Monkeys, Morality, and Information Cascade
Information cascades happen when people base their decisions off of others’ actions, without even knowing why those other people took the action they did. You experience this on a daily basis. For example, if you visit a new town and want to go out to eat, you may just people a place that seems busy. Unfortunately, it may only be busy because the current customers followed the same logic, and not because the restaurant is actually good. At first, this might seem like a cool, surprising result from probability theory, and nothing more. However, it has grave implications on how we behave from a morality standpoint. If we allow these information cascades to determine which actions are good versus bad, and morality tries to distinguish good versus bad behavior, it’s only a small jump before information cascades defines our morality. In his blog, John Stepper recounts a disturbing experiment that makes this possibility all too real. It’s called the Five Monkeys Experiment.
In this experiment, five monkeys are put in a room with a ladder. A hand of bananas is at the top of a ladder. Every time a monkey would try to reach the banana, all of the monkeys were sprayed with cold water. By doing this, the monkeys were trained to prevent one another from trying to climb the ladder. One of the monkeys was replaced with a new monkey. That monkey, upon seeing the bananas, started to climb the ladder. The other 4 monkeys immediately pulled the climbing monkey away and beat him.
When another one of the original monkeys was replaced with a new monkey, a similar scene ensued. The new monkey started to climb the ladder, but all of the other monkeys pulled him down and beat him. What’s terrifying is that the first new monkey also beat this new monkey. That monkey had learned not only that climbing the ladder was bad, but also that he should beat any other monkey that would try to climb the ladder. This process was repeated until all original monkeys were replaced. This information cascade resulted in 5 unhappy monkeys who wanted the banana, but would beat one another if any monkey tried to get it. Oddly enough, none of these new monkeys knew why there weren’t allowed to have bananas. They would be beaten by the other monkeys before even getting sprayed with cold water.
Although we like to, as humans, we are much more critical than these monkeys, we are still potential victim of information cascades. Stepper points out that in the work place, information cascade can manifest itself as innovation being suppressed in the work place. Taken to the extreme, information cascades can result in horrible things. Racism, slavery, genocide–many of these self-propagating phenomenons are only possible because people did not question the cause for people’s behavior. Asking what what people do rather than why leads to information cascades. We often associate racism with ignorance. We call slavery barbaric. We always ask, “How could people do that?” when we think of genocide. We do this because we can’t help but think that if only these people knew better–if they weren’t so easily influenced by the people surrounding them–then maybe they’d act different. It’s frightening to thing that ignorance and as a result information cascades can cause such horrible things, but by educating ourselves and asking why rather than what, we can combat becoming victims of and perpetuating information cascades.