Tomoo Tsukamoto & Naoko Aoki, 1995
1. Introduction
In this article I would like to describe how researchers in human cognition have changed their view of strategies in instruction/learning. The article will, I hope, also clarify how their view of learning has changed.
I shall first discuss how the concept of strategy changed our view of human capabilities. The argument over whether human intelligence is hereditary or determined by environment has had a long history. However, both positions were based on the assumption that there is no way one can intentionally improve one's own intelligence. Development of the concept of strategy enabled us to believe that one's intelligence is determined by what one does, from which emerged totally new conceptions of learners and education.
Secondly, I shall look at effects and limitations of teaching strategies. The discovery of strategies quite naturally led teachers and researchers to the idea that anyone can become a good learner by acquiring good strategies. In pointing out how strategy training research has come to recognize its own limitations, I would like to highlight problems relating to the concept of 'the learner' which strategy training research has been based on.
Finally, I shall briefly introduce some of the most recent research on learning within psychology. In this section the reader will find that strategies are hardly mentioned. This reflects the fact that recent research in learning gives only a very limited role to strategies.
2. The view of learning accompanying the concept of 'strategy'
This section deals with the views of human intelligence and learning accompanying the concept of strategy. For the sake of brevity, I shall take two models of human memory as examples, one implying a fixed human capacity and the other plasticity.
2.1 . Memory storage model
The memory storage model is such a popular concept that references to it may be found in any textbook of cognitive psychology (e.g. Mayer, 1981). The storage model assumes three different types of storage in human memory: sensory storage, short-term memory, and long-term memory.
One characteristic of this model is that the capacity of each type of storage can be measured quantitatively. Let me describe some methods of experiment to show how accurate measurements are possible. To investigate the capacity of sensory storage, for example, Sperling (1960) devised the following method:
(1) Present four alphabetical symbols per line, in three lines, to a subject;
(2) Sound a beep (high, medium, or low in tone), several hundred milliseconds after the symbols have disappeared;
(3) On hearing a beep of high, medium or low tone, the subject is asked to reproduce the top, middle or botton line, respectively.
The capacity of sensory storage is measured by changing the number of alphabetical symbols per line. By the same token, it is possible to measure how long a piece of information can be retained in the sensory storage by changing the interval between (disappearance of) the stimulus and the beep.
Such experiments showed that there are individual differences in the basic architecture of memory, including memory capacity and length of retention. Based on these results, attempts were made to explain differences in human intelligence with reference to architectural differences. Studies of the relationship between language ability and the speed of information transfer from long-term to short-term memory accorded with this paradigm. Transfer speed can be measured by differences of response time, as in the following two tests (Hunt, Lunneborg & Lewis, 1975) :
[Physical Matching Test]
Subjects are instructed to press a "yes" button if two symbols on a screen are identical. If not, they press a "no" button. For example, if the two symbols are "A" and "A", the subjects are supposed to press the "yes" button. If they see "A" and "B" or "A" and "a", they press the "no" button. To correctly answer this test, one needs to identify the information, i.e. the shape of a letter, in short term memory.
[Name Matching Test]
If two symbols on a screen represent the same letter, whether they be small letters or capital letters, subjects are asked to press a "yes" button. If different letters are presented, subjects press a "no" button. For example, if the symbols are either "A" and "A" or "A" and "a", the answer is "yes". If the symbols are "A" and "B", the answer is "no". In this test, when one symbol is written in a small letter and the other in a capital letter, subjects need to transfer the necessary information from long-term memory.
As name matching tests require a transfer of information from long-term memory, the difference in response time between a physical matching test and a name matching test corresponds to the transfer speed from long-term memory. To investigate relationships between this transfer speed and first language proficiency, the speed of subjects with high language ability and those with low ability were compared. The results indicated that people with high language ability transfer information from long term memory with higher speed than those with low language ability, implying that a higher-level ability such as language may be determined by a basic architectural characteristic such as memory (Hunt, Lunneborg & Lewis, ibid. ).
2.2 Levels of Processing Model
The processing level model was presented in counter-argument to the memory storage model. Proponents of the processing level model claim that the amount of information stored in memory is determined by how incoming information is processed (Craik & Lockhart, 1972). To clarify, let us consider the following tests, involving form-processing, phoneme-processing, and meaning-processing (Craik & Tulving, 1975) :
[Form-processing Test]
Subjects are asked to determine whether a word is written in capital letters. They need to identify the form of a word to correctly answer the test.
[Phoneme-processing Test]
Subjects are asked to determine whether a word presented in the test rhymes with a certain word. In order to do so correctly, they need to process the information phonemically.
[Meaning-processing Test]
Subjects are asked to determine whether a word presented in the test fits in a given sentence. In order to do so correctly, they need to process the meaning of the word.
It is claimed that the depth of processing required is smallest with form processing, and largest with meaning-processing. After the three kinds of tests, subjects were given a recognition test without any advance warning (i.e. they were asked to say whether given words had appeared in the previous tests). It was found that responses were more accurate with words processed during the tests on deeper levels. This seems to show that storage in memory is determined by the level of processing.
2.3 Views of Learning Implied by the Two Models
The basic difference between the two models of human memory described above lies in whether or not human memory is considered to be 'plastic' and controllable. In the storage model, it is assumed that the performance of the basic architecture, including storage capacity and transfer speed, does not change, at least for a considerable period of time. Learning efficiency is thus considered to remain fixed. The more you want to learn, the more time you need to spend on learning. The efficiency of learning is not expected to be improvable according to the learner's will.
In the levels of processing model, on the other hand, the efficiency of learning can be improved by changing the method of processing. Learning becomes more efficient with deeper processing, and learners can choose how deep they would like to go in their processing of incoming information. This means that human cognition is controllable and that it is possible to learn better by acquiring better strategies.
This is not just a matter of models of human memory. Underlying these two models are totally different views of human capacity. If human intelligence is determined by such fixed items as intelligence quotients or memory capacity, one can do nothing about one's intelligence. Improving academic performance, if considered possible at all, is seen as requiring an enormous amount of effort. The role of the teacher implied is one of teaching students within their current capacities. This view has led to the coining of humiliating terms such as "over achievement" and "under-achievement." (it might be argued that teachers and researchers should refer to their own "under-estimation" or "over-estimation" instead, when a student performs better or worse than they had expected!) The concept of strategy, however, implies that one can improve one's cognitive processes and make them more efficient. Put more bluntly, one can become smarter. The latter view of human capacity and implied concept of 'the learner' have quite naturally influenced our view of education. I shall discuss in the following section what kind of view of education and instructional activities was induced by an increased emphasis on the learner as 'user of strategies'.
3. Strategy instruction and its limitations
The discovery of strategy use stimulated research into what kind of strategies we use and, in due course, into how teachers can teach strategies in order to improve efficiency of learning. I would like to summarize here the development of research into the teaching of writing in order to show the limitations of strategy instruction and the concept of 'the learner' it entails.
3.1. Strategy Instruction
Scardamalia, Bereiter & Steinbach (1984) found strategy instruction to be effective in improving writing. Within the framework of a given theme, poor writers tend to just jot down various topics as these come to their minds. Good writers, on the other hand, carefully examine not only the content but the rhetoric of their composition : they go back and forth between the content and the rhetoric as they write. Based on this observation, the researchers used cue cards to teach poor writers a strategy for reexamining the content of their writing in terms of rhetoric. After the instruction, unskillful writers' compositions showed some improvement.
Bereiter and Scardamalia (1989), however, pointed out some problems they noticed after close examination of poor writers' behaviors, and suggested that students' strategies might in fact hinder their learning. Bereiter and Scardamalia had thought that cue cards would teach the poor writers a problem-solving approach to writing. The students themselves, however, used cue cards as a means to pad their compositions to meet the required length. School compositions in higher grades demand more pages as well as better quality in content. Simply writing what comes to one's mind does not make for a long enough composition. Students who had had a hard time fulfilling the length requirements in fact used the rhetorical hints on the cue cards as pointers for indiscriminate jotting, without examination of content.
The basic concept of strategy instruction involves a comparison of poor learners with good learners, and having poor learners learn the strategies used by good learners. Here, poor learners are defined as learners who do not do something. However, as Bereiter and Scardamalia (1989) observed, and as John Holt has described in his classic contribution, How Children Fail (Holt, 1964), poor learners do employ what they believe to be reasonable strategies. It is simply not true that poor learners do nothing : they do quite a lot of things to get by in the classroom. We have no way of knowing what poor learners are liable to do with the strategies of the good learner. A strategy is a powerful weapon, but it can be lethal if used improperly. It may be used for better learning, or for cutting corners....
How should we conceptualize learning, then, if the equation "poor learner + strategy = good learner" or, more plainly, "good learner - strategy = poor learner," does not hold true?
3.2 Intentional Learning
The crucial point in teaching strategies is what kind of objectives learners have in their mind when learning the target strategies. The teacher sets a target for the class so as to help them acquire knowledge or skills. The students, however, do not necessarily share the teacher's intention. Bereiter and Scardamalia (1989) give the example of a directed reading lesson to demonstrate this discrepancy between teacher and student expectations. In the lesson, students read part of a text chosen by the teacher, and respond to the teacher's questions. What kind of cognitive activities are the students engaged in? It seem clear, on the surface, that training in reading skills is involved. If one examines the lesson more carefully, however, one finds that all the important reading activities are done by the teacher. The teacher picks out what is important in the text, determines what is difficult and what is easy to understand, points out what inference is more important than others, and evaluates the correctness of interpretation. The major role of the students is to answer the teacher's questions. Thus, what is acquired in this lesson is likely to be no more than the skills or strategies needed in answering questions. As long as students consider the lesson as a task, their objective is likely to remain answering questions asked by someone else, and they will develop strategies to meet this objective.
In response to this kind of observation, Bereiter and Scardamalia (1989) developed the idea of "intentional learning". If learners consider learning to be a task, no strategies can make learning more efficient. These only serve to drive learners away from real learning. The point is: how can we change learners' view of "learning as task" to one of "learning as construction of knowledge" (i.e. "intentional learning")?
4. What is 'Learning'?
In the course of attempting to focus students' attention on the construction of knowledge itself in order to foster intentional learning, Bereiter, Scardamalia and their associates have come to claim that construction of knowledge is a social and not just an individual activity (Scardamalia, Bereiter and Lamon, 1994). "What is learning if it is not an individual activity?" is the key question which cognitive psychology is facing today. I shall briefly summarize some theoretical and practical studies in this area.
Situated Learning (Lave & Wenger, 1991) might be considered as representative of theoretical work on this topic. The work of Lave and Wenger is part of the reconstruction movement in psychology which started off from studies in "everyday cognition" or comparative culture studies (cf. Rogoff & Lave, 1984). Space does not allow me to give a detailed description of their claims and how these have developed. The reader is referred to Lave (1988) and Lave & Wenger (1991). Instead, I shall focus here on the view of learning criticized in these books.
Psychologists have tended to treat learning as a phenomenon occuring in individual minds, and have considered what is acquired as a result of learning only on the levels of knowledge and skills. This has led them to consider human beings as bundles of specific abilities. Specialists have further divided each kind of ability, e.g. linguistic or mathematical, into sub-categories. It was tacitly assumed that teaching mathematics, for example, means working on students' mathematical ability in order to change it. For this reason, typical experiments in instruction/learning have measured outcomes of instruction by giving tests to check the acquisition of particular knowledge or skills after a particular treatment. Learning was thought to consist in becoming able to answer a question which one was not able to answer previously, no more and no less.
In the paradigm described above, researchers never questioned what significance it might have for learners to be taught certain topics in mathematics. Neither was consideration given to how "subjects" perceived the act of participating in such instruction/learning experiments. Put bluntly, researchers failed to consider human subjects as "people". It did not matter to them at all who subjects were, so long as "variables" were well-controlled.
Cognitive psychologists nowadays focus their attention not so much on what and how we teach as on how the traditional view of learning can be outgrown and a new framework constructed. In recent years they have emphasised that we should view human beings in their entirety, and that learning should be seen as a form of enculturation which occurs in the context of each learner's relationship with the surrounding community. Learning is becoming a topic in sociology and anthropology rather than in psychology (Saeki, Fujita & Sato, 1995). Strategies are hardly discussed in this context.
One practical study worth mentioning here might be the work being done by the Cognition and Technology Group at Vanderbilt (cf. 1994). This group is involved in a large-scale experimental project teaching mathematics in a multimedia environment. They describe the change in their project as proceeding from "the curricular elaboration model" via "the classroom restructuring model" to "the learning communities model." This beautifully captures the general trend of change in studies of educational practice by cognitive psychologists. Like Bereiter and Scardamalia, the Vanderbilt researchers originally tried to apply the insights gained from studies of human understanding processes to the improvement of instruction and curriculum, then gradually shifted their attention to the reformation of classroom tasks, which in turn forced them to focus on the factor of learning communities. Recent years have seen a range of research focused on key words such as "culture", "practice" and "community" (e.g. Lampert, 1986; Healy, 1993; Rosebery, Warren & Conant,1992).
5. Concluding (personal) remarks
The research described in the previous section has led to a drastic alteration in my own attitude in studying what understanding of natural science involves. As a graduate student I used to think that learning science involves acquiring scientific knowledge and skills in order to solve scientific problems (cf. Murayama & Miyashita, 1986; Suzuki et.al., 1989). Lampart (1986), however, showed how we can do mathematics with primary school children through multiplication, in a way which is totally different from teaching the "right" procedures for multiplication. Rosebery et.al. (1992) demonstrated that learning was not just a matter of acquiring knowledge or a skill by focusing on a change in the use of the word "I" by learners, this never having been previously considered to be a legitimate criterion of learning outcome. These experimental educational practices forced me to rethink my own view of learning, and I have managed to grow out of the conventional curricular view of learning and teaching by reflecting not on learning by "subjects" but on my own learning of psychology (Murayama 1995a; 1995b). Of course, this is not the end of the story. As the discussion of the three topics I have focused on above shows, our understanding of learning is steadily changing (and hopefully improving).
References :
Bereiter, C & Scardamalia, M.(1989). Intentional learning as a goal of instruction. in Resnick, L. (Ed.). Knowing, Learning, and Instruction. Hillsdale, NJ.: LEA.
Craik, F.I.M. & Lockhard, R.S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior. 11. 671-684.
Craik, F.I.M. & Tulving, E. (1975). Depth of processing and the retention of word in episodic memory. Journal of Experimental Psychology: General. 104. 268 294.
Healy, C.C. (1993). Discovery courses are great in theory, but.... in Schwartz, J.L., Yershalemy, M. & Wilson, B. (Eds.), The Geometric Supposer: What is it a case of? Hillsdale, NJ: Lawrence Erlbaum Associates.
Holt, J. (1964). How Children Fail. New York: Pitman.
Hunt, E., Lunneborg, C. & Lewis, J. (1975). What does it mean to be high verbal? Cognitive Psychology. 7. 194-227.
Lampert, M.(1986). Knowing, doing and teaching multiplication. Cognition and Instruction. 3(4). 305-342.
Lave, J. (1988). Cognition in Practice: Mind, mathematics, and culture in everyday life. Cambridge: Cambridge University Press.
Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate peripheral participation. New York: Cambridge University Press.
Mayer, R.E. (1981). The Promise of Cognitive Psychology. CA: W.H.Freeman and Company.
Murayama, I. (1995a). Kagaku Kyoiku (Science Education). in Nihon Jido Kenkyu jo (Ed.). Jido Shinri-gaku no Shinpo: 1994 (Advances in child psychology: 1994). Tokyo: Kaneko Shobo.
-----------------. (1995b). Kagaku wa Ika ni shite Manabareru ka (How science is learned). in Saeki, Y., Sato, M. & Fujita, H. (Eds.). Series Manabi to Bunka. vol. 3. Kagaku Suru Bunka (Learning and Culture Series. vol. 3. Cultures that do science). Tokyo: The University of Tokyo Press.
Murayama, I. & Miyashita, T. (1986). Kagaku ni okeru Mondai Kaiketsu to Rikai (Problem Solving and Understanding in Science). in Azuma, H. & Hatano, G. (Eds.). Iwanami Koza Kyoiku no Hoho. vol. 6. Kagaku to Gijutu no Kyoiku (The Iwanami Course in Educational Methodology. vol. 6. Education of Science and Technology). Tokyo: Iwanami Shoten.
Rogoff, B. & Lave, J. (1984). Everyday Cognition. Cambridge, MA: Harvard University Press. Rosebery, A.S., Warren, B., & Conant, F.R. (1992). Appropriating scientific discourse: Findings from language minority classrooms. The Journal of the Learning Sciences. 2(1). 61-94.
Saeki, Y., Fujita, M. & Sato, H. (Eds.). (1995). Series Manabi to Bunka. vol. 1. Manabi e no Sasoi (Learning and Culture Series. vol.1. Invitation to Learning). Tokyo: The University of Tokyo Press.
Scardamalia, M., Bereiter, C. & Lamon, M. (1994). The CSILE Project: Trying to bring the classroom into World 3. McGilly, K. (Ed.). Classroom Lessons: Integrating cognitive theory and classroom practice. Cambridge, MA.: The MIT Press.
Scardamalia, M., Bereiter, C. & Steinbach, R.(1984). Teachability of reflective processes in written composition. Cognitive Science. 8.173-190.
Sperling, G. (1960). The information available in brief visual presentation. Psychological Monograph 74.
Suzuki, H., Suzuki, T., Murayama, I. & Sugimoto, T. (1989). Kyoka Rikai no Ninchi Shinrigaku (Cognitive Psychology for the Better Understanding of School Subjects). Tokyo: Shin'yosha.
The Cognition and Technology Group at Vanderbilt. (1994). From visual world problems to learning communities: changing conceptions of cognitive research. in McGilly, K. (Ed.). Classroom Lessons: Integrating cognitive theory and classroom practice. Cambridge, MA.: The MIT Press.
Translation by Tomoo Tsukamoto and Naoko Aoki
Originally published in Learning Learning 2/3 (December, 1995), 7-12. Tomoo Tsukamoto & Naoko Aoki, 1995
Isao Murayama
Back to the Learning Learning index page
Back to the LD N-SIG home page