The New Science of Learning

Cambridge University Press 0521845548 - The Cambridge Handbook of the Learning Sciences Edited by R. Keith Sawyer Excerpt More information

CHAPTER 1 Introduction

R. Keith Sawyer

r By the twentieth century, all major indus- Teachers know these facts and proce- trialized countries offered formal schooling dures, and their job is to transmit them to all of their children. When these schools to students. r took shape in the nineteenth and twenti- Simpler facts and procedures should be eth centuries, scientists didn’t know very learned first, followed by progressively much about how people learn. Even by the more complex facts and procedures. The 1920s, when schools began to become the definitions of “simplicity” and “complex- large bureaucratic institutions that we know ity” and the proper sequencing of mate- today, there still was no sustained study of rial were determined either by teachers, how people learn. As a result, the schools by textbook authors, or by asking expert we have today were designed around com- adults like mathematicians, scientists, or monsense assumptions that had never been historians – not by studying how children tested scientifically: actually learn. r r The way to determine the success of Knowledge is a collection of facts about schooling is to test students to see how the world and procedures for how to solve many of these facts and procedures they problems. Facts are statements like “The have acquired. earth is tilted on its axis by 23.45 degrees” and procedures are step-by-step instruc- This traditional vision of schooling is tions like how to do multidigit addition known as instructionism (Papert, 1993). by carrying to the next column. Instructionism prepared students for the r The goal of schooling is to get these industrialized economy of the early twen- facts and procedures into the student’s tieth century. But the world today is much head. People are considered to be edu- more technologically complex and econom- cated when they possess a large collection ically competitive, and instructionism is of these facts and procedures. increasingly failing to educate our students

2 the cambridge handbook of the learning sciences

to participate in this new kind of soci- procedures does not prepare a person to ety. Economists and organizational theorists perform as a knowledge worker. Factual have reached a consensus that today we are and procedural knowledge is only useful living in a knowledge economy, an economy when a person knows which situations that is built on knowledge work (Bereiter, to apply it in, and exactly how to mod- 2002; Drucker, 1993). In the knowledge ify it for each new situation. Instruction- economy, memorization of facts and pro- ism results in a kind of learning which is cedures is not enough for success. Edu- very difficult to use outside of the class- cated graduates need a deep conceptual room. When students gain a deeper con- understanding of complex concepts, and the ceptual understanding, they learn facts ability to work with them creatively to gen- and procedures in a much more useful erate new ideas, new theories, new prod- and profound way that transfers to real- ucts, and new knowledge. They need to be world settings. r able to critically evaluate what they read, Focusing on learning in addition to teach- to be able to express themselves clearly ing. Students cannot learn deeper concep- both verbally and in writing, and to be tual understanding simply from teach- able to understand scientific and mathemat- ers instructing them better. Students can ical thinking. They need to learn integrated only learn this by actively participating in and usable knowledge, rather than the sets their own learning. The new science of of compartmentalized and decontextual- learning focuses on student learning pro- ized facts emphasized by instructionism. cesses, as well as instructional technique. r They need to be able to take responsibility Creating learning environments. The job of for their own continuing, lifelong learning. schools is to help students learn the full These abilities are important to the econ- range of knowledge required for expert omy, to the continued success of participa- adult performance: facts and procedures, tory democracy, and to living a fulfilling, of course, but also the deeper conceptual meaningful life. Instructionism is partic- understanding that will allow them to ularly ill-suited to the education of cre- reason about real-world problems. Learn- ative professionals who can develop new ing sciences research has identified the knowledge and continually further their own key features of those learning environ- understanding; instructionism is an anachro- ments that help students learn deeper nism in the modern innovation economy. conceptual understanding. Beginning in the 1970s, a new science of r The importance of building on a learner’s learning was born – based in research emerg- prior knowledge. Learners are not empty ing from psychology, computer science, phi- vessels waiting to be filled. They come to losophy, sociology, and other scientific dis- the classroom with preconceptions about ciplines. As they closely studied children’s how the world works; some of them learning, scientists discovered that instruc- are basically correct, and some of them tionism was deeply flawed. By the 1990s, are misconceptions. The best way for after about twenty years of research, learning children to learn is in an environment scientists had reached a consensus on the fol- that builds on their existing knowledge; lowing basic facts about learning – a consen- if teaching does not engage their prior sus that was published by the United States knowledge, students often learn informa- National Research Council (see Bransford, tion just well enough to pass the test, and Brown, & Cocking, 2000): then revert back to their misconceptions r The importance of deeper conceptual under- outside of the classroom. r standing. Scientific studies of knowl- The importance of reflection. Students edge workers demonstrate that expert learn better when they express their knowledge includes facts and procedures, developing knowledge – either through but simply acquiring those facts and conversation or by creating papers,

introduction 3

reports, or other artifacts – and then are stand the problems with today’s cur- provided with opportunities to reflec- ricula, teacher education programs, and tively analyze their state of knowledge. standardized tests, and how to form a vision for the future. This handbook is an introduction to this r For professionals, reading about the new science of learning, and how researchers new science of learning can help you are using that science to lay the groundwork understand why many people are so for the schools of the future. This new sci- poorly informed about science, technol- ence is called the learning sciences because ogy, international relations, economics, it is an interdisciplinary science: it brings and other knowledge-based disciplines. together researchers in psychology, educa- r tion, computer science, and anthropology, And finally, education researchers can among others, and the collaboration among learn how their own studies relate to the these disciplines has resulted in new ideas, learning sciences, and can see how to par- new methodologies, and new ways of think- ticipate in building the schools of the ing about learning. Many people – parents, future. teachers, policy makers, and even many edu- This handbook is the second book that cational researchers – are not aware of the introduces a broad audience to the new sci- important discoveries emerging from the ence of learning. The first was the NRC learning sciences. Without knowing about report How People Learn, first published in the new science of learning, many people 1999 and with an expanded edition pub- continue to assume that schools should be lished in 2000 (Bransford, Brown, & Cock- based on instructionism. Parents and pol- ing, 2000). That book provides a higher-level icy makers remember being taught that way, overview of the learning sciences; this hand- and are often uncomfortable when their book goes into more depth, is more spe- children have different learning experiences. cific about exactly how to reform schools, Many teachers have spent an entire career and describes important work that has taken mastering the skills required to manage an place since 1999. In particular, this handbook instructionist classroom, and they under- describes how to use the new sciences of standably have trouble envisioning a dif- learning to design effective learning environ- ferent kind of school. The purpose of this ments, in classrooms and outside, often tak- handbook is to build on the new science ing advantage of new computer technology. of learning by showing various stakehold- Learning sciences is now over twenty years ers how to design learning environments old; the publication of this handbook is a sign and classrooms: that the scientific community has reached a r For teachers, reading about the new sci- consensus about some of the most impor- ence of learning can help you be more tant discoveries about learning. Redesigning effective in your classrooms. schools so that they are based on scientific r For parents, reading about the new sci- research is a mammoth undertaking, and it ence of learning can help you to be will require the participation of all of the an informed consumer of schools. The groups that read this book: teachers, parents, learning sciences explains why and when school leaders, policy makers, and education instructionism fails and which alternative researchers. learning environments are based in con- temporary science. r For administrators, reading about the new The Goals of Education and science of learning can help you to lead the Nature of Knowledge your school into the twenty-first century. r For policy makers, reading about the new The traditional role of educational research science of learning can help you under- has been to tell educators how to achieve

4 the cambridge handbook of the learning sciences

Table 1.1. Deep Learning Versus Traditional Classroom Practices Learning Knowledge Deeply Traditional Classroom Practices (Findings from Cognitive Science) (Instructionism) Deep learning requires that learners relate new Learners treat course material as unrelated to ideas and concepts to previous knowledge and what they already know. experience. Deep learning requires that learners integrate Learners treat course material as disconnected their knowledge into interrelated conceptual bits of knowledge. systems. Deep learning requires that learners look for Learners memorize facts and carry out patterns and underlying principles. procedures without understanding how or why. Deep learning requires that learners evaluate Learners have difficulty making sense of new new ideas, and relate them to conclusions. ideas that are different from what they encountered in the textbook. Deep learning requires that learners understand Learners treat facts and procedures as static the process of dialogue through which knowledge, handed down from an all-knowing knowledge is created, and they examine the authority. logic of an argument critically. Deep learning requires that learners reflect on Learners memorize without reflecting on the their own understanding and their own process purpose or on their own learning strategies. of learning.

their curriculum objectives, but not to help doing historical inquiry rather than memo- set those objectives. But when learning rizing dates and sequences of events: work- scientists went into classrooms, they discov- ing with primary data sources, and using ered that schools were not teaching the deep methods of historical analysis and argumen- knowledge that underlies intelligent perfor- tation that are used by historians (National mance. By the 1980s, cognitive scientists had Center for History in the Schools, 1996). discovered that children retain material bet- In science, the National Science Education ter, and are able to generalize it to a broader Standards calls for students to engage in the range of contexts, when they learn deep authentic practices of scientiﬁc inquiry: con- knowledge rather than surface knowledge, structing explanations and preparing argu- and when they learn how to use that knowl- ments to communicate and justify those edge in real-world social and practical set- explanations (National Research Council, tings (see Table 1.1). The notion of deep 1996,p.105). learning is explored by each learning sciences To better understand how to engage stu- researcher in a slightly different way, and dents in authentic practices, many learning most of the chapters in this handbook begin sciences reforms are based on studies of pro- by describing the type of deep knowledge fessional practice. studied. r One of the central underlying themes Professionals engage in a process of of the learning sciences is that students inquiry, in which they start with a driv- learn deeper knowledge when they engage ing question and then use discipline- in activities that are similar to the everyday speciﬁc methods to propose hypothetical activities of professionals who work in a dis- answers to the question, and to gather and cipline. Authentic practices are the keystone evaluate evidence for and against com- of many recent educational standards doc- peting hypotheses (Krajcik & Blumen- uments in the United States In history, for feld, this volume; Edelson & Reiser, this example, reforms call for learning history by volume).

introduction 5

r Professionals use complex representa- This combined research has led the learn- tions to communicate with each other ing sciences to a focus on how children learn during collaboration (as discussed in in groups (as discussed in the chapters in many of the chapters in Parts 3 and 4). Part 5). r Scientists and mathematicians work with Of course, students are not capable of concrete, visual models, so students doing exactly the same things as highly should too (Lehrer & Schauble, this trained professionals; when learning scien- volume). tists talk about engaging students in authentic practices, they are referring to develop- This focus on authentic practice is based mentally appropriate versions of the situated on a new conception of the expert knowl- and meaningful practices of experts. One of edge that underlies knowledge work in the most important goals of learning sciences today’s economy. In the 1980s and 1990s, research is to identify exactly what practices scientists began to study science itself, and are appropriate for students to engage in and they began to discover that newcomers learn, and how learning environments can become members of a discipline by learn- be designed that are age-appropriate with- ing how to participate in all of the prac- out losing the authenticity of professional tices that are central to professional life in practice. that discipline. And, increasingly, cutting- edge work in the sciences is done at the boundaries of disciplines; for this reason, stu- The Foundations of dents need to learn the underlying models, the Learning Sciences mechanisms, and practices that apply across many scientific disciplines, rather than learn- The learning sciences combines many disci- ing in the disconnected and isolated six- plinary approaches to the study of learning. week units that are found in instructionist Scholars in a range of university departments science classrooms – moving from studying conduct research in the learning sciences – the solar system to studying photosynthesis they are found in schools of education, of to studying force and motion, without ever course, but also in departments of com- learning about connections among these puter science and psychology. I review five units. early influences – constructivism, cogni- Studies of knowledge workers show that tive science, educational technology, socio- they almost always apply their expertise cultural studies, and studies of disciplinary in complex social settings, with a wide knowledge. array of technologically advanced tools along with old-fashioned pencil, paper, chalk, and Constructivism blackboards. These observations have led learning sciences researchers to a situativ- In the 1960s and 1970s, Jean Piaget’s writ- ity view of knowledge (Greeno, this vol- ings became widely influential in Ameri- ume). “Situativity” means that knowledge is can education. Before Piaget, most people not just a static mental structure inside the held to the commonsense belief that chil- learner’s head; instead, knowing is a process dren have less knowledge than adults. Piaget that involves the person, the tools and other argued a radically different theory: although people in the environment, and the activities children certainly possess less knowledge in which that knowledge is being applied. than adults, what’s even more important The situativity perspective moves beyond a to learning is that children’s minds contain transmission and acquisition conception of different knowledge structures than are in learning; in addition to acquiring content, adults’ minds. In other words, children dif- what happens during learning is that pat- fer not only in the quantity of knowledge terns of participation in collaborative activ- they possess; their knowledge is qualitatively ity change over time (Rogoff, 1990, 1998). different.

6 the cambridge handbook of the learning sciences

By the 1980s, researchers had con- vide much support to educators, because it firmed this fundamental claim that children focused on laboratory methodologies that think differently from adults. Educational removed learners from learning contexts, researchers had discovered, for example, and because it focused on static knowledge that children don’t get math problems like facts and procedures rather than the wrong only because they didn’t study hard processes of thinking and knowing (Kuhn, enough or because they forgot what they 1990,p.1). Around 1990, many key con- read in the textbook – they often got the cepts from cognitive science became central problems wrong because their minds were in the learning sciences; I discuss representa- thinking about the math problems in a differ- tion, expertise, reflection, problem solving, ent way than educators expected, and math and thinking. education wasn’t designed to correct these misconceptions. Cognitive scientists began representation to identify the cognitive characteristics of Central to cognitive science is the idea that children’s “na¨ıve math” and “na¨ıve physics,” intelligent behavior is based on representa- and began to accumulate an important body tions in the mind: “knowledge structures” of knowledge about the typical misconcep- such as concepts, beliefs, facts, procedures, tions that people have about these content and models. In the 1970s, cognitive scien- areas (diSessa, this volume; Linn, this vol- tists thought of representation in metaphors ume). This body of research allows design- drawn from computer memory techniques. ers of learning environments to connect A central feature of most computer lan- learning to students’ prior knowledge and guages is the pointer: a way for one mem- misconceptions. ory location to “point to” or “refer to” Constructivism explains why students another location. Building on the primitive often do not learn deeply by listening to a notion of a pointer, computer programmers teacher, or reading from a textbook. Learn- were able to develop hierarchically nested ing sciences research is revealing the deeper data structures – the highest level structure underlying bases of how knowledge con- could contain pointers to simpler, lower level struction works. To design effective learn- structures. For example, the simplest data ing environments, one needs a very good structure for a house would contain hun- understanding of what children know when dreds of variables, including the type of sink they come to the classroom. This requires in the kitchen and the color of the couch sophisticated research into children’s cogni- in the living room. But by using the nested tive development, and the learning sciences data structures that pointers made possible, draws heavily on psychological studies of a more sophisticated data structure for a cognitive development (e.g., Siegler, 1998). house could be constructed that would contain pointers to data structures for each room Cognitive Science in the house; and the room data structures would each contain pointers to multiple fur- Many learning scientists began their careers niture and fixture data structures. This pro- in the interdisciplinary field known as cog- vided a metaphor for how knowledge might nitive science. Cognitive science combines be modularized in the mind, and is an exam- experimental investigation of how the mind ple of the kind of metaphors of human cog- works (in the tradition of cognitive psy- nition that have emerged from computer chology) with computational modeling of science. proposed mental processes (in the tradition of artificial intelligence), taking into account what we know from sociology and the cognitive bases of expertise anthropology about how people use knowl- One of the most surprising discoveries of edge in everyday settings. Through the 1970s 1970s cognitive science was that every- and 1980s, cognitive science did not pro- day behavior was harder to represent

introduction 7

computationally than expert behavior. Some pare her process to that of the expert. The of the most successful artificial intelli- computer’s role was to record the expert’s gence (AI) programs simulated expert per- reasoning, making it available whenever it formance in knowledge-intensive domains could be useful and to whoever needed like medicine, manufacturing, telecommu- it. In this way, the computer was sup- nications, and finance (Liebowitz, 1998). As porting a kind of reflection that was diffi- a result of these efforts, cognitive science cult to do without a computer. Since then, developed a sophisticated understanding of several learning sciences projects empha- the cognitive bases of expertise. Everyday size computer support for reflection. WISE commonsense behavior remains beyond the (Linn, this volume) prompts students to abilities of AI computer programs, even think about evidence and its uses as they as some complex aspects of expert perfor- are creating a scientific argument. Recip- mance in knowledge-intensive domains like rocal teaching (Palincsar & Brown, 1984) medicine have been successfully simulated. helps students to recognize the questions A large body of cognitive science research they need to ask themselves as they are shows that expertise is based on: trying to understand something they are r reading. Knowledge Forum (Scardamalia & A large and complex set of representa- Bereiter, this volume) prompts students to tional structures r think about their actions and their discussion A large set of procedures and plans as they are having knowledge-building con- r The ability to improvisationally apply versations. Learning by Design (Kolodner, and adapt those plans to each situation’s this volume) integrates reflection into unique demands classroom activities. r The ability to reflect on one’s own cognitive processes while they are occurring problem solving Cognitive scientists have spent several reflection decades attempting to identify the cognitive bases of problem solving. One of the most Studies of experts show they are better persistent theories about problem solving is than novices at planning and criticizing their that it depends on a person having a men- work – both reflective activities. For example, tal representation of a problem space (Newell when expert writers are asked to describe & Simon, 1972) which contains beliefs and their thought processes out loud as they mental representations – of concepts, spe- write, their talk reveals that they develop cific actions, and the external world. Prob- goals and plans while writing, and they con- lem solving is then conceived of as searching tinually reflect on and modify those goals through the problem space until the desired and plans as they write (Flower & Hayes, goal state is reached. Because knowledge 1980). School-age writers don’t spend time work typically requires problem solving, planning and reflecting (Burtis, Bereiter, many learning sciences approaches to learn- Scardamalia, & Tetroe, 1983). Based on these ing are based on this research. For exam- findings, and similar findings regarding other ple, Koedinger’s cognitive tutors (this vol- school subjects, learning scientists often con- ume) assume that production rules are used ceive of the problem of learning as a prob- to move through the problem space, and lem of transforming novices into experts by Kolodner’s case-based reasoning (this vol- developing their ability to reflect on their ume) assumes that case lookup and match- own thinking in these ways. ing algorithms are used. Collins and Brown (1988) first suggested that the computer could be used to support reflection (Collins, this volume). Collins and thinking Brown talked about capturing an expert’s Educators often talk about the importance of process, then allowing the student to com- higher-order thinking skills, but educational

8 the cambridge handbook of the learning sciences

programs that emphasize thinking skills business community that it was essential to are often not based on scientific research. get computers into schools (Cuban, 2001). Instead, they are based on one or another During the 1990s, there was a major push intuitively based taxonomy of thinking skills, to install computers and the Internet in with almost no scientific justification of why schools – including federal government pro- this specific set of skills should be taught grams like E-rate that paid for schools to in schools (Kuhn, 1990,p.2). Beginning be connected to the Internet. By 2003, in the 1980s and 1990s, cognitive psychol- 95 percent of schools were connected to ogists began to study informal reasoning the Internet with high-speed connections, (Voss, Perkins, & Segal, 1991) – the good and 93 percent of all classrooms were con- and bad reasoning that people engage in nected to the Internet. On average, there everyday, when faced with real-life problems were 4.4 students for each computer with that don’t have simple solutions. They also Internet access; this was a dramatic drop began to study everyday decision making, from 12.1 students in 1998, when it was first discovering a wide range of common think- measured (Parsad & Jones, 2005). ing errors that most people make (Baron, However, the impact of all of this invest- 1985; Kahneman, Slovic, & Tversky, 1982). ment has been disappointing. By 2000,no Also during this time, developmental psy- studies had shown that computer use was chologists began to identify a range of good correlated with improved student perfor- and bad thinking strategies and how these mance. When researchers began to look strategies develop over the lifespan. They more closely at why computers were hav- extended Piaget’s original insight, showing ing so little impact, they discovered that how children’s thinking differs from that of computer use was not based on the learn- adults – information that is absolutely crit- ing sciences; instead, they were being used ical to education based on the learning sci- as quick add-ons to the existing instructional ences (Dunbar & Klahr, 1989; Kuhn, 1989; classroom (Cuban, 2001). Schauble, 1990). Learning scientists emphasize the power- ful role that computers can play in transforming all learning. But their vision rejects Educational Technology instructionism and behaviorism and the CAI In the 1950s, B. F. Skinner presented his systems based on it, and presents a new “teaching machines” and claimed that they vision of computers in schools. Learning sci- made the teacher “out of date” (Skinner, ences research explains why the promise 1954/1968,p.22). The first educational of computers in schools has not yet been software was designed in the 1960s and realized; because to date, educational soft- was based on Skinner’s behaviorist theories; ware has been based on instructionist the- these systems are known as Computer ories, with the computer performing roles Assisted Instruction or CAI, and such sys- that are traditionally performed by the tems are still in use today. In the 1970s, a teacher – with the software acting as an few artificial intelligence researchers started expert authority, delivering information to working in education, developing auto- the learner. In contrast, learning sciences sug- mated tutoring systems and other applica- gests that the computer should take on a tions (Bobrow & Collins, 1975; Sleeman & more facilitating role, helping learners have Brown, 1982; Wenger, 1987). In the 1980s, the kind of experiences that lead to deep cognitive scientists like Roger Schank and learning – for example, helping them to col- Seymour Papert made widely popularized laborate, or to reflect on their developing claims that computers would radically trans- knowledge. Many of the chapters in this form schools (see Papert, 1980; Schank, this handbook describe the next generation of volume). educational software, software that is solidly By the 1990s, a strong consensus had based on the sciences of learning, and that is formed among politicians, parents, and the designed in close collaboration with teachers

introduction 9

and schools. Computers are only used as how children learn in nonschool settings – part of overall classroom reform, and only how children learn their first language or where research shows they will have the the norms and conventions of their culture; most impact. Computer software is central how apprentices learn on the job. Some in the learning sciences because the visual of the most interesting work along these and processing power of today’s personal lines focused on informal learning in non- computers supports deep learning: Western societies without formal schooling 1996 1988 1990 r (Cole, ;Lave, ; Rogoff, ; Saxe, Computers can represent abstract knowl- 1991). Equally influential studies examined edge in concrete form the socially distributed nature of knowledge r Computer tools can allow learners to work – including studies of navy ship navi- articulate their developing knowledge in gation (Hutchins, 1995), of London Under- a visual and verbal way ground control rooms (Heath & Luff, 1991), r Computers can allow learners to manip- of office systems (Suchman, 1987), and of air ulate and revise their developing knowl- traffic control centers (Hughes et al., 1988). edge via the user interface, in a complex This research revealed that outside of for- process of design that supports simultane- mal schooling, almost all learning occurs in ous articulation, reflection, and learning a complex social environment, and learning r Computers support reflection in a com- is hard to understand if one thinks of it as a bination of visual and verbal modes mental process occurring within the head of r an isolated learner. Internet-based networks of learners can The sociocultural approach has been share and combine their developing widely influential in all of the disciplines par- understandings and benefit from the ticipating in the learning sciences: power of collaborative learning r Artificial intelligence began to emphasize “distributed cognition” in part because of Sociocultural Studies the rapidly evolving network technolo- gies of the 1980s and 1990s After the burst of activity associated with r 1970s artificial intelligence and cognitive Cognitive psychology began to study psychology, by the 1980s many of these teamwork, collaboration, group dynam- scholars had begun to realize that their goal – ics, and the role of social context in cognitive development to understand and simulate human intel- r ligence in the computer – was still very Education research began to study class- far off. The 1980s disillusionment with AI room collaboration, collaborative dis- was so severe that it was informally known course in student groups, and project as “the AI winter.” Researchers began to teams step back and think about why the cognitive sciences had not been more success- The Nature of Knowledge Work ful. The most influential answer was provided by a group of interrelated approaches Should we reduce auto emissions because including the sociocultural, situative, and dis- of global warming? Should we allow stem tributed cognition approaches (Greeno, this cell research to proceed? Should we teach volume; Salomon, 1993). Socioculturalists both evolution and creationism in schools? began with the observation that all intel- Today’s public debate about such con- ligent behavior was realized in a complex troversial issues shows a glaring lack of environment – a human created environ- knowledge about scientific practice. The ment filled with tools and machines, but U.S. National Science Education Standards also a deeply social environment with col- (National Research Council, 1996) observed laborators and partners. Some of the most that “Americans are confronted increasingly important studies in this tradition examined with questions in their lives that require

10 the cambridge handbook of the learning sciences

scientific information and scientific ways with their doctor the potential risks of an of thinking for informed decision making” upcoming surgery, or as they evaluate the (p. 11). health risks of a new development near their By the early 1900s, major industrial coun- neighborhood. tries had all realized the important role that This new view of expert knowledge has science and engineering played in their rapid been extended beyond science to other growth, and many scholars began to analyze forms of knowledge work. For example, liter- the nature of scientific knowledge. In the first acy scholars have discovered that advanced half of the twentieth century, philosophers literacy involves much more than knowing came to a consensus on the nature of sci- which sounds correspond to which letters; entific knowledge: scientific knowledge con- literacy involves knowing how to partic- sisted of statements about the world, and ipate in a complex set of literate prac- logical operations that could be applied to tices – like reading a recipe, scanning the those statements. This consensus was known classifieds for a specific product, or writ- as logical empiricism (McGuire, 1992; Suppe, ing an email to a colleague (Palincsar & 1974). Logical empiricism combined with Ladewski, this volume). Social science edu- behaviorism and traditional classroom prac- cators have discovered that historians are tice to form the instructionist approach to experts because they know how to engage education: disciplinary knowledge consisted in the complex practices of historical inquiry of facts and procedures, and teaching was and argumentation. thought of as transmitting the facts and procedures to students. Beginning in the 1960s, sociologists, psy- Processes Involved in Learning chologists, and anthropologists began to study how scientists actually did their work, The learning sciences are centrally con- and they increasingly discovered that sci- cerned with exactly what is going on in a entific knowledge was not simply a body learning environment, and exactly how it of statements and logical operations. In is contributing to improved student perfor- this new view, scientific knowledge is an mance. The learning environment includes understanding about how to go about doing the people in the environment (teachers, science, combined with deep knowledge learners, and others); the computers in the of models and explanatory principles con- environment and the roles they play; nected into an integrated conceptual frame- the architecture and layout of the room and work. The practice of science involves the physical objects in it; and the social experimentation, trial and error, hypothesis and cultural environment. Key questions testing, debate and argumentation. And sci- include: How does learning happen? How ence is not a solo endeavor; it involves fre- do different learning environments con- quent encounters with peers in the scientific tribute to learning, and can we improve the community. Scientists frequently talk about design of learning environments to enhance evaluating other scientists’ claims, and think learning? Some researchers work on specific about how best to support and present their components of the learning environment – claims to others. software design, the roles that teachers In this new view, scientific knowledge is should play, or specific activities each stu- situated, practiced, and collaboratively gen- dent performs. Others examine the entire erated. The traditional science classroom, learning environment as a system, and focus with its lectures and step-by-step lab exer- on more holistic questions: How much sup- cises, completely leaves out these elements port for the student should come from the of science. But this kind of knowledge would teacher, the computer software, or from be extremely useful to the general pub- other students? How can we create a cul- lic as they read reports of an experimen- ture where learners feel like a “learning tal drug in the daily paper, as they discuss community”? How can we design materials