Вы находитесь на странице: 1из 9

Learning through Others Author(s): Michael Tomasello Source: Daedalus, Vol. 133, No. 1, On Learning (Winter, 2004), pp.

51-58 Published by: The MIT Press on behalf of American Academy of Arts & Sciences Stable URL: http://www.jstor.org/stable/20027896 . Accessed: 19/02/2011 06:16
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=mitpress. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

The MIT Press and American Academy of Arts & Sciences are collaborating with JSTOR to digitize, preserve and extend access to Daedalus.

http://www.jstor.org

Michael

Tomasello Learning through others

is a biological adaptation. -L/earning on Earth The majority of organisms learn little or nothing during their in dividual lifetimes. On the other hand, are born in a highly mammals many state and so must immature indi they vidually survival. learn things crucial for their In order to find food reliably,

of cogni degree processes significant tion. But the modern view that learning assumes diverse forms in different spe cies and behavioral and oper domains, ates in concert with cognitive processes that may be specific to particular species or domains, has for the most part sus that search. pended For social species such as humans and an other mammals, important especially form of learning is social learning. Ob serving the activities ing about the world them enables of others and learn from or through to acquire in individuals

of foraging species must youngsters learn the spatial layouts of their local environments. In order to distinguish of friends from enemies, youngsters social species must learn to recognize who make up their social groups. at For several decades, behaviorists to find the laws of tempted learning that the individuals applied equally to all species, for any and all tasks, and that did not involve to any Max Michael Tomasello is codirector of the Planck Institutefor Evolutionary Anthropology and honorary prof essor of psychology at the University of Leipzig and atManchester Univer sity.His publications include "The Cultural Ori gins of Human Cognition" (1999), "TheNew : Psychology of Language Cognitive and Func tionalApproaches toLanguage Structure" (1998), A Case Study of Early Gram and "FirstVerbs : matical Development" (1992). Academy of Art

formation with less effort and risk than if they were forced to learn on their own. For instance, many species of rats learn to eat to avoid which foods and which rats eat and by observing what other to them sub then seeing what happens a safer strategy than clearly sequently new out foods for oneself.1 always trying an in the overall Despite similarity function of learning in the lives of differ ent species of mammals, social learning, learning, comes inmany different forms. In our empirical work over the past fifteen years, we have in like individual vestigated
i Bennett

forms of social
G. Galef,

learning

that

? 2004 by the American & Sciences

in Influences Jr., "Social of Norway Rats and Mate Choic es of International Journal Quail," Japanese of : 1 - 24. 14 (1 2) (2001) Comparative Psychology Food Choices

dalus

Winter

2004

51

Michael
Tomasello on learning

beings share with other primate as well as forms that are species, unique The human. forms unique ly mostly from some social derive, we believe, human cognitive
possess.2

processes

that only humans

arti because learning, most importantly facts and practices proto exemplified use of tools and the typically by linguis tic symbols invariably point beyond to the phenomena for which themselves To learn the they have been designed. use of a tool or a symbol, conventional an individual must therefore come to understand end, another mate why, toward what outside individual is using it.4 nearest pri humans' Chimpanzees,

In brief, because human beings per ceive the behavior of others in intention al terms that is, because they perceive a person or the table' 'cleaning 'opening the drawer,' rather than simply moving her limbs in a particular way they learn of others in unique from the behavior 'cul ways. We have called this process it from tural learning' to distinguish in social of processes learning general, and also to highlight the crucial role of of many hu culture in the acquisition man skills. and I have dis My colleagues kinds three of cultural learn tinguished : imitative learn instructed ing learning, and collaborative The ing, learning.3 to imagine them shoes' of other peo as beings ple, to understand conspecifics have intentional like themselves who ability selves of individuals in the 'mental lives like themselves, enables of these types of cultural learning. Most our empirical work has focused on only one type of cultural learning - imitation in children before about two years of age. So that will be my focus here. of others as intentional JLhe recognition is crucial in human beings like oneself
2 Michael Tomasello, Ann "Cultural Sciences C. Kruger, and

relatives, do not learn from one in this same way. In 1996,1 re another viewed all of the experimental studies tool use, and I concluded of chimpanzee are very that chimpanzees good at learn af ing from others about the dynamic fordances of objects, but are not skillful at learning from others new behavioral or intentional activities per strategies a if mother se.5 For example, rolls over a eats insects and the underneath, log her child will very likely follow suit. From her mother's learned act the child has that there are insects under this particular log but she did not learn from her mother how to roll or how learned to eat insects ; she these on her own.

and mental

over a Thus

log could have

the youngster would have learned the same thing if the wind, rather than her mother, had exposed the ants under the log. This is an instance of 'emulation concerns changes of learning,' which state in the environment rather than a or intentional activity conspecific's behavioral strategy. In some

H. Ratner, Hillary havioral and Brain Michael M. Tomasello, and

Be Learning," :495 - 552 ; 16 (1993)

in Cecilia "Do Apes Ape?" G. Galef, Jr., eds., Social :The Roots in Animals (San Learning of Culture : Calif. Academic ; Press, 1996) Michael Diego, The Cultural Tomasello, Origins Cogni of Human Heyes Bennett tion Press, (Cambridge, 1999). and Ratner, "Cultural Mass. : Harvard University

emulation circumstances, amore adaptive is strategy learning For example, than learning by imitation. and I pre Kathy Nagell, Kelly Olguin, sented chimpanzees and two-year-old human children with a rake-like tool and
4 Tomasello, Cognition. The Cultural Origins of Human

3 Tomasello, Learning."

Kruger,

5 Tomasello,

"Do Apes

Ape?"

52 D

dalus

Winter

2004

object.6 The tool could be used in either of two ways leading to the same end result of obtaining the ob one Within each group of species, ject. em a demonstrator subjects observed a of inefficient method ploy relatively tool use, while another group observed amore efficient method of tool use. The result: human the method in general copied of the assigned demonstra children

an out-of-reach

during the demonstration, perceiving the actions of the demonstrator just, in The effect, as other physical motions. states of the demonstrator, intentional as dis and thus her behavioral methods tinct entities, are simply not a part of their experience. story may be told about the communication of chimpan gestural zees. In a series of studies, we explored whether youngsters acquire their gestur a or al signals by imitative by learning In ritualization.7 process of ontogenetic two organisms ritualization, ontogenetic devise a communicatory signal through instances of a social interac repeated tion. For example, an infant may initiate nursing by going directly for the moth er's nipple, perhaps grabbing and mov arm in the process. ing her mother's in some future encounter the mother So A similar

Learning

through
others

tor (imitative learning), while chimpan zees used the same methods to obtain the object no matter which demonstra tion they observed (emulation learning). The interesting point is that many chil dren insisted on reproducing adult be havior even if it seemed inefficient a less successful leading to performance Imitation than that of the chimpanzees. is thus not a 'higher' or 'more intelli ; gent' learning strategy than emulation it is simply amore culturally mediated in some circumstances which, strategy and for some behaviors, has some ad
vantages.

sense, and respond to, her in might fant's hunger at the first touch of her arm, leading the infant to abbreviate her signal for hunger even further the next is presumably analo to most in the that human gous way fants learn the 'arms over head' gesture to request that adults pick them up attempt to crawl up the time. This

are very creative in using Chimpanzees about understand tools, and intelligent in environment the ing changes brought about by the tool use of others. But they do not seem to understand the instru mental
same

first as a direct
7 Michael C. Kruger, Development

behavior
way

of conspecifics
do. Humans

in the
per Tomasello, Ann Barbara George, and Andrea "The Evans, Jeff Farrar, in of Gestural Communication

as humans

the demonstrator's inten apparent un tion as centrally and important, they derstand this goal as something separate from the various behavioral means that to it. In be used the ab may accomplish sence of this to understand goal ability as separable in the and behavioral means actions of others, chimpanzees focus on the changes of state (including changes in the spatial position)
6 Katherine chael Nagell, Tomasello, in the Tool Use Raquel "Processes

ceive

Evolu Journal Young Chimpanzees," of Human : 175 -186 tion" 14 (1985) Michael ; Tomasello, Deborah and Thomas "A Longi Gust, Frost, of Gestural Communi Investigation in Young Primates 30 Chimpanzees," : 35 - 50 Michael ; Tomasello, (1989) Josep Call, Katherine and Malinda Nagell, Raquel Olguin, cation Carpenter, Signals "The by Young Learning and Use of Gestural : 137 : A Trans 37 (1994) Call, tudinal

of the objects
and Mi Olguin, of Social Learning

generational ; 154 Michael Warren,

Study,"

Chimpanzees Primates Josep Malinda

Tomasello, Thomas Frost, Nagell,

Jennifer

and Katherine panzee Gestural

dytes) 174-186.

of Chimpanzees (Pan troglo and Human Children (Homo sapiens),"

Carpenter, "The Ontogeny of Chim

Journal of Comparative Psychology 107 (1993)

'

Groups nication

: A Comparison Across Signals and Generations," Evolution of Commu 1 (1997) : 223 - 253.

dalus

Winter

2004

53

Michael Tomasello on learning

and then, as the adult the anticipates baby's desire and picks her up, as an abbreviated, ritualized ver sion of this crawling activity performed adult's body, purposes only.8 evidence suggests that not imitative ritualization, ontogenetic is for chimpanzees' learning, responsible of communicative gestures. acquisition a number use Individual chimpanzees of idiosyncratic signals that must have All available been individually - a ized that longitudinal analy finding ses have confirmed.9 cap Significantly, tive youngsters raised in peer groups to observe that have no opportunity use many older conspecifics frequently of the same gestures that are common In among other chimpanzee youngsters. an I and study, colleagues experimental an the individual removed from group and taught her two different arbitrary she could use to obtain desired gestures food from a human.10 turned When she re to her group and used these sig nals to obtain food from a human, not even one either chimpanzee reproduced - even new of the gestures though all of the ges the other individuals observed turer and were highly motivated for the food. thus acquire youngsters if not the totality, of their them gestures by individually ritualizing for with one another. The explanation to the is analogous this learning process Chimpanzee the majority,
8 Andrew in Andrew Lock, Lock, "The of Language," Emergence and Sym Gesture, ed., Action, : Aca York (New of Language

for emulation explanation learning the case of tool use. Like emulation

in

for communicative

ritualization does learning, ontogenetic to not require individuals analyze the behavior of others in terms of ends and means in the same way as does imitative an arm learning. Imitatively learning for nursing would touch as a solicitation an infant observe another require that infant using an arm touch and under that other infant's goal. Ritualiz arm touch, on the other hand, the ing the only requires the infant to anticipate in a con future behavior of a conspecific text in which the infant already has the ritualiza goal of nursing. Ontogenetic a tion is thus, like emulation learning, is im that useful very learning process in all social species - but it is not portant a individuals learning process by which attempt to reproduce or behavioral activities ers ; it is not cultural humans JLLuman imitation practice it. the intentional strategies of oth learning the way stand

invented

and ritual

beings begin to learn through at around the first birthday.

to experimentation the features of this unique distinguish form of learning from those of another. For example, if an adult takes the top off of a pen and a child then does the same, there are many possible explanations, and mimicking emulation including without knowing (copying movements are have what they for). Researchers therefore devised ingenious techniques But it takes clever for analyzing the different components of what the child perceives, understands, act. in a demonstrated and reproduces
:

bol :The Emergence demic Press,

1978). The Chimpanzees of Gombe : Mass. Har

See

Patterns vard

Jane Goodall, of Behavior

University

(Cambridge, Press, 1986).

to the tech For example, according Meltzoff fourteen devised, nique Andy month-old infants saw an adult illumi nate a box by bending down and touch to the top of it.11 Although ing her head
il Andrew One-Week Meltzoff, Delay: "Infant Long-Term Imitation Memory After for a

10 Tomasello,

Call, Warren,

and Nagell, "The Ontogeny Gestural Signals."

Frost, Carpenter, of Chimpanzee

54 D

dalus

Winter

2004

infants could more easily have solved this task by emulation (e.g., by touching the box with their hand), they instead as the chose to use the same means as was. it These infants adult, unusual the adult's could have been mimicking action without understanding the goal of turning on the light. But if they had been copying this action with their behavior the same goal inmind, would have been an instance of imitative unusual learning. In order to determine which of these was at work, Malinda and I tested nine- to Nagell,

on objects.13 For each object, the was followed by a actions of pair striking result the sudden illumination of col In the key ex ored lights, for example. one of the demonstrator's periment, as was actions marked paired verbally intentional the other ("There !")while was marked verbally as accidental ! the actions "), but otherwise ("Woops looked very similar. Instead of mimick actions ing both of the actions they observed, even the youngest infants reproduced as intentional the action marked signifi more often than the one marked cantly as accidental. that in study demonstrated to the able imagine goal to was even the adult ward which acting, saw any con never though they actually crete results. In a 1995 experiment, Meltzoff showed eighteen-month-olds an adult either completing successfully a task a (pulling apart two halves of Another fants were (because the adult's hands the ends of the dumbbell).

Learning

through
others

two mechanisms

infants on amodified of this task: we delayed the illu mination of the light after the infants' version of the action, and noted reproduction whether they looked in anticipation to the light.12 We found that infants and older looked to the in before it came on. anticipation light come If the light did not on, they often or looked quizzi their action repeated in the room. This sug cally to the people were the adult's that gests they adopting means in order to achieve the same goal of turning on the light. Infants thus were not just mimicking the adult's action, in imitative but were engaging learning to achieve a perceived of a novel means twelve-months
end.

Carpenter, fifteen-month-old

dumbbell) or trying but failing to do so

slipped off Infants were or not able to complete the task whether seen an adult com had they successfully it. Yet these eighteen-month-olds plete were not able to achieve the same result when they watched amachine either the same task successfully completing or trying, or so. Francesca to do failing, I and the basic replicated Bellagamba of this with twelveand findings study we but infants, eighteen-month-old found that twelve-month-olds could not reproduce the adult's intended action when unsuccessfully they only saw her trying to perform it.14
Nameera Akhtar, and In and and Develop

In another shown identical

experiment, identical actions

infants were

results, intentions. Carpenter, Na expressed meera Akhtar, and I showed fourteen a series of two to eighteen-month-olds
Novel and Multiple tal Psychology 24 (1988) 12 Malinda Michael Attention, from Acts Stimuli," :470 - 476. Katherine Developmen

that produced but with different

13 Malinda Michael fants

Carpenter, Tomasello,

"Sixteen-Month-Old Intentional Behavior

Carpenter, "Social Tomasello, and Communicative

Nagell,

and Joint

Cognition,

Competence Monographs of

Imitate Differentially Accidental Actions," Infant : 315 - 330. ment 21 (1998) 14 Francesca sello,

9 to 15Months

of Age,"

the Societyfor Research inChildDevelopment 63 (4) (1998).

"Reenacting

and Michael Toma Bellagamba : Acts Intended Comparing

dalus

Winter

2004

55

Michael Tomasello on learning

Other social

studies

learning influence what behavior children repro duce and so gain insight into what they as intentional action. Using interpret as a starting point, s Meltzoff study and George Gergely, Harold Bekkering, Ilday Kir?ly showed fourteen-month olds an adult touching her head to the top of a box to turn on a light.15 In their saw the adult study, half of the infants turn on the light while her hands were a blanket (she was holding occupied and half saw her around shoulders), on her turn it while her hands were free. Infants who saw the hands-free demonstration their heads touched the box with more often

have manipulated context in an effort

the to

own

similar

choice of a behavioral means or different circumstances. of older

in

A series of studies extends

children

For example, these findings. and his showed Bekkering colleagues three- to six-year-old children an exper a table in one of two imenter touching In one condition there were locations.16 and dots on the table in those locations, were no in another condition there dots. children usual condition, the adult's behavior matched exactly, ly or even straight copying her crossed arm positions because presumably was no other there apparent goal to her actions than these arm movements. In the dot condition, children however, as the exper touched the same locations her ex imenter, but often did not match act arm positions. This is presumably there were dots they in because when as 'touching terpreted the adult's goal there were no the dots,' whereas when dots be the only possible goal seemed to arms like this.' Bek one's 'moving In the no-dot

significantly than infants who saw the hands

Infants thus demonstration. occupied to in context situation of the used the terpret the adult's behavior, appearing to assume that if the adult's hands were free and she still chose to use her head, be a good reason for this the infants who saw choice. Meanwhile, in the other demonstration apparently then there must use of her head as neces terpreted the (and so sary given her circumstances as an inessential her of intention), part this action. and thus did not reproduce These of the interpretation across condi differed thus goal condition tions : in the hands-occupied on was the light' 'turn her apparent goal was it 'turn in the hands-free condition infants' adult's on the light with infants your head.' By fourteen a very thus evidence ac of intentional understanding

concluded that kering and his colleagues is imitation children's young guided by of adults' goals and their understanding of those goals, and that of the hierarchy children adults' main imitate what they perceive goal to be. studies have confirmed use context to interpret the

Subsequent that children

adults' actions, and that this influences an adult one what ; they learn. In study, to of children five demonstrated groups a a box.17 how to pull out pin and open
W. Wohlschlaeg i6 Harold Andrew Bekkering, "Imitation of Ges er, and Merideth Gattis, tures in Children is Goal-directed," Quarterly

months,

deep tion, of how it relates to the surrounding for their context, and of what this means
18-Month-Olds," 22 (1999) Development 12- and Infant :277 - 282. Behavior and

Journal of Experimental Psychology 53A (1) (2000) :


153 -164. 17 Malinda Tomasello, Intentions Learn (2002) a and Michael Prior

Carpenter,

Josep

and Harold 15 Gy?rgy Bekkering, Gergely, in Pre verbal Imitation Ildik? Kir?ly, "Rational Infants," Nature 415 (2002) : 755.

"Understanding Enables 2-Year-Olds Task," Complex : 1431 -1442.

Call, Others'

to Imitatively 73

Child Development

56 D

dalus

Winter

2004

What

differed among the groups was what the children experienced just prior to this demonstration. One group of children the adult's received information about goal ahead of time ;another none ; the three other received group received groups varying amounts of information about the adult's goal. In the demonstration, the adult either on the door of the tugged unsuccessfully box, or showed the box already open, or visited and opened three different boxes before demonstrating how to open the test box. Thus all of the children in all of saw a full demonstra these conditions to open the box, but only the in the three prior-information children could know what the adult conditions was about to do before she began this tion of how
demonstration. Twoand two-and-a

own a raspberry along her might blow same in her the moth arm, exactly place er did, or she might blow a alternatively, arm - in raspberry along her mother's in this case the behavior terpreting as on the reciprocally 'blowing partner's called this 'role reversal imi In an ongoing study, my col and I have found that eighteen leagues are more month-olds likely to employ than are this reciprocal interpretation At both ages, chil twelve-month-olds. in the dren are more likely to reciprocate situation where the adult, for example, pats her own head (and the child pats his the two own), than in the case where arm.' tation.'
partners act on one another.

Learning through others

I have

A similar process occurs in the learn use lin ing of language, since learning to is also reciprocal. Thus, guistic symbols an a uses when adult linguistic symbol a in communicative act, she directs the to something. Conse use a as the to to learn quently, symbol adult does, the child must learn to direct as the adult had the adult's attention child directed the child's.18 and I have Interestingly, my colleagues some offered evidence that recently on similar in children's goes thing early symbolic play. Before two years of age, adults children by watching imitatively learn symbolic behaviors with objects, the same way that they learn actions with artifacts. As instrumental they grow older, they look to the adult more often, and in some cases smile much more bolic the sym often, when producing behaviors. This is evidence that a children of this age are reproducing - a kind of kind special intentionality of mutually in reciprocal intentionality for the moment adult agree, for example, as if itwere a horse.
i8 Tomasello, Cognition. The Cultural

half-year-old significant at better the box themselves ly opening when knew the adult's they goal ahead of time. In this study, then, children the exact same behavior interpreted on whether differently depending they knew the adult's goal ahead of time cues in adult emo with no concurrent or the like. In other tional expression in the control conditions words, children were not able to provide tentional is doing,' the an in

children

were

to attend

of 'what the adult description in the prior informa whereas tion conditions they were able to under as the intentional stand the behavior action 'trying to open the box.'

in

are some kinds of actions that Ihere children observe and attempt to imitate that have a special structure because they involve people having goals toward one another a For example, reciprocally. mother might blow a raspberry along to im her child's arm ; if the child wants choice itate this behavior, she is faced with a that depends on how she inter action. Thus, she

which

the child and the to treat a pencil

prets her mother's

Origins

of Human

dalus

Winter

2004

57

Michael Tomasello on learning

the general ability to learn a symbol through role reversal imitation, it is still the case that in learning particu occasions chil lar words on particular dren often need to read the adults inten Given tions in order to connect the word referent. ap to its intended

meek a different children

character

studies acquisition that children as young as eighteen can combine months all of the types of we have discussed intention reading nov above while imitatively learning a in study of For example, el words. show
twenty-four-month-olds, an adult an

propriately Several language

objects, provide evi dence early age children come action. intentional - name And human it is what is learning because human ly, cultural learning even when beings, quite young, are these word learning that at a very to understand studies the intentional and states mental of other human beings. this cultural Through understanding, take in some human processes cognition not possible in other species directions an essential and make human cognition ly collective enterprise. able to understand

performed as intentional. adult had marked Like the studies of actions on

themselves, the action that the

(prior) target object by saying, "Let's go find the toma."19 She searched through several and rejecting with a buckets, extracting objects inside. She then novel object with an excited expression and stopped search a later In test, when ing. comprehension asked to go get the toma themselves, children chose the object the adult had identified as fulfilling her intention. This used amodified experiment procedure scowl extracted another to show that twenty-four-month-old children could identify the intended ref erent even when the adult was unable to open the container with the target object that is, when she had an unful inside filled intention. Another study investi use of their understand gated children's ac versus of intentional accidental ing a In novel tions when words. learning of the study twenty-four-month-olds, to her (prior) intention adult announced a target action "I'm by saying, perform " going to meek Big Bird ! She then per in counterbalanced order, one formed, she verbalized action, which " " one ! and intentional by saying Woops which she indicated action, by saying " were asked to "There ! Later, when they accidental
19 Michael "Learning Developmental Tomasello Words and Michelle Barton, Contexts," :639 - 650. in Non-ostensive 30 (1994)

nounced

her

intention

to find a

the novel

Psychology

58 D

dalus

Winter

2004

Вам также может понравиться