Skinner Pigeons

One of B.F.Skinner's experiments back in the early days of the discovery of learning theory involved pigeons in boxes. The boxes had a food pellet delivery chute that dropped a pellet to the pigeon on a completely random feeding schedule. They used a random numbers table so that no one--including the experimenter--knew when the food pellet was going to come rolling down the food chute.

Skinner's pigeons did a very interesting thing. Whatever behavior they were doing just before the food came rolling down, they did more of. When it came down the next time, whatever they were doing before that, they did some more of.

Within a short length of time the pigeons are walking around . . . bobbing their heads, shaking their tails . . . and checking that food chute.

Superstitious Behaviour

Accidental reinforcement of a response can lead to superstitious behaviour. Skinner demonstrated the conditioning of such behaviour using pigeons. He set the dispenser to deliver food to animals in an operant chamber at fixed time intervals, for example every 15 minutes. The pigeons associated whatever behaviour they were engaging in at the time of the food being dispensed with the delivery of the food. The likelihood of those behaviours occuring then increased. Skinner conditioned pigeons to spin around in circles, nod their heads, or to make swaying motions. (Nye, 1992)

III. Superstitous behavior in pidgeons

A. A superstition is a belief in something, and we usually do not attribute such beliefs to animals. Skinner attempted to demonstrate superstitious behavior in an animal we don't usually think of as having beliefs -- a pidgeon.

B. Example of superstitious behavior -- class? Why does this behavior exist? You presume there is a connection between the behavior and some reinforcing consequence, even though in reality there is not. This connection occurs because the behavior was "accidentally" reinforced. Skinner called this "non-contingent" reinforcement, a reward that is not contingent on any particular behavior. You believe that there is a causal relationship between the behavior and the reward when no such relationship exists.

C. Used a Skinner box, not too surprising! Typically there was a disk for the pidgeon to peck to get food. In this box, the food dispenser was rigged to drop food pellets into the tray at 15 second intervals regardless of what the animal was doing at the time. Non- contingent reinforcement.

D. Used 8 pidgeons, fed less than normal so they would be hungry when put in the cage. Put in the cages for several days. Then two observers recorded their behavior.

E. Results: Six birds showed very specific, different behaviors. Quoted from Skinner's report.

Bird 1 -- made counter clockwise turns in the cage, made two to three turns between reinforcements.

Bird 2 -- thrust its head into one of the upper corners of the cage.

Bird 3 -- Tossing response as if placing its head beneath an invisible bar and tossing it up.

Birds 4&5 -- pendulum motion of their heads and body -- head was extended forward and swung from right to left with a sharp movement followed by a somewhat slower return.

Bird 6 -- brushing movements directed toward but not touching the floor.

F. NONE of these behaviors had been observed in the birds prior to the conditioning procedure. Skinner claims they became superstitious.

G. Skinner wanted to see what would happen if the time interval between reinforcements was extended. With one head-swing bird, he slowly increased food delivery from 15 sec to one minute. As the interval increased, the bird's movements became more energetic, like he was doing a pidgeon food dance.

H. This behavior was extinquished by discontinuing reinforcement (stop food). One bird produced over 10,000 responses before extinction occured.

B. F. Skinner used to study the laws of behavior by placing pigeons in a special cage with an electrically actuated food tray and a lever they could peck. To train them to peck the lever he needed to "reinforce" successive approximations of the lever-pecking behavior. Food is a good behavioral reinforcer when the pigeons are hungry, but they quickly satiate. It turned out that the sound of the food tray opening also became a "secondary" reinforcer by its association with the availability of the "primary" reinforcer food. So Skinner's first step would be to leave individually caged pigeons overnight during which their food trays would click open every few minutes and make food available for a few seconds each time.

By morning Skinner observed that every pigeon was engaging in some sort of ritualistic behavior or dance. Each pigeon's behavior was unique but repetitive. The explanation for this "supertitious" behavior was that the reinforcing sound of the food tray opening strenthened whatever behavior the pigeon might have been engaged in at the time, making that behavior even more likely to be occurring on successive food tray openings. So even though nothing the pigeons were doing was "actually" related to when the food tray opened (it opened according to an external timer), the pigeons behaved as though they "thought" they had to so some particular thing to get the tray to open.

This empirical nature of learning is the basis of many human superstitions and myths. Along with many useful and survival-enhancing behavior patterns we also learn many useless and sometimes harmful patterns. I regard almost all beliefs in the efficacy of prayer, for example, to be superstitious as described above. While there are cultural reasons for being taught that prayer heals sickness or prevents drought or famine, a belief in prayer is reinforced and made "fervid" when an illness ends or the rains come right after one has prayed.

All these explanations are inadequate. While they may explain the attraction of religion, they do not explain the power of it. Why should it exert such a strong hold over people's psyches that they defy even their biological urges, practising celibacy and vegetarianism and self-abuse? Why, for most of human history, has religion been a dominant political force? What is the neurological basis of religion's cultural universality?

A possible answer to the last query comes from an experiment by psychologist BF Skinner. In 1948, he tried a variation on a standard experiment where pigeons enclosed in a box were able to get food by pecking a switch. Skinner set up the box to reward the pigeon with food no matter what the bird did.

The results were fascinating.

Instead of just sitting back and waiting for food, the pigeons developed what Skinner termed "superstitious'' behaviour. One bird spun itself round and round; another repeatedly thrust its head towards a particular corner of the box; a third swung its body from side to side like a pendulum. All of them continued these actions until the food appeared; it was as though the pigeons had made some causal connection in their bird-brains between these gestures and getting food.

Human beings are not bird-brains-well, most of us aren't-but neither are our belief systems as simplistic as the pigeons'. Yet the connection is clear: the machinery in our brain designed to link effect with cause may be just as easily derailed, leading to the creation of religion.

˙ltima modificaciˇn: 26-junio-2002