Download Analysis of Situative Learning: Activity Systems & Participation Structures and more Exams Reasoning in PDF only on Docsity!
C H A P T E R 6
Learning in Activity
James G. Greeno
This chapter discusses a program of research in the learning sciences that I call “situa- tive.” The defining characteristic of a sit- uative approach is that instead of focus- ing on individual learners, the main focus of analysis is on activity systems : com- plex social organizations containing learn- ers, teachers, curriculum materials, software tools, and the physical environment. Over the decades, many psychologists have advo- cated a study of these larger systems (Dewey, 1896 , 1929 /195 8; Lewin, 193 5 , 1946 / 1997 ; Mead, 193 4; Vygotsky, 1987 ), although they remained outside the mainstream of psy- chology, which instead focused on individ- uals. Situative analyses include hypotheses about principles of coordination that sup- port communication and reasoning in activ- ity systems, including construction of mean- ing and understanding. Other terms for the perspective I refer to as situative include sociocultural psychology (Cole, 1996 ; Rogoff, 1995 ), activity theory (Engestr ¨om, 1993 ; 1999 ), distributed cogni- tion (Hutchins, 1995 a), and ecological psy- chology (Gibson, 1979 ; Reed, 1996 ). I use the term “situative” because I was intro-
duced to the perspective by scholars who referred to their perspective as situated action (Suchman, 1985 ), situated cognition (Lave, 1988 ), or situated learning (Lave & Wenger, 1991 ). I prefer the term “situative,” a modifier of “perspective,” “analysis,” or “the- ory,” to “situated,” used to modify “action,” “cognition,” or “learning,” because the lat- ter adjective invites a misconception: that some instances of action, cognition, or learn- ing are situated and others are not. During the 1980 s and 1990 s these scholars and others provided analyses in which concepts of cog- nition and learning are relocated at the level of activity systems. For example, Hutchins ( 1995 b) studied remembering in the activity of flying commercial airplanes and gave an analysis of remembering to change the set- tings of flaps and slats during a descent as an accomplishment of the activity system of the cockpit, including the two pilots along with instruments and other informational resources. Goodwin ( 1996 ) studied percep- tion and comprehension in the activity of managing ground operations at an airport and gave an analysis of perceiving and com- prehending conditions at flight gates as an
79
80 the cambridge handbook of the learning sciences
accomplishment of the activity system of the ops room, including several human par- ticipants along with telemonitors that pro- vided images of planes at their gates, with interpretation of the images on telemonitors organized by the relevance of information to their practice. Studies of reasoning and problem solv- ing have been especially influential. Lave, Murtaugh, and della Rosa ( 1984 ) analyzed reasoning by grocery shoppers as a process in which their decisions were shaped jointly by their initial goals and preferences along with the objects and symbols in the aisles of the supermarket. Scribner ( 1984 ) ana- lyzed problem solving by workers in a dairy warehouse as a process in which their per- formance of placing requested numbers of items in containers for delivery was jointly determined by the workers’ reading of forms showing the numbers needed and the visible numbers of items and open spaces in con- tainers in the situation. The decisions and solutions produced in the shopping and the dairy product-loading systems were gener- ally optimal; that is, the the shoppers gen- erally chose products that had the best unit price, and the dairy product loaders gener- ally filled orders by moving the minimum number of items. When mathematical prob- lems equivalent to those solved by peo- ple in everyday activity are given to them in school-like tests, they generally perform poorly, which was documented particularly by Nunes, Schliemann, and Carraher ( 1993 ). The strong conclusion is that it is virtu- ally meaningless to ask whether someone has learned a particular topic of mathemat- ics, such as numerical multiplication, with- out taking into account the kind of activity system in which the person’s “knowledge” is to be evaluated. Learning that occurs in one kind of activity system can influ- ence what one does in a different kind of system, but explanations in terms of over- lapping aspects of activities in practice are much more promising than explanations in terms of the transfer of knowledge struc- tures that individuals have acquired (e.g., Beach, 1995 ; Greeno, Smith, & Moore, 1993 ; Saxe, 1990 ).
From the situative perspective, all socially organized activities provide opportunities for learning to occur, including learning that is different from what a teacher or designer might wish. We study learning when we choose to focus our observations and anal- yses on changes over time and experience in people’s activities. The study of learning in activity requires us to develop concepts and principles that can explain how and why activities in a setting result in changes in what people can do. Use of the situative per- spective in designing learning environments focuses on characteristics of activity systems that can result in learners increasing their capabilities for participation in ways that are valued. The situative perspective builds on and synthesizes two large research programs in the study of human behavior, both of which emerged as alternatives to behaviorism in the 1960 s and 1970 s. The first is cognitive science ; this research focuses on patterns of informa- tion that are hypothesized to be recognized or constructed in activity. Generally, this research focuses on individuals, although social interactions can be (and increasingly are) considered as contexts of individual cognition and learning. The second is inter- actional studies ; this research focuses on patterns of coordination in groups of indi- viduals engaged in joint action with material and informational systems in their environ- ments. Each of these two research programs has developed a considerable body of empir- ical findings, theoretical concepts, and meth- ods. Each of them has succeeded in devel- oping concepts and principles that explain significant aspects of learning, and each has played a key role in the formation of the con- temporary learning sciences. Although these two lines of research both provide important scientific knowl- edge about learning, until very recently they developed mainly in isolation from one another. Research in the individual cognitive perspective has analyzed information struc- tures but has had little to say about the inter- actions that people have with each other and with technological resources in practice. Research in the interactional approach has
82 the cambridge handbook of the learning sciences
discourse analysis, symbolic interactionism, and sociocultural psychology (compare to “interaction analysis” in Jordan & Hen- derson, 1995 ; Sawyer, this volume). This research focuses on how people talk to each other as they plan, evaluate, and coordinate their interactions with the material and tech- nological systems in their environment. The goal is to identify patterns of interaction in which the several components (human and nonhuman) of systems coordinate their behaviors as they participate in their joint activity. Such patterns have been called participation structures or participant struc- tures (Phillips, 1972 ).^1 A participation struc- ture describes the distribution of the func- tional aspects of activity, including agency, authority, accountability, leading and follow- ing, initiating, attending, accepting, ques- tioning or challenging, and so on. Partici- pation structures that are characteristic of a community or group are aspects of the com- munity’s or group’s practices , and learning to become more effective in one’s participa- tion corresponds to achieving fuller partic- ipation in a community’s practices (Lave & Wenger, 1991 ). The interactional approach focuses its study on the whole activity system, and it leads to conclusions about the princi- ples of coordination of interactive systems. This means that the researcher has to ana- lyze the whole activity system without yet having complete understanding of the indi- vidual components – particularly the indi- vidual human participants in the system. The tension between the individual cogni- tive and the interactional approaches thus represents a general difficulty facing scien- tists who study complex systems: whether to proceed by reduction to study of the com- ponents, or by holistic study of the entire system (Sawyer, 2005 , Simon, 1969 ). Interactional studies have identified im- portant patterns of conversational inter- action – patterns of turn taking, open- ing and closing of topics, and mechanisms of repair in response to apparent misun- derstanding have been reported and dis- cussed (Sawyer, this volume; Levinson, 1983 ). Patterns of differential participation
by different individuals can be analyzed; for example, in some classroom settings, stu- dents’ contributions almost always respond directly to a question by the teacher (e.g., Bellak et al., 1966 ; Cazden, 1986 ; Mehan, 1979 ), and in others, discourse is arranged so that students also respond to each other’s presentations and ideas (e.g., O’Connor & Michaels, 1996 ; Phillips, 1972 ). An important contribution of interac- tional studies relates to an intuition by many educators that goals for student learning can be informed by aspects of the prac- tices of professional scholars. For exam- ple, science education can include goals for students to be able to engage in reason- ing, problem solving, and argumentation in ways that reflect practices that scientists have developed – for example, distinguish- ing hypotheses from evidence (e.g., Kuhn, 1989 ), inferring specific implications of gen- eral principles, and recognizing relations between specific problems and general prin- ciples (e.g., Chi, Feltovich, & Glaser, 1981 ). In pursuing this approach to science edu- cation, learning scientists can use results of studies of scientific practice – how scientists do their work and what knowledge and prac- tices are involved in conducting that work. Influential studies of scientific practice were published in the mid- 1980 s (Latour & Wool- gar, 1986 ; Lynch, 1985 ). These early interac- tional studies did not examine information structures in the scientific subject-matter disciplines; in fact, Latour and Woolgar ( 1986 ) famously claimed that all of science could be explained in terms of sociocul- tural factors, with no appeal to cognition. More recent studies have included careful and detailed analyses of the conceptual and empirical contents of scientific practice and development (e.g., Fujimura, 1996 ; Kitcher, 1993 ; Nersessian, 1984 ; 2002 ).
Including Interaction in Cognitive
Analyses
One strategy for unifying individual cogni- tive and interactional concepts and methods
learning in activity 83
is to work from the cognitive side and extend its reach to include situations involving inter- action between more than one person. If an activity system can be decomposed into indi- viduals and their tools, then we can analyze the activity system by reducing it to a study of the individuals and the tools, and then aggregating these explanations back together to form an explanation of the entire activ- ity system. A strategic assumption of the individual cognitive approach is that groups can be explained by reduction to individual study in this way (Sawyer, 2005 ). However, to study individual learners, researchers cre- ate a new kind of activity structure – a labo- ratory experiment – and because we do not yet know how the properties of individuals depend on the social context, we have to make a factoring assumption : that the princi- ples that characterize behavior of the indi- vidual research subject do not depend sig- nificantly on the rest of the activity system. Without analyzing the activity system as a whole, we risk arriving at conclusions that we think are about the individual, but in fact depend on broader features of the activity system, and thus would not readily general- ize from the laboratory to real-world learn- ing environments. Some research by individual cognitive- science researchers has provided promis- ing findings for the program of extend- ing cognitive principles from individual to group activity. For example, Schwartz ( 1995 ), studying performance of middle- and high-school students on tasks involv- ing understanding mechanical or biological systems, found that pairs of students work- ing together included useful abstractions in their conversations more often than was the case for thinking-aloud protocols of indi- vidual students. Okada and Simon ( 1997 ), studying performance of college students in a simulation of scientific problem solving, also found that pairs of participants out- performed individual participants. The pairs had greater frequencies of generating pro- ductive hypotheses to test in the simulation. Dunbar ( 1995 ), studying the conversations of biology laboratory groups, found that the participants made productive use of analo-
gies in their joint reasoning. These results indicate that some processes known to be important in reasoning and problem solving by individuals – attending to general features of problem situations, generating hypothe- ses, constructing analogies – are also signif- icant in reasoning by groups. That some of these processes occur more frequently and, perhaps, more productively in group than in individual performance, could be explained as an effect of the presence of other people as a favorable aspect of the social context. Other findings, however, indicate that analyses of activity by groups may involve significant processes that are less evident in individual activity. Barron ( 2003 ), studying mathematical problem solving by sixth- grade students, concluded that their man- agement of joint attention was an impor- tant factor in their success. Sawyer ( 2003 ), studying performance by groups performing jazz music and improvisational theater, con- cluded “that both verbal and musical per- formance collectively emerge from interac- tional processes,” and that “the analysis of group creativity requires a fundamentally interactional semiotics, one which empha- sizes the indexical properties of sign usage” ( 2003 , p. 95 ). That semiotic interpretation in these improvisational activities is funda- mentally indexical has the consequence that understanding meanings and, therefore, the course of a performance requires analysis of the interactional system in a way that goes significantly beyond that of scripted perfor- mances and problems that have stable prob- lem spaces.
A Situative Approach: Including
Information Structures in
Interactional Analyses
Studies in what I call the “situative perspec- tive” use another strategy aimed to bring concepts and methods of cognitive and inter- actional studies together. In a situative study the main focus of analysis is on performance and learning by an activity system: a collec- tion of people and other systems.
learning in activity 85
about temperatures and the amount of insulation and its quality. Their design has to include spaces for work, sleeping, and recreation. I discuss an episode that involves a spe- cial assignment that the teacher gave: find the value of insulation quality, called the “R value,” that would minimize the total cost of construction and heating over two years. To solve the task, students kept their designs and temperature assumptions con- stant, and used the program to calculate the total cost of construction and the monthly heating cost for different R values. The stu- dent’s group had constructed a table of val- ues with each row showing the construction cost and the two-year heating cost for one R value. In their analysis, they focused on pairs of successive rows in the table, noting how much the construction cost increased and the monthly heating cost decreased from the lower to the higher R value. The R value that they selected as the one that would minimize the total two-year cost was 20 because between R = 10 and R = 20 , the increase in construction cost was less than the decrease in heating cost, but between 20 and 3 0, the increase in construction cost was greater than the decrease in heating cost. The teacher had expected a different form of analysis, in which the total costs (construc- tion plus two years heating) would be calcu- lated for each R value, then allowing a quick identification of the R value that minimized total cost. In the conversation, the teacher and student successfully constructed a shared understanding of the group’s analysis and why it was correct even though it was not what the teacher was expecting. A situative analysis of this episode has two components: an interaction analysis of the conversation, including close attention to its turn tak- ing, responses, and contributions (Sawyer, this volume); and the semiotic structures of information that they constructed in the conversation, which include structures of information represented in the students’ table. This latter analysis identified refer- ences of symbols to different versions of
the design, each of which had the numer- ical properties of R value, construction cost, and heating cost represented in a row in the students’ table. The meanings of these sym- bols were constructed as information struc- tures that the teacher and student generated jointly as they achieved a mutual under- standing of the group’s reasoning in their conversation. The teacher and student were attuned to several interactional and semiotic practices as well as constraints in the task domain: turn taking conventions, including the expecta- tion that the student would be given a chance to explain the group’s reasoning; conventions of constructing and interpreting symbolic representations in numeric tables; regularities in the domain of building design, including the importance of cost; and arith- metic operations, which they used to com- pare R values. However, at least initially, the teacher was not attuned to the method that the group created to identify the optimal R value, although she followed the student’s explanation and became attuned to this reasoning.
Data Are Records of Interaction, Rather Than “Verbal Reports” One way the situative approach differs from individual cognitive research is in the kind of data that are typically used to infer proper- ties of information structures. In individual cognitive research on problem solving, evi- dence about information structures is often in the form of thinking-aloud protocols pro- vided by individual subjects. These proto- cols are then interpreted as providing evi- dence about the nature of the problem space as represented in the individual’s mind, and the processes that the individual used to work in that problem space. A situative approach, in contrast, begins by noting that problem solving often occurs in group set- tings. When engaged in joint problem solv- ing, participants talk, gesture, and create vis- ible representations for each other as they interact. Using methods of interaction anal- ysis, researchers transcribe the participants’
86 the cambridge handbook of the learning sciences
activity, and the transcript provides a group- level analog of the thinking-aloud proto- cols that are analyzed in studies of prob- lem solving: collaborative discourse is group thinking made visible. The evidence that par- ticipants provide each other through their collaborative discourse informs them about their understandings, goals, intentions, and expectations, and it provides evidence to the researcher about semiotic structures that are being generated and used.
How Semiotic Structures Are Generated
The learning sciences are fundamentally concerned with identifying how structures of information are generated and used in learning activities, and with ways that infor- mation functions in activity. In an individ- ual cognitive approach, these processes are analyzed at the level of individual mental activities; in the situative approach, they are analyzed at the level of activity systems. If there is more than one person in the system, their conversation is joint action that con- structs shared information (Clark & Schae- fer, 1989 ). Clark and Wilkes-Gibbs ( 1986 ) have shown that reference can be under- stood as an achievement of joint action, rather than being a property of a symbol itself; the meanings of symbols are often interpreted in relation to problems that emerge in ongoing activity (e.g., Goodwin 1995 ). Even the referential meaning of a sin- gle word is a collaborative achievement that results from representational practice (Clark & Wilkes-Gibbs, 1986 ). Researchers in conversation analysis (e.g., Schegloff, 1991 ) and psycholinguistics (e.g., Clark, 1996 ) have analyzed ways in which participants in a conversation mutually con- struct meanings. The conventions whereby symbols are interpreted differ in different cultures. These interpretations are integral components of the ongoing activities that people are engaged in as they participate in activity systems. The situative perspec- tive considers meaning to be a relation between these joint actions of achieving mutual understanding, and the states of affairs or ideas that the participants them-
selves interpret their statements to be refer- ring to. Material and other informational resources also contribute to the construc- tion of information, in ways investigated in research on distributed cognition (e.g., Hutchins, 1995 a) and in social studies of sci- ence (e.g., Pickering, 1995 ). When researchers shift the analysis of knowledge construction to the level of the activity system, they include explanations about the various participants in the activ- ity, and they analyze ways that individuals are positioned in the participant structures of interaction and how that positioning con- tributes to generation of information struc- tures.
From Representation to Representational Practice In the individual cognitive approach, rep- resentations are thought to be structures of information that connect concepts with each other in a network of propositions. These networks and concepts are mental objects that are stored in people’s memories. In con- trast, situativity treats representation as a relation between signs and aspects of sit- uations, resulting from interpretations by people in their activity. The focus shifts to include both representations and representa- tional practices. The emphasis on representation as both mental and socially distributed in prac- tices is a synthesis of cognitive and inter- actional perspectives. The individual cogni- tive perspective emphasizes representations of information, and the interactional per- spective emphasizes representational prac- tices as distributed across groups of people and across material objects and systems in the environment. The representations and representational practices analyzed in the situative approach extend the scope of the typical context- free semantic-network representations of cognitive science to include indexical rela- tions between a symbol and the context of its use. In a situative analysis, mental states can be considered to be represen- tations, but there should be evidence that
88 the cambridge handbook of the learning sciences
contributions to their activity. A general schema of interaction in tasks is shown in Figure 6. 1 , adapted from Clark and Schaefer’s ( 1989 ) schema for contributions to discourse. Working on a task involves per- forming actions that contribute to achieving the task goals. The results of action can just be an addition of information to the com- mon ground – information about properties of the task or situation, or about evaluations, intentions, or goals. Or the results of action can also be a change in the material situa- tion of the task, by moving or constructing an object or by writing or drawing some kind of representation. The schema sketched in Figure 6. 1 pro- vides a way of thinking about and represent- ing participatory aspects of interaction at a turn-by-turn level. In addition, general pat- terns in the ways contributions are made in a group’s or community’s practices can be identified. For example, classroom practices differ in the extent to which they encour- age problematizing of ideas and issues (e.g., Engle & Conant, 2002 ) and in the ways that differences are resolved (Ball & Bass, 2000 ; Greeno, 2003 ). In Figure 6. 1 , these correspond to frequent occurrence of the NEGOTIATE nodes, with discussion about alternative ideas, actions, and approaches to understanding and working on tasks. In rela- tion to Figure 6. 1 , problematizing and recon- ciling correspond to encouraging responses to students’ contributions that consider alternative ideas, actions, and approaches.
Participant Structures
Practices vary in the ways that agency is distributed between the participants. In interaction, different individuals are posi- tioned differently regarding the compe- tence, authority, and accountability that are attributed to them by others and by themselves. These differences in position- ing mean that individuals are differentially entitled and expected to initiate propos- als for action or interpretation, to ques- tion or challenge other participants’ pro- posals, and to indicate that an issue has been settled (as in the Initiation-Response- Evaluation sequence discussed in Sawyer,
this volume, and in other patterns of class- room discourse interaction, such as revoic- ing, O’Connor & Michaels, 1996 ). Positioning in relation to other partic- ipants involves entitlements and expecta- tions of the individual for initiating topics or questions, making assertions and propos- als for actions, questioning or challenging others’ assertions and proposals, and so on. Positioning in relation to the subject-matter domain has been characterized by Picker- ing ( 1995 ) as a dance of agency, involv- ing material agency, disciplinary agency, and conceptual agency. Conceptual agency is involved when an individual or group inter- acts with the subject-matter constructively – interpreting meanings, formulating ques- tions, choosing and adapting a method, designing an apparatus, and so on. Material agency is involved when a system (such as an experimental apparatus) determines the outcome of an action. Disciplinary agency is involved when established methods such as algorithms or proof procedures determine the outcome of an action. School activities often position students with little concep- tual agency, teaching them instead how to perform algorithms correctly (disciplinary agency) or to set up apparatus to obtain known empirical results (material agency). The emphasis by constructivist educators on knowledge construction in authentic prac- tice is designed to grant students some con- ceptual agency. Alignment among participants depends on how differences in ways that individ- uals participate are understood and incor- porated into practice. Situative analyses include study of the participant structures of episodes of activity, particularly ways that individuals are positioned to take initiative or not, to question or challenge others’ pro- posals and assertions or not, to engage in the group activity attentively or not, and so on. In an individual’s participation in a group or community over time, he or she gener- ally has some ways of interacting that come to be characteristic and expected by her- or himself and others in the group. These char- acteristic patterns, which are coconstructed by the individual and others in the group, constitute that person’s positional identity
learning in activity 89
CONTRIBUTE
(INITIATE): ACTION:
(NEGOTIATE) ACCEPT: (NEGOTIATE) ACCEPT:
Suggest or propose an action
An intention for action is in common ground
The proposed action may be discussed and modified
The action or its results may be evaluated and changed
The action is taken to be finished, for now
The initiating participant or another participant performs an action
Figure 6. 1. A schema of negotiation.
in that activity system. An individual’s posi- tioning can be understood both in regard to the other participants in the group – her or his characteristic positioning in par- ticipant structures – and in regard to the subject-matter of the group’s activities – her or his characteristic positioning in relation to using and generating semiotic structures, for example, with or without conceptual agency. Individuals’ identities in a classroom differ, such that some individuals are more likely than others to engage in working on tasks that are assigned, or to work collabo- ratively with others and try to reach mutual understanding, or to engage in social interac- tion not related to an assigned task (Gresalfi, 2004 ). Of course, an individual’s participa- tion is not always consistent with her or his general tendencies. Situative researchers generally do not assume that our models of these con- ventions, practices, and identities necessar- ily correspond to cognitive representations inside participant’s minds. Of course, people often do construct internal representations of these conventions and practices, and these constructions and interpretations are critical in activity and are an important topic for the learning sciences (e.g., Hall, 1996 ).
Community Practices
Communities and groups have practices that constrain (but do not determine) the subject-matter contents of their dis- course and other activities in their subject- matter domains. These practices include
what counts as knowledge in the group’s domain, including use and interpretation of its terminology, meanings of its concepts and principles, and applications of its meth- ods. Communities of learners share stan- dards about what characterizes worthwhile problems to engage in, and what consti- tutes an adequate or excellent solution of such a problem. Many of these standards are implicit: ways of formulating arguments and explanations, and ways of judging the rele- vance and significance of questions, informa- tion, evidence, and conclusions. In the situative perspective, the institu- tional contexts of activity systems are impor- tant in understanding learning. Attitudes such as having a commitment to succeeding in schoolwork or not (Eckert, 1989 , 1990 ), and having a positive or negative orientation toward activities of learning and knowledge construction, involve issues of affiliation and identity. In schools, students’ affiliations in formally organized groups (band, chess club, gangs) as well as informal networks of friends are crucial in the development of their iden- tities, and these groups sometimes shape stu- dent’s identities in ways that oppose the school’s preferred participation structures, as well as in ways that facilitate student engagement in academic pursuits that are valued by the institution. Students’ motiva- tion to learn depends on whether the learn- ing activity supports the continual devel- opment of their personal identities. When learning environments do not support per- sonal identity, learners will not be deeply engaged, even if they manage to maintain focus long enough to complete a classroom activity (Blumenfeld, Kemplar, & Krajcik, this volume).
Example: A Study of Learning
a Classroom Practice
To illustrate how learning scientists can study the learning of practices using the sit- uative perspective, I describe a case study: Hall and Rubin’s ( 1998 ) analysis of how a representational practice became estab- lished in a classroom.^3 They documented
learning in activity 91
Nersessian et al. ( 2003 ) found that to understand how problems were solved in this laboratory, they had to expand the tra- ditional cognitive science notions of “prob- lem space” and “mental representation” to consider these as being distributed across the people and the technology in the laboratory – a defining feature of situa- tive research. The problem space comprised models and artifacts together with a reper- toire of activities in which simulative model- based reasoning played a key role (cf. Lehrer & Schauble, this volume). The problem- solving processes of the lab were distributed throughout the cognitive system, which comprised both the researchers and the cog- nitive artifacts that they use (cf. Hutchins, 1995 a). Nersessian et al. ( 2003 ) used a mixed- method approach, combining cognitive analyses of the problems and models used by the biomedical researchers with an ethno- graphic analysis of the situative activities and tools and how they are used in the ongoing activity of the laboratory. Their close ethno- graphic analysis allowed them to document temporary and transient arrangements of the activity system – the laboratory routines, the organization of the workspace, the cultural artifacts being used, and the social organi- zation of the team members. Their cogni- tive analysis allowed them to document how people and their relationships changed over time – as they evaluated and revised problem definitions (often working closely with tech- nological artifacts), as they revised models of phenomena, and as their concepts changed over time. Because it is impossible to test artificial blood vessels in a live human body, mod- eling practices were critical to the work. The researchers had to design working mod- els to use for experimentation. Each itera- tion of a model represented the lab’s col- lective understanding of the properties and behaviors of the human body. For exam- ple, the flow loop is a device that emu- lates the shear stresses experienced by cells within the blood vessels. The flow loop orig- inated in the research of the senior scien- tist, and was passed down through genera- tions of researchers, enabling each to build
on the research of others, as it was reengi- neered in the service of model-based rea- soning. The flow loop is constructed so that the test fluid will create the same kinds of mechanical stresses as a real blood vessel. But because the model is a mechanical sys- tem, its design is subject to engineering con- straints, and these often require simplifica- tion and idealization of the target biological systems being modeled. For example, in the body arterial wall motion is a response to the pulsating blood flow, but in the flow loop simulation, known as a bioreactor , the fluid doesn’t actually flow, although it does model the pulsating changes in pressure experi- enced by the arterial wall. In scientific laboratories, collaboration is often mediated by external represen- tations such as these mechanical models, as well as diagrams and sketches. In this lab, devices were external representations of the collective knowledge of the group. Model-based reasoning is a distributed phe- nomenon, involving both the internal men- tal models that a researcher holds, as well as the shared external model manifested in devices and other models. A situative analysis focuses on this dis- tributed nature of cognition in the labo- ratory, treating it as a process involving multiple people and the technological arti- facts that they create and modify together. In the situative view of an activity sys- tem, learning is conceived of as transforma- tions over time in the nature of the inter- actions among people and between people and their constructed artifacts. For example, when newcomers to the lab were first intro- duced to a device like the bioreactor, they assumed that its design was fixed. As they began to interact with these devices, they quickly learned the many problems: tubes leak, sutures don’t keep, reservoirs overflow, pumps malfunction. The newcom- ers soon realized that everyone else, includ- ing the most experienced old-timer, was always struggling to get things to work, always revising and modifying the devices. The newcomer’s learning was a process of coming to understand the contingent and changing nature of these devices – the new- comers built relationships with the devices
92 the cambridge handbook of the learning sciences
that Nersessian et al. ( 2003 ) called cognitive partnerships.
Conclusion
Analyses that use the situative perspective consider learning environments as activity systems in which learners interact with each other and with material, informational, and conceptual resources in their environment. The situative perspective is a synthesis of the two major scientific approaches to under- standing human behavior: cognitive science and interactional studies. It combines the strengths of each of these approaches with the goal of better understanding how learn- ing occurs and how to design learning envi- ronments. Situativity is a general scientific perspec- tive and as such does not say what edu- cational practices should be adopted. Even so, it is well suited for analyzing processes of interaction and learning in the types of learning environments recommended by many progressive educators – a move away from a transmission-and-acquisition style of instruction, toward more collaborative, active, and inquiry-oriented classrooms. By its focus on activity systems, the situa- tive perspective emphasizes that the activi- ties that take place in different learning envi- ronments are important, not only because of differences in how effectively they teach content knowledge but also because partici- pation in practice is a central part of what students learn. If an aim in education is for students to learn practices of inquiry and sense-making, then learning environ- ments must provide opportunities for them to participate in such practices. The situa- tive perspective is reflected in a wide range of learning sciences projects, such as math- ematics classrooms in which students par- ticipate in developing definitions, conjec- tures, representations, and arguments (e.g., Ball & Bass, 2000 ; Boaler, 2002 ; Fawcett, 193 8; Lampert, 2001 ; Moses & Cobb, 2001 ; Schoenfeld, 1994 ; Schwartz, Yarushalmy, & Wilson, 1993 ). In science classrooms, stu- dents develop and evaluate hypotheses and arguments in science (Brown & Campione,
1994 ; Goldman, 1996 ; Hatano & Inagaki, 1991 ; Reiner, Pea, & Shulman, 1995 ) and in social studies (Collins, Hawkins, & Carver, 1991 ; Scardamalia, Bereiter, & Lamon, 1994 ). This kind of practice was advocated by Dewey (e.g., 1910 / 1978 ) and is a major focus of learning sciences research and practice. The activities that contribute to these prac- tices encourage students to participate in processes that include conceptual inquiry and the use of skills in solving meaning- ful problems as part of authentic projects (Krajcik & Blumenfeld, this volume). These learning environments include activities such as formulating and evaluating con- jectures, conclusions, and arguments. In participation-oriented practices, class dis- cussions are organized both to foster stu- dent learning of content and also to sup- port their learning how to participate in the discourse practices that organize such dis- cussions. Students learn about content and also learn how to participate in collaborative inquiry, and how to use the concepts and methods of a discipline to solve authentic problems. They learn representational sys- tems, not only to express information in a domain but also to apply them in represen- tational practices as they develop and share their understandings of questions, hypothe- ses, and arguments in the domain. A chal- lenge for the learning sciences is to advance our theoretical understanding of learning to provide more coherent and definite expla- nations of learning in these environments, as well as more helpful guidance for the design of productive resources and practices. I have tried in this chapter to show that the situa- tive perspective can be a valuable resource in this effort.
Acknowledgments
My research and writing are supported by a grant from the Spencer Foundation. This chapter benefited from energetic editing by Keith Sawyer. The conclusions in the chap- ter are mine, but there is significant material here that would not have been without Keith’s contributions.
94 the cambridge handbook of the learning sciences
Dewey, J. ( 1896 ). The reflex arc concept in psy- chology. Psychological Review, 3 , 3 5 7–3 70.
Dewey, J. ( 1978 ). How we think. In How we think and selected essays, 1910 – 1911 , The middle works of John Dewey, 1899 – 192 4 , volume 6 (Jo Ann Boydston, ed.) (pp. 177 – 3 5 6). Carbondale, IL: Southern Illinois University Press (originally published 1910 ). Dewey, J. (195 8). Experience and nature. New York: Dover (original work published 1929 ).
Dunbar, K. ( 1995 ). How scientists really reason: Scientific reasoning in real-world laboratories. In R. J. Sternberg & J. E. Davidson (Eds.), The nature of insight (pp. 3 65 – 3 95 ). Cambridge, MA: MIT Press/Bradford.
Eckert, P. ( 1989 ). Jocks and burnouts. New York: Teachers College Press. Eckert, P. ( 1990 ). Adolescent social categories: Information and science learning. In M. Gard- ner, J. G. Greeno, F. Reif, A. H. Schoenfeld, A. diSessa, & E. Stage (Eds.), Toward a scien- tific practice of science education (pp. 203 – 218 ). Hillsdale, NJ: Lawrence Erlbaum Associates.
Engestr ¨om, Y. ( 1993 ). Developmental studies of work as a testbench of activity theory: The case of primary care medical practice. In S. Chaiklin & J. Lave (Eds.), Understanding practice: Per- spectives on activity and context (pp. 64 – 103 ). Cambridge: Cambridge University Press.
Engestr ¨om, Y. ( 1999 ). Activity theory and indi- vidual and social transformation. In Y. Engestrt ¨om, R. Miettinen, & R.-L. Punamaki (Eds.), Perspectives on activity theory (pp. 19 – 3 8). Cambridge: Cambridge University Press.
Engestr ¨om, Y. ( 2001 ). Expansive learning at work: Toward an activity theoretical reconceptual- ization. Journal of Education and Work, 14 , 13 3 – 15 6.
Engle, R. A., & Conant, F. R. ( 2002 ). Guiding principles for fostering productive disciplinary engagement: Explaining an emergent argu- ment in a community of learners classroom. Cognition and Instruction. 2 0, 3 99– 483.
Fawcett, H. P. (193 8). The nature of proof: A description and evaluation of certain procedures used in a senior high school to develop an under- standing of the nature of proof, the thirteenth yearbook of the National Council of Teachers of Mathematics. New York: Bureau of Publica- tions, Teachers College, Columbia University. Fujimura, J. H. ( 1996 ). Crafting science: A sociohis- tory of the quest for the genetics of cancer. Cam- bridge, MA: Harvard University Press.
Gibson, J. J. ( 1979 ). An ecological approach to visual perception. Boston: Houghton Mifflin. Goldman, S. V. ( 1996 ). Mediating microworlds: Collaboration on high school science activi- ties. In T. Koschmann (Ed.), CSCL: Theory and practice of an emegting paradigm (pp. 45 – 82 ). Mahwah, NJ: Lawrence Erlbaum Associates. Goldman, S., & Moschkovich, J. ( 1995 ). Environ- ments for collaborating mathematically: The middle-school mathematics through applica- tions project. CSCL ’ 95 Proceedings. Goodwin, C. ( 1995 ). Seeing in depth. Social Stud- ies of Science, 2 5 , 23 7– 274. Goodwin, C. ( 1996 ). Transparent vision. In E. Ochs, E. A. Schegloff & S. A. Thompson (Eds.), Interaction and grammar (pp. 3 70– 404 ). Cambridge: Cambridge University Press. Greeno, J. G. ( 1995 ). Understanding concepts in activity. In C. A. Weaver III, S. Mannes, & C. R. Fletcher (Eds.), Discourse comprehension: Essays in honor of Walter Kintsch (pp. 65 – 96 ). Hillsdale, NJ: Lawrence Erlbaum Associates. Greeno, J. G. ( 2003 , November). A situative per- spective on cognition and learning in interac- tion. Paper presented at a workshop, “Theoriz- ing learning practice,” University of Illinois. Greeno, J. G., & Engle, R. A. ( 1995 ). Combin- ing analyses of cognitive processes, meanings, and social participation: Understanding sym- bolic representation. Proceedings of the Seven- teenth Annual Conference of the Cognitive Sci- ence Society, Pittsburgh. Greeno, J. G., & the Middle-school Mathe- matics through Applications Project Group ( 1998 ). The situativity of knowing, learning, and research. American Psychologist, 5 3 , 5 – 26. Greeno, J. G., Smith, D. R., & Moore, J. L. ( 1993 ). Transfer of situated learning. In D. K. Detterman & R. K. Sternberg (Eds.), Transfer on trial: Intelligence, cognition, and instruction (pp. 99 – 167 ). Hillsdale, NJ: Lawrence Erlbaum Associates. Gresalfi, M. S. ( 2004 ). Taking up opportunities to learn: Examining the construction of mathemat- ical identities in middle school classrooms. Doc- toral dissertation, Stanford University. Hall, R. ( 1996 ). Representation as shared activity: Situated cognition and Dewey’s cartography of experience. Journal of the Learning Sciences, 5 , 209 – 23 8. Hall, R., & Rubin, A. ( 1998 ). There’s five lit- tle notches in here: Dilemmas in teaching and learning the conventional structure of rate.
learning in activity 95
In J. G. Greeno & S. V. Goldman (Eds), Thinking practices in mathematics and science learning (pp. 189 – 23 5 ). Mahwah NJ: Lawrence Erlbaum Associates. Hatano, G., & Inagaki, K. ( 1991 ). Sharing cogni- tion through collective comprehension activity. In L. B. Resnick, J. M. Levine, & S. D. Teasley (Eds.), Perspectives on socially shared cognition (pp. 3 1–3 48). Washington, DC: American Psy- chological Association. Hutchins, E. ( 1993 ). Learning to navigate. In S. Chaiklin & J. Lave (Eds.), Understanding practice: Perspectives on activity and context (pp. 3 5 – 63 ). Cambridge: Cambridge Univer- sity Press. Hutchins, E. ( 1995 a). Cognition in the wild. Cam- bridge, MA: MIT Press. Hutchins, E. ( 1995 b). How a cockpit remembers its speeds. Cognitive Science, 19 , 265 – 288. Jordan, G., & Henderson, A. ( 1995 ). Interaction analysis: Foundations and practice. Journal of the Learning Sciences, 4 , 3 9– 103. Kintsch, W. ( 1998 ). Comprehension: A paradigm for cognition. Cambridge: Cambridge Univer- sity Press. Kitcher, P. ( 1993 ). The advancement of science. Oxford: Oxford University Press. Kuhn, D. ( 1989 ). Children and adults as intuitive scientists. Psychological Review, 96 , 674 – 689. Lampert, M. ( 1990 ). When the problem is not the question and the solution is not the answer: Mathematical knowing and teaching. American Educational Research Journal, 2 7 , 29 – 64. Lampert, M. ( 2001 ). Teaching problems and the problems of teaching. New Haven, CT: Univer- sity Press. Latour, B., & Woolgar, S. ( 1986 ). Laboratory life: The construction of scientific facts. Princeton, NJ: Princeton University Press. Lave, J. ( 1988 ). Cognition in practice: Mind, math- ematics, and culture in everyday life. Cambridge: Cambridge University Press. Lave, J., Murtaugh, M., & de la Rosa, O. ( 1984 ). The dialectic of arithmetic in grocery shopping. In B. Rogoff & J. Lave (Eds.), Everyday cognition: Its development in social context (pp. 67 – 94 ). Cambridge, MA: Harvard University Press. Lave, J., & Wenger, E. ( 1991 ). Situated cognition: Legitimate peripheral participation. Cambridge: Cambridge University Press. Levinson, S. ( 1983 ). Pragmatics. Cambridge: Cambridge University Press.
Lewin, K. (193 5 ). Dynamic theory of personality. New York: Mcgraw-Hill. Lewin, K. ( 1997 ). Behavior and development as a function of the total situation. In Resolving social conflicts & Field theory in social science (pp. 3 3 7–3 81). Washington, DC: American Psychological Association. (Originally pub- lished 1946 ). Lynch, M. ( 1985 ). Art and artifact in laboratory science: A study of shop work and shop talk in a research laboratory. London: Routledge and Kegan Paul. Mead, G. H. (193 4). Mind, self, and society. Chicago: University of Chicago Press. Mehan, H. ( 1979 ). Learning lessons. Cambridge, MA: Harvard University Press. Moses, R. P., & Cobb, C. E., Jr. ( 2001 ). Radical equations: Math literacy and civil rights. Boston: Beacon Press. Nersessian, N. ( 1984 ). Faraday to Einstein: Constructing meaning in scientific theories. Dordrecht: Martinus Nijhoff/Kluwer. Nersessian, N. ( 2002 ). Maxwell and the “method of physical analogy”: Model-based reasoning, generic abstraction, and conceptual change. In D. Malament (Ed.), Reading natural philosophy: Essays in the history and philosophy of sci- ence and mathematics. Lasalle, IL: Open Court. Nersessian, N. J. ( 2005 ). Interpreting scientific and engineering practices: Integrating the cog- nitive, social and cultural dimensions. In M. Gorman, R. Tweney, D. Gooding, & A. Kincan- non (Eds.), Scientific and technological thinking (pp. 17 – 5 6). Mahwah, NJ: Erlbaum. Nersessian, N. J., Kurz-Milcke, E., Newstetter, W. C., & Davies, J. ( 2003 ). Research laborato- ries as evolving distributed cognitive systems. In R. Alterman & d. Kirsh (Eds.), Proceedings of the Twenty-Fifth Annual Conference of the Cog- nitive Science Society (pp. 85 7– 862 ). Erlbaum. Newell, A., & Simon, H. A. ( 1972 ). Human problem solving. Englewood Cliffs, NJ: Prentice Hall. Nunes, T., Schliemann, A. D., & Carraher, D. W. ( 1993 ). Street mathematics and school mathematics. Cambridge: Cambridge Univer- sity Press. O’Connor, M. C., & Michaels, S. ( 1996 ). Shifting participant frameworks: Orchestrating think- ing practices in group discussion. In D. Hicks (Ed.), Discourse, learning, and schooling (pp. 63 – 103 ). Cambridge: Cambridge University Press.
