it was
decided to verify Hyde and Jenkins’ con-
clusion using the present paradigm. Thus,
Experiment 4 was a replication of Experi-
ment 3, but with the difference that sub-
jects were informed of the final recall task
at the beginning of the session.

Experiment 4
Method. The material and procedures were

identical to those in Experiment 3 except that
subjects were informed of the final free recall
task. They were told that the memory task was
of equal importance to the initial phase and that
they should thus attempt to remember all words
shown in the tachistoscope. A 10-min period was
allowed for recall. The subjects were 20 college

DEPTH OF PROCESSING AND WORD RETENTION 277

CASE RHYME SENTENCE CASE RHYME SENTENCE
LEVEL OF PROCESSING

FIGURE 3. Proportion of words recalled as a function of the initial task
(Experiment 3).

students, none of whom had participated in Experi-
ments 1, 2, or 3.

Results and discussion. The response
latencies are shown in Table 3. These data
are very similar to those from Experiment
3, indicating that subjects took no longer to
respond under intentional learning instruc-
tions. Analysis of variance showed that
deeper levels were associated with longer
decision latencies, F(2, 38) = 27.7, p < .01, and that second presentations were re- sponded to faster, F(l, 19) = 18.9, p < .01. No other effect was statistically reliable.

With regard to the recall results, the
analysis of variance yielded significant
effects of processing level, F(2, 38) = 43.4,
p < .01, of repetition, F(\, 19) = 69.7, p < .01, and of response type (yes-no}, F(l, 19) = 13.9, p < .01. In addition, the Num- ber of Presentations X Level of Processing interaction, F(2, 38) = 12.4, p < .01, and the Number of Presentations X Response Type (yes^no) interaction, F(\, 19) = 7.93, p < .025, were statistically reliable. Figure 4 shows that these effects were attributable to superior recall of sentence decisions,

twice-presented words and yes responses.
Words associated with semantic questions
and with yes responses showed the greatest
enhancement of recall after a second presen-
tation.

To further explore the effects of inten-
tional versus incidental conditions more
comprehensive analyses of variance were
carried out, involving the data from both
Experiments 3 and 4. For the latency data,
there was no significant effect of the inten-
tional-incidental manipulation, nor did the
intentional-incidental factor interact with
any other factor. Thus, knowledge of the
final recall test had no effect on subjects’
decision times. In the case of recall scores,
intentional instructions yielded superior
performance, F(l, 38) = 11.73, p < .01, and the Intentional-Incidental X Number of Presentations interaction was significant, F(l, 38) = 5.75, p < .05. This latter ef- fect shows that the superiority of inten- tional instructions was greater for twice- presented items. No other interaction in- volving the incidental-intentional factor was significant. It may thus be concluded that the pattern of results obtained in the present

278 FERGUS I. M. CRAIK AND ENDEL TULVtNG

CASE RHYME SENTENCE CASE RHYME SENTENCE
LEVEL OF PROCESSING

FIGURE 4. Proportion of words recalled as a function of the initial task
(Experiment 4).

experiments does not depend critically on
incidental instructions.

The findings that intentional recall was
superior to incidental recall, but that deci-
sion times did not differ between intentional
and incidental conditions, is at first sight
contrary to the theoretical notions proposed
in the introduction to this article. If recall
is a function of depth of processing and
depth is indexed by decision time, then
clearly differences in recall should be asso-
ciated with differences in initial response
latency. However, it is possible that fur-
ther processing was carried out in the inten-
tional condition, after the orienting task
question was answered, and was thus not
reflected in the decision times.

Discussion of Experiments 1—4

Experiments 1-4 have provided empirical
flesh for the theoretical bones of the argu-
ment advanced by Craik and Lockhart
(1972). When semantic (deeper level)
questions were asked about a presented
word, its subsequent retention was greatly
enhanced. This result held for both recog-
nition and recall; it also held for both inci-

dental and intentional learning (Hyde &
Jenkins, 1969, 1973; Till & Jenkins, 1973).
The reported effects were both robust, and
large in magnitude: Sentence-^« words
showed recognition and recall levels which
were superior to case-wo words by a factor
ranging from 2.4 to 13.6. Plainly, the na-
ture of the encoding operation is an impor-
tant determinant of both incidental and
intentional learning and hence of retention.

At the same time, some aspects of the
present results are clearly inconsistent with
the depth of processing formulation outlined
in the introduction. First, words given a
yes response in the initial task were better
recalled and recognized than words given a
no response, although reaction times to yes
and no responses were identical. Either
reaction time is not an adequate index of
depth, or depth is not a good predictor of
subsequent retention. We will argue the
former case. If depth of processing (defined
loosely as increasing semantic-associative
analysis of the stimulus) is decoupled from
processing time, then on the one hand the
independent index of depth has been lost,
but on the other hand, the results of Experi-

DEPTH OF” PROCESSING AND WORD RETENTION 273

rttents 1-4 can be described in terms of
qualitative differences in encoding opera-
tions rather than simply in terms of in-
creased processing times. The following
section describes evidence relevant to the
question of whether retention performance
is primarily a function of “study time” or
the qualitative nature of mental operations
carried out during that time.

The results obtained under intentional
learning conditions (Experiment 4) are
also not well accommodated by the initial
depth of processing notions. If the large
differences in retention found in Experi-
ments 1-3 are attributable to different
depths of processing in the rather literal
sense that only structural analyses are acti-
vated by the case judgment task, phonemic
analyses are activated by rhyme judgments,
and semantic analyses activated by category
or sentence judgments, then surely under
intentional learning conditions the subject
would analyse and perceive the name and
meaning of the target word with all three
types of question. In this case equal reten-
tion should ensue (by the Craik and Lock-
hart formulation), but Experiment 4 showed
that large differences in recall were still
found.

A more promising notion is that retention
differences should be attributed to degrees
of stimulus elaboration rather than to differ-
ences in depth. This revised formulation
retains the important point (borne out by
Experiments 1-4) that the qualitative na-
ture .of encoding operations is critical for
the establishment of a durable trace, but
gets away from the notions that semantic
analyses necessarily always follow structural
analyses and that no meaning is involved in
shallow processing tasks.

Discussion of the best descriptive frame-
work for these studies will be resumed after
further experiments are reported; for the
moment, the term depth is retained to signify
greater degrees of semantic involvement.
Before further discussions of the theoretical
framework are presented, the following sec-
tion describes attempts to evaluate the rela-
tive effects of processing time and the qual-
itative nature of encoding operations on the
retention of words.

PROCESSING TIME VERSUS ENCODING
OPERATIONS

As a first step, the data from Experiment
2 were examined for evidence relating the
effects of processing time to subsequent
memory performance. At first sight, Ex-
periment 2 provided evidence in line with
the notion that longer categorization times
are associated with higher retention levels—
Figure 2 demonstrated linear relationships
between initial decision latency and sub-
sequent recognition performance. How-
ever, if it is processing time which deter-
mines performance, and not the qualitative
nature of the task, then within one task,
longer processing times should be associated
with superior memory performance. That
is, with the qualitative differences in pro-
cessing held constant, performance should
be determined by the time taken to make the
initial decision. On the other hand, if dif-
ferences in encoding operations are critical
for differences in retention, then memory
performance should vary between orienting
tasks, but within any given task, retention
level should not depend on processing time.

This point was explored by analyzing the
data from Experiment 2 in terms of fast and
slow categorization times. The 10 response
latencies for each subject in each condition
were divided into the 5 fastest responses
and the 5 slowest responses. Next, mean
recognition probabilities for the fast and
slow subsets of words were calculated across
all subjects for each condition. The results
of this analysis are shown in Figure 5;
mean medians for the response latencies in
each subset are plotted against recognition
probabilities. If processing time were
crucial, then the words which fell into the
slow subset for each task should have been
recognized at higher levels than words which
elicited fast responses. Figure 5 shows
that this did not happen; Slow responses
were recognized little better than fast
responses within each level of analysis. On
the other hand, the qualitative nature of
the task continued to exert a very large
effect on recognition performance, suggest-
ing again that it is the nature of the encod-

280 FERGUS I. M. CRAIK AND ENDEL TULVING

I.U

.9
Q
UJ 8
N
z |7
O 6
LJ

.5

i .4h-
CC 3

§ ‘Z

. 1

0

YES Decisions

_
A

–
–

• _„

•

– i

^

NO Decisions

.

–
A-̂ 4 SENTENCE
• — • RHYME A
D — D CASE _ -̂- ”

* —

•

• °~~ ̂ ^̂ — *
•

1 1 1 I 1 1 1 t
500 600 700 800 900 500 600 700

INITIAL DECISION TIME (msec)

800 900

FIGURE 5. Recognition of words as a function of task and initial decision
time: Data partitioned into fast and slow decision times (Experiment 2).

ing operations and not processing time which
determines memory performance.

For both yes and no responses, slow case
categorization decisions took longer than
fast sentence decisions. However, words
about which subjects had made sentence
decisions showed higher levels of recogni-
tion; 73% as opposed to 17% for yes re-
sponses and 45% as opposed to 17% for no
responses. No statistical analysis was
thought necessary to support the conclusion
that task rather than time is the crucial
aspect in these experiments. Since the
point is an important one, however, a fur-
ther experiment was conducted to clinch the
issue. Subjects were given either a com-
plex structural task or a simple semantic
task to perform; it was predicted that the
complex structural task would take longer
to accomplish but that the semantic task
would yield superior memory performance.

Experiment 5
Method. The purpose of Experiment S was to

devise a shallow nonsemantic task which was
difficult to perform and would thus take longer
than an easy but deeper semantic task. In this
way, further evidence on the relative contribu-
tions of processing time and processing depth to
memory performance could be obtained. In both
tasks, a five-letter word was shown in the tachisto-
scope for 200 msec and the subject made a yes-wo
decision about the word. The nonsemantic deci-
sion concerned the pattern of vowels and con-
sonants which made up the word. Where V =

vowel and C = consonant, the word brain could
be characterized as CCVVC, the word uncle as
VCCCV, and so on. Before each nonsemantic
trial the subject was shown a card with a partic-
ular consonant-vowel pattern typed on it; after
studying the card as long as necessary, the sub-
ject looked into the tachistoscope and the word
was exposed. The experiment was again described
as a perceptual, reaction time study concerning
different aspects of words and the subject was
instructed to respond as rapidly as possible by
pressing one of two response keys. The seman-
tic task was the sentence task from previous
studies in the series. In this case, the subject
was shown a card with a short sentence typed
on it; the sentence had one missing word, thus
the subject’s task was to decide whether the word
on the tachistoscope screen would fit the sentence.
Examples of sentence-jiM trials are: “The man
threw the ball to the ” (CHILD) and
“Near her bed she kept a ” (CLOCK).
On sentence-Mo trials an inappropriate noun from
the general pool was exposed on the tachistoscope.
Again the subject responded as rapidly as pos-
sible. The subjects were not informed of the
subsequent memory test.

The pool of words used consisted of 120 high
frequency, concrete five-letter nouns. Each sub-
ject received 40 words on the initial decision
phase of the task and was then shown all 120
words, 40 targets and 80 distractors mixed ran-
domly, in the second phase. He was then asked
to recognize the 40 words he had been shown on
the tachistoscope by circling exactly 40 words.
Two forms of the recognition test were typed with
the same 120 words randomized differently. In
all, 24 subjects were tested in the experiment.
The pool of 120 words was arbitrarily parti-
tioned into three blocks of 40 words; the first 8
subjects received one block of 40 as targets and

DEPTH OF PROCESSING AND WORD RETENTION 281

the remaining 80 words served as distractors;
the second 8 subjects received the second block
of 40 words as targets and the third 8 subjects
received the third block of 40—in all cases the
remaining 80 words formed the distractor pool.
Within each group of 8 subjects who received
the same 40 target words, 4 received one form
of the recognition test and 4 received the other
form. Finally, within each group of 4 subjects,
each word was rotated so .that it appeared (for
different subjects) in all four conditions: non-
semantic yes and no and semantic yes and no.

Each subject was tested individually. After
the two tasks had been explained, he was given a
few practice trials, then received 40 further trials,
10 under each experimental condition. The order
of presentation of conditions was randomized.
After a brief rest period the subject was given
the recognition list and told to circle exactly 40
words (those he had just seen on the tachisto-
scope), guessing if necessary. The subjects were
24 undergraduate students of both sexes, paid
for their services.

Results. The results of the experiment
are straightforward. Table 4 shows that the
nonsemantic task took longer to accomplish
but that the deeper sentence task gave rise
to higher levels of recognition. Decisions
about consonant-vowel structure of words
were substantially slower than sentence
decisions'(1.7 sec as opposed to .85 sec)
and this difference was significant statis-
tically, F(l, 23) = 11.3, p < .01. Neither the response type (yes-no) nor the inter- action was significant. For recognition, the analysis of variance showed that sentence decisions gave rise to higher recognition, P(\, 23) = 40.9, p < .001; yes responses were recognized better than no responses, F(l, 23) = 10.6, p < .01, but the Task X Response Type interaction was not signifi- cant.

Experiment 5 has thus confirmed the con-
clusion from the reanalysis of Experiment
2; that it is the qualitative nature of the task
—we argue, depth of processing—and not
the amount of processing time, which deter-
mines memory performance. Figure 2
illustrates that a deep semantic task takes
longer to accomplish and yields superior
memory performance, but when the two
factors are separated it is the task which is
crucial, not processing time as such.

One constant feature of Experiments 1-4
has been the superior recall or recognition
of words given a yes response in the initial

TABLE 4
DECISION LATENCY AND RECOGNITION PERFORM-

ANCE FOR WORDS AS A FUNCTION OF THE INITIAL
TASK (EXPERIMENT 5)

Response type
Level of processing

Structure Sentence

Response latency (sec)

Yes
No

1

.70

1.74

.83

.88

Proportion recognized
Yes
No

.57

.50

.82

.69

perceptual phase. This result has also
been reported by Schulman (1974). The
reasons for the better retention of yes re-
sponses are not immediately apparent; for
example, it is not obvious that positive
responses require deeper processing before
the initial perceptual decision can be made.
This problem invites a closer investigation
of the yes-no difference and may perhaps
force a further reevaluation of the concept of
depth.

POSITIVE AND NEGATIVE CATEGORIZATION
DECISIONS

Why are words to which positive re-
sponses are made in the perceptual-decision
task better remembered ? As discussed pre-
viously, it does not seem intuitively reason-
able that words associated with yes responses
require deeper processing before the deci-
sion is made. However, if high levels of
retention are associated with “rich” or
“elaborate” encodings of the word (rather
than deep encodings), the differences in
retention between positive and negative
words become understandable. In cases
where a positive response is made, the
encoding question and the target word can
form a coherent, integrated unit. This
integration would be especially likely with
semantic questions: for example, “A four-
footed animal?” (BEAR) or “The boy met a

— on the street” (FRIEND). How-
ever,, integration of the question and tar-
get word would be much less likely in the
negative case: “A four-footed animal ?”

282 FERGUS I. M. CRAIK AND ENDEL TULVING

(CLOUD) or “The boy met a on
the street” (SPEECH). Greater degrees of
integration (or, alternatively, greater de-
grees of elaboration of the target word)
may support higher retention in the sub-
sequent test. This factor of integration or
congruity (Schulman, 1974) between tar-
get word and question would also apply to
rhyme questions but not to questions about
typescript: If the target word is in capital
letters (a yes decision), the word’s encod-
ing would be elaborated no more than if the
word had been presented in lowercase type
(a no decision). This analysis is based on
the premise that effective elaboration of an
encoding requires further descriptive attri-
butes which (a) are salient, or applicable to
the event, and (b) specify the event more
uniquely. While positive semantic and
rhyme decisions fit this description, neg-
ative semantic and rhyme decisions and
both types of case decision do not. In line
with this analysis is the finding from Experi-
ments 1-4 that while positive decisions are
associated with higher retention levels for
semantic and rhyme questions, words elicit-
ing positive and negative decisions are
equally well retained after typescript judg-
ments.

If the preceding argument is valid, then
questions leading to equivalent elaboration
for positive and negative decisions should be
followed by equivalent levels of retention.
Questions which appear to meet the case
are those of the type “Is the object bigger
than a chair?” In this case both positive
target words (HOUSE, TRUCK) and negative
target words (MOUSE, PIN) should be en-
coded with equivalent degrees of elabora-
tion; thus, they should be equally well
remembered. This proposition was tested
in Experiment 6.

Experiment 6
Method, Eight descriptive dimensions were

used in the study: size, length, width, height,
weight, temperature, sharpness, and value. For
each of these dimensions, a set of eight concrete
nouns was generated, such that the dimension was a
salient descriptive feature for the words in each set
(e.g., size-ELEPHANT, MOUSE; value-DiAMOND,
CRUMB). The words were chosen to span the com-
plete range of the relevant dimension (e.g., from
very small to very large; very hot to very cold).

For each set an additional reference object was
chosen such that half of the objects represented by
the word set were “greater than” the reference ob-
ject and half of the objects were “less than” the
referent. The reference object was always used
in the question pertaining to that dimension;
examples were “Taller than a man?” (STEEPLE-
yes; CHILD-WO), “More valuable than $10?”
(JEWEL-JIM; BUTTON-WO). ” Sharper than a
fork?” (NEEDLE-JIM; CLUB-no). For half of the
subjects, the question was reversed in sense, so
that words given a yes response by one group of
subjects were given a no response by the other
group. Thus, “Taller than a man?” became
“Shorter than a man?” (STEEPLE-WO; CHILD-
yes).

Each subject was asked questions relating to
two dimensions; he thus answered 16 questions—
4 yielding positive responses and 4 yielding neg-
ative responses for each dimension. Four dif-
ferent versions of the questions and targets were
constructed, with two different dimensions being
used in each version. Four subjects received each
version—two received the original questions (e.g.,
“heavier than . . .” “hotter than . . .”) and two
received the questions reversed (“lighter than . . .”
“colder than . . .”). Thus each subject received
16 questions; both question type and response
type (yes-no) were randomized. Subjects were
16 undergraduate students of both sexes; they
were paid for their services.

On each trial, the subject looked into a tachisto-
scope; the question was presented auditorily, and
2 sec later the target word was exposed for 1
sec. The subject responded by pressing the ap-
propriate one of two keys. Subjects were again
told that they had to make rapid judgments about
words; they were not informed of the retention
test. After completing the 16 question trials,
subjects were asked to recall the target words.
Each subject was reminded of the questions he
had been asked. Thus, in this study, memory
was assessed in the presence of the original
questions.

Results. Again, the results are much
easier to describe than the procedure.
Words given yes responses were recalled
with a probability of .36, while words given
no responses were recalled with a probabil-
ity of .39. These proportions did not differ
significantly when tested by the Wilcoxon
test. Thus, when positive and negative
decisions are equally well encoded, the re-
spective sets of target words are equally well
recalled. The results of this demonstration
study suggest that it is not the type of
response given to the presented word that is
responsible for differences in subsequent
recall and recognition, but rather the rich-

DEPTH OF PROCESSING AND WORD RETENTION 283

ness or elaborateness of the encoding. It
is possible that negative decisions in Experi-
ments 1-4 were associated with rather poor
encodings of the presented words—they did
not fit the encoding question and thus did
not form an integrated unit with the ques-
tion. On the other hand, positive responses
would be integrated with the question, and
thus, arguably, formed more elaborate en-
codings which supported better retention
performance.

Experiment 7 was an attempt to manip-
ulate encoding elaboration more directly.
Only semantic information was involved in
this study. All encoding questions were
sentences with a missing word; on half of
the trials the word fitted the sentence (thus
all queries were congruous in Schulman’s
terms). The degree of encoding elabora-
tion was varied by presenting three levels
of sentence complexity, ranging from very
simple, • spare sentence frames (e.g., “He
dropped the “) to complex, elaborate
frames (e.g., “The old man hobbled across
the room and picked up the valuable
from the mahogany table”). The word
presented was WATCH in both cases. Al-
though the second sentence is no more
predictive of the word, it should yield a
more elaborate encoding and thus superior
memory performance.

Experiment 7
Method. Three levels of sentence complexity

were used: simple, medium, and complex. Each
subject received 20 sentence frames at each level
of complexity; within each set of 20 there were
10 yes responses and 10 no responses. The 60
encoding trials were randomized with respect
to level of complexity and response type. A
constant pool of 60 words was used in the experi-
ment, but two completely different sets of en-
coding questions were constructed. Words were
randomly allocated to sentence level and response
type in the two sets (with the obvious constraint
that yes and no words clearly fitted or did not
fit the sentence frame, respectively). Within
each set of sentence frames, two different ran-
dom presentation orders were constructed. Five
subjects were presented with each format thus
generated and 20 subjects were tested in all.

The words used were common nouns. Examples
of sentence frames used are: simple, “She cooked
the ” “The • is torn”; medium, “The

frightened the children” and “The ripe
tasted delicious”; complex, “The great bird

swooped down and carried off the struggling
” and “The small lady angrily picked up

the red .” The sentence frames were
written on cards and given to the subject. After
studying it he looked into the tachistoscope with
one hand on each response key. After a ready
signal the word was presented for 1.0 sec and
the subject responded yes or no by pressing the
appropriate key. The words were exposed for
a longer time in this study since the questions
were more complex. Subjects were again told
that the experiment was concerned with percep-
tion and speed of reaction and that they should
thus respond as rapidly as possible. No mention
was made of a memory test. The 20 subjects
were tested individually. They were undergrad-
uate students of both sexes, paid for their services.

After completing the 60 encoding trials, sub-
jects were given a short rest and then asked to
recall as many words as they could from the first
phase of the experiment. They were given 8 min
for free recall. After a further rest, they were
given the deck of cards containing the original
sentence frames (in a new random order) and
asked to recall the word associated with each
sentence. Thus there were two retention tests in
this study: free recall followed by cued recall.

Results. Figure 6 shows the results.
For free recall, there is no effect of sentence
complexity in the case of no responses, but
a systematic increase in recall from simple
to complex in the case of yes responses.
The provision of the sentence frames as
cues did not enhance the recall of no re-
sponses, but had a large positive effect on
the recall of yes responses; the effect of
sentence complexity was also amplified in
cued recall. These observations were con-

i.o
.9

Q
Ld .8

SIMPLE MEDIUM COMPLEX

SENTENCE TYPE

FIGURE 6. Proportion of words recalled as a
function of sentence complexity (Experiment 7).
(CR = cued recall, NCR = noncued recall.)

284 FERGUS I. M. CRAIK AND ENDEL TULVING

firmed by analysis of variance. In free
recall, a greater proportion of words given
positive responses were recalled than those
given negative responses, F(l, 19) = 18.6,
p < .001; the overall effect of complexity was not significant, F(2, 38) = 2.37, p >
.05, but the interaction between complexity
and yes-no was reliable, F(2, 38) = 3.78,
p < .05. A further analysis, involving posi- tive responses only, showed that greater sentence complexity was reliably associated with higher recall levels, F(2, 38) = 4.44, p < .025. In cued recall, there were sig- nificant effects of response type, F(\, 19) = 213, p < .001, complexity, F(2, 38) = 49.2, p < .001, and the Complexity X Re- sponse Type interaction, F(2, 38) = 19.2, p < .001. An overall analysis of variance, incorporating both free and cued recall, was also carried out and this analysis revealed significantly higher performance for greater complexity, F(2, 38) = 36.5, p < .001, for positive target words, F(\, 19) = 139, p < .001, and for cued recall rela- tive to free recall, F(\, 19) = 100, p < .001. All the interactions were significant at the p < .01 level or better; the descrip- tion of these effects is provided by Figure 6.

Experiment 7 has thus demonstrated that
more complex, elaborate sentence frames
do lead to higher recall, but only in the case
of positive target words. Further, the
effects of complexity and response type are
greatly magnified by reproviding the sen-
tence frames as cues.

These results do not fit the original simple
view that memory performance is deter-
mined only by the nominal level of pro-
cessing. In all conditions of Experiment 7
semantic processing of the target word was
necessary, yet there were still large differ-
ences in performance depending on sentence
complexity, the relation between target word
and the sentence context, and the presence
or absence of cues. It seems that other
factors besides the level of processing re-
quired to make the perceptual decision are
important determinants of memory perform-
ance.

The notion of code elaboration provides
a more satisfactory basis for describing the
results. If a presented word does not fit

the sentence frame, the subject cannot form
a unified image or percept of the complete
sentence, the memory trace will not rep-
resent an integrated meaningful pattern,
and the word will not be well recalled. In
the case of positive responses, such coherent
patterns can be formed and their degree of
cognitive elaborateness will increase with
sentence complexity. While increased elab-
oration by itself leads to some increase in
recall (possibly because richer sentence
frames can be more readily recalled) per-
formance is further enhanced when part of
the encoded trace is reprovided as a cue.
It is well established that cuing aids recall,
provided that the cue information has been
encoded with the target word at presenta-
tion and thus forms part of the same encoded
unit (Tulving & Thomson, 1973). The
present results are consistent with the find-
ing, but may also be interpreted as showing
that a cue is effective to the extent that the
cognitive system can encode the cue and the
target as a congruous, integrated unit.
Elaborate cues by themselves do not aid
performance even if they were presented
with the target word at input, as shown by
the poor recall of negative response words.
It is also necessary that the target and the
cue form a coherent, integrated pattern.

Schulman (1974) reported results which
are essentially identical to the results of
Experiment 7. He found better recall of
congruous than incongruous phrases; he
also found that cuing benefited congruously
encoded words much more than incongruous
words. Schulman suggests that congruent
words can form a relational encoding with
their context, and that the context can then
serve as an effective redintegrative cue at
recall (Begg, 1972; Horowitz & Prytulak,
1969). In these terms, Experiment 7 has
added the finding that the semantic richness
of the context benefits congruent encodings
but has no effect on the encoding of incon-
gruous words.

Is the concept of depth still useful in
describing the present experimental results,
or are the findings better described in terms
of the “spread” of encoding where spread
refers to the degrees of encoding elaboration
or the number of encoded features? These

DEPTH OF PROCESSING AND WORD RETENTION 285

questions will be taken up in the general
discussion, but in outline, we believe that
depth still gives a useful account of the
major qualitative shifts in a word’s encod-
ing (from an analysis of physical features
through phonemic features to semantic prop-
erties). Within one encoding domain, how-
ever, spread or number of encoded features
may be better descriptions. Before grap-
pling with these theoretical issues, three final
short experiments will be described. The
findings from the preceding experiments
were so robust that it becomes of interest
to ask under what conditions the effects of
differential encoding disappear. Experi-
ments 8, 9, and 10 were attempts to set
boundary limits on the phenomena.

FURTHER EXPLORATIONS OF DEPTH AND
ELABORATION

The three studies described in this sec-
tion were undertaken to examine further
aspects of depth of processing and to throw
more light on the factors underlying good
memory performance. The first experi-
ment explored the idea that the critical dif-
ference between case-encoded and sentence-
encoded words might lie in the similarity
of encoding operations within the group of
case-encoded words. That is, each case-
encoded word is preceded by the same ques-
tion, “Is the word in capital letters?”,
whereas each rhyme-encoded and sentence-
encoded word has its own unique question.
At retrieval, it is likely that the subject uses
what he can remember of the encoding
question to help him retrieve the target
word, Plausibly, encoding questions which
were used for many target words would be
less effective as retrieval cues since they
do not uniquely specify one encoded event
in episodic memory. This overloading of
retrieval cues would be particularly evident
for case-encoded words. It is possible to
extend the argument to rhyme-encoded
words also; although each target word
receives a different rhyme question, pho-
nemic differences may not be so unique or
distinctive as semantic differences (Lock-
hart, Craik, & Jacoby, 1975).

Some empirical support for these ideas
may be drawn from two unpublished studies
by Moscovitch and Craik (Note 1). The
first study used the same paradigm as the
present series and compared cued with non-
cued recall, where the cues were the original
encoding questions. It was found that cuing
enhanced recall, and that the effect of cuing
was greater with deeper levels of encoding.
Thus the encoding questions do help
retrieval, and their beneficial effect is
greatest with semantically encoded words.
The second study showed that when several
target words shared the same encoding
question (e.g., “Rhymes with train?” BRAIN,
CRANE, PLANE; “Animal category?” LION,
HORSE, GIRAFFE), the sharing manipulation
had an adverse effect on cued recall. Fur-
ther, the adverse effect was greatest for
deeper levels of encoding, suggesting that
the normal advantage to deeper levels is
associated with the uniqueness of the en-
coded question-target complex, and that
when this uniqueness is removed, the
mnemonic advantage disappears.

These ideas and findings suggest an
experiment in which a case-encoded word
is made more unique by being the one word
in an encoding series to be encoded in this
way. In this situation the one case word
might be remembered as well as a word,
which, nominally, received deeper process-
ing. Such an experiment in its extreme
form would be expensive to conduct, in that
one word forms the focus of interest. Ex-
periment 8 pursues the idea of uniqueness
in a less extreme form. Three groups of
subjects each received 60 encoding trials;
each trial consisted of a case, rhyme, or
category question. However, each group
of subjects received a different number of
trials of each question type: either 4 case,
16 rhyme, and 40 category trials; 16, 40,
and 4 trials; or 40, 4, and 16 trials, respec-
tively. The prediction was that while the
typical pattern of results would be found
when 40 trials of one type were given, sub-
sequent recognition performance would be
enhanced with smaller set sizes; this en-
hancement would be especially marked for
the case level of encoding.

286 FERGUS I. M. CRAIK AND ENDEL TULVING

TABLE 5
DESIGN AND RESULTS OF EXPERIMENT 8

Experimental
condition

Case Rhyme

Yes No Yes No

Category
Yes No

Design: Number of trials per condition

Group 1
Group 2
Group 3

2
8
20

2
8
20

8
20
2

8
20
2

20
2
8

20
2
8
Proportion recognized
Group 1
Group 2
Group 3

Set size 4
Set size 16
Set size 40

.50

.51

.49

.50
.51
.49
.36

.40

.43

.36
.40
.43

.73

.66

.90

.90
.73
.66

.47

.54

.70
.70
.47
.54
.88

.95

.91

.95
.91
.88
.70

.64

.68

.64
.68
.70

Experiment 8
Method. Three groups of subjects were tested.

Group 1 received 4 case questions, 16 rhyme
questions, and 40 category questions. Group 2
received 16, 40, and 4, respectively, while Group
3 received 40, 4, and 16, respectively. At each
level of encoding, half of the questions were de-
signed to elicit yes responses and half no responses.
Thus each group received 60 trials; question type
and response type were randomized. The design
is shown in Table 5.

The subjects were tested individually. Each
question was read by the experimenter while the
subject looked in the tachistoscope; the word was
exposed for 200 msec and the subject responded
by pressing one of two response keys. The sub-
jects were informed that the test was a perceptual-
reaction time task; the subsequent memory test
was not mentioned. After completing the 60 en-
coding trials, each subject was given a sheet
containing the 60 target words plus 120 distrac-
tors. He was told to check exactly 60 words-—
those words he had seen on the tachistoscope.

The same pool of 60 common nouns was used
as targets throughout the experiment. Within
each experimental group there were four pre-
sentation lists; in each case Lists 1 and 2 differed
only in the reversal of positive and negative deci-
sions (e.g., category-jiej in List 1 became cat-
egory-no in List 2 ) . Lsits 3 and 4 contained a
fresh randomization of the 60 words, but again
Lists 3 and 4 differed between themselves only
in the reversal of positive and negative responses.
In all, 32 subjects were tested in the experiment;
11 each in Groups 1 and 2, and 10 in Group 3.
Two or three subjects were tested under each
randomization condition.

Results. Table 5 shows the proportion
recognized by each group. Each group
shows the typical pattern of results already

familiar from Experiments 1-4; there is no
evidence of a perturbation due to set size.
Table 5 also shows the recognition results
organized by set size; it may now be seen
that set size does exert some effect, most
conspicuously on rhyme-yes responses.
However, the differences previously attri-
buted to different levels of encoding were
certainly not eliminated by the manipula-
tion of set size; in general, when set size
was held constant (across groups), strong
effects of question type were still found.

To recapitulate, the argument underlying
Experiment 8 was that in the standard ex-
periment, the encoding operation for case
decisions is, in some sense, always the same;
for rhyme decisions, it is somewhat similar
from word to word, and is most dissimilar
among words in the category task. If the
isolation effect in memory (see Cermak,
1972) is a consequence of uniqueness of
encoding operations, then when similar en-
codings (e.g., “case decision” words) are
few in number, they should also be encoded
uniquely, show the isolation effect, and thus
be well recalled. Table 5 shows that reduc-
ing the number of case-encoded words from
40 to 4 did not enhance their recall, thus
lack of isolation cannot account for their low
retention. On the other hand, a reduction
in set size did enhance the recall of rhyme-
encoded words, thus isolation effects may
play some part in these experiments,
although they cannot account for all aspects

DEPTH OF PROCESSING AND WORD RETENTION 287

of the results. Finally, it may be of some
interest that recall proportions for rhymes-
Set Size 4 are quite similar to category-Set
Size 40 (.90 and .70 vs. .88 and .70); this
observation is at least in line with the notion
that when rhyme encodings are made more
unique, their recall levels are equivalent to
semantic encodings.

Experiment 9: A Classroom Demonstration

Throughout this series of experiments,
experimental rigor was strictly observed.
Words were exposed for exactly 200 msec;
great care was exercised to ensure that
subjects would not inform future subjects
that a memory test formed part of the ex-
periment; subjects were told that the experi-
ments concerned perception and reaction
time; response latencies were painstakingly
recorded in all cases. One of the authors,
by nature more skeptical than the other, had
formed a growing suspicion that this rigor
reflected superstitious behavior rather than
essential features of the paradigm. This
feeling of suspicion was increased by the
finding of the typical pattern of results in
Experiment 9, which was conducted under
intentional learning conditions. Accord-
ingly, a simplified version of Experiment 2
was formulated which violated many of the
rules observed in previous studies. Sub-
jects were informed that the main purpose
of the experiment was to study an aspect of
memory; thus the final recognition test was
expected and encoding was intentional
rather than incidental. Words were pre-
sented serially on a screen at a 6-sec rate;
during each 6-sec interval subjects recorded
their response to the encoding question.
Indeed, the subjects were tested in one group
of 12 in a classroom situation during a course
on learning and memory; they recorded
their own judgments on a question sheet and
subsequently attempted to recognize the tar-
get words from a second sheet. Reaction
times were not measured.

The point of this study was not to attack
experimental rigor, but rather to deter-
mine to what extent the now familiar pat-
tern of results would emerge under these
much looser conditions. If such a pattern
does emerge, it will force a further examina-
tion of what is meant by deeper levels of

TABLE 6
PROPORTION OF WORDS RECOGNIZED FROM Two

REPLICATIONS OF EXPERIMENT 9

Response
type Case Rhyme Category

1st study

Yes
No

.23

.59

.28 .33
.81
.62

2nd study

Yes
No

.42

.37

.65

.50

.90
.65

processing and what factors underlie the
superior retention of deeply processed
stimuli.

Method. On a projection screen, 60 words were
presented, one at a time, for 1 sec each with a
S-sec interword interval. All subjects saw the
same sequence of words, but different subjects
were asked different questions about each word.
For example, if the first word was COPPER, one
subject would be asked, “Is the word a metal?”,
a second, “Is the word a kind of fruit?”, a third,
“Does the word rhyme with STOPPER?”, and so
on. For each word, six questions were asked
(case, rhyme, category X yes-no). During the
series of 60 words, each subject received 10 trials
of each question-response combination, but in a
different random order. The questions were pre-
sented in booklets, 20 questions per page. Six
types of question sheet were made up, each type
presented to two subjects. These sheets balanced
the words across question types. The subject
studied the question, saw the word exposed on the
screen, then answered the question by checking
yes or no on the sheet. After the 60 encoding
trials, subjects received a further sheet contain-
ing 180 words consisting of the original 60 target
words plus 120 distractors. The subjects were
asked to check exactly 60 words as “old.” Two
different randomizations of the recognition list
were constructed; this control variable was crossed
with the six types of question sheets. Thus each
of the 12 subjects served in a unique replication
of the experiment. Instructions to subjects
emphasized that their main task was to remember
the words, and that a recognition test would
be given after the presentation phase. The ma-
terials used are presented in the Appendix.

Result. The top of Table 6 shows that
the results of Experiment 9 are quite similar
to those of Experiment 2, despite the fact
that in the present study subjects knew of
the recognition test and words were pre-
sented at the rate of 6 sec each. The find-
ing that subjects show exactly the same pat-

288 FERGUS I. M. CRAIK AND ENDEL TULVING

tern of results under these very different
conditions attests to the fact that the basic
phenomenon under study is a robust one.
It parallels results from Experiment 4 and
previous findings of Hyde and Jenkins
(1969, 1973). Before considering the
implications of Experiment 9, a replication
will be mentioned. This second experiment
was a complete replication with 12 other
subjects. The results of the second study
are also shown in Table 6. Overall recog-
nition performance was higher, especially
with case questions, but the pattern is the
same.

The results of these two studies are quite
surprising. Despite intentional learning
conditions and a slow presentation rate,
subjects were quite poor at recognizing
words which had been given shallow encod-
ings. Since subjects in this experiment
were asked to circle exactly 60 words, they
could not have used a strict criterion of
responding. Thus their low level of recog-
nition performance in the case task must
reflect inadequate initial registration of the
information or rapid loss of registered infor-
mation. Indeed, chance performance in
this task would be 33%; we have not cor-
rected the data for chance in any experi-
ment. The question now arises as to why
subjects do not encode case words to a
deeper level during the time after their
judgment was recorded. It is possible that
recognition of the less well-encoded items is
somehow adversely affected by well-encoded
items. It is also possible that subjects do
not know how best to prepare for a memory
test and thus do no further processing of
each word beyond the particular judgment
that is asked. A third hypothesis, that sub-
jects were poorly motivated and thus simply
did not bother to rehearse case words in a
more effective way, is put to test in the
final experiment. Here subjects were paid
by results; in one condition the recognition
of case words carried a much higher reward
than the recognition of category words.

In any event, Experiment 9 has demon-
strated that encoding operations constitute
an important determinant of learning or
retention under a wide variety of experi-
mental conditions. The finding of a strong
effect under quite loosely controlled class-

room conditions, without the trappings of
timers and tachistoscopes, is difficult to
reconcile with the view that was implicit in
the initial experiments of the series: that
processing of an item is somehow stopped
at a particular level and that an additional
fraction of a second would have led to bet-
ter performance. This view is therefore
now rejected. It seems to be the qualitative
nature of the encoding achieved that is
important for memory, regardless of how
much time the system requires to reach
some hypothetical level or depth of encod-
ing.

Experiment 10
The final experiment to be reported was

carried out to determine whether subjects
can achieve high recognition performance
with case-encoded words if they are given
a stronger inducement to concentrate on
these items. Subjects were paid for each
word correctly recognized; also, they were
informed beforehand that a recognition test
would be given. Correct recognition of the
three types of word was differentially re-
warded under three different conditions.
Subjects know that case, rhyme, and cat-
egory words carried either a 1 ,̂ 3(f, or 6^
reward.

Method. Subjects were tested under the same
conditions as subjects in Experiment 9. That
is, 60 words were presented for 1 sec each plus
S sec for the subject to record his judgment.
Each subject had 20 words under each encoding
condition (case, rhyme, category) with 10 yes and
10 no responses in each condition. As in Experi-
ment 9, each word appeared in each encoding
condition across different subjects. After the
initial phase, subjects were given a recognition
sheet of 180 words (60 targets plus 120 distrac-
tors) and instructed to check exactly 60 words.

There were three experimental groups. All
subjects were informed that the experiment was
a study of word recognition, that they would be
paid according to the number of words they
recognized, and therefore that they should
attempt to learn each word. The groups differed
in the value associated with each class of word:
Group 1 subjects knew that they would be paid
10, 60, and 30 for case, rhyme, and category
words, respectively; Group 2 subjects were paid
30, 10, and 60, respectively; and Group 3 subjects
were paid 6tf, 30, and 10, respectively. These
conditions are summarized in Table 7. Thus,
across groups, each class of words was associated
with each reward. There were 12 undergraduate
subjects in each of three groups.

DEPTH OF PROCESSING AND WORD RETENTION 289

Results. Table 7 shows that while recog-
nition performance was somewhat higher
than the comparable conditions of Experi-
ment 9 (Table 6), the differential reward
manipulation had no effect whatever. An
analysis of variance confirmed the obvious;
there were significant effects due to type
of encoding, F(2, 22) = 90.7, p < .01, response type (yes-no), F(l, 11) = 42.4, p < .01, and the Encoding X Response Type interaction, F(2, 22) = 4.13, p < .05, but no significant main effect or interactions involving the differential reward conditions.

Although this experiment yielded a null
result, its results are not without interest.
Even when subjects were presumably quite
motivated to learn and recognize case-
encoded words, they failed to reach the per-
formance levels associated with rhyme or
category words. Subjects in Group 3
(6-3-1) reported that although they really
did attempt to concentrate on case words, .
the category words were somehow “simply
easier” to recognize in the second phase of
the study.

Thus, Experiments 8, 9, and 10, con-
ducted in an attempt to establish the bound-
ary conditions for the depth of processing
effect, failed to remove the strong superi-
ority originally found for semantically en-
coded words. The effect is not due to iso-
lation, in the simple sense at least (Experi-
ment 8), it does not disappear under inten-
tional learning conditions and a slow pre-
sentation rate (Experiment 9), and it re-
mains when subjects are rewarded more for
recognizing words with shallower encod-
ings (Experiment 10). The problem now
is to develop an adequate theoretical con-
text for these findings and it is to this task
that we now turn.

GENERAL DISCUSSION

The experimental results will first be
briefly summarized. Experiments 1-4
showed that when subjects are asked to
make various cognitive judgments about
words exposed briefly on a tachistoscope,
subsequent memory performance is strongly
determined by the nature of that judgment.
Questions concerning the word’s meaning
yielded higher memory performance than
questions concerning either the word’s

TABLE 7
PROPORTIONS OF WORDS RECOGNIZED UNDER

EACH CONDITION IN EXPERIMENT 10

Encoding
operation

Case

Yes
No

Rhyme
Yes
No

Category
Yes
No

Mean
Yes
No

Reward value
1 cent 3 cents 6 cents

.50
.51
.73

.53

.93

.72

.72
.59
.51
.50
.73
.50

.89

.75

.71

.58

.54

.52

.69

.60

.88

.77

.70
.63

M

.52
.51
.72
.54
.90
.75
.71
.60

sound or the physical characteristics of its
printed form. Further, positive decisions
in the initial task were associated with
higher memory performance (for more
semantic questions at least) than were
negative decisions. These effects were
shown to hold for recognition and recall
under incidental and intentional memoriz-
ing conditions. One analysis of Experi-
ment 2 showed that recognition increased
systematically with initial categorization
time, but a further analysis demonstrated
that it was the nature of the encoding op-
erations which was crucial for retention,
not the amount of time as such. Experi-
ment 5 confirmed that conclusion. Experi-
ments 6 and 7 explored possible reasons
for the higher retention of words given
positive’ responses; it was argued that en-
coding elaboration provided a more satis-
factory description of the results than depth
of encoding. Experiment 8 showed that
isolation effects could not by themselves
give an account of the results, Experiment
9 demonstrated that the main findings still
occurred under much looser experimental
conditions, and Experiment 10 showed that
the pattern of results was unaffected when
differential rewards were offered for remem-
bering words associated with different
orienting tasks.

This set of results confirms and extends
the findings of other recent investigations,

290 FERGUS I. M. CRAIK AND ENDEL TULVING

notably the series of studies by Hyde, Jenk-
ins, and their colleagues (Hyde, 1973; Hyde
and Jenkins, 1969, 1973; Till & Jenkins,
1973; Walsh & Jenkins, 1973) and by
Schulman (1971, 1974). It is abundantly
clear that what determines the level of recall
or recognition of a word event is not inten-
tion to learn, the amount of effort involved,
the difficulty of the orienting task, the
amount of time spent making judgments
about the items, or even the amount of
rehearsal the items receive (Craik & Wat-
kins, 1973) ; rather it is the qualitative
nature of the task, the kind of operations
carried out on the items, that determines
retention. The problem now is to develop
an adequate theoretical formulation which
can take us beyond such vague statements
as “meaningful things are well remem-
bered.”

Depth of Processing

Craik and Lockhart (1972) suggested
that memory performance depends on the
depth to which the stimulus is analyzed.
This formulation implies that the stimulus
is processed through a fixed series of ana-
lyzers, from structural to semantic; that
the system stops processing the stimulus
once the analysis relevant to the task has
been carried out, and that judgment time
might serve as an index of the depth reached
and thus of the trace’s memorability.

These original notions now seem unsatis-
factory in a number of ways. First, the
postulated series of analyzers cannot lie on
a continuum since structural analyses do not
shade into semantic analyses. The modified
view of “domains” of encoding (Sutherland,
1972) was suggested by Lockhart, Craik,
and Jacoby (1975). The modification
postulates that while some structural
analysis must precede semantic analysis,
a full structural analysis is not usually car-
ried out; only those structural analyses
necessary to provide evidence for subsequent
domains are performed. Thus, in the case
where a stimulus is highly predictable at
the semantic level, only rather minimal
structural analysis, sufficient to confirm the
expectation, would be carried out. The
original levels of processing viewpoint is
also unsatisfactory in the light of the present

empirical findings if it is assumed that yes
and no responses are processed to roughly
the same depth before a decision can be
made, since there are no differences in
reaction times, yet there are large differ-
ences in retention of the words.

Second, large differences in retention
were also found when the complexity of
the encoding context was manipulated.
Experiment 7 showed that elaborate sen-
tence frames led to higher recall levels than
did simple sentence frames. This observa-
tion suggests than an adequate theory must
not focus only on the nominal stimulus but
must also consider the encoded pattern of
“stimulus in context.”

Third, and most crucial perhaps, strong
encoding effects were found under inten-
tional learning conditions in Experiments 4
and 9; it is totally implausible that, under
such conditions, the system stops processing
the stimulus at some peripheral level.
Unless one assumes complete perversity of
subjects, it must be clear that the word is
fully perceived on each trial. Thus, dif-
ferential depth of encoding does not seem
a promising description, except in very gen-
eral terms. Finally, as detailed earlier,
initial processing time is not always a good
predictor of retention. Many of the ideas
suggested in the Craik and Lockhart (1972)
article thus stand in need of considerable
modification if that processing framework
is to remain useful.

Degree of Encoding Elaboration

Is spread of encoding a more satisfactory
metaphor than depth? The implication
of this second description is that while a
verbal stimulus is usually identified as a
particular word, this minimal core encoding
can be elaborated by a context of further
structural, phonemic, and semantic encod-
ings. Again, the memory trace can be con-
ceptualized as a record of the various pat-
tern-recognition and interpretive analyses
carried out on the stimulus and its context;
the difference between the depth and spread
viewpoints lies only in the postulated orga-
nization of the cognitive structures respon-
sible for pattern recognition and elabora-
tion, with depth implying that encoding
operations are carried out in a fixed

DEPTH OF PROCESSING AND WORD RETENTION 291

sequence and spread leading to the more
flexible notion’ that the basic perceptual
core of the event can be elaborated in many
different ways. The notion of encoding
domains suggested by Lockhart, Craik, and
Jacoby (1975) is in essence a spread theory,
since encoding elaboration depends more on
the breadth of analysis carried out within
each domain than on the ordinal position of
an analysis in the processing sequence.
However, while spread and elaboration may
indeed be better descriptive terms for the
results reported in this paper, it should be
borne in mind that retention depends
critically on the qualitative nature of the
encoding operations performed—a minimal
semantic analysis is more beneficial for
memory than an elaborate structural analysis
(Experiment 5).

Whatever the sequence of operations, the
present findings are well described by the
idea that memory performance depends on
the elaborateness of the final encoding.
Retention is enhanced when the encoding
context is more fully descriptive (Experi-
ment 7), although this beneficial effect is
restricted to cases where the target stim-
ulus is compatible with the context and can
thus form an integrated encoded unit with
it. Thus the increased elaboration provided
by complex sentence frames in Experiment
7 did’not increase recall performance in the
case of negative response words.- The same
argument can be applied to the generally
superior retention of positive response
words in all the present experiments; for
positive responses the encoding question
can be integrated with the target word and
a more elaborate unit formed. In certain
cases, however, positive responses do not
yield a more elaborately encoded unit; such
cases occur when negative decisions specify
the nature of the attributes in question as
precisely as positive decisions. For ex-
ample, the response no to the question “Is
the word in capital letters?” indicates
clearly that the word is in lowercase letters;
similarly a no response to the question “Is
the object bigger than a man?” indicates
that the object is smaller than a man. When
no responses yield as elaborate an encoding
as yes responses, memory performance
levels are equivalent. There is nothing

inherently superior about a yes response;
retention depends on the degree of elabora-
tion of the encoded trace.

Several authors (e.g., Bower, 1967; Tul-
ving & Watkins, 1975) have suggested that
the memory trace can be described in terms
of its component attributes. This viewpoint
is quite compatible with the notion of encod-
ing elaboration. The position argued in this
section is that the trace may be considered
the record of encoding operations carried out
on the input; the function of these opera-
tions is to analyze and specify the attributes
of the stimulus. However, it is necessary
to add that memory performance cannot be
considered simply a function of the num-
ber of encoded attributes; the qualitative
nature of these attributes is critically im-
portant. A second equivalent description
is in terms of the “features checked” during
encoding. Again, a greater number of fea-
tures (especially deeper semantic features)
implies a more elaborate trace.

Finally, it seems necessary to bring in the
principle of integration or congruity for a
complete description of encoding. That is,
memory performance is enhanced to the
extent that the encoding question or context
forms an integrated unit with the target
word. The higher retention of positive
decision words in Schulman’s (1974) study
and in the present experiments can be de-
scribed in this way. The question immedi-
ately arises as to why integration with the
encoding context is so helpful. One pos-
sibility is that an encoded unit is unitized
or integrated on the basis of past experience
and, just as the target stimulus fits naturally
into a compatible context at encoding, so at
retrieval, re-presentation of part of the
encoded unit will lead easily to regeneration
of the total unit. The suggestion is that at
encoding the stimulus is interpreted in
terms of the system’s structured record of
past learning, that is, knowledge of the
world or “semantic memory” (Tulving,
1972) ; at retrieval, the information pro-
vided as a cue again utilizes the structure
of semantic memory to reconstruct the initial
encoding. An integrated or congruous
encoding thus yields better memory per-
formance, first, because a more elaborate
trace is laid down and, second, because

292 FERGUS I. M. CRAIK AND ENDEL TULVING

richer encoding implies greater compatibility
with the structure, rules, and organization
of semantic memory. This structure, in
turn, is drawn upon to facilitate retrieval
processes.

Broader Implications

Finally, the implications of the present
experiments and the related work reported
by Hyde and Jenkins (1969, 1973), Schul-
man (1971, 1974) and Kolers (1973a;
Kolers & Ostry, 1974) will be briefly dis-
cussed. All these studies conform to the
new look in memory research in that the
stress is on mental operations; items are
remembered not as presented stimuli acting
on the organism, but as components of men-
tal activity. Subjects remember not what
was “out there” but what they did during
encoding.

In more traditional memory paradigms,
the major theoretical concepts were traces
and associations; in both cases their main
theoretical property was strength. In turn,
the subject’s performance in acquisition,
retention, transfer, and retrieval was held to
be a direct function of the strength of asso-
ciations and their interrelations. The deter-
minants of strength were also well known:
study time, number of repetitions, recency,
intentionality of the subject, preexperimental
associative strength between items, inter-
ference by associations involving identical
or similar elements, and so on. In the ex-
periments we have described here, these
important determinants of the strength of
associations and traces were held constant:
nominal identity of items, preexperimental
associations among items, intralist similarity,
frequency, recency, instructions to “learn”
the materials, the amount and duration of
interpolated activity. The only thing that
was manipulated was the mental activity of
the learner; yet, as the results showed,
memory performance was dramatically
affected by these activities.

This difference between the old paradigm
and the new creates many interesting re-
search problems that would not readily have
suggested themselves in the former frame-
work. For example, to what extent are
the encoding operations performed on an
event under the person’s volitional strategic

control, and to what extent are they deter-
mined by factors such as context and set?
Why are there such large differences be-
tween different encoding operations? In
particular, why is it that subjects do not, or
can not, encode case words efficiently when
they are given explicit instructions to learn
the words? How does the ability of one
list item to serve as a retrieval cue for
another list item (e.g., in an A-B pair)
vary as a function of encoding operations
performed on the pair as opposed to the
individual items? The important concept
of association as such, the bond or relation
between the two items, A and B, may
assume a different form in the new paradigm.
The classical ideas of frequency and recency
may be eclipsed by notions referring to
mental activity.

There are problems, too, associated with
the development of a taxonomy of encoding
operations. How should such operations
be classified ? Do encoding operations really
fall into types as implied by the distinction
between case, rhyme, and category in the
present experiments, or is there some
underlying continuity between different op-
erations ? This last point reflects the debate
within theories of perception on whether
analysis of structure and analysis of mean-
ing are qualitatively distinct (Sutherland,
1972) or are better thought of as continuous
(Kolers, 1973b).

Finally, the major question generated by
the present approach is what are the encod-
ing operations underlying “normal” learn-
ing and remembering? The experiments
reported in this article show that people do
not necessarily learn best when they are
merely given “learn” instructions. The
present viewpoint suggests that when sub-
jects are instructed to learn a list of items,
they perform self initiated encoding opera-
ions on the items. Thus, by comparing
quantitative and qualitative aspects of per-
formance under learn instructions with per-
formance after various combinations of in-
cidental orienting tasks, the nature of learn-
ing processes may be further elucidated.
The possibility of analysis and control of
learning through its constituent mental op-
erations opens up exciting vistas for theory
and application.

DEPTH OF PROCESSING AND WORD RETENTION 293

REFERENCE NOTE

1. Moscovitch, M., & Craik, F. I. M. Retrieval
eves and levels of processing in recall and
recognition. Unpublished manuscript, 1975.
(Available from Morris Moscovitch, Erindale
College, Mississauga, Ontario, Canada).

REFERENCES
Begg, I. Recall of meaningful phrases. Journal

of Verbal Learning and Verbal Behavior, 1972,
11, 431-439.

Bobrow, S. A., & Bower, G. H. Comprehension
and recall of sentences. Journal of Experi-
mental Psychology, 1969, 80, 55-61.

Bower, G. H. A multicomponent theory of the
memory trace. In K. W. Spence & J. T. Spence
(Eds.), The psychology of learning and moti-
vation (Vol. 1). New York: Academic Press,
1967.

Bower, G. H., & Karlin, M. B. Depth of pro-
cessing pictures of faces and recognition mem-
ory. Journal of Experimental Psychology,

‘ 1974, 103, 751-757.
Broadbent, D. E. Behaviour. London: Eyre &

Spottiswoode, 1961.
Cermak, L. S. Human memory: Research and

theory. New York: Ronald, 1972.
Craik, F. I. M., & Lockhart, R. S. Levels of

processing: A framework for memory research.
Journal of Verbal Learning and Verbal Be-
havior, 1972, 11, 671-684.

Craik, F. I. M., & Watkins, M. J. The role of
rehearsal in short-term memory. Journal of
Verbal Learning and Verbal Behavior, 1973,
12, 599-607.

Eagle, M., & Leiter, E. Recall and recognition in
intentional and incidental learning. Journal of
Experimental Psychology, 1964, 68, 58-63.

Horowitz, L. M., & Prytulak, L. S. Redinte-
grative memory. Psychological Review, 1969,
76, 519-531.

Hyde, T. S, Differential effects of effort and
type of orienting task on recall and organiza-
tion of highly associated words. Journal of
Experimental Psychology, 1973, 79, 111-113.

Hyde, T. S., & Jenkins, J. J. Differential effects
of incidental tasks on the organization of
recall of a list of highly associated words.
Journal of Experimental Psychology, 1969, 82,
472-481.

Hyde, T. S., & Jenkins, J. J. Recall for words
as a function of semantic, graphic, and syntactic
orienting tasks. Journal of Verbal Learning
and Verbal Behavior, 1973, 12, 471-480.

Jacoby, L. L. Test appropriate strategies in
retention of categorized lists. Journal of Verbal
Learning and Verbal Behavior, 1973, 12, 675-
682.

Kolers, P. A. Remembering operations. Mem-
ory & Cognition, 1973, 1, 347-355. (a)

Kolers, P. A. Some modes of representation. In
P. Pliner, L. Krames, & T. Alloway (Eds.),
Communication and affect: Language and
thought. New York: Academic Press, 1973.
(b)

Kolers, P. A., & Ostry, D. J. Time course of
loss of information regarding pattern analyzing
operations. Journal of Verbal Learning and
Verbal Behavior, 1974, 13, 599-612.

Lockhart, R. S., Craik, F. I. M., & Jacoby, L. L.
Depth of processing in recognition and recall:
Some aspects of a general memory system. In
J. Brown (Ed.), Recognition and recall. Lon-
don: Wiley, 1975.

Neisser, U. Cognitive psychology. New York:
Appleton-Century-Crofts, 1967.

Norman, D. A. (Ed.). Models of human mem-
ory. New York: Academic Press, 1970.

Paivio, A. Imagery and verbal processes. New
York: Holt, Rinehart & Winston, 1971.

Postman, L. Short-term memory and incidental
learning. In A. W. Melton (Ed.), Categories
of human learning. New York: Academic
Press, 1964.

Rosenberg, S., & Schiller, W. J. Semantic cod-
ing and incidental sentence recall. Journal of
Experimental Psychology, 1971, 90, 345-346.

Schulman, A. I. Recognition memory for targets
from a scanned word list. British Journal of
Psychology, 1971, 62, 335-346.

Schulman, A. I. Memory for words recently
classified. Memory & Cognition, 1974, 2, 47-52.

Sheehan, P. W. The role of imagery in incidental
learning. British Journal of Psychology, 1971,
62, 235-244.

Sutherland, N. S. Object recognition. In E. C.
Carterette & M. P. Friedman (Eds.), Hand-
book of perception (Vol. 3). New York:
Academic Press, 1972.

Till, R. E., & Jenkins, J. J. The effects of cued
orienting tasks on the free recall of words.
Journal of Verbal Learning and Verbal Be-
havior, 1973, 12, 489-498.

Treisman, A., & Tuxworth, J. Immediate and
delayed recall of sentences after perceptual
processing at different levels. Journal of Ver-
bal Learning and Verbal Behavior, 1974, 13,
38-44.

Tulving, E. Episodic and semantic memory. In
E. Tulving & W. Donaldson (Eds.), Organiza-
tion of memory. New York: Academic Press,
1972.

Tulving, E., & Thomson, D. M. Encoding
specificity and retrieval processes in episodic
memory. Psychological Review, 1973, 80, 352-
373.

Tulving, E., & Watkins, M. J. Structure of
memory traces. Psychological Review, 1975,
82, 261-275.

Walsh, D. A., & Jenkins, J. J. Effects of orient-
ing tasks on free recall in incidental learning:
“Difficulty,” “effort,” and “process” explana-
tions. Journal of Verbal Learning and Verbal
Behavior, 1973, 12, 481-488.

Waugh, N. C., & Norman, D. A. Primary
memory. Psychological Review, 1965, 72, 89-
104.

Wickelgren, W. A. The long and the short of
memory. Psychological Bulletin, 1973, 80, 425-
438.

(Received February 5, 1975)

294 FERGUS I. M. CRAIK AND ENDEL TULVING

APPENDIX

Each subject in Experiment 9 received the
same 60 words in the same order, but six dif-
ferent “formats” were constructed, such that
all six possible questions (case, rhyme, cat-
egory X yes-no) were asked for each word
(Table Al). Thus, for SPEECH, the questions
were (a) Is the word in capital letters? (b)
Is the word in small print? (c) Does the

word rhyme with each? (d) Does the word
rhyme with tense? (e) Is the word a form of
communication? (f) Is the word something
to wear? Each format contained 10 questions
of each type. Negative questions were drawn
from the pool of unused questions in that
particular format.

TABLE Al
WORDS AND QUESTIONS USED IN EXPERIMENT 9

Word

SPEECH
BRUSH
CHEEK
FENCE
FLAME
FLOUR
HONEY
KNIFE
SHEEP
COPPER
GLOVE

XMONK
DAISY
MINER
CART
CLOVE
ROBBER
MAST
FIDDLE
CHAPEL
SONNET
WITCH
ROACH
BRAKE
TWIG
GRIN
DRILL
MOAN
CLAW
SINGER

Rhyme
question

each
lush
teak
tense
claim
sour
funny
wife
leap
stopper
shove
trunk
crazy
liner
start
rove
clobber
past
riddle
grapple
bonnet
rich
coach
shake
big
bin
fill
prone
raw
ringer

Category question

a form of communication
used for cleaning
a part of the body
found in the garden
something hot
used for cooking
a type of food
a type of weapon
a type of farm animal
a type of metal
something to wear
a type of clergy
a type of flower
a type of occupation
a type of vehicle
a type of herb
a type of criminal
a part of a ship
a musical instrument
a type of building
a written form of art
associated with magic
a type of insect
a part of a car
a part of a tree
a human expression
a type of implement
a human sound
a part of an animal
a type of entertainer

Word

BEAR
LAMP
CHERRY

XROCK
EARL
POOL
WEEK
BOAT
PAIL
TROUT
GRAM
WOOL
CLIP
JUICE
POND
LANE
NURSE
LARK
STATE
SOAP
JADE
SLEET
RICE
TIRE
CHILD
DANCE
FIELD
FLOOR
GLASS
TRIBE

Rhyme
question

hair
camp
very
stock
pearl
school
peak
rote
whale
bout
tram
pull
ship
noose
wand
pain
curse
park
crate
rope
raid
feet
dice
fire
wild
stance
shield
sore
pass
scribe

Category question

a wild animal
a type of furniture
a type of fruit
a type of mineral
a type of nobility
a type of game
a division of time
a mode of travel
a type of container
a type of fish
a type of measurement
a type of material
a type of office supply
a type of beverage
a body of water
a type of thoroughfare
associated with medicine
a type of bird
a territorial unit
a type of toiletry
a type of precious stone
a type of weather
a type of grain
a round object
a human being
a type of physical activity
found in tne countryside
a part of a room
a type of utensil
a group of people

Vol. 74, No. 11 Whole No. 498, 196

0

Psychological M o n o g r a p h s : General and A p p l i e d

T H E I N F O R M A T I O N A V A I L A B L E I N
B R I E F V I S U A L P R E S E N T A T I O N S

1

GEORGE SPERLING2

Harvard University

H o w much can be seen in a single brief
exposure ? This is an important prob

–

lem because our normal mode of seeing
greatly resembles a sequence of brief ex-
posures. Erdmann and Dodge (1898)
showed that in reading, for example, the
eye assimilates information only in the brief
pauses between its quick saccadic move-
ments. The problem of what can be see

n

in one brief exposure, however, remains
unsolved. The difficulty is that the simple
expedient of instructing the observer of
a single brief exposure to report what he
has just seen is inadequate. When complex
stimuli consisting of a number of letters are
tachistoscopically presented, observers enig-
matically insist that they have seen more
than they can remember afterwards, that
is, report afterwards.3 The apparently
simple question: “What did you see ?” re-
quires the observer to report both what he
remembers and what he has forgotten.

1 This paper is a condensation of a doctoral
thesis (Sperling, 1959). For further details,
especially on methodology, and for individual data,
the reader is referred to the original thesis. It is a
pleasure to acknowledge my gratitude to George A.
Miller and Roger N. Shepard whose support made
this researcli possible and to E. B. Newman, J.
Schwartzbaum and S. S. Stevens for their many
helpful suggestions. Thanks are also due to Jerome
S. Brunei’ for the use of his laboratory and his
tachistoscope during his absence in the summer
of 1957. This research was carried out under
Contract AF 33(038)-14343 between Harvard Uni-
versity and the Operational Applications Labo-
ratory, Air Force Cambridge Research Center, Air
Research Development Command.

2 Now at Bell Telephone Laboratories, Murray
Hill, New Jersey.

3 Some representative examples are: Bridgin
(1933), Cattell (1883), Chapman (1930), Dallen-
bach (1920), Erdmann and Dodge (1898), Glanville
and Dallenbach (1929), Kulpe (1904), Schumann
(1922), Wagner (1918), Whipple (1914), Wil-
cocks (1925), Woodworth (1938).

The statement that more is seen than can
be remembered implies two things. First,
it implies a memory limit, that is, a limit
on the (memory) report. Such a limit on
the number of items which can be given
in the report following any brief stimulation
has, in fact, been generally observed; it is
called the span of attention, apprehension,
or immediate-memory (cf. Miller, 1956b).
Second, to see more than is remembered
implies that more information is available
during, and perhaps for a short time after,
the stimulus than can be reported. The
considerations about available information
are quite similar, whether the information is
available for an hour (as it is in a book
that is borrowed for an hour), or whether
the information is available for only a frac-
tion of a second (as in a stimulus which
is exposed for only a fraction of a second).
In either case it is quite probable that for a
limited period of time more information
will be available than can be reported. It
is also true that initially, in both examples,
the information is available to vision.

In order to circumvent the memory limita-
tion in determining the information that
becomes available following a brief ex-
posure, it is obvious that the observer must
not be required to give a report which
exceeds his memory span. I f the number of
letters in the stimulus exceeds his memory
span, then he cannot give a whole report
of all the letters. Therefore, the observer
must be required to give only a partial re-
port of the stimulus contents. Partial re-
porting of available information is, of
course, just what is required by ordinary
schoolroom examinations and by other
methods of sampling available information.

An examiner can determine, even in a
short test, approximately how much the

GEORGE SPERLING

student knows. The length of the test is
not so important as that the student not be
told the test questions too far in advance.
Similarly, an observer may be “tested” on
what he has seen in a brief exposure of a
complex visual stimulus. Such a test re-
quires only a partial report. The specific
instruction which indicates which part of
the stimulus is to be reported is then given
only after termination of the stimulus. On
each trial the instruction, which calls for
a specified part of the stimulus, is randomly
chosen from a set of possible instructions
which cover the whole stimulus. I>y re-
peating the interrogation (sampling) pro-
cedure many times, many different random
samples can be obtained of an observer’s
performance on each of the various parts
of the stimulus. The data obtained thereby
make feasible the estimate of the total in-
formation that was available to the observer
from which to draw his report on the aver-
age trial.

The time at which the instruction is given
determines the time at which available in-
formation is sampled. By suitable coding,
the instruction may be given at any time:
before, during, or after the stimulus presen-
tation. Not only the available information
immediately following the termination of
the stimulus, but a continuous function
relating the amount of information available
to the time of instruction may be obtained
by such a procedure.

Many studies have been conducted in
which observers were required to give
partial reports, that is, to report only on
one aspect or one location of the stimulus.
In prior experiments, however, the instruc-
tions were often not randomly chosen, and
the set of possible instructions did not
systematically cover the stimulus. The no-
tions of testing or sampling were not ap-
plied.’ .It is not surprising, therefore, (hat

” The experiments referred to are (cf. Sperling,
1959) : Kiilpc (1904), Wilcocks (1925), Chapman
(1932), Long, Henneman, and Reid (1953), Long
and Lee (1953a), Long and Lee (1953b), Long,
Reid, and Garvey (1954), Lawrence and Coles
(1954), Adams (1955), Lawrence and Laherge
(1956), liroadbent (1957a).

estimates have not been made of the total
information available to the observer fol-
lowing a brief exposure of a complex
stimulus. Furthermore, instructions have
generally not been coded in such a way as
to make it possible to control the precise
time at which they were presented. Con-
sequently, the temporal course of available
information could not have been quantita-
tively studied. In the absence of precise
data, experimenters have all too frequently
assumed that the time for wdiich informa-
tion is available to the observer corresponds
exactly to the physical stimulus duration.
Wundt (1899) understood this problem
and convincingly argued that, for extremely
short stimulus durations, the assumption
that stimulus duration corresponded t

o

the duration for which stimulus informa-
tion was available was blatantly false, but
he made no measurements of available
information.

The following experiments were con-
ducted to study quantitatively the informa-
tion that becomes available to an observer
following a brief exposure. Lettered stimuli
were chosen because these contain a rela-
tively large amount of information per item
and because these are the kind of stimuli
that have been used by most previous in-
vestigators. The first two experiments are
essentially control experiments; they at-
tempt to confirm that immediate-memory
for letters is independent of the parameters
of stimulation, that it is an individual
characteristic. In the third experiment the
number of letters available immediately
after the extinction of the stimulus is de-
termined by means of the sampling (partial
report) procedure described above. The
fourth experiment explores decay of avail-
able information with time. The fifth
experiment examines some exposure param-
eters. Tn the sixth experiment a technique
which fails to demonstrate a large amount
of available information is investigated.
The seventh experiment deals with the role:
of the historically important variable : order
of report.

THE INFORMATION AVAILABLE I N BRIEF VISUAL PRESENTATIONS

G e n e r a l M e t h o d

Apparatus. The experiments utilized a Gerbrands
tachistoscope.” This is a two-field, mirror tachisto-
scope (Dodge, 1907b), with a mechanical timer.
Viewing is binocular, at a distance of about 24
inches. Throughout the experiment, a dimly i l –
luminated fixation field was always present.

The light source in the Gerbrands tachistoscope
is a 4-watt fluorescent (daylight) bulb. Two such
lamps operated in parallel light each field. The
operation of the lamps is controlled by the micro-
switches, the steady-state light output of the lamp
being directly proportional to the current. How-
ever, the phosphors used in coating the lamp con-
tinue to emit light for some time after the cessation
of the current. This afterglow in the lamp follows
an exponential decay function consisting of two
parts: the first, a blue component, which accounts
for about 40% of the energy, decays with a time
constant which is a small fraction of a millisecond;
the decay constant of the second, yellow, component
was about IS msec, in the lamp tested. Fig. 1

ID
_ l
<

o

< Ul z -

J

1

>-
1 –

z
UJ
z

50 MILLISECONDS PER DIVISION

Fig. 1. A SO-millisecond light flash, such as was
used in most of the experiments. (Redrawn from
a photograph of an oscilloscope trace)

illustrates a SO-mscc. light impulse on a linear
intensity scale. The exposure time of 50 msec.
was used in all experiments unless exposure time
was itself a parameter. Preliminary experiments
indicated that, with the presentations used, ex-
posure duration was an unimportant parameter.
Fifty msec, was sufficiently short so that eye
movements during the exposure were rare, and
it could conveniently be set with accuracy.

Stimulus materials. The stimuli used in this
experiment were lettered 5×8 cards viewed at a
distance of 22 inches. The lettering was done
with a Leroy No. 5 pen, producing capital letters
about 0.45 inch high. Only the 21 consonants
were used, to minimize the possibility of Ss inter-
preting the arrays as words. In a few sets of cards
the letter Y was also omitted. In all, over 500
different stimulus cards were used.

There was very little learning of the stimulus
materials either by the .Ss or by the E. The only
learning that was readily apparent was on several
stimuli that had especially striking letter com-
binations. Except for the stimuli used for train-
ing, no .S” ever was required to report the same
part of any stimulus more than two or three times,
and never in the same session.

Figure 2 illustrates some typical arrays of
letters. These arrays may be divided into several
categories: (a) stimuli with 3, 4, 5, 6, or 7 letters
normally spaced on a single line; (b) stimuli with
six letters closely spaced on a single line (6-
massed) ; (c) stimuli having two rows of letters
with three letters in each row (3/3), or two rows

R N F

X V N K H

L Q D K K J

ZYVVFF

K L B

Y NX

XMR J

P NK P

T O R
SR N
FZR

7 1 V F
X L 5 3
B 4 W7

5 Ralph Gerbrands Company, 96 Ronald Road,
Arlington 74, Massachusetts.

Fig. 2. Typical stimulus materials. Col. 1 : 3, 5,
6, 6-masscd. Col. 2: 3/3, 4/4, 3/3/3, 4/4/4 L&N.

GEORGE SPERLING

of four letters each (4/4) ; (d) stimuli having
three rows of letters with three letters in each
row (3/3/3). The stimulus information, calculated
in bits, for some of the more complex stimuli is
26.4 bits (6-lettcrs, 6-massed, 3/3), 35.1 bits (4/4),
and 39.5 bits (3/3/3).

tn addition to stimuli that contained only letters,
some stimuli that contained both letters and num-
bers were used. These had eight (4/4 CAN,
35.7 bits) and twelve symbols (4/4/4 L&N, 53.6
bits), respectively, four in each row. Each row had
two letters and two numbers—the positions being
randomly chosen. The .9 was always given a sample
stimulus before L&N stimuli were used and told
of the constraint above. He was also told that
O when it occurred was the number “zero” and was
not considered a letter. Calculated with these con-
straints, the information in each row of Tour
letters and numbers (17.9 bits) on such a card
is nearly equal to the information in a row of four
randomly chosen consonants (17.6 bits), even
though there are different kinds of alternatives in
each case.

Subjects. The nature of the experiments made
it more economical to use small numbers of trained
,9s rather than several large groups of untrained ,9s.
Four of the five 5s in the experiment were ob-
tained through the student employment service.
The fifth 5 (RNS) was a member of the faculty
who was interested in the research. Twelve ses-
sions were regularly scheduled for each S, three
times weekly.

Instructions and trial procedures. S was in-
structed to look at the fixation cross until it was
clearly in focus; then he pressed a button which
initiated the presentation after a 0.5-sec. delay.
This procedure constituted an approximate be-
havioral criterion of the degree of dark adaptation
prior to the exposure, namely, the ability to focus
on the dimly illuminated fixation cross.

Responses were recorded on a specially prepared
response grid. A response grid appropriate to each
stimulus was supplied. The response grid was placed
on the table immediately below the tachistoscope,
the room illumination being sufficient to write by.
The .9s were instructed to fill in all the required
squares on the response grid and to guess when
they were not certain. The .9s were not permitted
to fill in consecutive X’s, but were required to guess
“different letters.” After a response, .9 slid the
paper forward under a cover which covered his last
response, leaving the next part of the response
grid fully in view.

Scries of 5 to 20 trials were grouped together
without a change in conditions. Whenever con-
ditions or stimulus types were changed, .9 was
given two or three sample presentations with the
new conditions or stimuli. Within a sequence of
trials, S set his own rate of responding. The -9s
(except ND) preferred rapid rates. In some
conditions, the limiting rate was set by the E’s
limitations in changing stimuli and instruction

tones. This was about three to four stimuli per
minute.

Each of the first four and last two sessions
began with and/or ended with a simple task:
the reporting of all the letters in stimuli of 3,
4, 5, and 6 letters. This procedure was under-
taken in addition to the usual runs with these
stimuli to determine if there were appreciable
learning effects in these tasks during the course
of the experiment and if there was an accuracy
decrement (fatigue) within individual sessions.
Very little improvement was noted after the second
session. This observation agrees with previous
reports (Whipple, 1914). There was little differ-
ence between the beginning and end of sessions.

Scoring and tabulation of results. Every report
of all ,9s was scored both for total number of
letters in the report which agreed with letters in
the stimulus and for the number of letters reported
in their correct positions. Since none of the pro-
cedures of the experiments had an effect on cither
of these scores independently of the other, only
the second of these, letters in the correct position,
is tabulated in the results. This score, which takes
position into account, is less subject to guessing
error,” and in some cases it is more readily inter-
preted than a score which does not take position
into account. As the maximum correction for
guessing would be about 0.4 letter for the 4/4/4
(12-letter) material—and considerably less for all
other materials—no such correction is made in the
treatment of the data. In general, data were not
tabulated more accurately than 0.1 letter.

Data from the first and second sessions were
not used if they fell below an 5″s average per-
formance on these tasks in subsequent sessions.
This occurred for reports of five and of eight
(4/4) letters for some ,9s. A similar criterion
applied in later sessions for tasks that were
initiated later. In this case, the results of the
first “training” session(s) are not incorporated in
the total tabulation if they lie more than 0.5
letter from ,9’s average in subsequent sessions.

Experiment. 1 : Immediate-

Memory

When an S is required to give a complete
(whole) report of all the letters on a briefly
exposed stimulus, he will generally not re-

8 I f there are a large number of letters in the
stimulus, the probability that these same letters
will appear somewhere on the response grid, irre-
spective of position, becomes very high whether or
not .9 has much information about the stimulus.
In the limit, the correspondence approaches 100%
provided only that the relative frequency of each
letter in the response matches its relative fre-
quency of occurrence in the stimulus pack. If the
response is scored for both letter and position, then
the percent guessing correction is independent of
changes in stimulus size.

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS

port all the letters correctly. The average
number of letters which he does report
correctly is usually called his immediate-
memory span or span of apprehension for
that particular stimulus material under the
stated observation conditions. An expres-
sion such as immediate-memory span
(Miller, 1956a) implies that the number
of items reported by S remains invariant
with changes in stimulating conditions. The
present experiment seeks to determine to
what extent the span of immediate-memory
is independent of the number and spatial
arrangement of letters, and of letters and
numbers on stimulus cards. I f this inde-
pendence is demonstrated, then the quali-
fication “for that particular stimulus
material” may be dropped from the term
immediate-memory span when it is used
in these experiments.

Procedure. Ss were instructed to write all the
letters in the stimulus, guessing when they were
not certain. All 12 types of stimulus materials
were used. At least IS trials were conducted with
each kind of stimulus with each S. Each 6″ was
given at least SO trials with the 3/3 (6-letter)
stimuli which had yielded the highest memory
span in preliminary experiments. The final run
made with any kind of stimulus was always a test
of immediate-memory. This procedure insured that
J?s were tested for memory when they were
maximally experienced with a stimulus.

Results. The lower curves in Fig. 3
represent the average number of letters
correctly reported by each S for each ma-
terial.7 The most striking result is that
immediate-memory is constant for each S,
being nearly independent of the kind of
stimulus used. The immediate-memory span
for individual vS*s ranges from approxi-
mately 3.8 for JC to approximately 5.2 for
NJ with an average immediate-memory
span for all ^s of about 4.3 letters. (The
upper curves are discussed later.)

The constancy which is characteristic of
individual immediate-memory curves of
Fig. 3 also appears in the average curve for
all Ss. For example, three kinds of stimuli
were used that had six letters each: six
letters normally spaced on one line, 6-

massed, and 3/3-letters (see Fig. 2). When
the data for all ^s are pooled, the scores
for each of these three types of materials
are practically the same: the range is 4.1-

4.3

letters. The same constancy holds for
stimuli containing eight symbols. The aver-
age number of letters correctly reported for
each of the two different kinds of eight
letter stimuli, 4/4, 4/4 L & N , is nearly the
same: 4.4, 4.3, respectively.

Most .S*s felt that stimuli containing both
letters and numbers were more difficult
than those containing letters only. Never-
theless, only NJ showed an objective deficit
for the mixed material.

I n conclusion, the average number of
correct letters contained in an S”s whole
report of the stimulus is approximately

AVERAGE
ALL SUBJECTS

l i l t I i i i

£>2
Eio
>,
r >

RNS / ‘

/

‘••• / ‘ s^
; / * * ~ ^ –

,-j-i i i i i i i.„
1 4 5 6 B 9 10 12

7 See Sperling (1959) f o r tables g i v i n g the
numerical values o f all points appearing in this and
i n all other figures.

3 4 5 6 8 9 10 12 3 4 5 6 8 9 10 12 ~
NUMBER OF LETTERS IN STIMULUS

Fie. 3. “Channel capacity curves.” Immediate-
memory and letters available (output i n f o r m a t i o n )
as functions of the number of stimulus letters
( i n p u t i n f o r m a t i o n ) . L o w e r curves = immediate-
memory ( E x p . 1) ; upper curves = letters available
immediately a f t e r termination o f the s t i m u l u s ;
diagonal lines = m a x i m u m possible score (i.e.,
input = o u t p u t ) . Code: X = letters on one l i n e ;
+ = 6-massed; o = 3 / 3 , 4 / 4 , 5 / 5 ; A = 3 / 3 / 3 ;
• = 4/4 L&N, 4/4/4 L&N.

6 GEORGE SPERLING

equal to the smaller of (a) the number of
letters in the stimulus or (b) a numerical
constant—the span of immediate-memory

—

which is different for each .S”. The use of
the term immediate-memory span is there-
fore justified within the range of materials
studied. This limit on the number of letters
that can be correctly reported is an individ-
ual characteristic, but it is relatively similar
for each of the five .S”s of the study.

Experiment 2: Exposure Duration

The results of Experiment 1 showed that,
regardless of material, .9s could not report
more than an average of about 4.5 items
per stimulus exposure. Jn order to deter-
mine whether this limitation was a peculiar
characteristic of the short exposure dura-
tion (0.05 sec), it was necessary to vary
the exposure duration.

Procedure. As in the previous experiment, .S’s
were instructed to report all the letters in the
stimulus. The stimuli were six letter cards (3/3).
NJ, who was able to report more than five correct
letters in Experiment 1, was given 4/4 stimuli in
order to make a possible increment in his accuracy
of responding detectable. The Ss were given

10

trials in each of the four conditions, .015-, .050-,
.150- (.200-), .500-scc. exposure duration, in the
order above, hi a later session, additional trials
were conducted at .015-sec. exposure as a control
for Experiment 5. These trials are averaged with
the above data.

Results and discussion. Figure A illus-
trates the number of letters correctly re-
ported as a function of the duration of

.015 .05
EXPOSURE DURATION (SEC)

F i g . 4. L e t t e r s c o r r e c t l y r e p o r t e d as a f u n c t i o n
o f exposure d u r a t i o n .

exposure. The main result is that exposure
duration, even over a wide range, is not an
important parameter in determining the
number of letters an S can recall correctly.
Both individually and as a group, S’s show
no systematic changes in die number of
letters correctly reported as the exposure
duration was varied from 0.015 to 0.500 sec.
The invariance of the number of letters
reported as a function of exposure dura-
tions up to about 0.25 sec. for the kind of
presentation used (dark pre- and post-
exposure fields) has long been known
(Schumann, 1904).

Experiment 3: Partial Report

Experiments 1 and 2 have demonstrated
the span of immediate-memory as an in-
variant characteristic of each .9. In Experi-
ment 3 the principles of testing in a
perceptual situation that were advanced in
the introduction are applied in order to
determine whether S has more information
available than he can indicate in his limited
immediate-memory report.

The S is presented with the stimulus as
before, but he is required only to make a
partial report. The length of this report
is four letters or less, so as to lie within
.V’s immediate-memory span. The instruc-
tion that indicates which row of the stimulus
is to be reported is coded in the form of a
tone. The instruction tone is given after the
visual presentation. The .9 docs not know
until he hears the tone which row is called
for. This is therefore a procedure which
samples the information that 5″ has avail-
able after the termination of the visual
stimulus.

Procedure. Initially, stimulus materials having
only two lines were used, that is, 3/3 and 4/4. The
.9 was told that a tone would be sounded, that this
tone would come approximately simultaneously
with the exposure, and that it would be either a
high tone (2500 cps) or a low tone (250 cps).
I f it were a high tone, he was to write only the
upper row of the stimulus; if a low tone, only
the lower row. He was then shown a sample card
of 3/3 letters and given several high and low
tones. It was suggested that he keep his eyes
fixated on the fixation point and be equally pre-
pared for either tone. \t would not be possible to
outguess the E who would be using a random
sequence of tones.

THE INFORMATION AVAILABLE I N BRIEF VISUAL PRESENTATIONS

The tone duration was approximately O.S sec.
The onset of the tone was controlled through
the same microswitch that controlled the off-go of
the light, with the completion of a connection from
an audio-oscillator to the speaker. Intensity of the
tone was adjusted so that the high (louder) tone
was “loud but not uncomfortable.”

In each of the first two sessions, each 5* re-
ceived 30 training trials with each of the materials
3/3, 4/4. In subsequent sessions 5s were given
series of 10 or more “test” trials. Later, a third,
middle (650 cps) tone was introduced to cor-
respond to the middle row of the 3/3/3 and 4/4/4
stimuli. The instructions and procedure were es-
sentially the same as before.

In any given session, each tone might not occur
with equal frequency for each type of stimulus.
Over several sessions, usually two, this unequal
frequency was balanced out so that an 5* had an
exactly equal number of high, medium, and low
tones for each material. If an 5 “misinterpreted”
the tone and wrote the wrong row, he was asked
to write what he could remember of the correct
row. Only those letters which corresponded to the
row indicated by the tone were considered.

Treatment of the Data. In the experiments
considered in this section, S is never required to
report the whole stimulus but only one line of a
possible two or three lines. The simplest treatment
is to plot the percentage of letters correct. This,
in fact, will be done for all later comparisons.
The present problem is to find a reasonable meas-
ure to enable comparison between the partial
report and the immediate-memory data for the same
stimuli. The measure, percent correct, does not
describe the results of the immediate-memory
experiments parsimoniously. In Experiment 1 it
was shown that 5s report a constant number of
letters, rather than a constant percentage of
letters in the stimulus. The measure, number of
letters correct, is inappropriate to the partial re-
port data because the number of letters which 5″
reports is limited by the E to at most three or
four. The most reasonable procedure is to treat
the partial report as a random sample of the
letters which the 5 has available. Each partial
report represents a typical sample of the number
of letters 5 has available for report. For ex-
ample, if an 5 is correct about 90% of the time
when he is reporting three out of nine letters, then
he is said to have 90% of the nine letters—about
eight letters—available for partial report at the
time the instruction tone is given,

In order to calculate the number of available
letters, the average number of letters correct in
the partial report is multiplied by the number of
equiprobable (nonoverlapping), partial reports. If
there are two tones and two rows, multiplication
is by 2.0; if three, by 3.0. As before, only the
number of correct letters in the correct position is
considered.

Results. The development of the final,
stable f o r m of the behavior is relatively
rapid for 5s giving partial reports. The
average for all 5s after 30 trials (first
session) w i t h the 3 / 3 stimuli was 4.5; on
the second day the average of 30 more
trials was S.l. On the t h i r d day 5s aver-
aged 5.6 out of a possible six letters. Most
of the improvement was due to just one 5 :
N D who improved f r o m 2.9 to 5.8 letters
available. I n the 3 / 3 / 3 stimulus training,
all 5s reached their final value after the
initial 40 trials on the first day of training.
The considerable experience 5s had ac-
quired w i t h the partial reporting procedure
at this time may account for the quick
stabilization. N J , whose score was 7.7
letters available on the first 20 trials, was
given almost 150 additional trials in an
unsuccessful attempt to raise this initial
score.

I n Fig. 3 the number of letters available
as a function of the number of letters in
the stimulus are graphed as the upper
curves. For all stimuli and for all 5s, the
available information calculated from the
partial report is greater than that contained
in the immediate-memory report. Moreover,
from the divergence of the two curves it
seems certain that, if still more complex
stimuli were available, the amount of avail-
able information would continue to increase.

The estimate above is only a lower bound
on the number of letters that 5s have avail-
able for report after the termination of the
stimulus. A n upper bound cannot be ob-
tained f r o m experiments utilizing partial
reports, since it may always be argued that,
w i t h slightly changed conditions, an im-
proved performance might result. Even the
lower-bound measurement of the average
available information, however, is twice
as great as the immediate-memory span.
The immediate-memory span for the 4 / 4 / 4
(12-letters and numbers) stimuli ranges
from 3.9 to 4.7 symbols for the 5s, with
an average of 4.3. Immediately after an
exposure of the 4 / 4 / 4 stimulus material,
the number of letters available to the 5s
ranged f r o m 8.1 ( N D ) to 11.0 ( R O R ) ,
w i t h an average of 9.1 letters available.

8 GEORGE SPERLING

This number of letters may be transformed
into the bits of information represented by
so many letters. For the 4 / 4 / 4 (12-letters
and numbers) material, the average number
of bits available, then, is 40.6, w i t h a range
from 36.2 to 49.1 (out of a possible 53.6
bits). These figures are considerably higher
than the usual estimates. For example, in
a recent review article Quastler (1956)
writes: ” A l l experimental studies agree
that man can . . . assimilate less than

25

bits per glance” (p. 32). The data obtained
in Experiment 3 not only exceed this m a x i –
mum, but they contain no evidence that the
information that became available to the
i’s following the exposure represented a
limit of ” m a n ” rather than a maximum
determined by the limited information con-
tained in the stimuli which were used.

Experiment 4 : Decay of Available
Information

Part 1 : Development of Strategies of
Observing

I t was established i n Experiment 3 that
more information is available to the .Js
immediately after termination of the
stimulus than they could report. I t remains
to determine the fate of this surplus i n –
formation, that is, the “forgetting curve.”
The partial report technique makes possible
the sampling of the available information at
the time the instruction signal is given. B y
delaying the instruction, therefore, decay of
the available information as a function of
time w i l l be reflected as a corresponding
decrease in the accuracy of the report.

Procedure. The principal modification from the
preceding experiment is that the signal tone, which
indicates to the S which row is to be reported, is
given at various other times than merely “zero
delay” following the stimulus off-go. The follow-
ing times of indicator tone onset relative to the
stimulus were explored: 0.0S sec. before stimulus
onset (-0.10 sec), ±0.0-, +0.1S-, +0.30-, +0.S0-,
+1.0-sec. delays after stimulus off-go. The stimuli
used were 3/3, 4/4.

The 5s were given five or more consecutive
trials in each of the above conditions. These trials
were always preceded by at least two samples in
order to familiarize S with the exact time of
onset. The particular delay of the instruction
tone on any trial was thus fixed rather than chosen

randomly. The advantages of this procedure are
(a) optimal performance is most likely in each
delay condition, if .S” is prepared for that precise
condition (cf. Klemmer, 1957), (b) minimizing
delay changes makes possible a higher rate of
stimulus presentations. On the other hand, a
random sequence of instruction tone delays would
make it more likely that 5 was “doing the same
thing” in each of the different delay conditions.

The sequence in which the different delay con-
ditions followed each other was chosen either as
that given above (ascending series of delay con-
ditions) or in the reverse order (descending
series). Within a session, a descending series
always followed an ascending series and vice versa,
irrespective of the stimulus materials used. At
least two ascending and two descending series of
delay conditions were run with each 5″ and with
each material after the initial training (Experi-
ment 3) with that material. This number of trials
insures that for each S there are at least 20 trials
at each delay of the indicator tone.

Results and discussion. The development
of the typical behavior is illustrated by the
S, R O R , in Figs. 5a, b, c. Figure 5a shows
ROR’s performance in a single session, the
first posttraining session. The upper and
lower curves represent the ascending” and
descending series of tone instruction delays,
respectively. The arrows indicate the order.
Although each point is based upon only
five trials, the curves are remarkably similar
and regular. Clearly, most of the letters
in excess of ROR’s memory span are for-
gotten within about 0.25 sec. The rapid for-
getting of these letters justifies calling this a
short-term memory and accounts for the
fact that it may easily be overlooked under
less than optimal conditions. I n the follow-
ing session ( F i g . 5b) the descending series

.150 .30 .50 -.10 O .150 .30
DELAY OF INSTRUCTION TONE (SEC.)

Fjg. 5. Partial report of eight (4/4) letters,
three consecutive sessions. Arrows indicate the
sequence in which conditions followed within a
session. The light flash is shown on same time
scale at lower left of each figure. Bar at right
indicates immediate-memory for this material. One
subject (ROR).

THE INFORMATION AVAILABLE I N BRIEF VISUAL PRESENTATIONS

was given first. Here orderly behavior dis-
integrates. I n the t h i r d session ( F i g . Sc)
two modifications were introduced: (a) the
number of trials in each delay condition was
increased to eight and ( b ) a new delay con-
dition was given—namely, a signal tone com-
ing 0.05 sec. before the stimulus onset. The
curves are again regular, but they are ob-
viously different for the ascending and de-
scending series. For the session indicated in
F i g . 5c, an analysis of the errors by position
shows that in the ascending series the errors
are evenly split between the top and bottom
rows of the stimulus; in the descending
series, the top row is favored 3 : 1 .

ROR’s performance is analyzable in
terms of two kinds of observing behavior
(strategies) which the situation suggests.
l i e may follow the instruction, given by E
prior to training, that he pay equal attention
to each row. I n this case, errors are evenly
distributed between rows. Or, he may t r y
to anticipate the signal by guessing which
instruction tone w i l l be presented. I n this
case, ^ is differentially prepared to report
one row. I f the signal and S’s guess coin-
cide, S reports accurately; if not, poorly.
Such a guessing procedure would lead to
the variability observed in Fig. 5b. On the
other hand, S may prefer always to antici-
pate the same row—in the case of R O R
( l u g . 5c, descending series), the top row.
This would again allow reliable scores, pro-
vided only that there are an equal number
of instruction signals calling for the top and
bottom rows. Concomitantly, a differential
accuracy of report for the two rows is
observed. (ROR’s preference for the top
row is again prominent in Experiment 7,
Figs. 10, 11.)

Equal attention responding is initially
reinstated on the t h i r d day. The obvious
change in procedure which is responsible
is the introduction of a tone 0.05 sec. before
the stimulus onset (—0.10 sec. ” a f t e r ” its
termination). This signal is sufficiently in
advance of the stimulus so that perfect
responding is possible by looking at only
the row indicated by the signal tone. R O R
scores 100%, both in this condition and i n
the succeeding zero delay. The whole
(ascending series) decay curve of F i g . 5c

is highly similar to that of Fig. 5a. A run
w i t h 3 / 3 stimuli was interposed between
the ascending and descending series shown
in Fig. 5c. Since the guessing procedures
were easily sufficient for a nearly perfect
score w i t h 3/3 materials, when the descend-
ing series of delay conditions was run, R O R
continued guessing. While guessing was
advantageous at the long delays, at the short
delays it was a decided disadvantage.

Figure 6 illustrates the performances of
R N S , a sophisticated observer. R N S de-
scribed the two strategies (equal attention,
guessing) to E. I n accordance with the
instructions to do as well as he could, R N S
said that he switched f r o m the first to the
second strategy at delays longer than 0.15
sec. Thus in the three short delay con-
ditions, R N S divided his errors almost
evenly (19:21) between the favored (top)
and the unfavored rows; in the two longer
delays, errors were split 4:26. The dip in
the curve indicates that R N S did not switch
strategies quite as soon as he could have,
for optimal performance. Such a dip is

l±J
_J
CD
<

I
in
a:
UJ

u j 2 –

.10 0 .15 .30 . 5 0

D E L A Y O F I N S T R U C T I O N S ( S E C . )

Fig. 6. Partial report of eight (4/4) letters,
last of three sessions. Arrows indicate the se-
quence in which conditions followed within session.
The light flash is shown on same time scale at
lower left. Bar at right indicates immediate-
memory for this material. ( R N S )

10 GEORGE SPERLING

characteristically seen in experiments of this
kind.

The other ^s exhibit similar curves.8
These are not presented, as the main fea-
tures have already been demonstrated by
ROR and RNS. In summary, the method
of delaying the instruction tone is a feasible
one for determining” the decay of the short-
term memory contents, but experience with
the difficult, long” delays causes a consider-
able increase in the variability of 5″s’ per-
formance which is carried over even to the
short delay conditions. This has been at-
tributed to ..Vs’ change from an equal atten-
tion to a guessing strategy in observing the
stimulus.0

Part 2: Pinal Level of Performance

The analysis of the preceding experiment
has indicated that two distinct kinds of
observing behavior develop when the in-
struction to report is delayed. The accuracy
of report resulting from the first of these
behaviors (equal attention) is correlated
with the delay of the instruction tone; it is
associated with the .9s initially giving equal
attention to all parts of the stimulus. The
accuracy of the other kind of report (guess-
ing) is uncorrected with the delay of the
instruction ; it is characterized by Ss’ differ-
ential preparedness for some part of the
stimulus (guessing). Equal attention ob-
serving is selected for further study here.
The preceding experiment suggests three
modifications that would tend to make equal
attention observing more likely to occur,
with a corresponding exclusion of guessing.

1. The use of stimuli with a larger number
of letters, that is, 3/3/3 and 4 / 4 / 4 . Dif-

8 Sec Sperling (1959) for tables containing indi-
vidual averages of all trials for each S at each
delay of instruction tone : 3/3, 4/4

9 Increase in variability (with consequent decre-
ment in accuracy and/or speed) is not unusual
after difficult conditions. For example, Cohen and
Thomas (1957) in a clinically oriented study have
reported an exactly analogous “hysteresis” phe-
nomenon in a study of discriminative reaction time.
Hysteresis refers to the fact that, when the diffi-
culty of an experimental task is changed, the cor-
responding change in accuracy of response lags
behind the change in task.

ferential attention to a constant small part
of the stimulus is less likely to be rein-
forced, the larger the stimulus. The use of
three tones instead of two diminishes the
probability of guessing the correct tone.

2. Training with instruction tones that
begin slightly before the onset of the
stimulus. Jt is not necessary for S to guess
in this situation since he can succeed by
depending upon the instruction tone alone.
This situation not only makes equal attention
likely to occur, but differentially reinforces
it when it does occur.

When delays of longer length are tested,
priority should be given to an ascending
series of delays so that S will, at the be-
ginning of a particular delay sequence, have
a high probability of entering with the de-
sired observing behavior. This probability
might be nearly 1.0 by interposing a series
of trials on which the instruction is given
in advance of, or immediately upon, termi-
nation of the stimulus and requiring that .S”
perform perfectly on this task before he
can continue to the particular delay being
tested.10 This tedious procedure was tried;
but, as it did not have an appreciable effect
upon the results, it was discontinued. The
problem is that J?s learn to switch between
the two modes of behavior in a small num-
ber of trials.

3. The Jl may be able to gain verbal
control over Ss’ modes of responding. In-
itially, however, even i” cannot control his
own behavior exactly. This suggests a
limit to what E can do. For example, fre-
quently >.9s reported that, although they had
tried to be equally prepared for each row,
after some tones they realized that they
had been selectively prepared for a par-
ticular row. This comment was made both
when the tone and the row coincided and
(more frequently) when they differed.

10 Jt takes, on the average, a very large number
of trials for ,S~ to get 10 consecutive perfect trials
even if he has 6 or 7 of 9 letters available or
knows with 2/3 probability what the tone will be.
Success in this task within a reasonable time limit
demands a level of excellence reached only with
“equal attention” observing, as judged by the
other criteria.

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS 11

Some verbal control is, of course, pos-
sible. An instruction that was well under-
stood was:

You will see letters illuminated by a flash that
quickly fades out. This is a visual test of your
ability to read letters under these conditions, not
a test of your memory. You will hear a tone
during the flash or while it is fading which will
indicate which letters you are to attempt to read.
Do not read the card until you hear the tone,
[etc.].

The instruction was changed at the midway
point in the experiment. The 5 was no
longer to do as well as he could by any
means, but was limited to the procedure
described above. Part 2 of this experiment,
utilizing 9- and 12-letter stimuli, was carried
out with the three modifications suggested
above.

Results. The results for 3/3/3 and 4 / 4 / 4
letters and numbers are shown for each
individual ^ in Figs. 7 and 8. The two

1 1
– < - ! \

1 N
– 1 o ^ . 1 1

1
• , 1 , ,

AVERAGE 4 SUBJECTS

^SK*° 1

J

A

1
\

-.10 0 .15 .30*.

50

!00 12

10

-.10 0 .15 .30

A
I N °–o-

100

I -o UJ

75

50 •

_

u -.10 0 .15 .30 .50
100 12r

i \

~I0 0 .15 .30 10 0 .15 .30 .50
D E L A Y OF I N S T R U C T I O N T O N E (SEC.)

Fig. 7. Decay of available information: nine
(3/3/3) letters. Light flash is shown on same
time scale at lower left. Bar at right indicates
immediate-memory for this material.

ordinates are linearly related by the equa-
tion :

percent correct
100 X

no. letters in stimulus — letters available

Each point is based on all the test trials in
the delay condition. The points at zero
delay of instructions for NJ and JC also
include the training trials, as these Ss
showed no subsequent improvement.

The data indicate that, for all kS’s, the
period of about one sec. is a critical one
for the presentation of the instruction to
report. I f J5″s receive the instruction

0.05

sec. before the exposure, then they give
accurate reports: 9 1 % and 82% of the
letters given in the report are correct for
the 9- and 12-letter materials, respectively.
These partial reports may be interpreted to
indicate that the S’s have, on the average,
8.2 of 9 and 9.8 of 12 letters available.
However, if the instruction is delayed until
one sec. after the exposure, then the ac-
curacy of the report drops 32% (to 69%)
for the 9-letter stimuli, and 44% (to 38%)
for the 12-letter stimuli. This substantial
decline in accuracy brings the number of

-.10 0 .15 .30 .50 1.0 ” ” -.10 0 .15
DELAY OF INSTRUCTION TONE (SEC.)

F j g . 8. Decay o f available i n f o r m a t i o n : t w e l v e
( 4 / 4 / 4 ) letters and numbers. L i g h t flash is s h o w n
on same time scale at l o w e r l e f t . B a r at r i g h t
indicates i m m e d i a t e – m e m o r y f o r t h i s m a t e r i a l .

12 GEORGE SPERLING

letters available very near to the number of
letters that 5s give in immediate-memory
(whole) reports.

The decay curves are similar and regular
for each 5 and for the average of all 5s.
Although individual differences are readily
apparent, they are small relative to the
effects of the delay of the instruction. For
example, when an instruction was given
with zero delay after the termination of
the stimulus, the least accurate reports by
any 5s are given by ND, who has 8.1 letters
available immediately after the termination
of the stimulus. With a one-sec. delay of
instructions, the most accurate reports were
given by }C, who has only 5.1 letters avail-
able at this time.

Tn Fig. 3, in which whole reports and
partial reports were compared, only that
particular partial report was considered in
which the instruction tone followed the
stimulus with zero delay. It is evident from
this experiment that the zero delay instruc-
tion is unique only in that it is the earliest
possible “after” instruction, but not because
of any functional difference.

In Fig. 9, therefore, the 0.15-, 0.50-, and
1.00-sec. instruction delays are also plotted
for one 5, ROR. Data for the six- (3/3)

3 4 5 6 8 9 10 12
NUMBER OF LETTERS IN STIMULUS

Fig. 9. Immediate-memory and available infor-
mation. The parameter is the time at which avail-
able information is sampled (delay of instruction).
Heavy line indicates immediate-memory for the
same materials. One subject (ROR).

and eight- (4/4) letter stimuli arc taken
from the two ascending delay series with
each material that yielded monotonic re-
sults. Figure 9 clearly highlights the sig-
nificance of a precisely controlled coded
instruction, given within a second of the
stimulus off-go, for the comparison of
partial and immediate-memory reports. One
second after termination of the stimulus,
the accuracy of ROR’s partial reports is no
longer very different from the accuracy of
his whole reports.

Experiment 5: Some Exposure Parameters

In Experiment 3 it was shown that the
number of letters reported correctly is al-
most independent of the exposure duration
over a range from 15 to 500 msec. It is
well known, however, that the relation be-
tween the accuracy of report and the ex-
posure duration depends upon the pre- and
post-exposure fields (Wundt, 1899).

In a technique developed in Helmholtz’s
laboratory (Baxt, 1871) the informational
(stimulus) field is followed, after a variable
delay, by a noninf ormational, homogeneous,
bright post-exposure field. Using this
method, Baxt showed that the number of
reportable letters was a nearly linear func-
tion of the delay of the bright post-exposure
field.11

Other combinations of pre- and post-
exposure fields have also been tried (Dodge,
1907a). The usual tachistoscopic presenta-
tion utilizes gray pre- and post-exposure
fields (Woodworth, 1938). Baxt’s pro-
cedure, however, is the most disadvan-
tageous for the observer. A similar
procedure was therefore selected, in order

1 x This important method was described by
Ladd (1899) and James (1890) in their textbooks,
but it is no longer well known. Consequently it
has been “rediscovered,” most recently, by Lindslcy
and Emmons (19S8). Baxt (1871) intended that
the bright second field would interfere with the
lingering image of (he first (informational) field.
Unfortunately, the effect depends in a complex
way upon the intensity of the two fields. Derived
time values must be used with caution. In some
cases the Baxt technique may actually result in
no loss of legibility, the second field producing
a negative “afterimage” instead of merely inter-
fering with the positive image (cf. Footnote IS).

THE INFORMATION AVAILABLE I N BRIEF VISUAL PRESENTATIONS 13

to study whole and partial reports in a
vastly different visual presentation f r o m
that of the previous experiments.

Procedure. 1. 6″s were instructed to write all
the letters of the stimulus; 3/3 stimuli were used.
After several sample presentations of a stimulus
card followed by a light post-exposure field, i”s
were given a random sequence of normal (pre-
and post-exposure fields dark) and Baxt (pre-
exposure dark, post-exposure field light) trials.
The Baxt trials do not correspond exactly to the
presentation that Baxt used. In this experiment,
the post-exposure field is the same intensity as
the stimulus (informational) field, whereas

Baxt

usually used more intense post-exposure fields;
also, the stimulus field always remains on until
the onset of the post-exposure field, whereas Baxt
used a fixed five-msec. duration for the stimulus
field. The post-exposure field itself remains on
for about one sec. The pre-exposure field is
always dark, as in all the previous experiments.
Two exposure durations, 0.01S and 0.050 sec, were
tested.

2. Three .9s were tested with the Baxt presenta-
tion of a 3/3/3 stimulus at an exposure duration
of 0.015 sec. The partial report procedure was
used to determine the effects of the post-exposure
field on the number of letters available.

3. The same three Ss were run as their own
controls. The procedure was exactly the same as
in Paragraph 2 above except that the post-exposure
field was normal (dark).

Results. The complete results are given
in Tables 1 and 2. I n all tests, the Baxt
procedure reduces the response accuracy of
all Ss to about one-half of their normal
score. This finding’ confirms the earlier
studies. However, a linear relation between
exposure duration and the number of letters
reported was not observed. The failure to
find a linear relation may be due to the
previously mentioned differences between
the presentations.

For R N S and N D , the number of letters
available is nearly the same (about two)
in the partial report of 9-letter stimuli as
the whole report of 6-letter stimuli. The
fact that in both procedures the number of
letters given by 5″s is the same suggests
that a Baxt presentation reduces the num-
ber, or the length of time that letters are
available, and that it does not directly affect
the immediate-memory span.

ROR’s partial reports of Baxt presenta-
tions are considerably more accurate than
those of the other 5″s, although they are not

TABLE 1

A Comparison of Response Accuracy with
Two Different Post-Exposure Fields

Subject

RNS

ROR

ND

NJ

JC

Mean

Exposure (sec.)

(0.015)
(0.050)

(0.015)
(0.050)
(0.015)
(0.050)
(0.015)
(0.050)
(0.015)
(0.050)
(0.015)
(0.050)

Normal

3.9
4.0

4.8

4.4

3.8
3.7

5.1
5.4

4.1
3.8

4.3
4.3

Baxt

2.5
2.2

3.5
2.8

1.9
2.3

2.4
3.4

2.7
3.4

2.6
2.8

Note.—Number of letters correctly reported. Whole report
of six (3/3) letter stimuli. Normal = pre- and poat-exposure
fields dark; Baxt = pre-expoKure field dark, post-exposure
field white.

as accurate as his reports i n control pre-
sentations. R O R seemed to show improve-
ment on successive Baxt trials. JC, another
S who seemed to show improvement, was
given additional Baxt trials on which he
continued to improve slowly. Unfortunately,
it was unfeasible to determine the asymp-
totic performance of these two 5s. Whether
the difference in performance between
R O R , and R N S and N D is attributable to
some overt response, such as squinting or
blinking, was not determined.

Table 2 also enables the comparison of
0.015- and 0.050-sec. exposures of 3 / 3 / 3
stimuli. A decrease in exposure duration
has only a slight effect on the number of
letters available. This suggests, as in the
immediate-memory experiments, that the
duration of a tachistoscopic exposure is
not as important a determinant of the num-
ber of letters available as the fields which
follow the exposure.

Experiment 6 : Letters and Numbers

I n Experiment 3 partial reports were
found to be uniformly more accurate than

14 GEORGE SPERLING

TABLE 2

A Comparison of Response Accuracy with Two D i f f e r e n t Post-Exposure Fields

Delay of Instruction (sec.):

Exposure Duration (sec):

Subject

RNS (N)
(B)

ROR (N)
(B)

ND (N)
(B)

Mean (N)
(B)

0

0.015

8.0
2.0

8.6

6.3

7.3
2.2

8.0
3.5

0
0.05

8.7

8.9

7.0

8.2

0.

0.015

5.4
2.2

8.3
5.4

5.8
1.7

6.5

3.1

IS

0.05

6.6

8.5

6.4

7.2

Note.—Number of letters available (fraction of letters correct in partial report X number of letters in stimulus). Stimuli: nine
(3/3/3) letters. (N) = normal (pre- and post-exposure fields dark), (B) = Baxt (pre-exposure field dark, post-exposure field
white).

whole reports. I n one case, stimuli of eight
letters were used and only one row of four
letters was reported. Designating the letters
to be reported by their location is only one
of a number of possible ways. I n the follow-
ing experiment, a quite similar set of stimuli
is used; each stimulus has two letters and
two numbers in each of the two rows. The
partial report again consists of only four
symbols, but these are designated either as
letters or as numbers rather than by row.
i n addition, a number of controls which
arc also relevant to .Experiment 3 are con-
ducted.

Procedure. I. Training: The .9s were given
practice trials with the instruction: “Write clown
only the numbers if you hear a short pip (tone
0.05-scc. duration) and only the letters if you
hear the long tone (0.50-sec. duration).” The
tones were then given with zero delay following
the stimulus off-go. The stimuli were 4/4 L&N.

If. In the following session, tests were con-
ducted with five different instructions:

1. Letters only—Instructions given well in ad-
vance of stimulus to write only the letters in the
following card(s). (8 trials)

2. Numbers only—Write only the numbers in
the following card(s). (8 trials)

3. Top only—Write only the top row in the
following card(s). (4 trials)

4. Bottom only—Write only (he bottom row
in the following card(s). (4 trials)

5. Instruction tone—Write either letters or num-
bers as indicated by tone. Tone onset 0.05 sec.
before stimulus onset. (16 trials) ROR was also
given additional trials at longer delay times.

Results. The results are illustrated in
Table 3. For purposes of comparison, the
number of correct letters is multiplied by
two when an instruction was used which
required 5″ to report only four of the eight
symbols of the stimulus. This includes
instructions given well in advance of the
stimulus. A l l measures, then, have 8.0 as
the top score and are thus equivalent within
a scale factor to percent correct measures.
The range is 0-8 instead of 0-100. Scores
which are based on partial reports are
therefore directly comparable to the partial
report scores (letters available) obtained in
Experiments 3, 4, and S.

When stimuli consist of letters and num-
bers, but 5s report only the letters or only
the numbers, then the SV partial reports
arc only negligibly more accurate than their
whole reports of the same stimuli. The
average number of letters available (cal-
culated from the partial report) is just 0.2
letter above the immediate-memory span

THE INFORMATION’ AVAILABLE I N BRIEF VISUAL PRESENTATIONS IS

TABLE 3

Comparison of Five Procedures

Subject

RNS
ROR
ND
NJ
JC

Mean

Letters
only

5.0
6.5
3.5
4.0
3.3

4.5

Numbers
only

4.5
6.5
3.8
5.0
4.0

4.8

Average
L&N

4.8
6.5
3.6
4.5
3.6

4.6

Instr. tone

-0.10

4.3
6.3
4.1
4.6
3.4

4.5

Immediate-
memory

4.6
4.5
4.1
4.3
4.1

4.3

One row
only

7.3
7.3
7.5
—
8.0

—

Note.—Average letters and/or numbers available (fraction of letters-
ia stimulus). Stimuli: eight (4/4) lettcis and numbers, -numbers—correct in partial report X number of symbols

for the same material. For practical pur-
poses, the partial report score is the same
score that 5s would obtain if they wrote
all the letters and numbers they could (that
is, gave a whole report) but were scored
only for letters or only for numbers, in-
dependently by the experimenter. The par-
tial report of letters only (or of numbers
only) does not improve even when the in-
struction is given long in advance verbally
instead of immediately before the exposure
by a coded signal tone.

The estimate of the number of available
letters and numbers which is obtained from
the partial report of letters (or numbers)
only is also the same as the estimate that
would be obtained if, on each trial, 5s wrote
only one row—either the top or the bottom
—according to their whim. Reporting only
one row of four letters and numbers is a
task at which the 5 s succeed w i t h over 90%
accuracy, liven if they are scored for the
whole stimulus, by arbitrarily reporting
only one row they would still achieve a
score of almost 50% correct or almost four
letters available. This is why no delay
series were conducted. I f 5s had ignored
the instruction to write only the letters (or
numbers) and had written only a single
row on each trial, they would have shown
less than a 0.5 letter decrement, no matter
what the delay of the instruction.

Only R O R showed a substantial improve-
ment when reporting only the numbers (or
letters). He was the only 5 w i t h whom it
made sense to conduct a systematic delay

series, although checks w i t h other 5s con-
firmed this conclusion. Table 4 indicates
that two extra symbols are available to
R O R for report only when the tone is given
before the stimulus, but not if it is given
immediately after. I t should be noted that
the information in the instruction tone
comes only after it has been on for 0.05 sec.
A t this time it either continues or is termi-
nated. The actual “instruction” is thus
given 0.05 sec. after the tone onset. R O R
therefore requires that the instruction be
given within 0.05 sec. of the stimulus
termination if any benefit of the partial
report procedure is to be retained.

Whether the 5s would have shown im-
provement with a large amount of addi-
tional training in the partial report of letters
or numbers cannot be stated. Table 3 shows
that, when 5s are required in advance to
report only one row, this task is trivial. The
substantial advantage of partial reports of
rows (report by position) over partial rc-

TABLE 4

Partial Reports of Letters or Numbers

Subj.

ROR

Prior
Verbal
Instr.

6.5

Delay of Instr.

-0.10
6.3

0.0

4.7

Tone
(sec.)

+0.25

4.4

Immedi-
ate

Memory
4.5

Note.—Average of symbols available (fraction of letters—
numbers—correct in partial report X number of symbols in
stimulus). Stimuli: eight (4/4) letters and numbers.

16 GEORGE SPERLING

ports of numbers or letters (report by
category) when the instruction is given
verbally long in advance of the exposure
is retained even when the instruction is
coded and given shortly after the exposure.

The failure in Experiment 6 to detect a
substantial difference in accuracy between
partial reports of only letters (or only
numbers) and whole reports clearly illus-
trates that partial reports by position are
more effective for studying the capacity of
short-term information storage than partial
reports by category.

Experiment 7: Order af Report

Interpretations of the effects of instruc-
tions upon the report following a single
brief visual exposure have often been con-
cerned with cither the perceptual sensitizing
effects of an instruction given before the
exposure or with the importance of for-
getting between the exposure and a post-
exposure instruction to report. The decay
curves of Experiment 4 include both of
these effects. Previous studies, however,
have usually assumed the order in which
the various parts (aspects, dimensions, etc.)
of the stimulus are reported to be the sig-
nificant correlate of post-exposure for-
getting. The possibility that information
might be well retained even though not
immediately reported has been mentioned
(Broadbent, 1958), but experimental in-
vestigations of such an effect by an inde-
pendent variation of the order of report
by Wilcocks (1925), Lawrence and Laberge
(1956), and Broadbent (1957a) have ap-
parently shown otherwise. Broadbent
(1957a) has also shown a case in which
independent variation of the order of report
did not reduce overall response accuracy.

In the present experiment, order of re-
port is introduced as a purely “nuisance”
variable for the S. The S is instructed to
get as many letters correct as possible, but
the E randomly manipulates the order in
which they are to be reported. The experi-
ment is a survey of how ..9s adapt to this
kind of interference with the normal order
of their report.

Procedure. The .S”s were instructed to write the
row indicated by the tone (high, low) first, then
to write the other row. They were to try to get
as many total letters correct as possible, it being
of no importance in which particular row the
correct letters might be. The instruction tone was
given with 0.0-, .30- (or .50-), and 1.0-sec. delay
after the termination of the stimulus.

Controls. In addition to the trials with a high
or low tone, two sets of 8 (or 10) trials were
given with a neutral, middle tone. The instruction
was : “Write all the letters in any order you wish,
but do not begin writing until you hear the tone.”
The tone was sounded with 0.0-sec. delay follow-
ing termination of the stimulus and also with
1.0-sec. delay. It bears repeating here that 5s
were not permitted just to mark X’s but were
required to guess various letters.

Results. Controls: The instruction which
required ^s to wait for 1.0 sec. before be-
ginning to write their answer was ignored
by the ^s, since it was almost physically
impossible to begin writing sooner. Con-
sequently the two different controls—trials
on which 5″ was required to “wait” for 1.0
sec. and trials on which 5* could begin his
report immediately—are grouped together.
These data, which are almost exactly the
same as the memory span data (Experi-
ment 1), are presented on the far right in
Fig. 10.

The Ss’ responses on the control trials
are analyzed in terms of the correlation
between the location of letters on the
stimulus and the accuracy of the report of
these letters in the response. The symbols
T and B above I-M in Fig. 10 represent the
percentage of the letters of the top and of
the bottom rows that 5s report correctly.
The middle point is the average percentage
of the letters of the top and bottom rows
that were correctly reported by 5″s. The
middle point is therefore also the average
percentage correct of all the letters that
were reported. Figure 10 shows that all -S”s
report the top row of the stimulus more
accurately than the bottom row, if they are
not instructed with regard to the order in
which they must report the rows.

The average accuracy with which 5″s
report the top and the bottom rows, when
instructions to report one or the other of
these rows are given with various delays
after the exposure, is also illustrated in

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS

100r

n
100

75

• ‘ s TOP

–
. 1 S T ST

– 2 ™ S s ^ ” ^ 3
AVERAGE N o

” ” ^ B O T

1

T
1
]
1
1

•
1
1 1
1

ROR
–
–
_
–
75
50
25
0

” J – ” TOP \

~ S ^ / °

! ” ^ b o t /
1 VV N°
I ,

T 1 1 1 1 1 1
4
i
i i
B

–
–
–

1.0 I-M
DELAY OF INSTRUCTION TONE (SEC)

Fig. 10. Accuracy of the first (second) row reported and of the top (bottom) row as a function of
the delay of the instruction to report one row first. Light flash is indicated at lower left; immediate-
memory (I-M) at right.

Fig. 10. The accuracy of reports of the top
row decreases slightly as the delay of the
instructions increases. In other respects,
however, the data show no systematic
changes in accuracy with changes in the
delay of instructions. The data clearly indi-
cate that the top row is generally reported
more accurately than the bottom row al-
though the instruction to report each row is
given with equal frequency.

The same data may also be analyzed with
regard to the accuracy of the row that must
be reported first and the row that is re-
ported last. A l l ^s except ROR are more
accurate when they report the first row
(the row called for by the instruction tone)
than the second row. For most 6″s, there-
fore, the order of report is correlated with
the accuracy of report.

There is a slight tendency for the accu-
racy of report of the row which is reported
first to decrease as the delay of the instruc-

tions increases. On the whole, however, the
overall accuracy of report decreases slightly
with the delay of the instruction to report
one row first. The experimental interfer-
ence with the normal order of report does
not change the overall number of letters
reported correctly by any ^ by more than
about 0.5 letter.

In this task, unlike the preceding ones,
individual differences are more striking
than the similarities. The pooled data are
highly untypical of three of the five 5s.
Figure 11 was devised as a two-dimensional,
graphical analysis of variance to compress
the details of Fig. 10 into one figure.12
Each coordinate represents the accuracy of

12 Figure 11 is based upon a suggestion by E. B.
Newman. A statistical analysis of variance was
not attempted since it would have had to be carried
out separately on each S. There was not enough
data to make this worthwhile, and Fig. 11 serves
the same purpose.

IX GEORGE SPERLING

5
o
IX.
n. o i-
l
. i –
7-UJ o
cr
hi a.

100
90

BO

to
60

hf)
40
30

20

10

=-R0R r \
— RNS f \ .
— ND , >>

~NJ 1 | » W R N S
* AVE ^*o

SEC. DELAY
o 0
• .30
• .50
• 1.0

i i l l

a c

i i i t
v 10 20 30 40 50 60 70 80 90 100

PER CENT OF TOTAL – FIRST ROW

[*’jg. 11. Graphical analysis of position on
stimulus vs. order of report as contributors to
response accuracy. Each point represents the
average of all trials of an S at a particular delay
of instruction. The order in which the points arc
connected corresponds to the magnitude of the
delays. Upper left: position preference in control
(immediate-memory) report,

one row of the report relative to the whole
report. Thus the ordinate represents the
number of letters that an individual o re-
ports correctly in the top row of the
stimulus (independently of order) divided
by the total number of letters (both rows)
that he reports correctly. Similarly, the
abscissa represents the percentage of the
total correct letters reported by 5 that are
contained in his report of the first row.
Since each coordinate is relative to 5’s own
accuracy, no point of the graph is inacces-
sible to 5 provided that, if necessary, he is
willing to sacrifice some accuracy. Since
the interference w i t h 5’s order of report
in this experiment had only slight effects
on the overall accuracy of 5’s report, this
method of presenting the data is justifiable.

F r o m Fig. 11 it is immediately evident
that, for example, 50% of the correct
letters that JC reports are f r o m the top
row and, by implication, 50% are f r o m the
bottom row of the stimulus. More than
70% of the correct letters that JC reports
are in the first row reported by him. R O R
represents the converse, preferring to re-
port the top row accurately, remaining

indifferent to whether it is called for first
or last. Other 5s lie between these ex-
tremes, each 5 maintaining approximately
the same relative accuracy for the top and
the first rows throughout the various delay
conditions. Each 5, therefore, operates
within a characteristic, limited area of the
graph. N D is an exception. A t zero delay
of the instruction tone, both position and
order account heavily for the correct letters
reported by N D . A t 0.5-sec. delay, N D
ignored the order (preferring to concen-
trate on the top r o w ) , and at 1.0-sec. delay
she lost her position (top row) preference
as well. I n these three conditions, the total
number of letters correctly reported by
N D remained approximately the same,
within 0.5 letter of the control condition.
At 1.0-sec. delay neither position nor the
order contribute to N D ‘ s accuracy of report.

A l l 5s operate in the upper right quadrant
of Fig. 11. This illustrates the finding that
no 5 consistently reported the bottom row-
more accurately than the top, nor the last
row better than the row first called for by
the instruction tone. I t docs not, of course,
indicate that 5s could not report the bottom
row or the last row more accurately under
other conditions. While 5s normally behave
quite consistently, the data of N D show that
they may try a number of different pro-
cedures. The instructions given the 5s prior
to the experiment were not restrictive. N o
specific procedure for making a report was
suggested to the 5s, because the purpose
of the experiment was to find out how 5s
respond when they are not given detailed
instructions. W i t h suitable instructions,
training, and reinforcements, 5s could
probably be induced to make most of the
possible kinds of reports that can be dia-
gramed by Fig. 11. This remains an
empirical problem.

The results obtained in this experiment
support the conclusions that both a position
preference and the order of report ordi-
narily correlate with the accuracy of re-
sponse, but that probably neither are
necessary conditions for response accuracy.
Some 5s can relinquish the position prefer-
ence and a favorable order of report with
no appreciable decrement in accuracy. This

THE INFORMATION AVAILABLE LIS! BRIEF VISUAL PRESENTATIONS 19

finding is in opposition to Lawrence and
Laberge’s (1956) contention that accuracy
is accounted for by the order of report.
Accuracy and order are often correlated,
but if a favorable order of report is not
necessary for accuracy, then it cannot be
the cause of accuracy.

When 5* is given a signal indicating which
row is to be reported first (Experiment 7),
the accuracy of report of the row indicated
by the signal (the first row reported) may
be compared to the accuracy of the partial
report (Experiment 3). The overt pro-
cedure on each trial is quite similar in
Experiments 3 and 7. The only difference
is that in the order of report experiment,
after the 5s have finished writing the row
indicated by the signal, they must also write
down the other row. In the partial report
procedure they do not have to write the
second row. The partial report and the
order of report experiments also share a
common dependent variable: the accuracy
of report of the row indicated by the in-
struction signal.

In view of the similarity in procedure,
it is surprising that the accuracy of this
common datum should be so different in
the two experiments. For example, when
5s give only the partial reports (the in-
struction signal being given immediately
after termination of the stimulus), then
they report 90% of the letters correctly in
one row of 4/4 stimuli. When they are
required to write the other row also, then
they report only 69% of the first four
letters correctly. Every 5″, individually,
gives a more accurate partial report (Ex-
periment 3) than a report of the first row—
of two rows to be reported (Experiment 7).
The consistent superiority of the partial re-
port over the first half of a whole report
prevails even when the instruction to report
is delayed for 0.5 (or 0.3) sec. In all cases
where data are available, each 5 reports a
row of four letters more accurately when
he does not have to write another row of
four letters afterwards. That what 5s must
write later should affect the accuracy of
what they write first must be explained—
if we disregard teleological explanation—by
the effect of prior instructions on the accu-

racy of the report. In other words, if order
of report is effective in determining the
accuracy of report, then this effect must be
a function of instructions given prior to
any report at all. For some 5s, no effect of
order of report upon response accuracy was
observed.

The two findings, that partial reports are
uniformly more accurate than whole reports
and that order of report may be uncorrelated
with accuracy, contradict Lawrence and
Laberge’s conclusion that “partial” reports
are essentially similar to “first” reports. In
fact, the second finding (that in some cir-
cumstances order of report and accuracy
of report are not correlated) provides a
direct counterexample to their conclusion.
Their different results may be in part due
to the vastly different stimuli which they
used. Lawrence and Laberge’s entire
stimuli each contained less information than
two randomly chosen letters.

Discussion

In all seven experiments, 5s were re-
quired to report the letters of briefly ex-
posed lettered stimuli. Two kinds of reports
were explored: partial reports, which re-
quired the 5s to report only a specified
part of the stimulus, and whole reports,
which required the 5s to report all the
letters of the stimulus. Experiment 3
demonstrated that the accuracy of partial
reports was consistently greater than the
accuracy of whole reports. Another im-
portant difference between partial and whole
reports is the correlation of accuracy with
the delay of the instruction to report. This
was shown in Experiment 4 in which the
time delay of the instruction signal, which
indicated the row of the stimulus to be
reported, was varied. The accuracy of the
partial report was found to be a sharply
decreasing function of the time at which
the instruction was given. I f the instruc-
tion signal was delayed for one sec. after
the exposure, the accuracy of the partial
report was no longer very different from
that of the whole report. In Experiment 7
it was shown that the accuracy of the whole
report does not change as the time of the
signal to report is varied—over the same

20 GEORGE SPERLING

range of time—up to one sec. after the
exposure.

The two kinds of report can also be
considered in terms of the information (in
these experiments, letters) which they indi-
cate the .S” has available for report. In the
whole report, the S reports all the informa-
tion that he can. When he gives a partial
report, the 5″ may have additional available
information that is not required for the
report. A calculation of the information
available to the ^s for their partial reports
indicates that between two and three times
more information is available for partial re-
ports than for the whole reports. This
discrepancy between the two kinds of report
is short-lived. Information in excess of
that indicated by the whole report was
available to the -Ss for only a fraction of a
second following the exposure. At the end
of this time, the accuracy of partial reports
is no longer very different from that of
whole reports.

The whole report has already been ex-
tensively studied by psychologists. The
maximum number of items an individual
can give in such a report is called his span
of immediate-memory; whole reports are
usually called immediate-memory reports.
Experiments 1 and 2 extend the well-known
conclusions that the span of immediate-
memory is an individual characteristic and
that it is constant over a wide range of
stimuli and exposure conditions. Although,
in immediate-memory experiments, items
are conventionally presented sequentially,
Experiments 1 and 2 illustrate that this is
not necessary—that a simultaneous presen-
tation may also give results characteristic
of immediate-memory experiments.

The main problems to be considered here
concern the partial-whole report discrep-
ancy: (a) Why is the partial report more
accurate than the whole report? (b) Why
does the partial report retain this added
accuracy only for a fraction of a second
after the exposure ?

The answers proposed are a systematic
elaboration of an observation that is readily
made by most viewers of the actual tachis-
toscopic presentation. They report that
the stimulus field appears to be still read-

able at the time a tone is heard which
follows the termination of the stimulus by
150 msec. In other words, the subjective
image or sensation induced by the light
flash outlasts the physical stimulus at least
until the tone is heard. The stimulus in-
formation is thus “stored” for a fraction
of a second as a persisting image of the
objective stimulus. As the visual image
fades, its legibility (information content)
decreases, and consequently the accuracy of
reports based upon it decreases.

There is other evidence, besides such
phenomenological accounts, that suggests
that information is available in the form
of an image for a short time after extinc-
tion of the physical stimulus. In the first
place, it is inconceivable that the observers
should stop seeing the stimulus at exactly
the moment the light is turned off. The
rise and fall of sensation may be rapid, but
they are not instantaneous. The question is
not whether the observer continues to see
the stimulus after the illumination is turned
on, but for how long he continues to see
the stimulus. A number of different kinds
of psychophysical measurements of the rise
and fall of sensation have been attempted.
These estimates of the persistence of the
visual sensation vary from a minimum of
0.05 sec. (Wundt, 1899) to almost one
sec. (McDougall, 1904). The most repre-
sentative estimates are in the neighborhood
of 1/6 sec. (cf. Pieron, 1934), a figure that
is in good agreement with the results.13

13 Measurements of the persistence of sensation
have almost invariably used techniques which have
at most questionable validity. Wundt’s method de-
pends upon masking, the effect of the persisting
stimulus upon another stimulus. The masking
power of a stimulus may be quite different from
its visibility. McDougall’s measurements, as well
as those cited by Pieron, depend upon motion of
a stimulus across the retina. Such measurements
are undoubtedly influenced by the strong temporal
and spatial interactions of the eye (Alpern,
1953). Schumann’s ingenious application of the
method of Baxt to the determination of persistence
is probably the only experiment that utilizes pat-
tern stimulation. The other methods have not
been tried with pattern stimuli although there is,
a priori, no good reason why they have not been.
The possibility that the persistence of pattern
information is quite different from persistence of
“brightness” has not been investigated.

THF. INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS 21

In Experiment 5 it was shown that the
post-exposure field strongly influences the
accuracy of both the partial and the whole
report. This experiment indicates that the
available information is sensitive to inter-
ference by noninfarmational visual stimuli
which follow the exposure. The dependence
of available information upon noninforma-
tional visual stimulation is just the depend-
ence that would be expected of a visual
image.

Finally, there are subtle aspects of the
sequence of letters reported by an S which
characterize the information that is avail-
able to him. In sequentially spoken letters,
for example, there is a limit—two—on the
number of letters that can be adjacent to
any given letter. Different limits apply to
a two-dimensional visual display. I f in-
formation is stored in a form topologically
similar to the stimulus, this may be detected
by noting the sequential dependencies that
limit successive responses to the stimulus.

Probably the kinds of sequential respond-
ing that would most clearly distinguish
visual from auditory information storage
would be (a) the ability of the S to read
the rows of the visual stimulus backwards
as well as forwards, or to report the columns
or the diagonals, and (b) his inability to
do an equivalent task when presented with
the information sequentially. (All these
procedures merely require the report of
adjacent letters if the stimulus is two
dimensional.) Unfortunately, these particu-
lar experiments were not conducted.

The foregoing experiments offer some
relevant evidence. In contrast to spoken
letters and numbers, which are most accu-
rately recalled if they occur at the beginning
or end of a sequence (Pollack, 1952),”
no obvious gradients of accuracy were
found in the foregoing experiments. The
middle row actually tended to be slightly
better reported than the other rows. There-
fore, it is unlikely that the entire visual •
stimulus (12-letters and numbers) was
transformed into an auditory (sequential)
representation for storage. Such an entire
transformation is also unlikely, though not

34 Summarized in Luce (1956).

impossible, because of the relatively small
time between the stimulus exposure and the
report.

An analysis of errors reveals numerous
cases of errors that may be classified as
“misreading” (for example, confusions be-
tween E and F, B and R) and as “mis-
hearing” (for example, confusions between
B and D, D and T—Miller & Nicely, 1955).
Still other confusions (for example, C
and G) are ambiguous. All of these types
of errors occurred whenever errors oc-
curred at all. The ubiquity of misreading
and mishearing errors, taken at face value,
suggests that both visual and auditory
storage of information are always involved
in both whole and partial reports. A non-
quantitative error analysis is therefore not
likely to shed much light on the question of
visual imagery. The frequent mishearing
errors suggest that the storage of letters,
just prior to a written report, may share
some of the characteristics of audition. Like
the preceding analysis of the constraints
upon successive responses, error analysis
requires considerable research before it can
be quantitatively applied to problems of
imagery.

This then is the evidence—phenomeno-
logical reports, the effects of the post-
exposure fields, the known facts of the
persistence of sensation, and the detailed
characteristics of the responses—that is
consistent with the hypothesis that informa-
tion is initially stored as a visual image and
that the ^s can effectively utilize this in-
formation in their partial reports. In the
present context, the term, visual image, is
taken to mean that (a) the observer behaves
as though the physical stimulus were still
present when it is not (that is, after it has
been removed) and that (b) his behavior
in the absence of the stimulus remains a
function of the same variables of visual
stimulation as it is in its presence. The
units of a visual image so defined are al-
ways those of an equivalent “objective
image,” the physical stimulus. It is as logi-
cal or illogical to compute the information
contained in a visual image (as was done in
Experiments 3 and 4) as it is to compute
the information in a visual stimulus.

22 GEORGE SPERLING

“Visual image” and “persistence of sen-
sation” are terms suggested by the asyn-
chrony between the time during which a
stimulus is present and the time during
which the observer behaves as though it
were present. Although asynchrony is in-
evitable for short exposure durations, there
is, of course, no need to use the term “visual
image” in a description of this situation.
One might, for example, refer simply to
an “information storage” with the charac-
teristics that were experimentally observed.
This form of psychological isolationism
does injustice to the vast amount of relevant
researches.

Imagery that reputedly occurs long after
the original stimulation (memory images,
eidctic images, etc.) is of interest as well
as imagery that occurs for only a few tenths
of a second following stimulation. Whether
the term “imagery,” as it has been used
here to describe the immediate effects of
brief stimulation, is an appropriate term for
the description of the lasting effects of
stimulation is an empirical problem. I t is
hoped, however, that the principles and
methods developed here will not be without
relevance to these traditional problems.

Persistence of Vision and Afterimages

Between the short persistence of vision
and the remembrance of a long-passed
event, there is an intermediate situation,
the afterimage, which requires considera-
tion. In discussing afterimages, it will be
useful to distinguish some phases of vision
that normally follow an intense or pro-
longed stimulus. First, there is the “initial”
(or primary, or original) “image” (or sen-
sation, or impression, or perception, or
response). Any combination of a term from
the first and from the second of these
groups may be used. The initial image is
followed by a latent period during which
nothing is seen and which may in turn be
followed by a complex sequence of after-
images. Afterimages may be cither positive
or negative; almost any sequence is possi-
ble, but the initial image is almost always

positive.35 Some authors distinguish the
initial image from a positive afterimage
(for example, McDougall, 1904) ; others
do not (for example, von Helmholtz, 1924-
25). It is often implicit in such distinctions
that the persistence of the initial image is
due to a continued excitatory process,
whereas afterimages arise from receptor
fatigue. I f there is no repeated waxing and
waning of sensation, but merely a single
rise and fall, one cannot distinguish two
phases in the primary image, one cor-
responding to the “initial image” and the
other to an identical “positive afterimage”
of it.

Although it is difficult to prove that
visual information is stored in the initial
image, there can be no gainsaying that an
afterimage may be a rich store of informa-
tion. Positive or negative afterimages may
carry many fine details, including details
that were not visible at the time of stimula-
tion (von Helmholtz, 1924-25). After-
images generally last for at least several
seconds, and following high energy stimula-
tion they normally last for several minutes
(Berry & Imus, 1935). The clarity of the
details, of course, deteriorates with the pas-
sage of time. Since afterimages appear to
move when the eye is moved, they usually
have been considered retinal phenomena.
Taken together, these facts imply that there
is a considerable capacity for visual in-
formation storage in the retina. I f the
illumination of the stimulus cards used in
the foregoing experiments had been suffi-
ciently intense to blaze the letters upon the
retina and thereby take maximum advantage
of its information storage capacity, there
would have been little doubt afterwards as
to the nature or location of most of the
available information. The stimulus presen-
tations actually used, however, rarely elicited
reports of afterimages; .S’s usually reported
seeing simply a single brief flash. The
problem is therefore to determine the pcr-

3BRidwell (1897) and Sperling (1960) describe
conditions for seeing a negative “after-image
without prior positive image.” The method in-
volves a presentation quite similar to that of
Baxt (1871).

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS 23

sistence of the image of the brief flash, or
equivalently, the duration of seeing (the
stimulus) or the persistence of vision (of
the stimulus), rather than the duration of
an afterimage. These terms are used to
suggest that the -S1 feels he is responding
directly to the stimulus rather than to after-
effects of stimulation.

Psychologists have often carelessly as-
sumed that the absence of discernible after-
images following a visual presentation was
sufficient to insure that the duration of sen-
sation will correspond to the duration of the
physical stimulus, that is, that there is no
persistence of vision at all. Wundt (1899)
was one of the first to take vigorous excep-
tion to this naive view. Wundt’s most com-
pelling example was drawn from VVeyer
(1899). Weyer had found that two 40-
microsecond light flashes had to be sepa-
rated by 40 to 50 msec, in order for them
to be seen as two distinct flashes; at smaller
separations they were seen as a single flash.
In the dark adapted eye, the minimum
separable interval that consistently yielded
reports of “two flashes” was 80 to 100
msec.

Wundt argued that the two flashes could
not be seen as distinct until the sensation
occasioned by the first flash had ceased.
Thus, under optimum conditions, the mini-
mum duration of the sensation of a short
flash was at least 40 msec, which, in this
case, was 1,000 times the stimulus duration.
W.aidt thought that, in order to determine
t’ne duration of a longer flash, one must
merely add the 40 msec, of fade-out time to
the actual physical light duration. While
these details of Wundt’s reasoning may be
questioned, his main point, based on the
example of the short flash, is indisputable:
one docs not directly control the time for
which information is visually available
simply by manipulating exposure duration.
The experiments reported here provide a
direct proof of this assertion.

An Application of the Results to
“Before and After” Experiments

The previous experiments showed that
more information is available to i’s for a

few tenths of a second after the exposure
than they can give in a complete report of
what they have seen. It was suggested that
the limit on the number of items in the
memory report is a very general one, the
span of immediate-memory, which is rela-
tively independent of the nature of the
stimulus. Evidence was offered that infor-
mation in excess of the immediate-memory
span is available to the J? as a rapidly fading
visual image of the stimulus. I f more in-
formation is available to him than he can
remember, the 5″ must “choose” a part of
it to remember. In doing so, he has chosen
the part to forget. In Experiments 3 and 4,
i’s exercised only locational choices, that is,
portions of the stimulus were remembered
only on the basis of their location. Loca-
tional choices are probably not the only
effective choices that the 51 can make. Dur-
ing the short time that information is avail-
able to him, the S may process it in any
way in which he normally handles informa-
tion. Usually, what he does, or attempts
to do, is determined by the instructions.
The .S”‘s (unobservable) response to the
stimulus is probably the same whether the
instruction to make this response is given
before the stimulus presentation or after
i t ; the difference between the two cases lies
in the information that the response can
draw upon. I f the stimulus contains more
information than the S’s immediate-memory
span, and if the post-exposure instruction is
delayed until the 6* has little of this extra
information available, then a difference in
the accuracy of the responses with prior-
and post-exposure instructions will be ob-
served. I f the stimulus does not contain
more information than can be coded for
immediate-memory, or if the post-exposure
instruction is given soon enough so that
the S can utilize the still available informa-
tion effectively, then only minor differences
in the accuracy of responses with prior- and
post-exposure instructions will be observed.
If the stimulus is destitute of information
(for example, a single, mutilated, dimly
illuminated letter of the alphabet) then a
host of other factors which are normally
insignificant may become crucial. In this
case, the “stimulus” itself ma)’ well be

24 GEORGE SPERLING

irrelevant (Goldiamond, 1957), and the
effects of instructions given before or after
the exposure must be predicted on some
other basis.

There are some simple experiments in
which it is known a priori that the effects
of instructions given either before or after
the exposure will be exactly the same. This
degenerate situation can be illustrated by a
stimulus which is exposed for one micro-
second and with sufficient energy to be
clearly visible. By suitable coding, the pre-
exposure and post-exposure instructions
can be separated by only two microseconds.
The example serves to emphasize that what
is implicitly referred to by “before and
after” is not the exposure but something
else: traditionally, the sensitization and/or
forgetting that presumably occur in con-
junction with the exposure. Thus, the
theory that has been presented here merely
gives an explicit statement of assumptions
that have long been implicit.

Unobscrvable Responses and
the Order of Report

The subjective response to the high signal
tone is “looking up.” Since eye movements
cannot occur in time to change the retinal
image with any of the presentations used
(Diefendorf & Dodge, 1908) a successful
looking-up must be described in terms of a
shift in “attention.” Nonetheless, such a
shift in attention can be quantitatively
studied by means of a stabilized retinal
image (Pritchard, 1958) although Wundt
(1912), who did not use this modern, tech-
nically difficult technique, was able to give
many essential details. The reaction time
for the attentional response, like the reaction
time for more observable responses, is
greater than zero. Therefore, if the -S* is
given an instruction before the presentation,
he can prepare for, or sensitize himself to,
the correct row of the stimulus even though
there is not time enough for a useful eye
movement. The response to an instruction
which is given 0.05 sec. before the stimulus
is probably the same as the response to a
similar instruction that is given 0.1 sec.
later, immediately after the exposure. The

short time difference, 0.1 sec, accounts for
the similar accuracy of responding in these
two conditions.

Once his attention is directed to the ap-
propriate row, the S still has to read the
letters. This, too, takes time. Baxt’s (1871)
data indicate that the time required to read
a letter is about 10 msec. Baxt’s experiment,
with some modifications, was repeated by the
author, and similar results were obtained.16

How is all this relevant to the order of
report ? The order in which the letters are
finally reported can be an important variable
because of (a) purely temporal factors
(letters that arc reported first will be more
accurately reported only because they are
reported sooner after the exposure, the
actual order of reporting the letters per se
being relatively unimportant) or (b) inter-
action effects (the report of some letters is
detrimental to the report of the remaining
letters, that is, letters reported later suffer
from proactive interference by the letters
reported earlier).

That purely temporal factors cannot be
very important can be seen from the slope
of the curves describing available informa-
tion as a function of time. In the foregoing
experiments, the amount of available in-
formation approached the immediate-
memory span at about 0.5 to 1.0 sec. after
the exposure; further decrements in avail-
able information as a function of time are
slight. The report of the letters usually
does not begin until 1.0-1.5 sec. after the
exposure. The passage of time during the
actual time that letters are being reported,
therefore, cannot account for appreciable
accuracy changes as a function of the order
of report.

The second possible effect of order of
report—the interfering effect of the letters
reported first upon unreported letters—
cannot be so readily discounted. Proactive
interference would imply that partial re-
ports are more accurate than whole reports
by an amount dependent upon their relative
lengths. The results of Experiment 4 tend
to support this view. At delays of the in-

16 Sperling, G. Unpublished experiments con-
ducted at the Bell Telephone Laboratories, 1958.

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS 25

struction signal greater than 0.3 sec, partial
reports of three letters (from stimuli of
nine letters) indicate more available letters
than do partial reports of four letters (from
stimuli of twelve letters). On the other
hand, in Experiment 7 one S, ROR, does
not show decreased accuracy as a function
of the length of his report. ROR is able to
report eight letters as accurately when he
begins his report with three or four incor-
rect letters as when he ends his report with
three or four incorrect letters. Other 5s did
not systematically attempt to report incor-
rect letters first. Had they been required to
report incorrect letters first, they might well
have been able to do so.

The choice of what part of the stimulus
to attend to or of which letters to read is
the choice of what fraction of the stimulus
information to utilize. This choice can be
made successfully only while the informa-
tion is still available. The 5s prefer to
report what they remember first, but this
does not imply that they remember it be-
cause they report it first. It is difficult to
disentangle the many factors that determine
precisely what stimulus letters will appear
in the response, but important choices of
what information is to be recalled must
occur while there is still something to
choose from. Since the actual report begins
only when there are available but a few
letters in excess of the immediate-memory
span, the order of report can at most play
only the minor role of determining which
of the few “excess” letters will be for-
gotten.17

The Questions of Generality
and Repeatability

To what extent arc the results obtained
limited to the particular conditions of the
foregoing experiments ? The possibility that
the actual physical fading of the light
source is important to the availability of

17 There are many ways in which proactive
interference might occur. For example, if letters
are stored “sequentially” prior to report (cf.
Broadbent, 1957b), then the importance of order
of report may lie in the agreement of the two
sequences : storage and report.

information can be rejected not only on
prior grounds (see Apparatus) but also by
the empirical findings. For example, in
Fig. 9, the curve representing the number
of letters available 0.15 sec. after exposure
is quite similar to the 0.0-sec. delay curve.
There is no visible energy emitted by the
light source 0.15 sec. after its termination.

In the present case, the answer to the
problem of repeatability of the results is
made less speculative by three separate
investigations that have since been con-
ducted with similar techniques to those re-
ported here.

The experiments have been repeated by
the author18 with a different tachistoscope,
timer, and a light source that has only a
negligible afterglow. All the main findings
were reproducible.

Klemmer and Loftus (1958) confirmed
the existence of a short-term, high informa-
tion storage. They used a display consisting
of four discrete line patterns, the 5 being
required to report only one of these. The
instruction was coded either as a signal
light or verbally. Decay curves obtained
when the instruction is delayed are similar
to those reported above. A similar experi-
ment has also been conducted by Aver-
bach,10 who used a television tachistoscope
to present stimuli containing up to 16 letters.
A pointer appeared above the letter to be
reported. Initially, 5″s had about twelve
letters available for report, but the number
decreased rapidly when the visual instruc-
tion was delayed.

It is usually technically more difficult to
code instructions visually than acoustically.
Although the principle of sampling in order
to determine available information is com-
mon to both kinds of instructions, visually
coded instructions differ in some interesting-
ways from acoustically coded instructions.
For example, the time taken to “interpret”
—or even to find—a visual instruction may
well depend on its location relative to the
fixation point. Moreover, there may be

18 Sperling, G. Unpublished experiments con-
ducted at the Bell Telephone Laboratories, 19S8.

10 Averbach, E. Unpublished experiments con-
ducted at the Bell Telephone Laboratories, 1959.

26 GEORGE SPERLING

spatial interactions between the visual “in-
struction” and the symbols to be reported.
On the other hand, prior to training, the
task of interpreting a visual marker is
easier for Ss than the equivalent task with
an acoustically coded instruction. Ulti-
mately, such differences are probably only
of secondary importance since the two
kinds of experiments agree quite well.

Three main findings emerge from the ex-
periments reported here: a large amount of
information becomes available to observers
of a brief visual presentation, this informa-
tion decays rapidly, the final level is ap-
proximately equal to the span of immediate-
memory. Although the exact, quantitative
aspects of information that becomes avail-
able following a brief exposure unquestion-
ably depend upon the precise conditions of
presentation, it seems fair to conclude that
the main results can be duplicated even
under vastly different circumstances in
different laboratories.

Summary and Conclusions

When stimuli consisting of a number of
items are shown briefly to an observer,
only a limited number of the items can be
correctly reported. This number defines
the so-called “span of immediate-memory.”
The fact that observers commonly assert
that they can see more than they can report
suggests that memory sets a limit on a
process that is otherwise rich in available
information. In the present studies, a
sampling procedure (partial report) was
used to circumvent the limitation imposed
by immediate-memory and thereby to show
that at the time of exposure, and for a
few tenths of a second thereafter, observers
have two or three times as much informa-
tion available as they can later report. The
availability of this information declines
rapidly, and within one second after the
exposure the available information no
longer exceeds the memory span.

Short-term information storage has been
tentatively identified with the persistence
of sensation that generally follows any brief,
intense stimulation. I n this case, the per-
sistence is that of a rapidly fading, visual

image of the stimulus. Evidence in support
of this hypothesis of visual information
storage was found in introspective accounts,
in the type of dependence of the accuracy
of partial reports upon the visual stimula-
tion, and in an analysis of certain response
characteristics. These and related problems
were explored in a series of seven experi-
ments.

An attempt was first made to show that
the span of immediate-memory remains
relatively invariant under a wide range of
conditions. Five practiced observers were
shown stimuli consisting of arrays of sym-
bols that varied in number, arrangement,
and composition (letters alone, or letters
and numbers together). It was found (Ex-
periments 1 and 2) that each observer was
able to report only a limited number of
symbols (for example, letters) correctly.
For exposure durations from 15 to 500
msec, the average was slightly over four
letters; stimuli having four or fewer letters
were reported correctly nearly 100% of the
time.

In order to circumvent the immediate-
memory limit on the (whole) report of
what has been seen, observers were required
to report only a part—designated by loca-
tion—of stimuli exposed for 50 msec.
(partial report). The part to be reported,
usually one out of three rows of letters,
was small enough (three to four letters) to
lie within the memory span. A tonal signal
(high, middle, or low frequency) was used
to indicate which of the rows was to be
reported. The ^ did not know which signal
to expect, and the indicator signal was not
given until after the visual stimulus had
been turned off. In this manner, the infor-
mation available to the J? was sampled im-
mediately after termination of the stimulus.

Each observer, for each material tested
(6, 8, 9, 12 symbols), gave partial reports
that were more accurate than whole reports
for the same material. For example, follow-
ing the exposure of stimuli consisting of 12
symbols, 76% of the letters called for in the
partial report were given correctly by the
observers. This accuracy indicates that the
total information available from which an
observer can draw his partial report is

THE INFORMATION AVAILABLE IN BRIEF VISUAL PRESENTATIONS 27

about 9.1 letters (76% of 12 letters). This
number of randomly chosen letters is
equivalent to 40.6 bits of information, which
is considerably more information than pre-
vious experimental estimates have suggested
can become available in a brief exposure.
Furthermore, it seems probable that the
40-bit information capacity observed in
these experiments was limited by the small
amount of information in the stimuli rather
than by a capacity of the observers.

In order to determine how the available
information decreases with time, the instruc-
tion signal, which indicated the row of the
stimulus to be reported, was delayed by
various amounts, up to 1.0 sec. (Experi-
ment 4). The accuracy of the partial report
was shown to be a sharply decreasing func-
tion of the delay in the instruction signal.
Since, at a delay of 1.0 sec, the accuracy
of the partial reports approached that of
the whole reports, it follows that the in-
formation in excess of the immediate-
memory span is available for less than a
second. In contrast to the partial report, the
accuracy of the whole report is not a func-
tion of the time at which the signal to
report is given (Experiment 7).

The large amount of information in ex-
cess of the immediate-memory span, and
the short time during which this informa-
tion is available, suggests that it may be
stored as a persistence of the sensation re-
sulting from the visual stimulus. In order
to explore further this possibility of visual
information storage, some parameters of
visual stimulation were studied. A decrease
of the exposure duration from 50 to IS
msec, did not substantially affect the accu-
racy of partial reports (Experiment 5). On
the other hand, the substitution of a white
post-exposure field for the dark field
ordinarily used greatly reduced the accu-
racy of both partial and whole reports. The
ability of a homogeneous visual stimulus
to affect the available information is evi-
dence that the process depends on a per-
sisting visual image of the stimulus.

Whether other kinds of partial reports
give similar estimates of the amount of
available information was examined by ask-
ing observers to report by category rather

than by location. The observer reported
numbers only (or the letters only) from
stimuli consisting of both letters and num-
bers (Experiment 6). These partial reports
were no more accurate than (whole) re-
ports of all the letters and numbers. The
ability of observers to give highly accurate
partial reports of letters designated by loca-
tion (Experiment 3), and their inability to
give partial reports of comparable accuracy
when the symbols to be reported are desig-
nated as cither letters or numbers, clearly
indicates that all kinds of partial reports
are not equally suitable for demonstrating
the ability of observers to retain large
amounts of information for short time
periods.

In the final study (Experiment 7), the
order of report was systematically varied.
Observers were instructed to get as many
letters correct as possible, but the order in
which they were to report the letters was
not indicated until after the exposure. An
instruction tone, following the exposure,
indicated which of the two rows of letters
on the stimulus was to be reported first.
This interference with the normal order of
report reduced only slightly the total num-
ber of letters that were reported correctly.
As might be expected, the first row—the
row indicated by the instruction tone—was
reported more accurately than the second
row (order effect). There was, however,
a strong tendency for the top row to be
reported more accurately than the bottom
row (position effect). Although, as a group,
the observers showed both effects, some
failed to show either the order or the posi-
tion effect, or both. The fact that, for some
observers, order and position are not cor-
related with response accuracy suggests that
order of report, and position, are not the
major causes of, nor the necessary con-
ditions for, response accuracy. The high
accuracy of partial report observed in the
experiments does not depend on the order
of report or on the position of letters on
the stimulus, but rather it is shown to
depend on the ability of the observer to read
a visual image that persists for a fraction
of a second after the stimulus has been
turned off.

28 GEORGE SPERLING

REFERENCES

Adams, J. S. The relative effectiveness of pre-
and post-stimulus setting as a function of stimu-
lus uncertainty. Unpublished master’s disserta-
tion, Department of Psychology, University of
North Carolina, 1955.

Ai.pern, M. Metacontrast. / . Opt. Soc. Amer.,
1953, 43, 648-657.

Baxt, N. “Obcr die Zeit welche notig ist damit
ein Gesichtseindruck zum Bewusstsein kommt
und iiber die Grosse (Extension) der bewussten
Wahrnehmung bei einem Gesichtseindrucke von
gegebener Dauer. Pfliiger’s Arch. ges. Physiol.,
1871, 4, 325-336.

Berry, W., & Imus, H. Quantitative aspects of
the flight of colors. Amer. J. Psychol., 1935,
47, 449-457.

Bidwell, S. On the negative after-images follow-
ing brief retinal excitation. Proc. Roy. Soc.
Lond., 1897, 61, 268-271.

Bridgin, R. L. A tachistoscopic study of the
differentiation of perception. Psychol. Monogr.,
1933, 44(1, Whole No. 197), 153-166.

Broadbent, D. E. Immediate memory and simul-
taneous stimuli. Quart. J. exp. Psychol., 1957,
9, 1-11. (a)

Broadbent, D. E. A mechanical model for human
attention and memory. Psychol. Rev., 1957, 64,
205-215. (b)

Broadbent, D. E. Perception and communication.
New York: Pergamon, 1958.

Cattell, J. McK. t)ber die Tragheit dcr Netzhaut
und des Sehcentrums. Phil. Stud., 1883, 3,
94-127.

Chapman, D. W. The comparative effects of
determinate and indeterminate aufgaben. Un-
published doctor’s dissertation, Harvard Uni-
versity, 1930.

Chapman, D. W. Relative effects of determinate
and indeterminate “Aufgaben.” Amer. J. Psychol,
1932, 44, 163-174.

Cohen, L. D., & Thomas, D. R. Decision and
motor components of reaction time as a function
of anxiety level and task complexity. Amer.
Psychologist, 1957, 12, 420. (Abstract).

Dallenbach, K. M. Attributive vs. cognitive
clearness. / . exp. Psychol., 1920, 3, 183-230.

Diefendorf, A. R., & Dodge, R. An experimental
study of the ocular reactions of the insane from
photographic records. Brain, 1908, 31, 451-489.

Dodge, R. An experimental study of visual fixa-
tion. Psychol. Monogr., 1907, 8(4, Whole No.
35). (a)

Dodge, R. An improved exposure apparatus.
Psychol. Bull., 1907, 4, 10-13. (b)

Erdmann, B., & Dodge, R. Psychologische Unter-
suchungen iiber das Lesen auf experimenteller
Grundlage. Halle : Niemeyer, 1898.

Glanvillk, A. D., & Dallenbach, K. M. The
range of attention. Amer. J. Psychol., 1929, 41,
207-236.

Goldiamond, I. Operant analysis of perceptual
behavior. Paper read at Symposium on Experi-
mental Analysis of Behavior, APA Annual Con-
vention, 1957.

James, W. The principles of psychology. New
York: Holt, 1890.

Klemmer, E. T. Simple reaction time as a func-
tion of time uncertainty. / . exp. Psychol., 1957,
54, 195-200.

Klemmer, E. T., & Loftus, J. P. Numerals,
nonsense forms, and information. USAF Cam-
hridge Research Center, Operational Applications
Laboratory, Boiling Air Force Base, 1958. (Astia
Doc. No. AD110063)

Kulpe, O. Versuche iiber Abstraktion. In, Bericht
iiber den erste Kongress fur experimentelle
Psychologic. Leipzig: Barth, 1904. Pp. 56-68.

Ladd, G. T. Elements of physiological psychology:
A treatise of the activities and nature of the
mind. New York: Scribner, 1889.

Lawrence, D. H., & Coles, G. R. Accuracy of
recognition with alternatives before and after
the stimulus. /. exp. Psychol., 1954, 47, 208-214.

Lawrence, D, H., & Laberge, D. L. Relationship
between accuracy and order of reporting stimulus
dimensions. /. exp. Psychol., 1956, 51, 12-18.

Lindsley, D. B., & Emmons, W. H. Perception
time and evoked potentials. Science, 1958, 127,
1061.

Long, E. R., Henneman, R. H., & Reid, L. S.
Theoretical considerations and exploratory in-
vestigation of “set” as response restriction: The
first of a series of reports on “set” as a deter-
miner of perceptual responses. USAF WADC
tech. Rep., 1953, No. 53-311.

Long, E. R., & Lee, W. A. The influence of
specific stimulus cuing on location responses:
The third of a series of reports on “set” as a
determiner of perceptual responses. USAF
WADC tech. Rep., 1953, No. 53-314. (a)

Long, E. R., & Lee, W. A. The role of spatial
cuing as a response-limiter for location re-
sponses : The second of a series of reports on
“set” as a determiner of perceptual responses.
USAF WADC tech. Rep., 1953, No. 53-312. (b)

Long, E. R., Reid, L. D., & Garvey, W. D. The
role of stimulus ambiguity and degree of re-
sponse restriction in the recognition of distorted
letter patterns: The fourth of a series of reports
on “set” as a determiner of perceptual responses.
USAF WADC tech. Rep., 1954, No. 54-147.

Luce, D. R. A survey of the theory of selective
information and some of its behavioral applica-
tions. New York: Bureau of Applied Social
Research, 1956.

T H E INFORMATION AVAILABLE I N BRIEF VISUAL PRESENTATIONS 29

McDougall, W. The sensations excited by a
single momentary stimulation of the eve. Brit.
J. Psychol, 1904, 1, 78-113.

Miller, G. A. Human memory and the storage of
information. IRE Trans. Information Theory,
1956, I T – 2 , No. 3, 129-137. (a)

Miller, G. A. The magic number seven, plus or
minus two: Some limits on our capacity for
processing information. Psychol. Rev., 1956, 63,
81-97. (b)

Miller, G, A., & Nicely, P. E. An analysis of
perceptual confusions among some English con-
sonants. /. Acoust. Soc. Amer., 1955, 27, 338-
352.

Pieron, H. L’evanouisscment de la sensation
lumineuse: Persistance indifferenciable et per-
sistance totale. Ann. psychoi, 1934, 35, 1-49.

Pollack, T. The assimilation of sequentially-
encoded information. Hum. Resources Res. Lab.
memo Rep., 1952, No. 25.

Pritchard, R. M. Visual illusions viewed as
stabilized retinal images. Quart. J. cxp. Psychol.,
1958, 10, 77-81.

Quastler, H. Studies of human channel capacity.
In H. Quastler, Three survey papers. Urbana,
111.: Control Systems Laboratory, Univer. Illinois,
1956. Pp. 13-33.

Schumann, F. Die Erkennung von Buchstaben
und Wortcn bei momentaner Beleuchtigung. In,
Bericht iiber den erste Kongress fur experi-
mentelle Psychologie. Leipzig: Barlh, 1904. Pp.
34-40.

Schumann, F. The Erkennungsurteil. Z. Psychol.,
1922, 88, 205-224.

Sperling, G. Information available in a brief
visual presentation. Unpublished doctor’s disser-
tation, Department of Psychology, Harvard Uni-
versity, 1959.

Sperling, G. Afterimage without prior image.
Science, 1960, 131, 1613-1614.

von Helmholtz, H. Treatise on physiological
optics. Vol. I I . The sensations of vision.
(Transl. from 3rd German ed.) Rochester, New
York: Optical Society of America, 1924-25.

Wagner, J. Experimented Beitrage zur Psy-
chologie des Lesens. Z. Psychol, 1918, 80, 1-75.

Weyer, E. M. The Zeitschwellcn gleichartiger und
disparater Sinneseindrucke. Phil Stud., 1899,
15, 68-138.

Whipple, G. M. Manual of physical and mental
tests. Vol. I. Simpler processes. Baltimore:
Warwick & York, 1914.

Wilcocks, R. W. An examination of Kiilpe’s ex-
periments on abstraction. Amer. J. Psychol,
1925, 36, 324-340.

Woodworth, R. S. Experimental psychology.
New York: Holt, 1938.

Wundt, W. Zur Kritik tachistosckopischer Ver-
suche. Phil. Stud. 1899, 15, 287-317.

Wundt, W. An introduction to psychology. Lon-
don : Allen & Unwin, 1912.

3arly publication received April 28, 1960. (Ear

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Make my memory: How advertising can change our memories of the past
Braun, Kathryn A;Ellis, Rhiannon;Loftus, Elizabeth F
Psychology & Marketing; Jan 2002; 19, 1; ProQuest Central
pg. 1

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.