Skip to main content

Monkeys and information cascades

In 1967, psychologists conducted an interesting experiment where monkeys demonstrated a surprising fact about group behavior. A group of monkeys were placed in a room, with a ladder and a banana at the top. If a monkey attempted to climb up the ladder, the rest of the monkeys were showered with cold water. Soon, every time one monkey climbed the ladder, the rest of the monkeys beat it up. After a while, none of them tried climbing again.

Then, a monkey was replaced with a new monkey from outside. It naturally went for the banana, but was beat up by the rest. Once the second of the original monkeys was replaced, the first replacement monkey joined in the beatings even though it never experienced the cold shower. This process was repeated, and after all of the original monkeys that experienced the actual cold shower was replaced, the monkeys still participated in the beatings whenever one went for the banana even though there was no cold shower.

It is certainly surprising that a group of social agents can perpetuate a behavior that they do not know the reason for. In a way, the result sounds similar to our model of information cascades; once a cascade is set going, even without the lack of a signal, new individuals keep making the same choice.

However, the difference in the experiment is that the initial group of monkeys all shared the same experience of being cold showered (describing using the cascade model, they also knew about each other’s signals). As much as this difference will make us deviate from the analysis in class, this points to an interesting observation: if the initial group of members in an information cascade collaborate, they can choose the direction of the cascades. It is plausible that the initial group of monkeys could have let the first monkey grab the banana despite their cold shower, and could have shared the banana. Then, since the monkeys were never showered after the initial round, all the subsequent monkeys would have enjoyed the bananas. In a way, the selfishness of the initial group of monkeys ended up starving the entire sequence of monkey population.


Original source:

Stephenson, G. R. (1967). Cultural acquisition of a specific learned  response among rhesus monkeys. In: Starek, D., Schneider, R., and Kuhn,  H. J. (eds.), Progress in Primatology, Stuttgart: Fischer, pp. 279-288.

Good diagrams to see the experiment:

– bo burger


Leave a Reply

Blogging Calendar

November 2012
« Oct   Mar »