Friendship Evolved Behind The Bars

I was surfing the web to understand the origin of coronavirus. I was aware that the bats are the natural reservoir of a range of coronaviruses and serendipitously I landed on an article about a bizarre behaviour of Vampire Bats. 

Vampire bats often feed the hungry nestmates by regurgitating blood when they return to their roosts after successful foraging trips. One might think that’s a pervasive behaviour for many species in the animal kingdom to bring food for other members of their group. But in the case of vampire bats, that’s certainly not so simple. An individual bat will only survive for 70 hours without blood; therefore, sharing blood can cost the bat its life, and the receiver gets more benefit than the donor. Interestingly, it has been found that this act of generosity is more towards those individuals that have previously shared the blood with the donor. 

This triggered a very trivial but an unknown question for me that Why Do We Show Kindness? And How Animals Including Humans Evolve To Be Kind?

We see in everyday life that we show kindness and altruistic behaviour not only with those we share genes with but even with strangers and other animals. And not just humans, animals also exhibit this behaviour. You might have seen monkeys grooming each other, the jungle safari guide trying to hear warning cries of birds to locate and show magnificent predators to the tourists, or devoted papa penguins sharing the task of child care and so on. 

In our everyday life, we connected with people, with our friends, romantic partners who we don’t share genes with and yet we have a strong bond. We also show kindness to strangers and how that comes about is altogether a different story, but let us discuss a beautiful relationship called friendship. 

So, a way to understand friendship is in terms of what the biologist Robert Trivers has stated “reciprocal altruism”. In simple terms, it is a behaviour where one organism acts altruistically in a manner that temporarily reduces its fitness but increases the fitness of other, with the expectation that the receiver will act similarly in the future. 

We pay for our friends when they don’t have money, and we often don’t ask to pay back with the understanding and expectation that the friend will pay when we don’t have money in the future. But unfortunately, society is not so ideal, and we have a “problem of cheating”. Suppose I post a charming photo on my friends birthday, taking a risk (my parents can find out my id), with the hope to find a notification of my picture, in her status, on my birthday, but what if that never happens, she betrays.

When a bird in the jungle gives out a loud warning cry on seeing a predator, it risks its life by revealing its location to the predator, but it benefits all the other animals. But, do you know who would be best off? That “free rider” who just took advantage of the warning cries, who only listen to them and save its own life, but never give out warning cry for others by risking its life. So by the evolutionary perspective, this selfish animal always has a better payoff; should outproduce and reciprocal altruism should not be there in our behaviour, but here is a twist.

Reciprocal altruism evolved because animals do punish cheaters; they penalise free riders. Punishment can be kicking their ass, causing them pain or killing them. Shunning and ostracising them can also be an effective mechanism in cases where animals are mutually dependent and need interaction with others for survival.

Now even if we accept that these animals have evolved a mechanism for punishment. But that’s not all; this behaviour certainly requires a lot of psychological fitness and cognitive strength. The animals should learn how to find the cheaters, they have to record their identity and hold it in the memory across time, and had to be motivated to punish. 

One very general way to understand how to solve this problem in the evolutionary sense is through what’s known as the “Prisoner’s Dilemma”. 

Imagine two friends got arrested and interrogated by the police officer in a separate room for a suspected crime. The officer explains each friend separately about different years of imprisonment (payoffs) based on whether they chose to cooperate and remain silent or confess that another friend has committed the crime. Well, of course, in everyday life, no cop will sit and explain prisoners dilemma to the suspects. But, it is not so dissimilar to what happens with prisoners.

If both of them cooperate, both will get a reasonably mild sentence of one year in prison. If however, anyone squeals but other remains silent. The confessor will walk away free of charges, and the other will be punished severely for 20 years. If both of them squeal on each other, both will get a prison sentence for five years. 

So, the police officer further says, “So, look at this and be aware that your friend is looking at the same sheet, with same choices and same offers “and explains to them separately, “Look, its best to give up on your friend and you’ll get a lesser sentence.”

Now it matters on the trust between the friends and how much they know each other. Well, if friend A cooperates, that’s the bottom two squares, the best thing for friend B is to defect. That’s the bottom left corner and walk away. Suppose friend A defects, well, what’s the best option for friend B? Its again is to defect and get 5 years in prison, which is better than 20 years. So, when faced with this option, friend A says, “Fine, I defect, and I want to blame on my partner.” But Friend B got the same offer, and he also defects. 

In short, the best case is to defect while the other guy cooperates. The worst case is to cooperate when other defects. The best overall is if both cooperate and the worst overall is when both defect. The puzzle is, as the cop has reminded each one of them, regardless of what your friend does, it would be best to defect, but if both friends defect, both are worse off. 

Prisoners Dilemma is everywhere, and we consciously or unconsciously play it to decide our payoffs, and in the end, we are worse off as its the equilibrium of the game. Now, there may not be any solution to the one iteration of prisoner’s dilemma. Where you face another individual who’s a total stranger, you have interacted once, and you make your choice based on that single interaction. Now imagine yourself playing the game over and over and over again with the same individual, and you are asked to decide your best strategy to maximise payoff. Would you always defect or always cooperate? Would you select your choices randomly? What would you do? 

Robert Axelrod many years ago had organised a competition. The aim was to find the best strategy to play this game large numbers of times using computer programs. Many brains got involved and applied sophisticated algorithms and techniques. But the cool thing that happened was that the winning program was the most straightforward program they had with just four lines of code. It was based on a simple “tit for tat” strategy with two elementary rules. The first is, the computer always cooperates on the first move. The second is, on each successive trial, it imitates the behaviour of the opponent in the previous trial. You see, this is pure genius. It starts friendly. It’s not a sucker. If it starts friendly and opponent defects on it, it will defect back, but it’s forgiving. Once the opponent is kind and cooperates again, it’s going to be nice right back. It elegantly rewards niceness with niceness and meanness with meanness, and it’s transparent. 

The lesson is pretty clear, and it solves our problem on how to work together for mutual benefit. If you are going to play with a tit for tat machine a series of games, you will cooperate all the time because you know your cooperation will be rewarded equally. 

What’s fascinating is, it seems like our social emotions are calibrated to different aspects of the prisoner’s dilemma. So, we feel grateful and love those who cooperate with us; it motivates us to behave nicely to them in the future, we dislike and distrust those who betray us, as they drive us to avoid them in the future. Furthermore, we feel that guilt inside us, whenever we cheat those who had cooperated us. This emotion of guilt is a feedback which motivates us to be a better human in the future. 

The evolutionary and game-theoretic considerations provide some fascinating insights into the nature of this emotion. This simple, beautiful logic plays a critical role in understanding how not only humans but many other animals and even microbes evolved with this behaviour. 


  1. Wilkinson, G. (1988). “Reciprocal Altruism in Bats and Other Mammals”. Ethology and Sociobiology. 9 (2–4): 85–100. DOI:10.1016/0162-3095(88)90015-5
  2. Trivers, R.L. (1971). “The evolution of reciprocal altruism”. Quarterly Review of Biology. 46: 35–57. DOI:10.1086/406755

About the author: Akarsh is a fifth-year BS-MS student at IISER Bhopal and a KVPY fellow, majoring in Earth and Environmental Sciences. He is especially interested in Atmospheric and Aerosol Sciences. Apart from this, he likes to understand human behavior. He loves to interact, help people and mingle with strangers. In the free time, he loves to eat, play badminton, TT and read books.

He would love you to share your feelings and thoughts. You can approach him anytime on @ralhanakarsh ( or feel free to Whatsapp on +91-9424600656, Email – .

Other articles by Akarsh:

B for Behaviourism!

Leave a Reply

Your email address will not be published. Required fields are marked *