B.F. Skinner, the psychologist and social philosopher, had an idea about how to bring about world peace. It involved teaching pigeons to guide missiles.
Skinner isn’t the big name today that he was in the middle 20th century. He was a Harvard professor then, a pioneer of the study of behaviorism. But he was also a gifted animal trainer. By day, in his lab, he worked out the fastest ways to train rats and pigeons to pull levers and push buttons. By night, he trained his children’s kittens to play the piano. He taught the family beagle hide and seek.
An advocate of positive reinforcement training, he focused on rewarding behaviors that he wanted; by calculating what is rewarding to an animal and then doling out and withholding reward strategically, it’s possible to coax an endless array of behaviors without any kind of force. It’s a method of animal training that has long been used by marine animal trainers who work with creatures that are too big and powerful to be easily coerced with punishment.
Today, most positive reinforcement dog training is done using a clicker, which is a handheld noise maker, or some other signal that’s been repeatedly paired with a reward. I train dogs by using the sharp clicking noise to pinpoint the moment that something has been done correctly. The noise is always backed up with a reward, just like a poker chip has meaning because it can always be cashed in for a dollar amount. Desired behaviors are positively reinforced; non-reinforced responses tend to go away, just like you’d probably stop showing up at work if your paychecks were cut off.
The word “positive” simply means that something is added to the equation. It has nothing to do with keeping a good attitude. Indeed, it has nothing to do with actual good doing. Jesus used positive reinforcement to influence people, but so did Hitler.
In the 1930s, when B.F. Skinner first started outlining the workings of this kind of conditioning, he founded a school of philosophy called Radical Behaviorism. At its foundation it’s the idea that it is easier to manipulate surroundings in order to control how we will behave than it is to try to peer inside the brain and psychoanalyze in order to affect change in ourselves or others. It’s about looking outward, not inward, to solve problems. In his science fiction novel Walden Two, Skinner produced a template for how to create a small utopian community governed by behaviorists implementing the methods of positive reinforcement. Critics called it “fascism without tears.” Others compared his community to a really big dog obedience class.
People were scared because the implications of manipulating people through conditioning meant bad things for the American idea of free will. No one wants to think they’re easily manipulated. But we are. Sure, we all make choices, but those choices are usually made because they’ve been reinforced in some way, or because it’s a choice that will avoid punishment. We’re always going towards what feels good. By controlling the outcomes of people’s behaviors through calculated reinforcement, Skinner believed people could be helped to live happier, safer, healthier lives.
PIGEONS LEARN JUST LIKE WE DO
In Skinner’s perfect world, war would be unnecessary. But his ideas about using positive reinforcement to manipulate people in order to levy world peace were largely dismissed when he first voiced them in the early 1940s. There were some factions of individuals who attempted to build their own Walden Twos, but the government wasn’t offering them subsidies. However, when the young professor suggested that positive reinforcement could be used to train pigeons to guide missiles, the National Defense Research Committee took notice.
The three screens of the pigeon missile nose cone. A pigeon was placed behind each screen
With the aforementioned clicker system now widely used to train dogs, Skinner taught pigeons to peck at a small, moving point placed under a glass screen. He suggested that the birds could be put in missiles with a screen inside; a distant enemy plane would look like a speck on the glass, and would cause the birds to peck. The movement of their necks could then be translated into navigational directions for the missiles. As long as their pecks stayed in the center of the screen, the missile would fly straight; off-center pecks would cause the screen to tilt, which would then cause the missile to change course. If properly trained, the birds could aim at their targets with machine-like precision, even if exposed to extreme noise or atmospheric pressure.
While working on the secretive Project Pigeon (later called Project Orcon, for “organic control”), Skinner taught pigeons to reliably peck at a spot – one bird was trained to peck at an image more than 10,000 times in 45 minutes – but that was the least of his feats. He trained them to read, hit ping-pong balls, and play the piano. After the attack on Pearl Harbor, the NDRC contracted General Mills to work on the idea with Skinner, but while they contributed $25,000 to the research, they neglected to give Skinner’s team any sample target images or specifications about what kind of missile or pigeon carriers would be used. In short, the humans screwed up.
Skinner’s biographer Daniel Bjork notes that the failure of Project Pigeon was likely linked to the fact that it was happening concurrently with the Manhattan Project; compared to the power of a potential nuclear weapon, animal-led warfare seemed anachronistic and silly. The kamikaze pigeons never flew. Instead, we dropped the bomb on Hiroshima.
Today, several Skinner disciples teach periodic workshops to demonstrate the power of operant conditioning by showing people how to “clicker train” a chicken to peck at a spot. Here, a chicken has just been taught to differentiate between two shapes