The Friendly Orange Glow

by Brian Dear

  Skinner’s 1954 machine Credit 3

  Skinner’s machine is a plywood box some fifteen inches high, wide, and deep. The hinged top opens so that a scroll of pleated paper tape can be loaded inside. Typewritten onto the paper are arithmetic problems to be solved by the student; the answers to each problem are encoded as small holes, like in a player piano, punched in very specific places to denote a corresponding value. When the box top is shut, you can read the math problems, one at a time, through a small opening in the top of the box, showing a small, exposed area of the paper. You might see a math problem, such as “3 + 2,” through the viewing window on the top side of the box. To answer the problem, you manipulate a series of what Skinner called “sliders” that can be moved up or down through slits in the wood, serving as number scales. This particular math problem is looking for a one-digit answer, so only one slider need be moved into place. The correct slider to move is the one that causes numbers to appear in the hole under the “3 + 2.” Pull the slider toward yourself and the number increases up to 9. Push it away and it drops to zero. Once you’ve formed your answer you can attempt to turn a big black knob on the front side of the box. If the answer is correct, the knob turns freely and the scroll advances to the next problem. This is how you know your answer is correct. If the knob is locked and the scroll won’t advance, you know you’re wrong.

  Unlike a written test, which requires a teacher to laboriously correct and grade afterward, in theory this device freed her from that labor, at least the labor of grading—she still had to compose the questions. When a student finished an entire scroll of problems, the teacher could assume the student finally got all the answers right.

  Skinner’s accompanying paper at the conference was subsequently reprinted in the Harvard Educational Review and elsewhere, and quickly became one of the most widely cited papers in the field for many years. Like the provocative “Baby in a Box” article from 1945, it sparked discussion and controversy, but most of all great interest. In 1968, he finally extended the ideas born out of that paper into a book-length work entitled The Technology of Teaching. He dedicated the book to Mary Graves.

  —

  Pressey read Skinner’s 1954 paper and contacted him. “He wrote me and asked if we could get together,” Skinner told the author. “So we had breakfast and…he brought me his papers. Later he sent me one of his machines. He was very cordial. And he saw the point of it all.”

  But Skinner never considered Pressey’s machine to be a teaching machine: instead, he felt Pressey’s device was more of a testing machine, not something that taught something new. Still, Pressey’s thinking clearly influenced Skinner, and the eventual outcomes Pressey hoped for in the long term were arguably things Skinner would have agreed with: an “industrial revolution” in education, in which, Pressey had once written, “educational science and the ingenuity of educational technology combine to modernize the grossly inefficient and clumsy procedures of conventional education. Work in the schools of the future will be marvelously though simply organized, so as to adjust almost automatically to individual differences and the characteristics of the learning process. There will be many labor-saving schemes and devices, and even machines—not at all for the mechanizing of education, but for the freeing of teacher and pupil from educational drudgery and incompetence.”

  As Skinner’s ideas spread, and the field of teaching machines and “programmed instruction” exploded over the next few years, a long-retired Pressey began to have a change of heart. “He became unhappy,” Skinner told the author. “He felt that I was wrong, that a teaching machine should be a testing machine. You should instruct, and then use the machine. Not use the machine to instruct.” Which was exactly the pivot Pressey hoped Skinner and the teaching machine practitioners would undertake in their strategy.

  Pressey’s concerns were largely ignored.

  —

  The Shady Hill School had been founded out of necessity by parents and educators in the Harvard community who had been sending their children to the Agassiz public school until 1914, when the building was condemned. Concerned parents, realizing that a new school building was not likely, took matters into their own hands, starting their own Cooperative Open Air School, initially situated on the back porch of Agnes and William Ernest Hocking, who along with other Harvard faculty taught the students. William Ernest Hocking was a noted philosophy professor at Harvard whom later Skinner met during graduate school. In a few years the Hockings found a more permanent location in the nearby Shady Hill neighborhood, and the school took that name. Students arrived in the newly constructed buildings in 1917. It has grown and expanded ever since, and is a highly regarded private school today.

  The school was founded on progressive educational principles, and the faculty were encouraged to study the work of William James, John Dewey, Francis Wayland Parker, Alfred North Whitehead, and Jean Piaget. Katharine Taylor, Shady Hill’s founding director, who had been handpicked from the Francis Parker School in Chicago, would write, in 1937, regarding teaching, “The more you think of teaching the more you realize that it can never be classified as science. It is nearer to being an art. It makes use of the findings of science, but the ability to teach can never be fully acquired through scientific pursuit alone, and the procedure of teaching can never be organized purely as a science.”

  Unlike B. F. Skinner’s own childhood school experience that he looked upon with affection and gratitude, such was not the case for daughter Deborah, the proverbial Baby in the Box, who by the time she was eight and attending Shady Hill School was not happy being there at all. Shady Hill itself was apparently not the problem. She simply felt she wasn’t prepared. “I hated it,” she said, “because I was a poor student and not popular. The reason, I like to think, is that most of the other kids had been taught to read in the womb, but my parents thought school would be the best place to be taught. When I was eight, I was still a poor reader and the teacher complained to my parents!”

  “She wasn’t the most popular of kids,” admits Mary Eliot, “but she wasn’t really disliked.” She was, however, behind a bit in her skills. Eliot would learn that Deborah’s father was not just unhappy with what he saw during that Father’s Day classroom sit-in. There was more. “He was upset about the teaching of arithmetic, math, and later he was very much worried about how spelling was taught.”

  Skinner wanted his daughter to get some extra help. He believed that the breakthroughs he was making with teaching machines and “programmed instruction” could help Deborah. She could be a guinea pig for his ideas. “The school was asked,” says Eliot, “to provide a special time for her to go in a closet, a special little room, and work on spelling, and it was just about programmed learning.” Eliot remained skeptical. She felt Skinner’s approach was too “controlling,” and in the ensuing years she would never warm to his idea of applying behaviorist principles to learning. “[Deborah] had things that he had sketched out for her to learn to spell, and I wasn’t very much impressed by that, because I found that if children wrote and read and had their compositions corrected, they learned to spell just about as fast as if you had spelling lists.”

  —

  A contemporary of John B. Watson’s, the psychologist E. L. Thorndike, in his book Education, wrote,

  If, by a miracle of mechanical ingenuity, a book could be so arranged that only to him who had done what was directed on page one would page two become visible, and so on, much that now requires personal instruction could be managed by print. Books to be given out in loose sheets, a page or so at a time, and books arranged so that the student only suffers if he misuses them, should be worked in many subjects.

  Thorndike’s vision was remarkable given that he wrote these words in 1912. Hypertext, which serves as the basis of today’s World Wide Web thanks to its HTML language, might arguably claim to be one application of Thorndike’s vision, the notion extended far beyond simply education to all knowledge. It’s also notable that Thorndike’s vision wa
s published more than forty years before Skinner built his first machine, and yet Skinner’s machine lacked that “miracle of mechanical ingenuity,” limiting his machines to the same linear presentation of problems and questions for all students who sat down to use the machine. The idea of a machine that can achieve that miracle, breaking from its linear program to address the truly individual needs of each particular student, was an educational Holy Grail that eluded Skinner’s work.

  Over the course of his career, Skinner would extend this notion of reinforcement in shaping behavior from rats and pigeons to learning in animals in general, including human learning. But Skinner’s theories rubbed some scientists, educators, and psychologists the wrong way. As the decades wore on enterprising researchers realized it might be possible to not only think about brains as computers but even try to model—using a computer—how brains worked. The process of collecting observational data to support “how” the brain worked would have made Watson blow a fuse. Behaviorism’s dismissal of “looking under the hood” to see what was going on in the brain—how consciousness worked, how learning worked—actually spurred entirely new and different “cognitive” theories, unafraid to consider human learning that involved brain functions for which external, observable behavior alone could not account.

  Controversy would follow Skinner and his ideas for decades. In academic corners, his views became less and less popular in the 1960s and 1970s as, coincidentally, computers became more and more available. Noam Chomsky, an MIT professor of linguistics, was one of Skinner’s many rivals and one of the leaders of the cognitive revolution aimed to combat behaviorism with what cognitivists believed was a better explanation of what was going on inside the brain. For years, Chomsky would attack Skinner’s entire thesis, whether it be the behaviorist model or Skinner’s later treatises on broader themes like freedom and dignity.

  Some researchers with a passion for understanding learning and improving education found Skinner a perplexing enigma. Seymour Papert was one. Papert, who was greatly influenced early on in his career by the work of Jean Piaget, helped create and was a major proponent of the LOGO educational computing movement in the late 1960s and onward, and also a major proponent of the “cognitive science” revolution, an entirely different view of the mind and an approach that vigorously attacked, some might even add ridiculed, behaviorism. Instead of a machine teaching a student, Papert was in favor of children teaching machines, and in so doing, learning about mathematical concepts, not to mention gaining skills in computer programming.

  “I find Skinner somewhat of a contradiction,” Papert once confessed to this author, “because as a person he’s intellectually very rich and multi-sided and very literate and likes poetry and I think is a great person. When he thinks about children and education, there’s a lot of richness. The form in which it takes when it gets out into the world is extremely,” he said, pausing for a moment as if to choose the next word carefully, “pernicious. He has a very pernicious doctrine. The pernicious doctrine being that you can break up knowledge into fragments and guide children toward acquiring the knowledge like you might involve the behavior of a rat or a pigeon. I find that a contradiction. I find when you think of Skinner as a whole person, he’s so far away from this kind of thinking and practice of education, well, I’m full of wonderment that he isn’t the main critic of the way that his ideas are being used in the world.”

  —

  In the years following the 1954 debut of Skinner’s teaching machine, he iterated repeatedly on the design, producing newer, more complex devices. His ideas spread around the world, and many others took up the task of tinkering with mechanical devices that could, in theory, teach.

  One of these more advanced machines, heavier, sleeker, enclosed in sheet metal, Skinner designed for use in his Natural Science 114 classes at Harvard University during the middle to late 1950s.

  To operate, an instructor had to first load it with what Skinner would call the “program,” by opening the top and placing a cardboard disk, about the size of an LP record, onto a platter. The disk would have room for about thirty questions, each typed inside a very thin “pizza slice” of the disk. Each slice provided enough room for maybe two or three lines of text. After placing the disk in the machine, the instructor would shut the top and lock it. With the top of the box shut, the student could only read one question frame at a time through a rectangular viewing hole exposing one of those thirty slices of the disk and could only remove the disk once all questions had been answered correctly. The upper-right-hand corner of the little window had a shield, a kind of metallic eyelid, hiding part of the text underneath. This is where the question’s answer was printed.

  Skinner’s 1958 teaching machine Credit 4

  Located next to the rectangular window, another area existed for writing your answer on an exposed section of a paper tape. The machine’s innards revealed a fantastic assortment of clocklike gears, pulleys, levers, and other mechanisms working together in a harmonious logic that, through a careful choreography of mechanical motions, permitted the student to advance to the next frame by pulling a lever to rotate the disk slightly.

  As teachers, these machines were not mechanical equivalents of Mary Graves, to be sure. The machine’s elaborate gears and pulleys and knobs, though impressive, had no means of “knowing” whether a student’s response to a question was really correct or not. Nor, for that matter, did it know if a response was inappropriate. If asked for the name of the first president, the student could just as easily write “Mickey Mouse” as he could write “George Washington” and move the lever horizontally, telling the machine that he’d answered correctly. The machine had no way to verify the student’s answer. Skinner’s design placed the responsibility of judging the correct response entirely on the student. Here, the limitations of mechanical engineering forced Skinner to fall back on the safety net of Harvard’s Honor Code prohibiting cheating. Another weakness of this machine, in fact all of the machines Skinner would design over the years, is that the “program” is linear. Each student gets the same questions in the same order; everything is uniform. There is no possibility of branching; no possibility for the machine to think, This here is one bright student, how about we skip to some harder questions.

  Skinner also designed machines that would be manufactured by IBM. One was based on the 1954 “slider” design, built to teach spelling or arithmetic. The “spelling” version of this machine employed a dozen sliders with the alphabet imprinted upon them. Thus, one might be asked to spell the word “manufacture” by positioning eleven sliders with the correct letters until the word is formed.

  A set of six frames is shown in the illustration below. This example shows what at the time was called the “tiny steps” approach Skinner took, based on his experience with training rats and pigeons. The instructional “program” guided the student, through incremental steps, how to “behave” accordingly—in this case, spell the word “manufacture” correctly and understand what it means. The student must correctly spell the first word before being presented with the second, and so on:

  A set of frames designed to teach a third- or fourth-grade pupil to spell the word “manufacture” Credit 5

  Skinner actively pursued relationships with commercial firms, from IBM to, improbably, a company called Rheem, a boiler and air conditioner maker, which viewed the teaching machine market as a new craze attractive enough for them to try to cash in on. But after a few frustrating years he learned that American companies simply could not deliver on what they promised. “It was really a terrible thing,” Skinner told this author. “It was shocking. Rheem was the worst one. They obviously took this on, gave me some kind of an arrangement, [but] I don’t remember getting much money out of it….[They] kept me around for several years, never did what they said they were going to do at all. It was a shock.”

  IBM seemed a little bit more hopeful as a partner but in the end disappointed Skinner as well. “They built this machine, which was ver
y good, it was doing what they said it would do, and as a matter of fact they then redesigned it and got a patent based on my patent. But then they stopped it. And one of the engineers there who was dying to see it developed tried hard to get [IBM president] Thomas Watson’s children to try my machine. They had given it to the typewriter division because they had contacts with high schools and he simply said, ‘I can use this money to stay ahead of my competition in typewriters’ and that was that. I don’t think there’s anybody within IBM who knows that that machine is now in the Smithsonian.”

  In a classic example of the right hand not knowing what the left hand was doing, another corporate group within IBM enthusiastically contacted Skinner a few years later inquiring about his machines and how they might be commercialized. They were unaware of IBM’s previous dealings with Skinner. Skinner did not pursue the new opportunity.

  —

  Interest in teaching machines and what was by now called “programmed instruction” was growing. Across the country, in corporate, university, government, and military settings, a variety of mechanical devices were starting to be utilized to teach and train people on numerous subjects.

  Beyond the already cited weaknesses of Skinner’s machines, some experts became concerned with another perceived drawback of Skinner’s design, that students were not given an opportunity to learn why their answers were wrong. The design emphasis of his devices focused on reinforcing and rewarding correct answers. But in the real world, students often get things wrong, and while it is useful to know when one has gotten something wrong, it’s even more useful to know why, so one can learn to get things right in the future.