B.F. Skinner Demonstrates His “Teaching Machine,” the 1950s Automated Learning Device

The name B.F. Skin­ner often pro­vokes dark­ly humor­ous ref­er­ences to such bizarre ideas as “Skin­ner box­es,” which put babies in cage-like cribs, and put the cribs in win­dows as if they were air-con­di­tion­ers, leav­ing the poor infants to raise them­selves. Skin­ner was hard­ly alone in con­duct­ing exper­i­ments that flout­ed, if not fla­grant­ly ignored, the eth­i­cal con­cerns now cen­tral to exper­i­men­ta­tion on humans. The code of con­duct on tor­ture and abuse that osten­si­bly gov­erns mem­bers of the Amer­i­can Psy­cho­log­i­cal Asso­ci­a­tion did not exist. Rad­i­cal behav­ior­ists like Skin­ner were redefin­ing the field. His work has come to stand for some of its worst abus­es.

But Skin­ner has been mis­char­ac­ter­ized in the pop­u­lar­iza­tion of his ideas — a pop­u­lar­iza­tion, it’s true, in which he enthu­si­as­ti­cal­ly took part. The actu­al “Skin­ner box” was cru­el enough — an elec­tri­fied cage for ani­mal exper­i­men­ta­tion — but it was not the infant win­dow box that often goes by the name. This was, instead, called an “air­crib” or “baby-ten­der,” and it was loaded with crea­ture com­forts like cli­mate con­trol and a com­ple­ment of toys. “In our com­part­ment,” Skin­ner wrote in a 1945 Ladies Home Jour­nal arti­cle, “the wak­ing hours are invari­ably active and hap­py ones.” Describ­ing his first test sub­ject, his own child, he wrote, “our baby acquit­ted an amus­ing, almost ape­like skill in the use of her feet.”

Skin­ner was not a soul­less mon­ster who put babies in cages, but he also did not under­stand mam­malian babies’ need for phys­i­cal touch. Like­wise, when it came to edu­ca­tion, Skin­ner had ideas that can seem con­trary to what we know works best, name­ly a vari­ety of meth­ods that hon­or dif­fer­ent learn­ing styles and abil­i­ties. Edu­ca­tors in the 1950s embraced far more reg­i­ment­ed prac­tices, and Skin­ner believed humans could be trained just like oth­er ani­mals. He treat­ed an ear­ly exper­i­ment in class­room tech­nol­o­gy just like an exper­i­ment teach­ing pigeons to play ping-pong. It was, in fact, “the foun­da­tion for his edu­ca­tion tech­nol­o­gy,” says edu­ca­tion jour­nal­ist Audrey Wat­ters, “that we’ll build machines and they’ll give stu­dents — just like pigeons — pos­i­tive rein­force­ment and stu­dents — just like pigeons — will learn new skills.”

To this end, Skin­ner cre­at­ed what he called the Teach­ing Machine in 1954 while he taught psy­chol­o­gy at Har­vard. He was hard­ly the first to design such a device, but he was the first to invent a machine based on behav­ior­ist prin­ci­ples, as Abhishek Solan­ki explains in a Medi­um arti­cle:

The teach­ing machine was com­posed of main­ly a pro­gram, which was a sys­tem of com­bined teach­ing and test items that car­ried the stu­dent grad­u­al­ly through the mate­r­i­al to be learned. The “machine” was com­posed of a fill-in-the-blank method on either a work­book or on a com­put­er. If the stu­dent was cor­rect, he/she got rein­force­ment and moved on to the next ques­tion. If the answer was incor­rect, the stu­dent stud­ied the cor­rect answer to increas­ing the chances of get­ting rein­forced next time.

Con­sist­ing of a wood­en box, a met­al lid with cutouts, and var­i­ous paper discs with ques­tions and answers writ­ten on them, the machine did adjust for dif­fer­ent stu­dents’ needs, in a way. Skin­ner “not­ed that the learn­ing process should be divid­ed into a large num­ber of very small steps and rein­force­ment must be depen­dent upon the com­ple­tion of each step. He believed this was the best pos­si­ble arrange­ment for learn­ing because it took into account the rate of learn­ing for each indi­vid­ual stu­dent.” He was again inspired by his own chil­dren, com­ing up with the machine after vis­it­ing his daugh­ter’s school and decid­ing he could improve on things.

The method and means of learn­ing, as you’ll see in the demon­stra­tion films above, were not indi­vid­u­al­ized. “There was very, very lit­tle free­dom in Skin­ner’s vision,” says Wat­ters. “Indeed Skin­ner wrote a very well-known book, Beyond Free­dom and Dig­ni­ty in the ear­ly 1970s, in which he said free­dom does­n’t exist.” While Skin­ner’s machine did­n’t itself become wide­ly used, his ideas about edu­ca­tion, and edu­ca­tion tech­nol­o­gy, are still very much with us. We see Skin­ner’s machine “tak­ing new forms with adap­tive teach­ing and e‑learning,” writes Solan­ki.

And we see the dark­er side of his design in class­room tech­nol­o­gy, says Wat­ters, in an indus­try that prof­its from alien­at­ing, one-size-fits all ed-tech solu­tions. But she also sees “stu­dents who are resist­ing and com­mu­ni­ties who are build­ing prac­tices that serve their needs rather than serv­ing the needs of engi­neers.” Skin­ner’s the­o­ries of con­di­tion­ing were and are incred­i­bly per­sua­sive, but his reduc­tive views of human nature seem to leave out more than they explain. Learn more about the his­to­ry of teach­ing machines in Wat­ters’ new book, Teach­ing Machines: The His­to­ry of Per­son­al­ized Learn­ing.

Relat­ed Con­tent: 

The Lit­tle Albert Exper­i­ment: The Per­verse 1920 Study That Made a Baby Afraid of San­ta Claus & Bun­nies

Her­mann Rorschach’s Orig­i­nal Rorschach Test: What Do You See? (1921)

A Brief Ani­mat­ed Intro­duc­tion to Noam Chomsky’s Lin­guis­tic The­o­ry, Nar­rat­ed by The X‑Files‘ Gillian Ander­son

Josh Jones is a writer and musi­cian based in Durham, NC. Fol­low him at @jdmagness