In 1964 I moved from one world to another. For the previous five years I had lived in Alpine villages near Geneva, Switzerland, where I worked with Jean Piaget. The focus of my attention was on children, on the nature of thinking, and on how children become thinkers. I moved to MIT into an urban world of cybernetics and computers. My attention was still focused on the nature of thinking, but now my immediate concerns were with the problem of artificial intelligence: How to make machines that think?
Two worlds could hardly be more different. But I made the transition because I believed that my new world of machines could provide a perspective that might lead to solutions to problems that had eluded us in the old world of children. Looking back I see that the cross-fertilization has brought benefits in both directions. For several years now Marvin Minsky and I have been working on a general theory of intelligence (called “The Society Theory of Mind”) which has emerged from a strategy of thinking simultaneously about how children do and how computers might think.
Minsky and I, of course, are not the only workers to have drawn on the theory of computation (or information processing) as a source of models to be used in explaining psychological phenomena. On the contrary, this approach has been taken by such people as Warren McCulloch, Allen Newell, Herbert Simon, Alan Turing, Norbert Wiener, and quite a number of younger people. But the point of departure of this book is a point of view—first articulated jointly with Minsky—that separates us quite sharply from most other members of this company: that is to say, seeing ideas from computer science not only as instruments of explanation of how learning and thinking in fact do work but also as instruments of change that might alter, and possibly improve, the way people learn and think.
The book grew out of a project designed to explore this concept by giving children access to “the best of computer science” including some of its best technology and some of its best ideas. At the heart of the project was the creation of a children’s learning environment in the same building that houses MIT’s Artificial Intelligence Laboratory and Laboratory for Computer Science (Project MAC). We hoped that by bringing children and people interested primarily in children into this world of computers and computerists, we would create conditions for a flow of ideas into thinking about education.
I shall not try to describe all that happened in the course of this project or all that was learned from it, but I shall concentrate on some personal reflections. Readers who want to know more about the project itself will find pointers to other publications in the notes at the end of the book.
The project is really an experiment in cultural interaction. It set out to grow a new “education culture” in an environment permeated with a particular form of “computer culture.” Too many people were involved for me even to know all their names. The interchanges of ideas took place much more in conversations in the quiet of after-midnight hours (for this is a computer culture that does not respect the conventional clock cycles) than in organized seminars or written papers. In early drafts I attempted to chronicle the growth of the culture. But it proved too difficult, and in the end I wrote the book in a very personal style. This has a certain advantage in allowing me to give freer reign to my own personal interpretations of ideas and incidents that other participants might well see very differently. I hope that it does not obscure my sense of belonging to a communtity and of expressing a set of shared ideas. I regret that space does not permit me to show how some of these ideas have been picked up by others and elaborated into much more advanced forms.
Marvin Minsky was the most important person in my intellectual life during the growth of the ideas in this book. It was from him that I first learned that computation could be more than a theoretical science and a practical art: It can also be the material from which to fashion a powerful and personal vision of the world. I have since encountered several people who have done this successfully and in an inspirational way. Of these, one who stands out because he has so consistently turned his personal computational vision to thinking about children is Alan Kay. During the whole decade of the 1970s, Kay’s research group at the Xerox Palo Alto Research Center and our group at MIT were the only American workers on computers for children who made a clear decision that significant research could not be based on the primitive computers that were then becoming available in schools, resource centers, and education research laboratories. For me, the phrase “computer as pencil” evokes the kind of uses I imagine children of the future making of computers. Pencils are used for scribbling as well as writing, doodling as well as drawing, for illicit notes as well as for official assignments. Kay and I have shared a vision in which the computer would be used as casually and as personally for an even greater diversity of purposes. But neither the school computer terminal of 1970 nor the Radio Shack home computer of 1980 have the power and flexibility to provide even an approximation of this vision. In order to do so, a computer must offer far better graphics and a far more flexible language than computers of the 1970s can provide at a price schools and individuals can afford.
In 1967, before the children’s laboratory at MIT had been officially formed, I began thinking about designing a computer language that would be suitable for children. This did not mean that it should be a “toy” language. On the contrary, I wanted it to have the power of professional programming languages, but I also wanted it to have easy entry routes for nonmathematical beginners. Wallace Feurzeig, head of the Educational Technology Group at the research firm of Bolt Beranek and Newman, quickly recognized the merit of the idea and found funding for the first implementation and trial of the language. The name LOGO was chosen for the new language to suggest the fact that it is primarily symbolic and only secondarily quantitative. My original design of the language was greatly improved in the course of discussions with Daniel Bobrow, who had been one of the first graduate students in the MIT Artificial Intelligence group, Cynthia Solomon, and Richard Grant, all of whom were working at that time at Bolt Baranek and Newman. Most subsequent development of the LOGO language, which has gone through several rounds of “modernization,” took place at MIT. Of the very many people who contributed to it, I can list only a few: Harold Abelson, Bruce Edwards, Andrea diSessa, Gary Drescher, Ira Goldstein, Mark Gross, Ed Hardeback, Danny Hillis, Bob Lawler, Ron Lebel, Henry Lieberman, Mark Miller, Margaret Minsky, Cynthia Solomon, Wade Williams, and Terry Winograd. For many years Ron Lebel was the chief systems programmer in charge of LOGO development. But the people who worked directly on LOGO form only the tip of an iceberg: The influence of the MIT community on LOGO went much deeper.
Our Artificial Intelligence Laboratory has always been near the center of a movement, strongly countercultural in the larger world of computers, that sees programming languages as heavily invested with epistemological and aesthetic commitments. For me this “Whorfian” view has been best articulated in the work of three computer scientists who were graduate students at the time LOGO was in formation: Carl Hewitt, Gerald Sussman, and Terry Winograd. But it goes back to the founders of the MIT Artificial Intelligence group, Marvin Minsky and John McCarthy, and owes much to the tradition of “hackers” of whom I feel most directly the influence of William Gosper and Richard Greenblatt. In the cultural atmosphere created by such people it was as unacceptable for children to enter the computer culture by learning computer languages such as BASIC as it would be to confine their access to English poetry to pidgin English translations.
I have always considered learning a hobby and have developed many insights into its nature by cultivating a sensitivity to how I go about doing it. Thus, I have perhaps engaged in deliberate learning of a wider range of material than most people. Examples of things I have learned in this spirit include chapters of science (like thermodynamics), reading Chinese characters, flying airplanes, cooking in various cuisines, performing circus arts such as juggling, and even two bouts of living for several weeks with distorting spectacles (on one occasion left-right reversing glasses, on the other a rather complex prismatic distortion of the visual field). Part of what I found so attractive about the artificial intelligence community was a shared interest in this approach to using one’s self as a source of insight into psychological processes and a particular interest in observing oneself engaged in skilled activities. Here again I owe debts to many people and am able to single out only those whose contributions were most salient: Howard Austin, Jeanne Bamberger, Ira Goldstein, Bob Lawler, Gerald Sussman, and the graduate students who took part in my “loud thinking seminars” where such methods were explored. My approach to “loud thinking” acquired greater sophistication during a period of collaboration with Donald Schon and Benson Snyder and in interaction with a number of psychologists including Edith Ackermann, Daniel Bobrow, Howard Gruber, Annette Karmiloff-Smith, and Donald Norman.
All these influences entered into the emergence of a learning/teaching methodology in the computational environments we were building for children. The person closest to me in this work was Cynthia Solomon. As in the case of Marvin Minsky, my collaboration with her was so close over so long a period that I find it impossible to enumerate the substantial contributions she made. Solomon was also the first to develop an intellectually coherent methodology for training teachers to introduce children to computers and is still one of the few people to have approached this problem with the seriousness it deserves.
Many people contributed ideas about teaching children LOGO. Ira Goldstein undertook the difficult problem of developing a theoretical framework for the instructional process and was followed in this work by Mark Miller. Others approached teaching in a more pragmatic spirit. Special contributions have been made by Howard Austin, Paul Goldenberg, Gerianne Goldstein, Virginia Grammar, Andree Green, Ellen Hildreth, Kiyoko Okumura, Neil Rowe, and Dan Watt. Jeanne Bamberger developed methods for using LOGO in musical learning and in increasing teachers’ sensitivity to their own thinking.
A central idea behind our learning environments was that children would be able to use powerful ideas from mathematics and science as instruments of personal power. For example, geometry would become a means to create visual effects on a television screen. But achieving this often meant developing new topics in mathematics and science, an enterprise that was possible only because we were working within an institution rich in creative mathematical talent. The task is of a new kind: It consists of doing what is really original research in mathematics or science but in directions chosen because they lead to more comprehensible or more learnable forms of knowledge and not for the kinds of reasons that typically motivate mathematical research. Many students and faculty members at MIT contributed to this work, but two stand out as professionals in the area: Harold Abelson, a mathematician, and Andrew diSessa, a physicist.
Many LOGO workers contributed to the aesthetic of the Turtle drawings. Those who most influenced me were Cynthia Solomon, Ellen Hildreth, and Ilse Schenck (who arranged the garden and birds in this book).
In this book I write about children but, in fact, most of the ideas expressed are relevant to how people learn at any age. I make specific references to children as a reflection of my personal conviction that it is the very youngest who stand to gain the most from change in the conditions of learning. Most of the children who collaborated with us were of mid-elementary school age. Radia Perlman was the first to explore techniques for working with much younger children, as young as four years of age. Abelson and diSessa have specialized in work with older students of high school and college age. Gary Drescher, Paul Goldenberg, Sylvia Weir, and Jose Valente are among those who have pioneered teaching LOGO to severely handicapped children. Bob Lawler carried out the first, and so far the only, example of a different kind of learning experiment, a kind that I think will become very important in the future. In Lawler’s study, a child was observed “full time” during a six-month period so as to capture not only the learning that took place in contrived situations but all the overt learning that took place during that period. I have also been influenced by another study on “natural learning” now being conducted as part of research by Lawrence Miller for his thesis at Harvard. Both Lawler and Miller provided data for a general intellectual position that underlies this book: The best learning takes place when the learner takes charge. Edwina Michner’s Ph.D. thesis was a learning study of a very different sort, an attempt to characterize some of the mathematical knowledge that the mathematical culture does not write down in its books.
I have acknowledged intellectual obligations to many people. I have to thank most of them for something else as well: for support and for patience with my too often disorganized working style. I am deeply grateful to everyone who put up with me, especially Gregory Gargarian who had the very difficult jobs of maintaining the organization of the LOGO Laboratory and of entering and updating many successive versions of this book in the computer files. In addition to his competence and professionalism, his friendship and support have made easier many moments in the writing of this book.
MIT has provided a highly stimulating intellectual environment. Its administrative environment is also very special in allowing out-of-the-ordinary projects to flourish. Many people have helped in an administrative capacity: Jerome Wiesner, Walter Rosenblith, Michael Dertouzos, Ted Martin, Benson Snyder, Patrick Winston, Barbara Nelson, Eva Kampits, Jim McCarthy, Gordon Oro, and George Wallace come to mind but I am sure there are many others. Of these I owe a very special debt to Eva Kampits who was once my secretary and is now Dr. Kampits.
The LOGO project could not have happened without support of a different kind than I have mentioned until now. The National Science Foundation has supported the work on LOGO since its inception. I want also to mention some of the Foundation’s individuals whose imaginative understanding made it possible for us to do our work: Dorothy Derringer, Andrew Molnar, and Milton Rose. The value of the support given by such people is moral as well as material, and I would include in this category Marjorie Martus at the Ford Foundation, Arthur Melmed at the National Institute of Education, Alan Ditman at the Bureau for the Education of the Handicapped, and Alfred Riccomi of Texas Instruments. I would also most especially include three individuals who have given us moral and material support: Ida Green, Erik Jonsson, and Cecil Green all from Dallas, Texas. It has been a particularly rich experience for me to work closely with Erik Jonsson on developing a project using computers in the Lamplighter School in Dallas. I have come to appreciate his clarity of thought and breadth of vision and to think of him as a colleague and a friend. His support for my ideas and intolerance of my disorganization helped make this book happen.
John Berlow contributed beyond measure to the writing of this book. He came into the picture as an unusually intelligent editor. At every phase in the manuscript’s development, his critical and enthusiastic readings led to new clarity and new ideas. As the project developed he became, for me, more than an editor. He became a friend, a dialog partner, a critic, and a model of the kind of reader I most want to influence. When I met John he was without computer expertise, although his knowledge in other areas provided him with an immediate base from which to generate his own ideas concerning computers and education.
There are many people whose contributions cannot be categorized. Nicholas Negroponte is a constant source of inspiration, in part precisely because he defies categorization. I also wish to thank Susan Hartnett, Androula Henriques, Barbel Inhelder, A. R. Jonckheere, Duncan Stuart Linney, Alan Papert, Dona Strauss, and I. B. Tabata. And there are a few people with whom disagreements about how computers should be used have always been valuable: John Seeley Brown, Ira Goldstein, Robert Davis, Arthur Leuhrman, Patrick Suppes. If the book can be read as an expression of positive and optimistic thinking, this must be attributed to my mother, Betty Papert. Artemis Papert has helped in so many ways that I can only say: Merci.
Everyone concerned with how children think has an immense general debt to Jean Piaget. I have a special debt as well. If Piaget had not intervened in my life I would now be a “real mathematician” instead of being whatever it is that I have become. Piaget invested a lot of energy and a lot of faith in me. I hope that he will recognize what I have contributed to the world of children as being in the spirit of his life enterprise.
I left Geneva enormously inspired by Piaget’s image of the child, particularly by his idea that children learn so much without being taught. But I was also enormously frustrated by how little he could tell us about how to create conditions for more knowledge to be acquired by children through this marvelous process of “Piagetian learning.” I saw the popular idea of designing a “Piagetian Curriculum” as standing Piaget on his head: Piaget is par excellence the theorist of learning without curriculum. As a consequence, I began to formulate two ideas that run through this book: (1) significant change in patterns of intellectual development will come about through cultural change, and (2) the most likely bearer of potentially relevant cultural change in the near future is the increasingly pervasive computer presence. Although these perspectives had informed the LOGO project from its inception, for a long time I could not see how to give them a theoretical framework.
I was helped in this, as in many other ways, by my wife Sherry Turkle. Without her, this book could not have been written. Ideas borrowed from Sherry turned out to be missing links in my attempts to develop ways of thinking about computers and cultures. Sherry is a sociologist whose particular concerns center on the interaction of ideas and culture formation, in particular how complexes of ideas are adopted by and articulated throughout cultural groups. When I met her she had recently completed an investigation of a new French psychoanalytic culture, of how psychoanalysis had “colonized” France, a country that had fiercely resisted Freudian influence. She had turned her attention to computer cultures and was thinking about how people’s relationships with computation influence their language, their ideas about politics, and their views of themselves. Listening to her talk about both projects helped me to formulate my own approach and to achieve a sufficient sense of closure in my ideas to embark on this writing project.
Over the years Sherry has given me every kind of support. When the writing would not work out she gave me hours of conversation and editorial help. But her support was most decisive on the many occasions when I fell out of love with the book or when my confidence in my resolution to write it flagged. Then, her commitment to the project kept it alive and her love for me helped me find my way back to being in love with the work.
SEYMOUR PAPERT
Cambridge, Massachusetts
April 1980