PSYCHOLOGICAL SCIENCE
General Article
HAS PSYCHOLOGY A
.FUTURE?
By Eleanor J. Gibson
I consider that psychologists have a duty to explore their own
field, to create hypotheses about the behavior of animals, es-
pecially Imman, and to test these and look for possible causal
relationships at the level ofacting, thinking, perceiving humans
in their environmelltal colltext. Our agenda should be at this
level, not one that depends all waiting for reductionist theories
at some other level, be it neural, genetic, nuclear, or especially
artificial. To go about this work, a developmelltal approach, in
a systems setting-the organism-environment system-holds
great promise and has the dignity of successful precedents in
other sciences.
All of behavior-perceiving, acting, problem solvillg, commu-
nicatillg with others-should be our provillce. The hallmarks of
human behavior call alld should be studied ill all ofthese areas.
It is their development that offers a road to understandillg.
We should continue to look for theories of considerable gener-
ality, however elegant a model may be in a tiny realm. If we
cannot find universal laws, we may still come up with unifying
principles.
Not so long ago, in allother keynote address, George Miller
told us that it was time to "give psychology away" (Miller,
1969). I am afraid that that is exactly what we are doing, though
not in the sense that Miller intended. Let us keep psychology,
and recapture the old excitement, looking for the causes in
behavior itself, and not giving up on the grand unified theory.
Over the last 5 or 6 years we have been celebrating,
one university after another, the founding of laboratories
of psychology. There have even been quarrels about
which was the first. Presumably, the theme for rejoicing
that motivated these celebrations was the separation of
our own discipline from philosophy. We accepted an
agenda that entailed the use of scientific methods of re-
search in the service of investigating the predictable re-
lationships in human (and animal) behavior, both cogni-
tion and action. At the time of psychology's birth as a
science, I believe it was accepted, too, that humans and
animals live and behave in a world, so the processes that
concern us should be studied in their environment, both
physical and social.
Despite the fact that early German psychologists took
physics as their model, rather than biology, a kind of
psychology known as functionalism soon flourished in
this country. The adaptiveness of behavior was empha-
Address correspondence to Eleanor J. Gibson, RRI, Box 265A,
Middlebury, VT 05753.
VOL. 5, NO.2, MARCH 1994
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
sized; both comparative and applied psychology were es-
tablished; and especially, from Thorndike on, the topic of
learning was a focus of theoretical development.
I suppose very few of you remember the excitement in
the air about the future of psychology in the 1930s. Young
apprentices envisioned unlimited possibilities for new
discoveries; it was unthinkable for a student not to be
well schooled in earlier theories that foreshadowed a new
science of psychology; and the grand theory, a truly psy-
chological one, was eagerly awaited, certain to make its
appearance during our lives as scientists. There were
some visible portents, the great men of the day; Hull and
Tolman are the ones I remember best, of course. There
was Piaget, too. Although I never had any special rever-
ence for him, I think now that his design for psychology
was the boldest and his intentions the most ambitious. In
any case, being a graduate student in a psychology lab-
oratory then was a highly sought after privilege, and a
great future lay ahead.
WHAT'S WRONG?
Something went wrong with this youthful dream,
about halfway into the century. Enthusiasm waned, and
wariness about the grand general theories grew. It was
supposed by a fair number of psychologists, remember-
ing perhaps the ecumenism of William James's Princi-
ples, that breaking the shackles of radical behaviorism
would set things right. Just let us use words like thinking,
perceiving, imagining, and reasoning, and a new day
would dawn. That happened. We had the so-called cog-
nitive revolution. It was spurred by events outside the
mainstream of psychology, partly the hiatus caused by
World War II, but principally by computers and artificial
intelligence, and by architectural discoveries in neuro-
science. Information processing came into its own. So
did "modeling," composing very precise formulations
("minitheories") about limited problem areas. It cannot
be denied that progress resulted from the work that was
thereby inspired. But, as time went on, some very bright
psychologists felt a little uneasy about what was happen-
ing, perhaps because of the narrowness of the lines that
began to be drawn. They made a mighty intellectual
thrust, and a new discipline resulted, called cognitive sci-
ence. I believe the hope was that bringing together cog-
Copyright © 1994 American Psychological Society 69
 PSYCHOLOGICAL SCIENCE
Has Psychology a Future?
nitive psychologists, philosophers, linguists, computer
scientists, and neuroscientists would bring back the life,
broaden the thinning stream, resuscitate the flagging sci-
ence, and regenerate the excitement that every young
psychologist felt 60 years earlier.
First-rate minds have gathered in many seminars seek-
ing to meld the ideas from these quarters so as to generate
something new and all-encompassing. The individual
contributions are generally praiseworthy, but I cannot
find any emergent principle rising from the meld, at least
any emergent theory that sounds to me like the view of
psychology for which I have waited so long. The new
combination is not more than the sum of its parts, nor
does it represent psychology. It will not be surprising to
you that I am disappointed. But so are some of the people
who were the most hopeful. A recent review of An Invi-
tation to Cognitive Science (Osherson & Lasnick, 1990)
proclaimed the individual chapters exceptional and even
exciting, but lacking in broader implications, and con-
cluded that "this volume as a whole does not fully con-
vey why cognitive science is indeed a coherent and vi-
brant discipline" (Keil, 1991, p. 292). Another reviewer
remarked that the book is dominated by the view that
behavior relies "largely or entirely on language," and
that the book has a "remarkably narrow biological base"
(J.A. Anderson, 1991, p. 287). _ 1940, I did not give dualism a thought, but in recent years
I interpret the combination of praise and uneasiness as
reflecting that, although there is nothing wrong with the
book itself (everyone says "buy it"), this interdiscipli-
nary approach does not, and perhaps cannot, lead to an
"integrated and coherent discipline" that exposes and
clarifies "deep underlying issues." I think that Keil
(1991) scored a direct hit in diagnosing some of the trou-
ble when he pointed out that a developmental view of any
of the systems under discussion might well have revealed
connections with one or more of the areas examined in
other chapters; and that restricting discussion of cogni-
tion to humans overlooks totally the importance of an
animal's encounters with its environment and the niche it
occupies.
Perhaps there is danger in "big science," danger that
one can be spread all over the map, without ever reaching
an encompassing theory, or the heart of the whole body.
There is certainly real danger in giant federal subsidies
that concentrate on a single domain, enlisting hundreds of
technicians, discouraging individual creativity, and by-
passing other needy domains. One politician of science
even suggested that universities might be requested to cut
research programs that are not "world class," while pur-
suing increased collaboration. But ideas are generated in
the small laboratories, and scientists are trained there.
- If we do not endorse the grand-seale-program ap-
proach, must we give up hope for the grand idea, the
universal laws? That is big science, but in a different way.
70
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
It is a search for encompassing principles, ones that make
sense of all the underlying relationships. I think many
psychologists have given up on this, a hope we still
shared 60 years ago.
AN AGENDA FOR PSYCHOLOGY
After listening to me play Cassandra, you may well
ask: "If you are woeful about the plight of (or the flight
from) psychology today, what do you suggest?" I do
have suggestions for a brighter future. They contain some
general strictures that apply to the whole field. And then,
of course, I have my own agenda, an agenda inspired by
attributes of living beings that are of concern to no other
science, but are our responsibility as psychologists.
First, my general strictures begin with some negatives.
I think we must rid ourselves of three ancient traps that
we have been caught in, dichotomies that have plagued
us over and over.
The first is dualism. Is not the uncertainty of dualism
the reason why there was rejoicing when psychology was
freed from philosophy? The endless argument left psy-
chology outside the arena of natural science, threw doubt
on the nature of our subject matter, and confused the
young people who were trying to identify the field. In
I have watched it creep back, the ghost in the machine.
We do not need it.
The concept of mechanism is one of the consequences
of dualistic thinking. Descartes concluded that animal be-
haviorcould be explained in terms of mechanisms; mech-
anisms also controlled human bodily functions, but
humans were gifted in addition with a mind-a super-
ordinate controlling force. The behaviorists accepted the
mechanism notion without buying the mind. But that led
in the end to dissatisfaction because of the dismissal of
mental (cognitive) functions. When someone asks me (as
they quite often do), "But what is the mechanism?" my
answer is that I am not a mechanist and I do not believe
in separation of mental processes and action. Accepting
the notion of mechanism implies dualism, that there is a
mind separate from the mechanism. This pernicious dis-
tinction leads subtly to reductionism, in the search for the
so-called mechanism.
The second thing we do not need is reductionism.
Meaningful explanations should be at their own level of
function. Causal relations do not exist between levels,
only at their own. To look for correspondences between
levels is most certainly a major interest of scientists; but
to be meaningful, a correspondence must exist between
appropriate units of each of two levels. Correspondence
as such does not in itself indicate a causal relationship,
but in the final analysis we may discover more general
laws that cover both levels. However impressed I am by
VOL. 5, NO.2, MARCH 1994
 PSYCHOLOGICAL SCIENCE
- Eleanor J. Gibson
recent progress in neuroscience (and I am-who would
not be excited by the imaging techniques that now ex-
ist?), we must not let our own science be driven by that
field. It is not true, as someone said to me at a recent
conference, that perhaps psychology has no subject mat-
ter of its own. We are badly needed as the scientists who
know how to study behavior, who can describe the intri-
cate intertwining of perceiving and acting in the adaptive
life of a human animal, and who can observe the devel-
opment of this activity with insight into the constraints,
opportunities, and environmental offerings that underlie
the dynamics of change.
Understanding the dynamics of change is the goal that
requires us to relinquish the third hobgoblin that has
haunted us-the nature-nurture dichotomy. Everyone
ostensibly agrees with the proposition that nature and
nurture are not separable, that we cannot attribute cau-
sality to one or the other alone, that the gene itself re-
quires interaction with supportive and dynamic media in
order to influence development-and yet, my sophisti-
cated colleagues can ask (and have asked), "Where do
you stand on the nature-nurture dimension?" There is no
such dimension; there is always interaction of many fac-
tors in development. I have been deeply distressed in
recent years by people who refer to certain achievements
of infancy as "innate," and then drop the problem as if it
were solved. This is a futile, evasive tactic.
What is the positive side then? We must study behav-
ior at its own level, looking for causes, predictions, and
laws (may we be so lucky!) at that level. I quote a phys-
icist who wrote recently, "The problem with these micro
levels is that one is not able to describe the process as a
whole, but only piece wise, in very, very small chunks at
a time. These pieces are devoid of meaning. Because the
meaning of behavior is to be found on the scale of the
body of the agent, one has to pick 'a level of abstraction
that allows one to express causal connections on that
scale" (Koenderink, 1992, p. 124). It is the functioning
(both cognitive and active) of the whole animal in the
environment in which it evolved that has to concern us.
To be more specific, I offer you my agenda. It deals
with the whole creature functioning adaptively in a dy-
namic exchange with a world of events and places and
people. I go for the big picture, and I have not given up
hope on big principles..
THE HALLMARKS OF HUMAN BEHAVIOR
I think there is just one way to understand behavior,
and that is to take a developmental approach. There is no
typical or standard moment of maturity; besides, we gain
Our understanding from change and becoming. (That
seems to be true of all science, from cell biology to cos-
mic physics.) The study of behavior should begin before
VOL. 5, NO.2, MARCH 1994
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
birth, and we are beginning to be able to do that, even for
a cognitive activity like remembering. I am not talking
about the ends of development, nor an ideal or norm, nor
traditional subdivisions that have been domains of spe-
cialization.
My program is organized around what I call the hall-
marks of human behavior. These fundamental properties
of human behavior are the ones that we, as psychologists,
need to account for. They are our subject matter. All of
them are present, in at least primitive form, at birth, but
they also develop, differentiating and elaborating, giving
us, as they shift gears or change, a way of studying the
factors that interact in making them what they are (and
we are), early or late.
I present each of the hallmarks very briefly, show that
it is present in rudimentary form early in life, and indicate
the kind of change that we may observe in the course of
development. Accounting for fundamental hallmarks of
behavior is not new-in a way, that is what Tolman was
challenging Hull to do years ago. But we now have new
technologies and a continuity of naturalistic observation
to spur the task. Certainly, the flourishing of research on
infancy over the past 20 years has made this approach a
more viable one, though our study only starts there.
I began with four hallmarks and have now added a
fifth, urged on me by my longtime friend Ulric Neisser:
• Agency (the self in control)
• Prospectivity (the forward-looking character of behav-
ior)
• Flexibility (transferability of means)
• Communicative creativity (multiplication of means of
communicating)
• Retrospectivity (the backward-looking character of be-
havior)
I begin with agency, or control, the property that
James (1879) had in mind when he wrote his famous pa-
per "Are We Automata?" What is our answer to this
question now? We can answer that activity is not reflex
even in an infant. Babies can act spontaneously, observe
the consequences of their actions, and show some selec-
tivity with regard to further action. Actions have envi-
ronmental consequences that can be observed. At the
same time, an action provides information about oneself.
This combination of intermodal information, about both
oneself and a change in the external world, is perceived by
the infant, and is a perfect source of knowledge for the self
as a causal agent. Piaget made observations of this process
on his own son at 2+ months, but we now have research
galore to attest to it. High-amplitude, nonnutritive sucking
has been used for years to investigate speech percep-
71
 PSYCHOLOGICAL SCIENCE

Has Psychology a Future?

tion in infants. Whether or not this procedure was professor spoke of the directedness of behavior. We
thought of as instrumental conditioning by its early users, could even take the notion back to Aristotle. Are we any
it is clear now that the baby can use the action to control wiser about it now?
what it wants to hear or see. Kalnins and Bruner (1973) All of these terms refer to the forward-looking char-
used sucking at high amplitude to produce clearing of acter of behavior. And I believe all of them must take for
focus of a movie presented to infants. When sucking re- granted the notion of control that I began with (though I
duced blur, amplitude rose to a peak; when the experi- certainly never heard Hull use the term control nor men-
menter altered the arrangement, sucking was discontin- tion intention or agency). In any case, activity is prospec-
ued. Infants learned to control the rate of sucking and tive; it must be prepared for any affordance of the envi-
breathing to maintain the picture in constant focus. ronment to be realized, or any task to be completed.
In other well-known experiments, strings have been Affordonce, a term coined by J.J. Gibson (1979), refers
tied to a baby's wrist (as Piaget did) or to an ankle to be to a reciprocal relationship between an animal and its
yanked so as to produce interesting spectacles, such as a environment. The environment provides opportunities,
mobile twirling (Rovee-Collier & Gekoski, 1979). Infants and the animal provides (or does not provide) an action
learn to control the mobile's motion, literally provided system that can utilize what the environment offers. Per-
with a tool that controls an external event. If the event is ceiving the relationship is perceiving an affordance, an
made noncontingent on the baby's pulls, action ceases animal-environment fit. To perceive an affordance im-
abruptly. So do expressions of enjoyment, such as smil- plies prospectivity. It requires detection of both the op-
ing and vocalization. Emotional expressions accompany- portunity offered by the environment (e.g., an object, its
ing learning about control and loss of it through the ex- size, its distance from the observer) and the dimensional
perimenter's manipulation are convincing evidence of the and dynamic properties of the observer (e.g., how far the
baby's discovery of causal efficacy, of itself as an agent arm can be stretched, how heavy a load can be lifted).
(Lewis, Alessandri, & Sullivan, 1990). The earliest age I Spontaneous exploratory activity of an animal gives rise
have heard cited in a relevant experiment was 10 days to to perceptible consequences that are utilized in ongoing
4 weeks (Meer, Weel, & Lee, 1993). Weights put on a activities, leading to predictable and economical utiliza-
baby's wrists allowed the experimenter to determine that tion of affordances. Information about the environment,
the baby's apparently random waving of an arm was not the actor~s bodily effectiveness, and the fit between the
so much random as exploratory. When a movement two is constantly coming in; as changes in the informa-
brought one of the baby's hands within its sight, the arm tion are observed, better and earlier detection of affor-
was dragged into view despite the weight, but not so the dances is possible.
nonviewable hand. The event ceased when the erstwhile Consider a baby reaching for something. An action
viewable hand was screened off. Infant behavior is more like reaching is always effected upon a background of
than mechanical responses to stimuli; it includes events postural stability. The posture must be prepared before
initiated by the perceiver-actor that inform the perceiver the explicit action can be performed. Postural prepared-
about action capabilities and their outcome. ness for reaching is demonstrably prospective from early
Two-year-olds have come a' long way in discovering infancy, showing intentionality in postural readiness for
agency and seeking to exercise control over objects and actions. Hofsten (1993) showed, by means of electromyo-
people, sometimes frustrating themselves and frequently graphic (EMG) recordings, that preparations for the
exasperating their caretakers. They are learning what reach occur earliest in the back muscles and then in the
they can do with a vengeance. Control of an observable abdominal muscles, all predating the arm extension itself.
event by one's own action is the essence of agency, even- This example may be a far cry from the long jumper
tually exercised ,in elaborate and strategic ways by skilled setting himself to clear a pit (Lee, Lishman, & Thomson,
planners. The detection of oneself as a potential agent 1982), but the organization of behavior over time extend-
and controller of observable external consequences fore- ing ahead is no less present.
shadows and leads to knowledge of causal relations and a Exploratory activity of all kinds is characteristic of
concept of oneself as agent, as Mandler (1992) has so early development, as new actions and new powers be-
persuasively argued in her article on "conceptual primi- come available and as the actor's dimensions change
tives. " (E.J. Gibson, 1988). Exploratory actions inform young
My second hallmark of behavior is prospectivity. I perceiver-actors about their own effective properties in
owe the word to David Lee and Claes von Hofsten both relation to present (and potential future) conditions to
of whom have written about it. But the concept i~ very make possible, prospectively, ensuing consummatory ac-
old indeed. Tolman wrote about "purposive behavior"; tions.
Hull wrote about "goal attraction and directing ideas" Detection of predictable, prospective relationships
(as habit mechanisms, to be sure). My Psychology 101 may also be the result of observation of sequences of

72 VOL. 5, NO.2, MARCH 1994

Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
-
PSYCHOLOGICAL SCIENCE
external events that are not necessarily the consequence
of an action of one's own. Haith (1993) and his colleagues
have conducted numerous experiments on :what he terms
"future-oriented processes in infancy." Experiments
with infants as young as 2 months have demonstrated that
visual expectations are formed for pictorial displays that
appear discretely and successively in predictable places
in the infant's field of view. Expectation of a presentation
was demonstrated by anticipatory eye movements. Haith
interpreted the anticipations as more than mere classical
conditioning because the character of the pictures (their
familiarity, e.g.) influenced the occurrence of anticipa-
tion; the babies apparently extracted content information
in the course of detecting predictability. Haith's finding
that transfer occurred when different actions were impli-
cated is further evidence of the nonreflex character of the
expectant behavior. Infants were presented first with se-
quences that required horizontal eye movements to ob-
serve the extended events. That procedure was followed
with presentations requiring vertical movements, impli-
cating a quite different musculature. Predictable alterna-
tive presentation in both cases resulted in facilitation of
expectant behavior in the second set of presentations. As
infants learn more about "what leads to what," about
predictable relations within events both self-perpetrated
and external, the prospective, forward-looking character
of behavior increases.
Finding transfer in this primitive case brings me to my
third hallmark, the flexibility of human behavior. I have
described flexibility as the transfer of means and strate-
gies, but typically more modest generalization occurs in
early development, especially before behavior and
events in the world are differentiated by a young per-
ceiver into the embedded sequences that characterize
most event relations. Events are typically nested, small
ones within larger ones. Predecessor events may have
different relations to .later events and to encompassing
ones, such as a task that involves some means to reach a
goal. The ability to select from varied means to realize an
affordance is a mark of flexible behavior. A need to ac-
count for such flexibility probably motivated Hull (60
years ago) to write his papers on the habit-family hierar-
chy (1934a, 1934b) and on "the assembly of behavior
segments in novel combinations suitable for problem so-
lution" (1935). There is a conspicuous interest today in
problem solving at an adult level-for example, J.R.
Anderson's (1993) work on problem solving and what he
calls "means-end operators" in a problem space, during
search for and selection of means. But that analysis in-
volves very high level cognition.
There is little research in progress on transfer of any
kind in children of 2 or 3 years (but see Brown & Cam-
pione, (984), and so far as we know, transfer seems to be
rather limited in neonates. Piaget (1954) observed some
VOL. 5, NO.2, MARCH 1994
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
Eleanor J. Gibson
flexibility in 2-month-old Laurent, who was tweaking a
toy with the aid of a ribbon tied to his right wrist. After
Laurent succeeded in tugging the arm reliably, making
the toy dance, Piaget tied the ribbon to Laurent's left
wrist. Laurent then tugged with the left arm, to the ex-
clusion of the right one.
Rovee-Collier and her colleagues have used a similar
task (only with an ankle instead of a wrist) and observed
the effect of changing the context in which the action
occurred, for instance, substituting a different crib lining
or a different mobile to be tweaked. Performance at 3
months seemed to be quite specific to the context. But if
the context was changed during training, the action was
resumed under a still different context several days later
(Rovee-Collier & DuFault, (991). Changing the context
during learning may increase attention to an invariant
property of the events, reminiscent of the old "learning
to learn" research (Harlow, 1959; Stevenson, (972).
Babies learn to use simple tools that involve separate
actions as means in a task during the second half of the
1st year (Willatts, (989). Exploratory activity in the
transfer of means seems to play an essential role in tasks
for which several means might be possible. An example is
new research by Adolph (1993) on locomotion of novice
crawlers and novice walkers presented with sloping sur-
faces to maneuver themselves over. The babies faced a
parent, with a slope of 0° to 40° intervening. Both novice
crawlers and novice walkers, given slopes to go up,
would have a try on all the slopes. There is little risk in a
failed attempt when attempting an upgrade, so that find-
ing is not very surprising. But downgrades are a different
matter. Novice crawlers frequently set off attempting to
crawl down slopes that were too steep for safety, and had
to be rescued. As they gained skill in crawling, they
learned new approaches to the steeper downhill slopes,
such as sliding, prone or sitting, or simply refusing to go,
seemingly recognizing the risk. Surprisingly, a few
months later, when the same babies shifted their mode of
locomotion to walking, many plunged downward at first
and went through exploratory trials a second time before
the strategies learned as crawlers were adopted. Flexibil-
ity in transferring means apparently requires some learn-
ing itself.
We expect a high level of transferability and flexibility
in selecting means to a goal from adult experts in a field.
But there may be a long road leading to strategic behav-
ior, including learning to differentiate tasks into sub-
events, acquiring a repertory of potential variations, and
developing selectivity with reference to goals and affor-
dances. Transfer of means from task to task with chang-
ing context of surroundings and affordances may also be
the road to the use of analogy-using applicability of
means to detect some deep similarity of process or basic
dynamics, a kind of creative insi~ht.
73
 PSYCHOLOGICAL SCIENCE
Has Psychology a Future?
Flexibility, which we have traditionally measured by
generalization and transfer in the laboratory, changes de-
velopmentally as exploratory activities; observation of
outcomes, contextual changes, task differentiation, and
the dawning of potential new affordances bring greater
variation of behavioral strategies and more differentiated
selection, with recognition of invariant aspects of differ-
ent tasks. Learning in development need not lead to au-
tomaticity and "habit," but rather to potentially greater
variability and selectivity. I think only developmental re-
search will tell us how this happens.
My fourth hallmark has to do with communication, but
not just communication per se. All animals communicate
in some way, to some extent, the means ranging from
pheromones to bird song and body language, but humans
have the potential to extend the means of communication
to an amazing degree. From gesture to vocalization to
language to reading print to mathematical symbols and
computerized codes, there seems to be no end.
Infants communicate with caretakers from day one.
Eye contact and mimicking of grimaces and tongue ex-
tension soon after birth prepare us (and them) for quick
development of more elaborate means via alerting par-
ents through their cries. By 2 months, a real interaction is
beginning. A wealth of research has shown that infants
and their mothers have facial and vocal interchanges that
are truly reciprocal, a kind of turn taking sometimes re-
ferred to as "protocommunication" (Trevarthen, 1979).
The adult smiles and speaks to the baby, then stops,
maintaining eye contact so the baby can detect an oppor-
tunity to make facial and other gestures and vocalize.
Experiments in which the exchange is interrupted make
clear that the baby anticipates its turn and expects a re-
sponse that is appropriately timed and controlled-
another example of prospectivity (Murray & Trevarthen,
1987).
Progress in our knowledge of how infants learn lan-
guage has been rapid and intensive in recent years. In-
fants discriminate phonemic differences of all languages,
apparently, to begin with; but by the second half of the
1st year, they are shaping their repertoire of expected
sounds to the language they hear. Parents do attempt to
teach children language in bouts of so-called joint atten-
tion, but language learning comes easily and naturally to
most children. Our literature is now rich in observations
of how this happens. Speech in quite young children in-
dicates intentionality, prospectivity, and flexibility of us-
age. They do not just repeat what they hear or are taught.
Speech also paves the way for reading, a further mode of
communication.
My own years of research on reading (I regret to re-
port) led to no grand theories, but they gave me a chance
to observe its development. Children vary enormously in
case of learning to read. Many children, if books are put
74
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
in their way, learn quickly and spontaneously, before we
even realize it. When demonstrating a form of the Stroop
test once, I gave it to my 4-year-old grandson as a com-
parison with adults to show how a preschool child would
rapidly name the color of the ink whatcver word was
inscribed. To my total surprise, he did not say "blue,"
but read the word "green," which had been printed in
blue. But sometimes children find the task of reading
hard and need a lot of help. However it happens, children
learn to read only by reading themselves. Flexibility of
approach and eventual transferability of subskills mark
progress, as does anticipation of the emoedded structure
of text.
Codes of other kinds seem to come quite easily to
older children and can even serve as a kind of play, some-
thing we might investigate further. Most interesting, how-
ever, is the ability of speakers to recode linguistic expres-
sions of events. If what we are trying to communicate is
not understood, or we do not like the way a sentence
turns out, we can try it other ways, using other words or
other means. Communicating, by language or other
means, is potentially very flexible, again illustrating the
organized interaction of these prime characteristics of
human behavior.
What gets recoded, in an effort to communicate about
a particular cvent, is apt to be information about an event
that has already happened. I am now warily approaching
my fifth hallmark, retrospectivity. Young children do not
usually talk about past events-they talk about what is
going on here and now and what will happen, or what
they want to happen, soon. Perhaps retrospectivity is
achieved only as we grow old enough to weave some kind
of story of our past lives. Yet babies remember things
from as early a beginning as technology has allowed psy-
chologists to probe. Newborns differentiate the voices of
their mothers from those of other persons on the basis of
events that took place before birth (DeCasper & Fifer,
1980). Newborns also remember a specific sound (a
word) and differentiate it from another word after a 24-hr
period, when the word is again presented (Swain, Zelazo,
& Clifton, 1993). Evidence is plentiful that remembering,
in some form, is going on.
"Memory" is one of the oldest topics in psychology. I
remember a speaker being told by his host after a collo-
quium years ago, "Well, Ebbinghaus said it all!" But
what a change we have seen in the way memory is
treated! It was once thought of as a kind of bank from
which we drew items (images?), the ease of doing this
depending on the amount of repetition, so that one item
provoked the next in a kind of list. Nowadays many kinds
of memory are differentiated, and no one would think of
confusing generic or generalized knowledge of the world
and what to do about it with remembering specific
events; nor would they confuse "implicit" memory with
VOL. 5, NO.2. MARCH 1994
 PSYCHOLOGICAL SCIENCE
- Eleanor J. Gibson
"explicit." I wonder if the many distinctions that have
been made are telling us that the term memory was a
categorical mistake in the first place. It i.s too static a
concept, and it fails to conjure up the functional quality of
remembering, its intentionality and flexibility.
Remembering has one thing in common with my other
four hallmarks: The way that we remember changes de-
velopmentally. Yes, babies can remember a face, a ges-
ture, a word, at least in the sense of responding to it
differentially (implicit or procedural memory, I suppose).
But it is only after considerable time that little children
appear to mark particular, individual events and are able
to recall them later at will. Generic recall of common
happenings at a birthday party or at the pediatrician's or
lunch at McDonald's is known to occur for a 3-year-old,
but remembering a particular birthday party or luncheon
is unlikely until later (Nelson, 1986). It is equally unlikely
that any of us remembers a particular event that occurred
when we were only 3 years old-so-called infant amne-
sia.
Retrospectivity certainly seems to demand language,
an ability to recode past events, and to relate them to
future anticipated ones. It implies a developing sophisti-
cation in all the other hallmarks, even flexibility in the
sense of considering what might have happened, not just
what did. Nelson (1993) has suggested that autobiograph-
ical memory-a kind of putting together the story of
one's own life-begins ·as "infant amnesia" ends. The
change, she thinks, has to do with communicating with
other people about shared events, detecting that the same
event can be perceived from both points of view, and
discussed from one's own. But even autobiographical
memories, however freighted with episodes they may be,
can change and be recoded as development proceeds and
new affordances are offered by changes in a person's
environment, means of acting, and way of living. Re-
membering is a dynamic process, and we should look for
its development to be adaptive. Retrospectivity is as
loaded with developmental change as prospectivity, and
thus is as potentially informative about the factors that
interact in creating a never-finished, perceiving, acting,
thinking human being.
'Koenderink, J.J. (1992). Wechsler's vision: An essay review of "Computational
I will only add that I hold these truths to be self-
evident: that agency, prospectivity, flexibility, commu-
nicative creativity, and retrospectivity are, and always
have been recognized as, qualities that mark human be-
havior, however often the names given them have been
altered over history. They still give us our agenda, and
because they are continuously developing, they offer us a
strategy for research, a way to investigate the factors,
both environmental and internal, that interact to drive the
change. If we observe the ontogeny of these properties
with all the new research methods at hand, are we at all
likely to find any general laws or trends? We might. Bi-
VOL. 5, NO.2, MARCH 1994
Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015
ology recognizes that evolution requires variation and
selection. So does ontogenetic development. Both are
essential for learning to occur, as Siegler and Munukata
(1993) have pointed out. Differentiation of species, or-
gans, and cells is characteristic of evolution and of indi-
vidual biological development as environmental and ge-
netic factors interact. These thought-provoking analogies
may help us understand development of the properties
that mark human behavior.
Acknowledgments-This article is in substance the author's keynote
address to the Fifth Annual Convention of the American Psycho-
logical Society, Chicago, June 1993.
REFERENCES
Adolph, K.E. (1993). Perceptual-motor de~'elopment in infants' locomotion over
slopes. Unpublished doctoral dissertation, Emory University, Atlanta.
Anderson, J.A. (1991). Comments on foundations of cognitive science. Psycho-
logical Science, 2, 283,285-287.
Anderson, J.R. (1993). Problem solving and learning. American Psychologist, 48,
35-44.
Brown, A.L., & Campione, J.C. (1984). Three faces of transfer: Implications for
early competence, individual differences, and instruction. In M.E. Lamb,
A.L. Brown, & B. Rogoff (Eds.), Ad~'ances in developmental psychology
(pp. 143-192). Hillsdale, NJ: Erlbaum.
DeCasper, A.J., & Fifer, W.P. (1980). Of human bonding: Newborns prefer their
mothers' voices. Science, 208, 1174-1176.
Gibson, E.J. (1988). Exploratory behavior in the development of perceiving, act-
ing, and the acquiring of knowledge. Annual Review of Psychology.- 39,
1-44.
Gibson, J.J. (1979). The ecological approach to ~'isual perception. Boston:
Houghton-Miffiin.
Haith, M.M. (1993). Future-oriented processes in infancy: The case of visual
expectations. In C.E. Granrud (Ed.), Visual perception and cognition in
infancy (pp. 235-264). Hillsdale, NJ: Erlbaum.
Harlow, H.F. (1959). Learning set and error factor theory. In M.S. Koch (Ed.),
Psychology: A study of a science (Vol. 2). New York: McGraw-Hill.
Hofsten, C. von. (1993). Prospective control: A basic aspect of action develop-
ment. Human De~'elopment, 36, 253-270.
Hull, e. (l934a). The concept of the habit-family hierarchy and maze learning:
Part L Psychological Re~'iew, 41, 33-52.
Hull, C. (l934b). The concept of the habit-family hierarchy and maze learning:
Part II. Psychological Review, 41, 134-152.
Hull, C.L. (1935). The mechanism of the assembly of behavior segments in novel
combinations suitable for problem solution. Psychological Re~'iew, 42,219-
245.
James, W. (1879). Are we automata? Mind, 4, 1-22.
Kalnins, LV., & Bruner, J.S. (1973). The coordination of visual observation and
instrumental behavior in early infancy. Perception, 2,307-314.
Keil, F.e. (1991). On being more than the sum of its parts: The conceptual co-
herence of cognitive science. Psychological Science, 2, 283,287-293.
Vision" by Harry Wechsler. Ecological Psychology, 4, 121-128.
Lee, D.N., Lishman, J.R., & Thomson, J.A. (1982). Regulation of gait in long
jumping. Journal ofExperimental Psychology: Human Perception and Per-
formance, 8, 448-459.
Lewis, M., Alessandri, S.M., & Sullivan, M.W. (1990). Violation of expectancy,
loss of control, and anger expressions in young infants. De~'elopmental
Psychology, 26, 745-751.
Mandler, J.M. (1992). How to build a baby: II. Conceptual primitives. Psycho-
logical Review, 99, 587-604.
Meer, A.L.H. van der, Weel, F.R. van der, & Lee, D.N. (1993). Gra~'ilational
know-how in neonates. Unpublished manuscript, Department of Psychol-
ogy, University of Edinburgh, Edinburgh, Scotland.
Miller, G.A. (1969). Psychology as a means of promoting human welfare. Amer-
ican Psychologist, 24, 1063-1075.
Murray, L., & Trevarthen, C. (1987). Emotional regulation of interactions be-
tween two-month-olds and their mothers. In T.M. Field & N .A. Fox (Eds.),
Social perception in infants (pp. 177-197). Norwood, NJ: Ablex.
75
 PSYCHOLOGICAL SCIENCE

Has Psychology a Future?

Nelson. K. (1986). Ewmt knowledge: Structure and function in del·elopment.
Hillsdale, NJ: Erlbaum.
Nelson, K. (1993). The psychological and social origins of autobiographical memo
ory. Psychological Science, 4, 7-14. . Crofts.
Osherson, D.H., & Lasnick, H. (Eds.). (1990). An invitation to cognitive science.
Cambridge. MA: MIT Press.
Piaget. J. (1954). The construction ofreality in the child. New York: Basic Books.
Rovee-Collier, C.K., & DuFault, D. (1991). Multiple contexts and memory reo
trieval at three months. Del'elopmental Psychobiology, 24, 3~9.
Rovee·Collier. C.K•• & Gekoski, M. (1979). The economics of infancy: A review
of conjugate reinforcement. In W.H. Reese & L.P. Lipsitt (Eds.), Ad,'ances
in child del'elopment and behal'ior (Vol. 13. pp. 195-255). New York:
Academic Press.
Siegler, R.S.• & Munukata, Y. (1993. Winter). Beyond the immaculate transition:
Advances in the understanding of change. SRCD Newsletter.
Stevenson, H.W. (1972). Children's learning. New York: Appleton-Century.
Swain, I.U., Zelazo, P.R., & Clifton, R.K. (1993). Newborn infants' memory for
speech sounds retained over 24 hours. Developmental Psychology. 29,312-
323.
Trevarthen, C. (1979). Communication and cooperation in early infancy: A de·
scription of primary intersubjectivity. In M. Bullowa (Ed.). Before speech:
The beginning of interpersonal communication (pp. 321-347). Cambridge.
England: Cambridge University Press.
Willatts, P. (1989). Development of problem-solving in infancy. In A. Slater & G.
Bremner (Eds.).lnfant del'elopment (pp. 143-182). Hillsdale. NJ: Erlbaum.

Forthcoming Papers

Feature Review
Circling Creativity
Jennifer J. Freyd

Research Articles
Going In and Out of Languages: An Example of Bilingual Flexibility
Franf:ois Grosjean and Joanne L. Miller

On the Dangers of Averaging Across Subjects When Using Multidimensional Scaling or the
Similarity-Choice Model
F. Gregory Ashby. W. Todd Maddox. alld W. William Lee

Structural Alignment in Comparison: No Difference Without Similarity

Dedre Gentner and Arthur B. Markman

Why Do People Use Figurative Language?

Richard M. Roberts and Roger J. Kreuz

Research Reports
Adults Listen Selectively; Infants Do Not
Jill Y. Bargones and Lynne A. Werner

Memory Processes in Classification Learning: An Investigation of Amnesic Performance in

- Categorization of Dot Patterns and Artistic Styles
Jonathan A. Kolodny

Technical Commentaries
Using Cluster Analysis to Discover and Characterize Covert Strategies
Shui-l Shih and George Sperling

Informed Consent for Psychological Research: Do Subjects Comprehend Consent Forms

and Understand Their Legal Rights?
Traci Mann

76 VOL. 5, NO.2, MARCH 1994

Downloaded from pss.sagepub.com at LAKEHEAD UNIV on March 12, 2015