Scribd Logo
Upload

Welcome to Scribd!

  • 1996-Cognition and Learning

    Uploaded by

    Carolina
    20 views32 pages

    Original Title

    1996-Cognition_and_learning

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    20 views32 pages

    1996-Cognition and Learning

    Uploaded by

    Carolina

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 32
    x COGNITION AND LEARNING James G. Greeno “STANFORD UNIVERSITY AND THE INSTITUTE FOR RESEARCH ON LEARNING Allan M. Collins BOLT, BERANEK AND NEWMAN, INC, CAMBRIDGE, MA, AND [NORTHWESTERN UNIVERSITY, EVANSTON Lauren B. Resnick UNIVERSITY OF PITTSBURGH INTRODUCTION Cognition and leaming are cental concepis in educational peychology. Research on these topics has been productive both, for advancing fundamental scentfic understanding and for in- forming educational practice. In this chapter, we review re ‘search accomplishments that have influenced the character of ‘educational practice significanly. We also review research that has important practical implications bur that has only begun to snform practices of education, ‘We believe dat educational research s undergoing a major advance that will further deepen our theoretical understanding cof fundamental processes of cognition, learning, and teaching and further strengthen cur abilities t contbute to educational practice. This advance is leading toward a psychology of cogni- tion and learning that includes individual, social, and environ mental factors in a coherent theovetical and practical unde standing. Accomplishing this change will require merging and ‘extending concepts and methods that, until recently, have de veloped relatively separately in cognitive science, in ecological psychology, and in ethnographic anthropology and sociology. ‘The relationship between theoretical and practical unde sanding is one of the important aspects Of our science that is ‘currently in transition. One of the promising ideas is that re search can provide more articulate and more valid principles that serve as assumptions of practice (A. L. Brown, 1994; A L Brown & Campione, 1994; J. $. Brown, 1991). To develop the principles of a practical theory, several groups of researchers are conducting studies that we sefer to a8 design experiments (AL Brown, 1992; Collins, 1992) In these suadies, researchers and practitioners, paricularly teachers, collaborate in the de- sign, implementation, and analysis of changes in practice. Re sults provide case studies that can serve as instructive models bout conulitons that need to be satisfied for reforms of the same kind to be successful, and about conditions that impede success. Results also contribute to an accumulating body of ‘theoretical principles about processes of cognition and learning inthe social and material environments of schools and other set- ings “There are distinct tradons.n educational theories and pric- tices that derive from differing perspectives oa the phenomena ‘of the domain. We organize our discussion with three general perspectives that have developed in psychological research, ‘We recognize tha other organizing principles could be chosen, and thar many of our colleagues would characterize the field in different tems. Our version groups together many research contributions that could be distinguished in important ways, We have arrived at this grouping, however, in our own effort ‘o understand broad trends and issues in educational esearch, and we hope that this characterization is helpful to readers in thet efforts to grasp general characteristics ofthe field. ‘The perspectives correspond to three general views of knowing and leaming in European and North American thought, which, generally following Case (1992, 1992) and Packer (1985), we refer to as empiricist, rationalist, and prag- ‘mats soctobistoric. Por the third view, Case used the simpler label "sociohisiorie,” but we use the adminedly more cumber- Preparation of this chapter was supported by National Scence Foundation (NSE grants No, MDOIS400 and ESIO450522 to the Trae For Research on Learing ().6.G.); by NSF grant No. MDIG0S3609 to Bot, Beranck and Newman, Inc. (AC); bythe National Resezch Center on Stodent Learing of the Leming Research and Development Ceate, University of Pitsbugh, with funds from the Office of Educational Research and lmpcovement, OERI Award No. RI7G1O003, United Sates Deparment of Fdcation (LR); and By a rant fom the Andrew W. Mellon Foundation BR), Weare grateful for comments by Rober Cafe, Robbie Case, and Richard Snow on dfs ofthis chapter, and for conversations ‘with many colleagues, parculaly with Giyoo Hatano, a8 we developed these ideas. 16 © COGNITION AND MotaTION some term to emphasize the largely separate origins of the vier in American though. Packer's discussion, focused on the hemenestie perspective, exemplified hy Heidegesr provides ‘eackground forthe siualive perspective (Winograd & Flores 1986) descried below. Empiricism, pied by Jacke and Thorndike, emphasizes consistency of knowledge wath exper nee, Rationalism, ypfed by Descanes and Piaget, emphasiass oncepival coherence and formal rtesa of tuth. Praga ‘ypiied by Dewey and Mead, and sociohistocism, typified by Yegotsky, emphasize that knowledge i constructed in racial sciviis 51 aca OF pale a6 they interact wah each other andtheirmatenal environments, Current manifesations ofthese {ice perspectives ae the bebanionst perspective the cograine perspective, and the siuatve perpectve ‘all three ofthese perspecives have contributed, and con tinue to canbe, important insights to fundamental siete ienowledge and understanding of cognition and feaming and have influenced educational practices sigificandy. While each perspective is valuable, they frame theorcical and practical ‘sues in distinctive and complementary ways, somewhat inthe sway that physics, chemistry, and biology frame issues sur sounding processes Sich as genetic replication in different but complementary ways. We hope, in this chapter, to convey the considerable strengths of all three ofthe perspectives snd the ‘value and importance of using their resources phuralisically in considering eiveational problems In the scond section ofthis chapter we discuss theoretical developments within the three perspectives, The section isorga- rized around thee theoretieal issues: the nature of knowing, the nature of learning and transfer, and the nature of motivation and engagement. We discuss research regarding each ofthese issues from the three perspectives. Tn the third seetion we discuss ways in which the three perspectives contibute to understanding and carrying out ed ‘atonal practices, The section s organized around vee practi cal issues: design of leaming environments, analysis and form Intion of caricula, and ascssmnen, which we discuss fom the three theoretical perspectives. We discuss these as examples of issues in eiucational practice tn which recent ané current design experiments have begun to develop a coherent body of principles in practice, OF course, these ate a small eubset of the practical issies that must be addressed and understood in the broad effort strengthen the educational system, and we discuss some additional sues brily inthe last section oF this chapter ce Cancaumons, p. 39) ISSUES OF THEORETICAL CONCEPTUALIZATION ‘This section considers thee thematic issues in the theory of| ‘cognition and leaming: > + the nature of knowing, > + the nature of leaming and transfer, and + the nature of motivation and engagement. he three general perspectives, the behaviorisvempircst view, the cognitive/rationalist view, and the situative/pragma- These ace coNeah Poragaphs chet Summarize. the Ste of dens tuscsociohistone view, frame each of these issues in diincive and complementary ways CD Inthe benanoravempiiis view, knowing is an organized ecumlation of associations and components of sll. eam ing isthe process in which associations and sil ate acquired and transfer occurs to the extent that Behaviors leamed in one situation are uiized ia another situation. Mouivation i a sate of the leamer that favors Formation of new associations and skis, pimanly volving incentives for arending to relevant aspett ofthe station and for responding appropriately. There ‘ce three traditions that we consider contribute to this view “Asoctatonis, which goes back to Locke and Hume, viewed Knowing 2s the associations berween ideas and learning 2 building new associations. Bebavioriom took the postion that aowing could be characterized oniy ia ters of observable connections between stimslt and responies and leaming ia terns of forming and strengthening or weakening and extn {ishing those connections through reinforcement or sonzen- forcement. Connecionisn (or neural networks) treats know. ledge as the patiem of connections berween neuroniike ements and learning as the suengthening or weakening of those connections “The commiive/ratonalss penpecsive on knowledge empha- Siees understanding of concepts and theories indifferent subject matter domains and general cognitive abilities, sich as reacon ing, planning, solving problems, and comprehending angusse. ‘There ae thre tadons of research that we consider o be branches ofthe cognitive perspective. The oldest ofthese is Gestalt psychology, hich emphasized the structural nature of knowledge and the imponance of insight in leaming. A socond tradition, consiuciism, was ongially developed by Paget and is focused on characterizing the cognitive growth of chi dren, especialy thee growth in concept understanding. The third wadiion, symbolic formation processing was developed ‘in American cognitive science by Chomsky, Simon, Newel and ‘others und i Focused on characterizing processes of language ‘understanding, reasoning, and problem solving. (Case (1992) ‘assed symbolic information processing & an empitici r- dition because ofits forus on knowledge asa set of associative networks and procedures. We locate it in the constrictvist category because of ts emphasis on the organization of inform tion in cognitive suuctures and procedures. This is but one ‘example of ways in which a classification has to ince cel Lively arbitrary Boundaries. AkhoUgh thece ae significant fer ‘eaces of emphasis between these research tadiions, they share imporant framing assumptions, especially the consmuctvst ‘and information processing waditions. All dee traditions em ‘hasize the importance of organied pattersin cognitive ativ- fry. The constructivist and information-processing traditions also focus on procedures ane operations for representing anc reasoning about information. Lesming i Undersiood asa con- structive process of concepeval growth, often involving £60" nization of concepts in the leamer’s uaderstanding, aod growth in general cognitive abilities such as problem solving strategies and metacognitive processes, Discussions of motivation often ‘emphasize that much lesing apparcdly occurs without the need for extrinsic sncentves, as in the case of learning one's firs language, ad instead focus on ways to foster the intinsic rest of leamers in ideas and concepts G The_sinative/pragmatsisociohixoric_perpecive_ views sch ete SF SAL knowledge as distributed among people and theic envicon- ments, including the objects, anifacts, tools, books, and the ‘communities of which they are 2 par. Analyses of activity in this perspective focus on processes of iteraction of individuals ‘with ocher people and with physical and technological systems Indeed, the term interactive Bickhard & Richie, 1983) isa close synonym forthe term stuative, Several research traditions have contributed to the situaive perspective, The best esablished ofthese is etbnograpiy, including the study of culeual practices and patterns of social interactions, as well as discourse analysis and conversation analysis in acsiviy theory, sociolinguistic, anthropology, and sociology. Another research tradition is eco logical psychology, which studies behaviors as physical interac tions in which animals, including people, paricipate in physical and technological systems (eg, Turvey, 1990, 1992). A third search tradition is situation tbeor, in logic and the philoso- EE § phty of mind and language, which analyzes meaning and action 4s relational systems and is developing a reformulation of logic tosupportthese relational analyses (eg, Barwise & Perry, 1983; Devlin, 1991). Knowing, in this perspective, is both an atibute ‘of groups that carry out cooperative activities and an attribute ‘of individuals who participate ia the communities of which they are members. A group or individual with knowledge is aituned to the requlariies of acuvities, which include the con- SMraints and affordances of social practices and of the mates! and technological systems of environments. Learning by 2 s0up of individual involves becoming attuned to constraints and affordances of mateval and socal systems with which they injeract, Discussions of motivation in this perspective often emphasize engagement of individuals with the functions and goals of the community, including interpersonal commitments and ways in which individuals’ identities are enhanced or dimin- ‘shed by their paicipaton, Views of Knowing ‘The main reason for schooling is that students should in- ‘crease in what chey know. But what is knowing? A major our ‘come of esearch in educational psychology isthe development of theories, grounded in empirical evidence, thathelp us under- stand what knowing is, as well as how it develops in seudents learning activities Different beliefs about the nature of knowing, undetie diferent priorities, values technologies, and practices in educational activity Knowing as Having Assoctattons: The Bebavtorist/Empiricst View. 4 serong tradition in psychology seeks to characterize knowing as having an organized collection of connections ‘among elementary mental or behavioral units. These units may be elementary sensory impressions that combine to form per cepts and concepts, of stimulus-response associations, or ab- siract elements of parallel, distbuted networks, This empiricist view emphasizes that what someone knows is often a reflection of that person's experience, and indeed, that coming know something requires an experience in which that knowledge ‘can be acquired, irl Rey sion A thoroughly de- ‘eloped version ofthe behaviors view was accomplished be- sinning inthe 19305, Key figures in this development were COGNITION AND LEARNING © 17 ‘Tolman (1982), Guthrie (1935), Skinner (1938), and Hull (1943), and the theoretical issues continue to he developed in current research (eg, Rescorla & Wagnes, 1972). All of these theories are framed by the assumption that behavior is be understood 4s the responses of an organism to stim io the sation, and {they make varying assumptions about the processes by sibich stimulus-response associations are strengthened and weakened Inthe events of an organism's acuity and experience. Although ‘most ofthe systematic theoretical development is based on the results of experiments on learning by animals, especially eats ‘and pigeons, the theoretical ideas of stimulus response associa- tions were also developed in analyses of human learning, espe- ially those involving rote memorization (eg. Estes, 1959; Underwood & Schulz, 1960). A major influence of stimulus response theory in education has been its support of a view ‘of knowledge as an assembly of speciic responses, a form of Knowledge often expressed as detailed behavioral objectives in curricula and assessment An important general technique of task analysis has been built on the assumption of associative knowing, Associationist theoris of learning called for analysis of school subjects into collections of stimulus-response connections (e.g, Thoendike, 1931). Under the influence of behaviodists such as Skinner (958), further proposal that he collections of spectficassocia- tions be expressed as behavioral objectives was added, and Gagné (1965) developed an elaborate system of carrying out aalyses of school tsks into discriminations, classifications, and response sequences. This approach has han enormous influence on the design of curricula, where learning tasks are suranged in sequences based on their relative complexity ac- ‘cording toa task analysis, wit simpler components treated as prerequisites for more complex tasks in which the analysis indicates that the prerequisites are included as components (eg, Gagné, 1968). Parallel Distributed Connectionism, The parallo-dicr- uied nehworkor neural network appreach characterizes know ing in terms of patemns of activation of units that excite or Inhibit each other (ef Rumelhar, McClelland, & PDP Research Group, 1986). These networks differ from networks of associa- ‘ons in traditional behavior theory, which have units of stimuli and responses. They also differ from the structures and proce- dures of cognitive theory, which have units that receive and transmit symbols, In paralle!-disuibuted connectionism, cog Live tates are represented as patterns of activation in a network Of elementary units. Each unit has only a level of activation and connections with other units that wanstnit ether excitation fr inhibition. In recognizing a pattern in the situation, the net ‘work setles into a characters pater of active and inactive nodes that is relatively stable, and tha is different from the ‘activation pattem into which it setles under different stimulus ‘conditions. In acting in the situation, a pattern of acuvation| ‘occurs that results ina specific patter of movement. Different ‘patterns that can be perceived, and diferent actions that can be performed, correspond to diferent patems of activation {volving the same wits, rather than to cifferent units, Although connectionist theories have no: yet been applied “exensively to educational questions, the approach is potentially very significant. t suggests an analysis of knowledge in terms ‘of attunement to regulates in the pattems of environmental 18 © cognition AND MomvaTION events and activities rather than in terms of components, as in behaviorist task analyses Knowing as Concepts and Cognitive abilities: The Cognttive/ Rationalist View. A second view teats knowing as bving structures of information and processes that recognize an con- suuct patterns of symbols in order to understand concepts and ‘oexhibit general abilities, such s reasoning, solving problems, and using and understanding language. This approach provides a bass for analyzing concepts and procedures of subject matter ‘curicla in terms information structures that have been spec= fied in considerable detail This has provided much stronger contact berween cognitively oriented educational psychologists and educators cancemed with the curricula and teaching of ‘subject matter domains than there was with behaviorsuc educa tional psychologists, Conceptual growth and the growth of reasoning have long been active research topics in developmental psychology, and these studies have provided characterizations of genera abil ties and understandings that change as children grow older. Tnformation-processing theories have also provided ways © look at general cognitive abilities 2s general srategies for han- ding information and as metacognitive processes. General Schemsts for Undewsanding and Reasoning. Pia sets extensive bod of Work on children's cogaive develop ‘ent was constructed over several decades, but became infki- ental in American educational psychology inthe 19608. His textly work (ea, 1927/1972, 1929, 1932) had focused on the specific knowledge stnictires that children develop— Itnowledge about physical and social causality, about the or {808 of rues, avs, and moral obligation, about how machines ‘work, Beginning inthe 1940s, however, Piaget began ta form Jace a theory of the development of logical structures and, alhough be actively rejected notions of bstoglcal determinism ‘n hurman development, he argued thatthe capaci to compre hend cenain concepts was limited by the childs level of general logicedeductive development. Piaget's influence on edi: sional practice has been considerable, especially in reinforcing olahanueatuneo ouunse seas eniabs wa ta involves students’ discovery of principles and concepis. csanding. Research on children's undes oncepts continues bea significant copie in developmental psychology. Recent research has focused on the growth of chdres understanding ir domains such a concepts of mnie , Gehman & Galle, 1978 1B. Res nick, 198, biologcl coneepts about living and noniing things (eg, Carey, 1985 Hatano & fagaki, 1967; Kel, 999), and psychologcal concep about mental functioning Ce Flavel, crovn,& Flavel, 1986, Wellman, 1890) This research is developing accounts ofthe rch intitve concepal under. Standing st chien have, and that undergoes signet change 28 they grow oer. The research empluszes that eens laring mus be viewed a ansforming significa un- derstanding that they already have, rater than a simple aequ sions wen on blank lates The resus svages that chien’ Understanding in the domain of concepts ofa subject matter provides amore important guide forthe organization of carc- Sa and teaching than does the stage they have reached in developing their general operational abiives in reasoning ‘There s considerable evidence that a chiklen gov, they are able to handle more complicated tasks (Case, 1985), but we doubt that educational pracuce needsto be guided very swongly by ideas about the development of general schemata of ogio- educive operations in children's reasoning. ‘Alber line of research bas examined conceptual under standiog where people display conceptial misunderstandings that deviate from accepted scenic concepts. These ateratve understandings have been characterized by some as "miscon- ceptions’ (e.g, McCloskey, 1983), and educators have been oncemes to find ways to combat them. Mote recent analyses have characterized the results in terns of stadens use of nr sve conceptions that deed 1 be further refined to apply cor realy inthe situations tht evoke misconceptions (eg, Chi Slots, & de Leeuw, 19% JP. Smith, dSeasa, & Roschelle 3993/1994, s00 alo chapter 15, this Yolume). This view suggests that innative understanding provides the basis for new under. ssandings that develop and should be weated as an esscria resource in students leaming eae Came al chenge teach, Reading and Weng. A major achievement ofthe informa~ sion processing approach to cognition has been the analysis oF language abies such as reading and wetting, Readiog has been analyzed asa combination of abilities to encode informa ‘on from tea into mental representations of eters and words, to recognize the words and activate representations of thet reanings, to combine representations of word ito the pat {ems of phrases and sentences anc to form representations of propositions dha they express, nd to combine represenations of propositions into coherent representations of information ‘coeweyed by texts (a, Just & Carpenter, 1980; Kinisch & van Dik, 197% AM. Lesgoid & Psfet, 1978). The pporance in ‘these models of recognizing and representing relations among the components of text las led to revised measures of txt readability (Miller & Kiotsch, 1980) and methods of systemati- cally mpeoving tes So they ae exsier ‘understand Briten & Gale, 951, Clin & Cale, 19%, Kitsch, 1998), Seek precy Problem Solving and Reasoniog. The cognitive theory de valoped in the 1970s and 1980s inluded information pro cessing models of problem solving and reasoning, Using con expis and programming meltads from the Gheory of txt comprehension (eg, Kintsch & van Dik, 1978) and problem solving (Newell Simon, 1972), sever] analyses of understand tng and solving text problems, especially in mathematics and science, have heen developed (See Grecno & Simoa, 1989; ‘anit, 198, for reviews). The most popular programming {oriathas been the production system, where cach componeat ‘of knowledge is represented asa condition action pairia which the condiion is» pater of symbols and dhe ation i another pattem of symbols that i constnicted by the program ifthe pattem in the condition is matched inthe situation. These mod: is include simulations of text comprehension that constevct, tepresentations ofthe given information of the problem using schemata for general paras. Based on the question ofthe problem, x model simulates seuing a goal to find that answe, find applies operators that ransforminormation in the problem representation, seting subgoalsif necessary, to construct solu tion ofthe problem. 17 Johnson-lLaid, 1983). In reasoning wit Researchers have also investigated reasoning and under- sanding that depends on mental representations, called montal models that provide a kind of simulation of events rather than, “Gescriptions of eveats (Gentner & Stevens, 1985, Halford, 199, descriptive represen tation, a8 Newell & Simon (1972) hypothesized, an operation Js applied to an expression, such as an equation or propos \on, that descnbes a situation. The operation produces 2 new ‘expression describing the siuation, In reasoning wath a mental simulation, a model represents properties of the system, and ‘operating on the model changes some of those properties in ‘ways that correspond to changes in properties of the system, GenenlandSpecicSeuedesandCompcencics Theids of general pobem-sotig heures ao payed a or owt lei the engi vew of knowing st ering Newel (980) node ter at math an rng metbodses bes forthe tncton beeen gen sll Ito specie doa y song Newel eam a person wha et el of leva weltorgmied ome Soni! be able ose anew problem ent pa by reegnlng friar pater new aay asin the need otto, ep ep anaes Ba song Teor requ dinapec ine, nd everyone helo encounter problems for which hey do ot ve the pops domainapee Knowledge In those case, ey tt eon move genera bt wake ioe mens ess cable) gene hess ‘Aspeliethcrtcal version of weak problem ecg th sway expres inthe Gener Problem Ser GP) dee peda contin o ba al lige Cem & Newel 1909) and cgnite pochology Geel & Sinn, 1972. he prblemsoling method Cranes aan) Propane in GPS a peer ert ocd hat hit Gov problem, gees Prteee sob y oh RET SL General competences for tinking have been sie and dase ets In developmental pychology fe ur caller dicsion of Pag, nthe deveoprent ef aca fr develope of aking ln pa, ad io he py chology of individual ferences Gee caper 8 and 9) To Arenal psychology, here a lgatnding ae ver eter there a Signe ctor of ene lige eg Shean 1900, & wher llerenes song dl const of pe competences i domaine uh ea Spat char nd tee ool abo Tony 1938. The later ew hasbeen deol et rsh tnd dicssion by HOt O58, any wes concer with wing fs specie sje have alo eps heed for se ado ger pees of thnkig and problen saving at are prosacive Ice domains A wel own eae mates the workotPaa Gq, 198) who chanted Neste ce for sling ical problens In way ha cn ed oven: of undesanding Schocnfls (385) hs oon ths in of hinking with ssematie search on ata rete sting Recognition of the power ¢ knowe fic ‘methods in problem solving was part of what has heen called. the knowledge revolution within cognitive science (Feigen- COGNITION AND LEARNING © 19 baum, 1989), Cognitive research began to focus heavily on mapping th ‘of the knowledge that suppods strong. oblem solving and reasoning. In educational psychology. sudy after study showed that stpdents ability © understand ‘exis, to solve mathematical problems, orto lear new concepts 2 inthe social or natural sciences depended heavily on what the suudenis already knew (Glaser, 1984). People need org schemata in order to understand and use new information. The richer and more appropnate to the problem these schemata are, the faster and more effectively will people be able to solve the problema, We discuss research conceming general strategic aspects of knowing and the contents of subject matter domains farher in the thi section of ths chapter, Isurs oF Pracricat Conceervauzanion, p. 26 Metacogniive Processes. Another important theme in the cognitive view of knowing i the concept of metavogntion, the capacity to reflect upon one's own thinking, and thereby {© monitor and manage t. These strategies have been studied under many labels, al pointing to the importance of sel-con: scious management of one's own leaning and thinking pro- ‘This heme was introduced by developmental psychologists (ex, A.L. Brown, 1975, Flavell & Wellean, 1977), who noted that a reflective, seifmonioring capacy discminated devel ‘opmentaly advanced chikiren from ther less advanced peers. For example, research with children who have special difeulty in reading has showin that they differ panticulaay froma moce able readers in being less likely to monitor their comprehension and actively generate expectations about the information in the passage (A. L Brown & Campione, 1981). Research comparing excellent adult leamers with less capa ble ones also confirmed tha the most sucessful leamers cabo rate what they read and construct explanations fr themselves Chi, Bassok, Lewis, Riemann, and Glaser (1989) provided a particularly clear demonstration in study of physics students Jeaming from worked-out example problems. Students were

    You might also like