Skinner, B.F. (1948). ‘Superstition’ in the Pigeon.
Arhe Vaninetti
Research Methods and Statistics, PSY 301, Pacific University, Oregon
February 14, 2023
In Skinner’s (1948) classical example of behaviorism, the question of random conditioning in pigeons is explored in relation to consistency and extinction of the behavior over time. While Skinner did not set out to explicitly condition a certain behavior or combination of behaviors, he wound up exploring the threat of extinction and re-conditioning on a particular behavior. During this study, Skinner set up a random interval system to feed hungry pigeons. Because the pigeons were on rationed food, they were eager to continually eat for the duration of the experiment, with Skinner stating that they were on 75% of their typical diet. When the food was presented at random intervals, the pigeons began expressing random behaviors that they associated with receiving food. This classical conditioning created various random behaviors in 6 of the 8 experimental pigeons, including spinning, rocking, flapping, and head jerking.
These behaviors became quite elaborate, and often included multi-step processes that the pigeons would complete repeatedly until they received their food from the machine. Skinner received consistent results on the random, complex behavior, and so he decided to test how long the pigeons could go performing the behaviors until they began indicating the extinction of the behaviors. One pigeon went for a full two minutes without food while still exhibiting consistent behavior, while most pigeons lasted around a minute. Finally, Skinner let the behaviors go extinct in the pigeons (I.e., they wouldn’t exhibit the conditioned behavior for 10 to 15 minutes straight, or more) and then restarted the experiment. The pigeons were able to become conditioned again quite easily, but they were not exhibiting the same behavior as before. Skinner concluded that, to regain the initial behavior, it would require necessary training procedures, and not random interval feedings. After the experiment had concluded, Skinner suggested that the pigeon’s behavior was not so different from human behavior—even giving examples of bowling and card games.
While this study was conducted as an experimental examination of behaviorism within pigeons, there are some notable flaws in the methodology and ethics. The experiment was not necessarily an experiment, and more of an exploration and application of previous research. An easy comparison would be to a layman conditioning his dog to shake hands—which is not phenomenal research. While important to the era of behaviorism, the subject does not still stand today as a valid field of thought, and has been lost to history in some respects. Through experimental research, we now understand that classical conditioning has limited applications. While Skinner does include the examples of classical conditioning in human behavior (the bowler and the card player), these are not necessarily examples of conditioning—but may be examples of other fields, like cultural or social psychology.
With regard to the pigeons, there was a low sample size, and the power was limited in this experiment. Even with the 8 pigeons in the sample, 2 still displayed no examples of classical conditioning throughout the experimental process. Regarding ethics, this study was conducted far before the industry standards of animal welfare, and so the rationing of food to 75% may be questionable in the case of pigeons. Furthermore, there was no mention of the pigeons’ cages and whether or not the pigeons were allowed to socialize with each other. This study was sparse with regard to set up, method, and replicability, and along with the issues of sample size, lessens both its validity and reliability.
Finally, with respect to generalizability, this study seems lacking. As mentioned previously, the school of thought of behaviorism has essentially extinguished in contemporary times, and while this study was important during 1948, it had no targeted effects on society as a whole. As for animal behavior, it contributed greatly to the development of random conditioning in pigeons and other creatures, but lacked generalizability across other species, including humans.