Two Ways to Elaborate Vygotsky's Concept of Mediation

Implications for Instruction

Yuriy V. Karpov H. Carl Haywood

Touro College Touro College and Vanderbilt University

L. S. Vygotsky's (1978, 1981, 1983, 1934/1988)writings suggest 2 major types of mediation as the main mechanism of children's learning and development. Metacognitive mediation refers to the acquisition of semiotic tools of selfregulation. Cognitive mediation refers to the acquisition of scientific concepts representing the essence of some class of phenomena. Some approaches taken by American researchers, characterized as guided discovery in a community of learners, are relevant to L. S. Vygotsky's concept of metacognitive mediation but are in sharp contrast to his concept of cognitive mediation. The "theoretical learning" approach taken by Russian followers of L. S. Vygotsky incorporates his concept of cognitive mediation but fails to emphasize adequately the concept of metacognitive mediation. Analysis of these approaches shows that it is advisable to develop an instructional procedure that incorporates both of L. S. Vygotsky's types of mediation.
isthin the past 20 years--long after Vygotky's death--some of his concepts have ome to exert a strong influence on western cognitive psychology and, through that medium, on education. In this article, we analyze Vygotsky's concept of mediation, show how this concept has influenced recent research in learning and instruction, and attempt to integrate derivative approaches to instruction.

of semiotic tools of self-regulation: self-planning, selfmonitoring, self-checking, and self-evaluating. Thus, this type of mediation facilitates the development of some processes that are designated in contemporary psychological literature as metacognitive (executive) processes (see, e.g., Borkowski & Kurtz, 1987; Brown, 1987; Brown, Bransford, Ferrara, & Campione, 1983; Brown & DeLoache, 1978; Flavell, 1976; Kluwe, 1987). In the remainder of this discussion, we refer to this type of mediation as metacognitive mediation. According to Vygotsky, metacognitive mediation of children's psychological processes has its roots in interpersonal communication. In the course of joint activity with children, adults or more mature peers use semiotic tools to regulate the children's behavior. For example, when a mother says no to a child to prevent the child from doing something dangerous or undesirable, she is both regulating the child's behavior and supplying the child with a tool of self-regulation. The child starts to say no aloud to himself or herself to overcome a temptation to do something inappropriate (so-called egocentric or private speech), sometimes even imitating the mother's voice (Luria, 1961). The use of egocentric speech for self-regulation is an intermediate step toward the develEditor's note. Author's note.

James L. Pate served as action editor for this article.

Mediation as the Main Mechanism of Learning and Development

The central concept in Vygotsky's (1978, 1981, 1983, 1934/1988) cognitive psychology is mediation. According to Vygotsky, all specifically human psychological processes (so-called higher mental processes) are mediated by psychological tools such as language, signs, and symbols. Adults teach these tools to children in the course of their joint activity, the children internalize them, and these tools then function as mediators of the children's more advanced psychological processes. By analyzing Vygotsky's works, one can distinguish two major types of mediation: metacognitive and cognitive. ~

Yuriy V. Karpov, Graduate School of Education and Psychology, Touro College; H. Carl Haywood, Graduate School of Education and Psychology, Touro College, and Department of Psychology and Human Development, Vanderbilt University. Correspondence concerning this article should be addressed to Yuriy V. Karpov, Graduate School of Education and Psychology, Touro College, 350 Fifth Avenue, Suite 1700, New York, NY 10118. Electronic mail may be sent via Internet to carlh@touro.edu.

Metacognitive Mediation
The first of Vygotsky's (1978, 1981, 1983, 1934/1988) two types of mediation refers to children's acquisition January 1998 American Psychologist

I We are aware of the relativity of our distinction between metacognitive and cognitive mediation in Vygotsky's works. This relativity is due partially to the fact that cognitive and metacognitive processes are not clearly differentiated in cognitive science (Brown, 1987) and also because a sharp separation of these processes may not be justifiable. Vygotsky himself did not differentiate explicitly between metacognitive and cognitive mediation. Thus, discussing the acquisition of scientific concepts (the major type of cognitive mediation of school pupils that we analyze in this article), Vygotsky (1934/1988) also pointed out "metacognitive" outcomes of such acquisition: "Reflective consciousness comes to the child through the portals of scientific concepts" (p. 171).

Copyright 1998by the AmericanPsychologicalAssociation,Inc. 0003-066X/98/$2.00 Vol. 53, No. I, 27-36

27

the specifically human capacity for language enables children to provide for auxiliary tools in the solution of difficult tasks, to overcome impulsive action, to plan a solution to a problem prior to its execution, and to master their own behavior. (Vygotsky, 1978, p. 28) 2

Cognitive Medlation
The second type of mediation that Vygotsky (1934/1988) described refers to children's acquisition of cognitive tools that are necessary for solving subject-domain problems. In the rest of this discussion, we refer to this type of mediation as cognitive mediation. Systematic cognitive mediation of children begins when they go to school and begin to acquire "scientific concepts." Vygotsky (1934/ 1988) contrasted scientific concepts with the "spontaneous concepts" of preschoolers. Spontaneous concepts are the results of generalization and internalization of everyday personal experience in the absence of systematic instruction. That is why such concepts tend to be unsystematic, empirical, not conscious, and often wrong. For example, a three-yearold child, having observed a needle, a pin, and a coin sinking in water, comes to the wrong conclusion that all small objects sink and begins to use this concept for predicting the behavior of different objects in water (Zaporozhets, 1986, p. 207). In contrast to spontaneous concepts, scientific concepts represent the generalization of the experience of humankind that is fixed in science and that children are taught in the course of systematic instruction. Once scientific concepts have been acquired and internalized, they come to mediate children's problem solving. In the example given here, the scientific concept that would make it possible to predict the behavior of objects in water is Archimedes's law. Vygotsky held, contrary to Hall's (1907) notion that "ontogeny recapitulates phylogeny," that children's development should not repeat the history of development of humankind. He held further, in contrast to Piaget's (1970) and Dewey's (1902) constructivist notions (discussed later), that children should not and cannot be required to come to understand the world by way of rediscovery of the principal explanatory laws already discovered by humankind. The development of children is so special, in large part, because adults teach them these laws ready-made. That is why, according to Vygotsky (1934/1988), the acquisition of scientific concepts should arise from their presentation to students in the form of precise verbal definitions (p. 148). Comparing Vygotsky's idea of cognitive mediation with contemporary psychological views on cognition and its development, one can see a strong resemblance of the 2There is evidence that chimpanzees are able to acquire at least some elementsof computer-basedlanguage (Rumbaugh, 1977). Furthermore, chimpanzees that have acquired these elements of language appear to demonstrate better self-regulation than do those that have not (Savage-Rumbaugh, 1986, 1996). In view of these reports, Vygotsky might have been wrong in assuming that the capacity for developing language is "specificallyhuman" but was right in his emphasis on the role of the acquisition of languagein the developmentof self-regulation. January 1998 American Psychologist

opment of inner (nonvocal) speech, which later becomes the child's internalized tool for self-regulation. According to experimental data obtained by Vygotsky (1934/1988) that were consistent with some of Piaget's (1923/1959) observations, egocentric speech almost completely disappears by the age of seven to eight years. The disappearance of egocentric speech by this age does not mean, however, that all eight-year-old children have internalized the tools of self-regulation. Recent research results, discussed in more detail below, suggest that selfregulation can be a problem even at a much older age and that Vygotsky's general model of intervention (from being regulated by another person to self-regulation) is applicable to older children as well. An important point that is often overlooked in reviews of Vygotsky's theory is that a powerful facilitator of children' s acquisition of semiotic tools of self-regulation is their use of such tools for regulating the behavior of others in the course of their joint activity. Vygotsky (1981) was very specific in formulating this proposition: " A sign is always originally a means used for social purposes, a means of influencing others, and only later becomes a means of influencing oneself" (p. 157). "Regulation of others' behavior by means of the word gradually leads to the development of verbalized behavior of the people themselves" (p. 159). Vygotsky's ideas on the development of self-regulation in children may be summarized as follows: (a) Someone regulates the child's behavior by the use of external speech, (b) the child regulates someone else's behavior by the use of external speech and regulates his or her own behavior by using egocentric speech, and (c) the child regulates his or her own behavior by using inner speech. Thus, 28

Vygotskian notion of scientific concepts to the component of declarative knowledge that has been designated "defined concepts" (Gagnt, 1985). The main limitation of such concepts is that, in themselves, they are not sufficient for the solution of subject-domain problems. For example, students may know the scientific concept of a right-angle triangle, but when they are given pictures of different triangles and asked which of them are right-angle ones, they may not use their knowledge of this scientific concept for solving the problems (Davydov, 1972). That is why contemporary cognitive psychologists, recognizing the importance of declarative knowledge, hold that another type of knowledge, namely procedural knowledge (the knowledge of methods and strategies of solving problems), is critically important for solving subject-domain problems (Anderson, 1983, 1993). They insist that the only way to have children develop meaningful knowledge is by "marrying concepts to procedures" (Bruer, 1993, p. 95), that is, teaching by demonstrating the links between concepts and procedures. Vygotsky (1934/1988) himself clearly understood that "the difficulty with scientific concepts lies in their verbalism" (p. 148). He pointed out that "scientific conc e p t s . . , just start their development, rather than finish it, at a moment when the child learns the term or wordmeaning denoting the new concept" (Vygotsky, 1934/ 1988, p. 159). Vygotsky did not describe how to link scientific concepts to procedures, but that was done by his contemporary Russian followers whose approach we discuss in the Cognitive Mediation in Russian Instructional Psychology section.

Metacognitive Mediation in American Instructional Psychology

Among the innovative approaches to instruction developed in contemporary psychology, some of those develJanuary 1998 American Psychologist

oped by American researchers are the most relevant to Vygotsky's (1978, 1981, 1983, 1934/1988) idea of metacognitive mediation. In this section, we discuss two groups of such approaches. The first group, defined as mediated-learning approaches, is explicitly based on Vygotsky's ideas of metacognitive mediation and on very similar ideas of the Israeli psychologist Feuerstein and colleagues (Feuerstein, Rand, Hoffman, & Miller, 1980). The main accomplishment of the American mediated-learning approaches has been the design of programs for metacognitive intervention with preschoolers (Bright Start; Haywood, Brooks, & Burns, 1986, 1992) and with elementary school students (COGNET; Greenberg, 1994). One of the major principles underlying these intervention programs is Vygotsky's and Feuerstein's idea of gradual transfer of the responsibility for planning, directing, monitoring, checking, and evaluating from the adult to the child in the course of their collaborative activity. For example, in Bright Start, preschool children play games in which they are taught to bring the tempo of their bodily movements under the control of an external stimulus (i.e., walk slowly when the accompanying music is slow, walk fast when the music is fast, and stand still when the music stops). Then, they are taught to verbalize this rule and to use it for regulating the behavior of other children as well as their own behavior (Haywood et al., 1992). This strategy of intervention has turned out to be very efficient for facilitating children's self-control (see, e.g., Chatelanat & Haywood, 1995; Haywood, 1996; Paour, Cbbe, & Haywood, 1997). The second group of approaches to instruction that is the most relevant to Vygotsky's idea of metacognitive mediation is the most important for our further discussion. The main accomplishment of these approaches has been the development of innovative programs for teaching academic subjects in school (Brown & Campione, 1990, 1994; Brown, Campione, Reeve, Ferrara, & Palincsar, 1991; Chang-Wells & Wells, 1993; Cobb, Wood, & Yackel, 1993; Cobb, Yackel, & Wood, 1992; Cognition and Technology Group at Vanderbilt, 1990, 1992, 1994; Schoenfeld, 1985, 1992; Wells, Chang, & Maher, 1990). Although they are different in certain aspects, these approaches have a major point in common. This major point can be designated, using Brown and Campione's (1994) terminology, as "guided discovery in a community of learners" (GDCL). From this designation, it is clear that all these approaches are based on two principal ideas. The first idea is that the course of instruction should be organized as students' cooperative, shared activity, under mutual control. For example, one of the instructional procedures within GDCL programs is "reciprocal teaching" (Palincsar & Brown, 1984; Palincsar, Brown, & Campione, 1993). Students and the teacher take turns leading discussions. The responsibility of the leader is summarizing, questioning, clarifying, and predicting, in other words, regulating and monitoring the students' problem solving. 29

This idea is directly relevant to Vygotsky's (1981, 1983) principle that the basis for development of children's self-regulation is their experience in regulating the behavior of others. This idea is also supported by experimental data that an advantageous situation for development of children's self-regulation is one of collaborative problem solving in which children "assume separate but complementary social roles. One child may perform an observing, guiding, and correcting role while the other performs the task procedures" (Forman & Cazden, 1995, p. 341). A community of learners has distinct advantages in this respect because regulation of each other's performance is a much more natural process in a peergroup situation than in a teacher-student interaction (Forman & Cazden, 1995). The second idea that underlies GDCL approaches is the idea of guided discovery. This idea is consistent with constructivist theories of learning influenced by the works of Dewey (1902) and Piaget (1970) but is in sharp contrast to Vygotsky's (1934/1988) views on cognitive mediation in the school setting. According to the idea of guided discovery, scientific knowledge should not be taught ready-made, as Vygotsky believed, but rather should be constructed by students themselves in the course of discussions, sharing their personal experiences and carrying out a kind of research activity. A group of students involved in guided discovery is similar to a group of research collaborators solving a scientific problem (Cobb et al., 1992), whereas the role of the teacher is to guide and to orchestrate students' discovery processes (Brown & Campione, 1994). Some adherents of GDCL clearly contrast Vygotsky's emphasis on "the importance of formal definitions and of the teacher's explicit explanations" with their own emphasis on "inquiry mathematics, [which is] interactively constituted in the classroom" (Cobb et al., 1993, p. 100). The authors of GDCL approaches usually contrast their instructional programs with "didactic teaching," which is typical in the traditional system of school instruction. They criticize didactic teaching for making students passive recipients of information; for not taking into consideration students' interests and needs in the course of instruction; for not developing students' abilities for critical thinking and independent learning; and for supplying students with inert, nontransferable knowledge that is, at best, used in the classroom but not in solving real-life problems. It turns out that GDCL programs have several advantages. Students acquire and transfer knowledge better, their planning and monitoring activities improve, and their learning motivation is much higher than in the traditional system of school instruction (Brown & Campione, 1994; Chang-Wells & Wells, 1993; Cobb et al., 1991; Cognition and Technology Group at Vanderbilt, 1994; Schoenfeld, 1992). However, the reasons for the successes of GDCL programs are still unclear. First, these successes may be due to the guided-discovery component of GDCL programs, that is, due to the fact that students are not taught ready-made scientific knowledge but rather discover this knowledge. A second possibility is that the 30

successes of GDCL programs are due to the cooperativelearning component of these programs, which implies students' involvement in the process of regulation of each other's performance. Third, the successes of GDCL programs may be due to both of the aforementioned components. The question of which of these possibilities is correct should be studied specifically. At the same time, there are indications that the successes of GDCL programs are mainly due to facilitating students' metacognition in the course of instruction (Borkowski & Muthukrishna, 1995). As we discussed earlier, facilitating metacognitive processes is an expectable and predictable outcome of engaging children in the process of regulating each other's performance. From this point of view, it is collaborative problem solving, rather than guided-discovery activity, that is mainly responsible for the successes of GDCL programs. There are, to be sure, certain limitations of GDCL programs that obviously are related to their guided-discovery component. First, guided discovery of new knowledge is extremely time-consuming. Second, there are empirical data showing that children do not always develop the correct concepts by themselves in the course of such instruction. Heller and Gordon (1992), who have worked on Brown and Campione's (1990, 1994) project, described what could be seen as a typical lesson organized in accordance with GDCL principles. The children were discussing the concept of an animal. There was a rather long discussion, during which the children were exchanging their understanding of this concept. Then, one of the children opened a dictionary and read aloud the scientific definition of the concept of an animal. One could agree that, during the children's discussion, they came to understand the limitations of their own spontaneous concepts and the necessity of turning to the dictionary for help, that their learning motivation increased, and that their metacognitive processing improved. But, the fact is, they took the ready-made scientific concept of an animal rather than discovering this concept by themselves. Finally, the problem of errors in guided-discovery learning deserves special discussion. GDCL advocates seem to support enthusiastically students' right to make errors in the course of learning: "As long as you're in my class it is okay to make a mistake" (Cobb et al., 1993, p. 98). Sometimes they even claim that " w e especially want students to recognize that there is no right or wrong side in most decisions" (Heller & Gordon, 1992, p. 10). From the point of view of GDCL advocates, this attitude about making errors will not prevent the students from developing, under the teacher's guidance, the correct scientific knowledge. Even if this is true, the danger that the students will come to the wrong knowledge under GDCL procedures should not be underestimated. These procedures involve many unguided students' activities when, for example, they are working as a research group on a project. As Chang-Wells and Wells (1993) pointed out, " A great deal of the learning . . . takes place as students work together (more or less) collaboratively, without the involvement of the teacher" (p. 84). Some January 1998 American Psychologist

experimental data show that such unguided learning activity by a group of peers may result in their rejection of the correct concept in favor of an incorrect concept that one of them has formulated (Tudge, 1992). Theoretically, sooner or later, the teacher in a GDCL classroom will have an opportunity to guide students to reject their misconception and to develop the correct concept, but in practice it might not happen. As Brown and Campione (1994) observed, guided discovery is difficult to orchestrate.

Cognitive Mediation in Russian Instructional Psychology

Vygotsky' s (1934/1988) doctrine of spontaneous and scientific concepts has not influenced substantially the work of American instructional psychologists (see Panofsky, John-Steiner, & Blackwell, 1992, and Schmittau, 1993, for two of a few exceptions). Moreover, his idea of the role of scientific concepts in classroom instruction has been criticized by those who consider this idea to be close to that of didactic teaching (see, e.g., Cobb et al., 1993). By contrast, Vygotsky's Russian students and followers have enthusiastically accepted and elaborated his doctrine, and this doctrine remains the dominant one in the Vygotskian tradition in Russian instructional psychology. One of the main innovations of Vygotsky's followers has been the contention that acquisition of psychological tools, including scientific concepts, involves not only the acquisition of certain verbal knowledge (such as signs and symbols) but also the mastery of relevant procedures (Galperin, 1957, 1966, 1969; Galperin, Zaporozhets, & Elkonin, 1963; Leontiev, 1935/1983). As we noted earlier, this contention is consistent with the ideas of contemporary American cognitive psychologists of "marrying concepts to procedures" (Bruer, 1993, p. 95). The next innovation of Vygotsky's followers has been their elaboration of the idea of the differences between spontaneous and scientific concepts. Their research has shown that the acquisition of spontaneous and scientific concepts is the result of fundamentally different types of learning (Davydov, 1972, 1986; Galperin, 1985; Talyzina, 1975/1981). Following Davydov, we refer to the type of learning that results in spontaneous concepts as empirical learning and to the type of learning that results in scientific concepts as

theoretical learning.
Empirical learning is based on children comparing several different objects or events; picking out their common salient features; and formulating, on this basis, a "general concept" about this class of objects or events. This type of learning can be defined as discovery learning by induction. The main deficiency of empirical learning is that the common salient features of objects or events as the child defines them might not reflect their significant, essential characteristics as formulated in science at the current level of its development. This explains why the preschooler in the example given earlier, having observed a needle, a pin, and a coin sinking, came to the wrong January 1998 American Psychologist

conclusion that all small objects sink. Small size is a common salient feature of a needle, a pin, and a coin, but it is not the essential characteristic that makes them sink. Theoretical learning is based on students' acquisition of methods for scientific analysis of objects or events in different subject domains. Each of these methods is aimed at selecting the essential characteristics of objects or events of a certain class and presenting these characteristics in the form of symbolic and graphic models. Teachers teach methods of scientific analysis, and the students then master and internalize these methods in the course of using them. These methods serve as cognitive tools that mediate students' further problem solving. Thus, theoretical learning is aimed not only at the development of declarative scientific knowledge in a certain subject domain but at mastery of procedural knowledge in this domain as well. Having analyzed the traditional system of school instruction from these positions, Russian followers of Vygotsky, just as the American GDCL advocates did, came to the conclusion that students in this system of instruction often do not master scientific knowledge. The reasons for this situation as formulated by Russian followers of Vygotsky, however, are very different from those formulated by GDCL advocates. Adherents of GDCL claim that students do not achieve mastery of scientific knowledge that is taught in class because this knowledge has been taught ready-made rather than having been constructed by the students themselves. Russian followers of Vygotsky claim that school students often are taught rote skills rather than scientific knowledge or that they are taught scientific knowledge at a purely verbal level without linking this declarative knowledge to the appropriate procedural knowledge (Davydov, 1972). This claim is in agreement with the recent results of the U.S. Department of Education's National Assessment of Educational Progress. These results "indicate that current curricula and teaching methods successfully impart facts and rote skills to most students but fail to impart high-order reasoning and learning skills" (Bruer, 1993, p. 5). In both cases (whether students have mastered a rote skill or have learned the verbal definition of a scientific concept), the acquired knowledge cannot be flexibly applied for solving problems in the relevant subject domain. Rote skills are meaningless and nontransferable, and pure verbal knowledge is inert. That is why students, as a matter of fact, are forced to develop their own empirical knowledge with which to deal with the subject-domain problems. Russian and American psychologists have collected many examples of such empirical knowledge (often wrong) that students hold. For example, having mastered computational operations at the level of rote skills, students have difficulty with figuring out which of these operations should be applied in solving a given word problem. To overcome such difficulty, they 31

look for a key word that reveals which operation to use. For example, "altogether" means add, "take away" means subtract, and "each" means multiply. Students pick an operation on the basis of the key word and apply it slavishly to every number in the problem, whether it makes sense or not. (Bruer, 1993, p. 100) Another observation shows that, having learned at a verbal level the scientific concepts of mammals, birds, and fish, students do not use these concepts for solving problems. Instead, they use their empirical knowledge of common salient features of representatives of each species and say that a whale is a fish because it looks like a fish (Davydov, 1972). DiSessa (1982) made similar observations about the broad use of spontaneous, empirical physical concepts by school students. Thus, traditional school instruction is full of "discoveries" that result in students developing spontaneous, empirical knowledge that often is wrong. Considering the theoretical learning approach to be the best alternative to traditional school instruction, Russian followers of Vygotsky have designed a large number of instructional programs based on the ideas of this approach. These programs have been used for more than 30 years to teach students of different ages (from fiveyear-old children through college students) a variety of subjects, including elementary mathematics, algebra, geometry, physics, chemistry, biology, language, and history (Aidarova, 1978; Davydov, 1966; Dyachenko, 1986; E1konin, 1976; Elkonin & Davydov, 1966; Galperin, 1977, 1985; Galperin & Talyzina, 1961, 1968, 1972; Pantina, 1957; Salmina & Sokhina, 1975; Semenyuk, 1970; Venger, 1986; Zhurova, 1978; and many others). Following are brief descriptions of two of them. 3 Both examples deal with teaching six-year-old children certain knowledge and skills that, under traditional school instruction, are often taught by drill-and-practice and rote memorization. That is why it is especially revealing to show how the theoretical learning approach works for teaching such "nonscientific" knowledge. The program of Pantina (1957) was designed to teach six-year-old children how to write letters of the Russian alphabet. Any letter (or, in general, any contour) can be described by using a set of dots, each of which is placed in a position where a change occurs in the direction of the contour. Such a set of dots would represent the essence of the given contour, that is, its model. Pantina's program is based on supplying students with the method for analysis of any contour and constructing its model on this basis. This method includes the following steps: (a) analyzing the letter to be copied to determine where the direction of the contour changes, (b) placing dots in those positions where a change occurs in the direction of the contour, and (c) reproducing the same set of dots in another place on the page (i.e., constructing the model of the letter) Having constructed the model of a letter, the students connect the marked dots (i.e., write the required letter). Initially, the students perform all of these steps at the practical (sensorimotor) level under the teacher's supervision, but as their analysis is 32

internalized, its major parts start to occur at the visualimagery level. The students visually analyze the letter to be copied, make a mental image of its model, and reproduce the letter. Dyachenko (1986) designed a program for teaching six-year-old children to retell the plot of a story In the course of instruction, the children master the method for analyzing a story in which they identify and model its essential episodes. Having mastered this method, a child, while listening to a story, draws its model (a set of squares connected with arrows) and fills each of the squares with symbols that he or she has invented in accordance with the characters and the actions of each episode. After that, the child retells the story, using the drawn model as the basis for retelling. As a child's analysis becomes internalized, he or she no longer requires drawn models of a story as a basis for retelling. While listening to a story, children make a mental image of its model and retell the story on the basis of this model. The long-term use of these programs, as well as of other instructional programs based on the ideas of the theoretical learning approach, has shown that both the course and the outcomes of instruction are improved considerably under these programs. Because these programs do not require mechanical drill and memorization, learning becomes more interesting (Davydov, 1986). The formation of knowledge proceeds almost faultlessly from the very beginning; mistakes do not usually exceed 5% to 6%, and as a rule, they are corrected by students themselves (Talyzina, 1973, 1975/1981). The time spent on instruction is considerably less than in the traditional system of instruction (Talyzina, 1973, 1975/1981). The main features of knowledge acquired in the course of theoretical learning are a high level of mastery and maintenance, broad transfer, and intentional use by students. Students learn to answer why questions, to substantiate the way in which they have solved a problem, and to defend their way of solving problems and the results obtained (Aidarova, 1978; Davydov, 1986; Dyachenko, 1986; Elkonin & Davydov, 1966; Galperin, 1985; Pantina, 1957; Talyzina, 1975/1981; Volovich, 1967). An American researcher who studied Russian elementary school students after they had been taught mathematics for three years under the theoretical learning approach stated that they evidenced mathematical understanding typically not found among U.S. high school and university students . . . . [she] found it refreshing to observe the degree to w h i c h . . , children understood mathematics concepts at their most abstract level and were likewise able to generalize them to new and unfamiliar situations (Schmittau, 1993, p. 35)
. .

An even more important outcome of the systematic use of theoretical learning is a change in the students' general strategy of dealing with new problems. When encountering a new problem, they try from the very be3For a more detailed description of these programs in English, see Karpov (1995). January 1998 American Psychologist

ginning to find the essential characteristics of this problem without attending to its salient surface features (Davydov, Pushkin, & Pushkina, 1972). It is worthy of note that, as a rule, under traditional instruction, even adults demonstrate such a strategy only when dealing with problems in their area of expertise (Chi, Feltovich, & Glaser, 1981; Novick, 1988). Thus, theoretical learning is a very effective way of mediating students' cognitive processes. At the same time, the theoretical learning approach has at least one major shortcoming, namely, underestimation of the role of students' collaborative problem-solving activity. Students' activity under this approach is often organized as an individual activity under the supervision of a teacher (for several exceptions, see Rubtsov, 1987; Tsukerman, 1993). We discussed earlier Vygotsky' s (1981, 1983) idea of the importance of children's experience in regulating the behavior of others for the development of their own self-regulation and the empirical support for this idea. Students' collaborative problem-solving activity is the optimal situation in this respect, but this is not true of their individual activity under the supervision of a teacher. As Forman and Cazden (1995) mentioned, when discussing the advantage of peer interaction in comparison with teacher-child interaction, "children never give directions to teachers" (p. 344). Thus, the situation of theoretical learning demands that children change from being regulated by the teacher to self-regulation, skipping the intermediate step of regulating others' behavior. This is obviously not very advantageous for the development of students' self-regulation.

A Third Alternative
GDCL approaches by American researchers are highly relevant to the Vygotskian idea of metacognitive mediation. Students' collaborative problem-solving activity, which is a major component of these approaches, is the optimal situation for their transition from otherregulation to self-regulation. It can be reasonably assumed that it is the situation of collaborative problem solving that is mainly responsible for the successes of these approaches. However, the second major idea in GDCL approaches, namely, guided discovery, is in sharp contrast to the Vygotskian principle of teaching scientific concepts as a means of cognitive mediation in school. This idea also raises the following theoretical and practical concerns. Social progress, in general, comes about when every new generation receives ready-made the essence of knowledge accumulated by previous generations. Why, when dealing with particular students, should the previous generation require them to reinvent this knowledge, even if such reinvention is guided by the teacher? As Bruner (1966) noted, " C u l t u r e . . . is not discovered; it is passed on or forgotten" (p. 101). From the practical point of view, as we discussed earlier, in the course of guided discovery, students may not come to the correct scientific knowledge, or they may develop the wrong knowledge. January 1998 American Psychologist

GDCL advocates often draw an analogy between a group of students involved in guided discovery and a group of scientists solving a scientific problem. This analogy, in our opinion, has two weak points. First, the process of solving a scientific problem may take years, which is hardly acceptable in the case of school instruction. Second, research scientists possess methods of scientific research and analysis that were taught to them in special university courses or that they developed during many years of research experience. School students are unlikely to possess these methods. Actually, as we showed earlier, the principal method of "scientific research" that students use in the situation in which they need to solve a problem in the absence of necessary scientific knowledge is empirical learning, which is based on taking into consideration common salient features of phenomena but not their essential characteristics. The Vygotskian idea of cognitive mediation, elaborated by his Russian followers into the theoretical learning approach, is a good alternative to the idea of guided discovery and the idea of didactic teaching. Under this approach, students "should not be cast in the role of a research scientist under the pretext of 'inventing' already existing knowledge" (Schmittau, 1993, p. 34), as happens under guided discovery, nor should they be taught concrete skills or mere verbal definitions, as often happens under didactic teaching. They are taught methods of scientific analysis aimed at identifying and modeling the essential characteristics of the objects or events within certain classes. Having been internalized, these methods become cognitive tools that mediate students' independent problem solving. At this point, students' problem solving starts to involve discoveries, but these discoveries deal, in a deductive mode, with the application of a general method of scientific analysis to the analysis of a concrete phenomenon rather than with an attempt to derive a general scientific concept or a rule from concrete experiences. 4 Thus, within the theoretical learning approach, "the child as an independent learner is considered to be a result, rather than a premise of the learning process" (Kozulin, 1995, p. 121). The theoretical learning approach is very effective for classroom teaching and for facilitating students' cognitive development but is less effective in promoting their self-regulation because of underestimation of the role of collaborative problem-solving activity. Thus, it would seem advisable to develop an instructional procedure that incorporates the strong features of both GDCL approaches and the theoretical learning approach. Such a procedure, which could be designated "theoretical learn4 Wittrock (1966) differentiatedtwo types of discovery learning: inductive and deductive. "Discovery learning is commonly equated with inductivelearning where the subjectproceedsfrom the specificto the general. It is just as plausible to assume that the learnerbegins with a higher order generalization, from which he derives more specific conclusions and thus discovers answers and even generalizations" (p. 42). From the point of view of this differentiation,GDCL approaches involve inductive discoveries, and the theoretical learning approach involves deductiveones. 33

ing in a c o m m u n i t y o f l e a r n e r s , " w o u l d b e b a s e d on s u p p l y i n g students w i t h m e t h o d s o f scientific analysis in subject d o m a i n s l e a d i n g to d e v e l o p m e n t o f r e l e v a n t scientific k n o w l e d g e . Students w o u l d m a s t e r these m e t h ods b y using t h e m for s o l v i n g c o n c r e t e p r o b l e m s in the c o u r s e o f c o l l a b o r a t i v e activity in w h i c h they w o u l d take turns s o l v i n g a p r o b l e m and p l a n n i n g , m o n i t o r i n g , c h e c k ing, and e v a l u a t i n g the p r o c e s s o f the p r o b l e m solving. For e x a m p l e , i f o n e w i s h e d to t e a c h six-year-old c h i l d r e n h o w to w r i t e letters o f the alphabet, the p r e v i o u s l y des c r i b e d p r o g r a m o f Pantina (1957), w i t h s o m e m o d i f i c a tion, c o u l d be used. T h e c h i l d r e n w o u l d w o r k in g r o u p s , e a c h o f t h e m in turn a n a l y z i n g the letter to be c o p i e d and c o n s t r u c t i n g its m o d e l , w h e r e a s the others w o u l d be p l a n n i n g , m o n i t o r i n g , c h e c k i n g , and e v a l u a t i n g that child's performance. Such a procedure would successfully i m p l e m e n t b o t h t y p e s o f m e d i a t i o n specified in the w o r k s o f Vygotsky.
REFERENCES

Aidarova, L. I. (1978). Psikhologicheskie problemy obucheniya mladshikh shkolnikov russkomu yazyku [Psychological problems of teaching Russian language to elementary-school pupils]. Moscow: Prosveschenie. Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press. Anderson, J. R. (1993). Problem solving and learning. American Psychologist, 48, 35-44. Borkowski, J. G., & Kurtz, B. E. (1987). Metacognition and executive control. In J. G. Borkowski & J. D. Day (Eds.), Cognition in special children (pp. 123-152). Norwood, NJ: Ablex. Borkowski, J. G., & Muthukrishna, N. (1995). Learning environment and skill generation: How contexts facilitate regulatory processes and efficacy beliefs. In E Weinert & W. Schneider (Eds.), Memory performance and competencies (pp. 283-300). Hillsdale, NJ: Erlbaum. Brown, A. L. (1987). Metacognition, executive control, self-regulation, and other more mysterious mechanisms. In E E. Weinert & R. H. Kluwe (Eds.), Metacognition, motivation, and understanding (pp. 64-116). Hillsdale, NJ: Erlbaum. Brown, A.L., Bransford, J.D., Ferrara, R.A., & Campione, J.C. (1983). Learning, remembering, and understanding. In J. H. Flavell & E. M. Markman (Eds.), Handbook of child psychology: Vol. 3. Cognitive development (pp. 77-116). New York: Wiley. Brown, A. L., & Campione, J. C. (1990). Communities of learning and thinking: Or a context by any other name. Contributions to Human Development; 21, 108-126. Brown, A. L., & Campione, J. C. (1994). Guided discovery in a community of learners. In K. McGilly (Ed.), Classroom lessons: Integrating cognitive theory and classroom practice (pp. 229-270). Cambridge, MA: MIT Press. Brown, A. L., Campione, J. C., Reeve, R. A., Ferrara, R. A., & Palincsar, A. S. (1991). Interactive learning and individual understanding: The case of reading and mathematics. In L. T. Landsman (Ed.), Culture, schooling, and psychological development (pp. 136-170). Hillsdale, NJ: Erlbaum. Brown, A. L., & DeLoache, J. S. (1978). Skills, plans, and self-regulation. In R. S. Siegler (Ed.), Children thinking: What develops? (pp. 3-35). Hillsdale, NJ: Erlbaum. Bruer, J. T. (1993). Schools for thought: A science of learning in the classroom. Cambridge, MA: MIT Press. Bruner, J. S. (1966). Some elements of discovery. In L. S. Shulman & E. R. Keislar (Eds.), Learning by discovery: A critical appraisal (pp. 101-113). Chicago: Rand McNally. Chang-Wells, G. L. M., & Wells, G. (1993). Dynamics of discourse: Literacy and the construction of knowledge. In E. A. Forman, N.

Minick, & C. A. Stone (Eds.), Contexts for learning: Sociocultural dynamics in children's development (pp. 58-90). New York: Oxford University Press. Chatelanat, G., & Haywood, H. C. (1995). Bright Start, ou programme d'education cognitive pour jeunes enfants [Bright Start: Cognitive curriculum for young children]. In E Btichel (Ed.), L'gducation cognitive: Le dgveloppement de la capacitg d'apprentissage et son gvaluation (pp. 217-241). Neuch~tel, Switzerland: Delachaux & Niestl6. Chi, M. T H., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5, 121-152. Cobb, P., Wood, T., & Yackel, E. (1993). Discourse, mathematical thinking, and classroom practice. In E. A. Forman, N. Minick, & C. A. Stone (Eds.), Contexts for learning: Sociocultural dynamics in children's development (pp. 91-120). New York: Oxford University Press. Cobb, P., Wood, T., Yackel, E., Nicholls, J., Wheatley, G., Trigatti, B., & Perlwitz, M. (1991). Assessment of a problem-centered second grade mathematics project. Journal for Research in Mathematics Education, 22, 3-29. Cobb, P., Yackel, E., & Wood, T. (1992). A constructivist alternative to the representational view of mind in mathematics education. Journal for Research in Mathematics Education, 23, 2-33. Cognition and Technology Group at Vanderbilt. (1990). Anchored instruction and its relationship to situated cognition. Educational Researcher, 19(6), 2-10. Cognition and Technology Group at Vanderbilt. (1992). The Jasper experiment: An exploration of issues in learning and instructional design. Educational Technology Research and Development, 40, 6580. Cognition and Technology Group at Vanderbilt. (1994). From visual word problems to learning communities: Changing conceptions of cognitive research. In K. McGilly (Ed.), Classroom lessons: Integrating cognitive theory and classroom practice (pp. 157-200). Cambridge, MA: MIT Press. Davydov, V. V. (1966). Psikhologicheskie osobennosti "dochislovogo" perioda obucheniya matematike [Psychological characteristics of the "pre-number" period in teaching mathematics]. In D. B. Elkonin & V. V. Davydov (Eds.), Vozrastnye vozmozhnosti usvoeniya znanii (pp. 104-190). Moscow: Prosveschenie. Davydov, V. V. (1972). Vidy obobscheniya v obuchenii [Types of generalization in learning]. Moscow: Pedagogika. Davydov, V. V. (1986). Problemy razvivayuschego obucheniya [Problems of development-generating learning]. Moscow: Pedagogika. Davydov, V. V., Pushkin, V. N., & Pushkina, A. G. (1972). Zavisimost razvitiya myshleniya mladshikh shkolnikov ot kharaktera obucheniya [Dependence of the development of elementary-school students' thinking on the type of instruction]. Voprosy Psikhologii, 6, 124132. Dewey, J. (1902). The child and the curriculum. Chicago: University of Chicago Press. DiSessa, A.A. (1982). Unlearning Aristotelian physics: A study of knowledge-based learning. Cognitive Science, 6, 37-75. Dyachenko, O.M. (1986). Formirovanie sposobnosti k naglyadnomu modelirovaniyu pri oznakomlenii s detskoi khudozhestvennoi literaturoi [Formation of graphic modeling in the course of becoming acquainted with children's literature]. In L. A. Venger (Ed.), Razvitie poznavatelnykh sposobnostei v protsesse doshkolnogo vospitaniya (pp. 94-113). Moscow: Pedagogika. Elkonin, D. B. (1976). Kak uchit detei chitat [How to teach children to read]. Moscow: Znanie. Elkonin, D. B., & Davydov, V. V. (Eds.). (1966). Vozrastnye vozmozhnosti usvoeniya znanii [Age-dependent potentialities of acquiring knowledge]. Moscow: Prosveschenie. Feuerstein, R., Rand, Y., Hoffman, M. B., & Miller, R. (1980). Instrumental Enrichment: An intervention program for cognitive modifiability. Baltimore: University Park Press. Flavell, J. H. (1976). Metacognitive aspects of problem solving. In L. B. Resnick (Ed.), The nature of intelligence (pp. 231-235). Hillsdale, NJ: Erlbaum. Forman, E. A., & Cazden, C. B. (1995). Exploring Vygotskian perspectives in education: The cognitive value of peer interaction. In J. V.

34

J a n u a r y 1998 A m e r i c a n P s y c h o l o g i s t

Wertsch (Ed.), Culture, communication, and cognition: Vygotskian perspectives (pp. 323-347). New York: Cambridge University Press. Gagn6, R. M. (1985). The conditions of learning and theory of instruction. New York: Holt, Rinehart & Winston. Galperin, P. Y. (1957). Umstvennoe deistviye kak osnova forrnirovaniya mysli i obraza [Mental act as the basis for formation of thought and image]. Voprosy Psikhologii, 6, 58-69. Galperin, P. Y. (1966). K ucheniyu ob interiorizacii [On the concept of internalization]. Voprosy Psikhologii, 6, 25-32. Galperin, P. Y. (1969). Stages in the development of mental acts. In M. Cole & I. Maltzman (Eds.), A handbook of contemporary Soviet psychology (pp. 34-61). New York: Basic Books. Galperin, P.Y. (Ed.). (1977). Upravlyaemoe formirovanie psikhicheskikh protsessov [Guided formation of the mental processes]. Moscow: Izdatelstvo MGU. Galperin, P.Y. (1985). Metody obucheniya i umstvennoe razvitie rebenka [Methods of instruction and the child's mental development]. Moscow: Izdatelstvo MGU. G alperin, P. Y., & Talyzina, N. E (1961). Formation of elementary geometrical concepts and their dependence on directed participation by the pupils. In N. O'Connor (Ed.), Recent Soviet psychology (pp. 247-272). New York: Liveright. Galperin, P. Y., & Talyzina, N. E (Eds.). (1968). Zavisimost obucheniya ot tipa orientirovochnoi deyatelnosti [Dependence of instruction on the type of orienting activity]. Moscow: Izdatelstvo MGU. Galperin, P. Y., & Talyzina, N. E (Eds.). (1972). Upravlenie poznavatelnoi deyatelnostyu uchaschikhsya [Guidance of cognitive activity of students]. Moscow: Izdatelstvo MGU. Galperin, P. Y., Zaporozhets, A. V., & Elkonin, D. B. (1963). Problemy formirovaniya znanii i umenii u shkolnikov i novye metody obucheniya v shkole [The problems of formation of knowledge and skills in school-children and the new methods of instruction at school]. Voprosy Psikhologii, 5, 61-72. Greenberg, K.H. (1994). COGNET: Teacher handbook. (Available from the COGNET Administrative Staff, University of Tennessee, 445 Claxton Addition, Knoxville, TN 37996) Hall, G. S. (1907). Aspects of child life in education. Boston: Ginn. Haywood, H. C. (1996). Educaci6n cognitiva temprana: Una clave para el exito escolar [Cognitive early education: A key to school success]. In S. Molina-Garcia & M. Fandos-Igado (Eds.), Educacidn cognitiva (pp. 167-192). Zaragoza, Spain: MIRA. Haywood, H. C., Brooks, P. H., & Burns, S. (1986). Stimulating cognitive development at developmental level: A tested, nonremedial preschool curriculum for preschoolers and older retarded children. In M. Schwebel & C. Maher (Eds.), Facilitating cognitive development: Principles, practices, and programs (pp. 127-147). New York: Haworth Press. Haywood, H.C., Brooks, P.H., & Burns, S. (1992). Bright Start: Cognitive curriculum for young children. Watertown, MA: Charlesbridge. Heller, J. I., & Gordon, A. (1992). Lifelong learning. Educator, 6(1), 4-19. Karpov, Y. V. (1995). L. S. Vygotsky as the founder of a new approach to instruction. School Psychology International, 16, 131-142. Kluwe, R.H. (1987). Executive decisions and regulation of problem solving behavior. In E E. Weinert & R. H. Kluwe (Eds.), Metacognition, motivation, and understanding (pp. 31-63). Hillsdale, NJ: Erlbaum. Kozulin, A. (1995). The learning process: Vygotsky's theory in the mirror of its interpretations. School Psychology International, 16,
117-129.

Leontiev, A. N. (1983). Ovladenie uchaschimisya nauchnymi poniatiyami kak problema pedagogicheskoi psikhologii [Mastering scientific concepts by students as a problem of educational psychology]. In A. N. Leontiev, lzbrannye psikhologicheskie proizvedeniya, Tom 1 (pp. 324-347). Moscow: Pedagogika. (Original work published 1935) Luria, A. R. (1961). The role of speech in the regulation of normal and abnormal behavior. Oxford, England: Pergamon Press. Novick, L. R. (1988). Analogical transfer, problem similarity, and expertise. Journal of Experimental Psychology: Learning, Memory, and Cognition, 14, 510-520.

Palincsar, A. S., & Brown, A. L. (1984). Reciprocal teaching of comprehension-fostering and comprehension-monitoring activities. Cognition and Instruction, 1, 117-175. Palincsar, A. S., Brown, A. L., & Carnpione, J. C. (1993). First-grade dialogues for knowledge acquisition and use. In E. A. Forman, N. Minick, & C. A. Stone (Eds.), Contexts for learning: Sociocultural dynamics in children's development (pp. 43-57). New York: Oxford University Press. Panofsky, C. P., John-Steiner, V., & Blackwell, P. J. (1992). The development of scientific concepts and discourse. In L.C. Moll (Ed.), Vygotsky and education: Instructional implications and applications ofsociohistorical psychology (pp. 251-267). New York: Cambridge University Press. Pantina, N. S. (1957). Formirovanie dvigatelnogo navyka pisma v zavisimosti ot tipa orientirovki v zadanii [Formation of writing skills depending on the type of task orientation]. Voprosy Psikhologii, 4, 117-132. Paour, J.-L., C~be, S., & Haywood, H. C. (1997). Learning to learn in preschool education: Effects on later school achievement. Manuscript submitted for publication. Piaget, J. (1959). The language and thought of the child. London: Routledge and Kegan Paul. (Original work published 1923) Piaget, J. (1970). Genetic epistemology. New York: Columbia University Press. Rubtsov, V. V. (1987). Organizatsiya i razvitie sovmestnykh deistvii u detei v protsesse obucheniya [Organization and development of joint actions in children in the course of instruction]. Moscow: Pedagogika. Rumbaugh, D. M. (1977). Language learning by a chimpanzee: The LANA project. New York: Academic Press. Salmina, N. G., & Sokhina, V. P. (1975). Obuchenie matematike v nachalnoi shkole [Teaching mathematics in elementary school]. Moscow: Pedagogika. Savage-Rumbaugh, E. S. (1986). Ape language: From conditioned response to symbol. New York: Columbia University Press. Savage-Rumbaugh, E. S. (1996). Kanzi. New York: Wiley. Schmittau, J. (1993). Vygotskian scientific concepts: Implications for mathematics education. Focus on Learning Problems in Mathematics, 15(2&3), 29-39. Schoenfeld, A. H. (1985). Mathematical problem solving. New York: Academic Press. Schoenfeld, A. H. (1992). Learning to think mathematically: Problem solving, metacognition, and sense making in mathematics. In P. H. Grouws (Ed.), Handbook of research on mathematics teaching: A project of the National Council of Teachers of Mathematics (pp. 334-370). New York: Macmillan. Semenyuk, L.A. (1970). Formirovanie umstvennykh deistvii v protsesse obucheniya istorii [Formation of mental actions in the course of teaching history]. In N. E Talyzina (Ed.), Kprobleme upravleniya obucheniem i vospitaniem (pp. 40-64). Moscow: Izdatelstvo MGU. Talyzina, N. E (1973). Psychological bases of programmed instruction. Instructional Science, 2, 243-280. Talyzina, N. E ( 1981). The psychology of learning. Moscow: Progress. (Original work published 1975) Tsukerman, G. A. (1993). Vidy sotrudnichestva v obuchenii [Types of cooperation in education]. Tomsk, Russia: Peleng. Tudge, J. (1992). Vygotsky, the zone of proximal development, and peer collaboration: Implications for classroom practice. In L.C. Moll (Ed.), Vygotsky and education: Instructional implications and applications of sociohistorical psychology (pp. 155-172). New York: Cambridge University Press. Venger, L.A. (Ed.). (1986). Razvitie poznavatelnykh sposobnostei v protsesse doshkolnogo vospitaniya [Development of cognitive abilities in the course of preschool education]. Moscow: Pedagogika. Volovich, M. B. (1967). Formirovanie obschikh priemov raboty s ponyatiyami [Formation of general methods of dealing with concepts]. Unpublished doctoral dissertation, Moscow State University, Moscow, Russia. Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

J a n u a r y 1998 A m e r i c a n P s y c h o l o g i s t

35

Vygotsky, L. S. (1981). The genesis of higher mental functions. In J. V. Wertsch (Ed.), The concept of activity in Soviet psychology (pp. 144188). Armonk, NY: Sharpe. Vygotsky, L. S. (1983). Istoriya razvitiya vysshikh psikhicheskikh funktsii [The history of the development of higher mental functions]. In L. S. Vygotsky, Sobranie sochineniL Tom 3 (pp. 6-328). Moscow: Pedagogika. Vygotsky, L. S. (1988). Thought and language. Cambridge, MA: MIT Press. (Original work published 1934) Wells, G., Chang, G. L., & Maher, A. (1990). Creating classroom

communities of literate thinkers. In S. Sharan (Ed.), Cooperative learning: Theory and research (pp. 95-121). New York: Praeger. Wittrock, M. C. (1966). The learning by discovery hypothesis. In L. S. Shulman & E. R. Keislar (Eds.), Learning by discovery: A critical appraisal (pp. 33-75). Chicago: Rand McNally. Zaporozhets, A. V. (1986). Razvitie myshleniya [The development of thinking]. In A. V. Zaporozhets, Izbrannye psikhologicheskie trudy, Tom 1 (pp. 200-215). Moscow: Pedagogika. Zhurova, L. E. (1978). Obuchenie gramote v detskom sadu [Teaching reading and writing at the kindergarten]. Moscow: Pedagogika.

36

J a n u a r y 1998 A m e r i c a n P s y c h o l o g i s t