Imagine a society without division of labor. The marksman would have no use for his talent in shooting and capturing game, and the scholar would not be able to pursue his intellectual interests. In allowing each and every person to do their best work, division of labor benefits all [1]. However, in order for this division of labor to exist, work must yield some reward. If a certain occupation does not allow for a person to make a living, that occupation will have no appeal, and few will be willing to do it when presented with alternatives, thereby robbing society of the benefit that it would bring. An acute case of this was the Great Depression. As a result of this sudden and steep economic downturn, manufacturing and production rates went down almost 72% in 1930 and 1931, and unemployment skyrocketed. The trust that one may always find a niche or position for oneself to prosper in one’s own craft is fundamental to society. However, upon delving deeper, it almost seems paradoxical. If an employer can get an employee to work for no wage, why shouldn’t they try to swindle their workers every time by being dishonest and cruel?
Let’s map this problem into another to try and gain some insight. This simple, illustrative, and general example is often used as an anchor point for introductory game theory. Let’s say we seat two people across from one another. There is a total of $10,000 between them. Each player can either split the money, or steal it for themselves. Now, let’s formalize. Player one’s set of choices to comprise set A, and player two’s to be set B. Therefore, set A= {Split, Steal} and, equivalently, set B = {Split, Steal}. Then, let S=AB. This is the “Cartesian product” of A and B, and returns a set which contains every single possible combination of the elements of sets A and B. Therefore, S= ({Split, Split}, {Split, Steal}, {Steal, Split}, {Steal, Steal}). S is known as the set of strategy profiles [3].
Fig. 1 – The two players’ set of outcomes. Image: Nikolai Morozov
In this case, each strategy profile bears different consequences. When both players split, each receives $5,000. When one steals and the other splits, the stealer gets all $10,000. Finally, when both steal, neither gets any money. When A goes up to the plate, they have two options: steal, and get either $10,000 or none, or split, and get either $5,000 or none. The choice seems obvious for player A – steal every time! The same conclusion we reached earlier with the employer.
This, of course, is not how we run society. If everyone defected constantly and never cooperated, our modern society would crumble in an instant. Still, if both players are rational and competent, then defection seems like the most logical choice. The reason why this crumbles is because of a hidden assumption. In saying that each player wants the money purely for themselves, we assume that they are minimally benevolent and maximally shortsighted. Over the long run, cooperators beat out defectors by playing each other, even if they lose money to defectors most of the time—the only reason why this one-time case is inaccurate is that it is a single moment. When a cooperator plays a cooperator, each interaction is a guaranteed $5,000, whereas a defector’s opponent catches on quickly and defects in turn, resulting in no gain for either one. Therefore, society stands upon the mutual understanding that defection leaves progress stagnant. Now, coming back to our original problem, we see why the employer doesn’t exploit his employees. If an employer defects by robbing his workers of their pay, except in extreme circumstances, few will be willing to work for him anymore. Eventually, with no workers, he’s out of a job. This is known as reciprocity, where all societally relevant actions cause an equivalent societal reaction [4].
However, because of the scale of human society, why we should cooperate still remains obscure. After all, nations have millions of inhabitants who may share different motives and beliefs, where some groups may appreciate different things than others. For example, in the community of employers, those who turn the best profit are lauded, whereas in the community of employees, it’s those who achieve a better work-life balance while still getting the best wage. The objectives of these two highly interdependent groups are contradictory. One way to explain this is to turn to species which cooperate similarly to us. Social insects such as bees, for example, are able to create beehives, which are complicated, organized societies [5]. At first, these species’ altruism was believed to accord with kin selection theory. This theory rests on inclusive fitness, which is the sum of the fitness one organism gains, and the fitness gain to the other organism it interacts with, scaled by relatedness, in an interaction. Thus, increasing the fitness of a highly related organism to oneself effectively increases one’s own fitness. This theory then posits that this is the origin of cooperation [6]. This result is expressed by equation (1), where if one organism is sufficiently related to another, it is favorable for them to choose to cooperate.
(1) R > c/b, where R is relatedness, c is cost, and b is benefit, as formalized by W.D. Hamilton in 1964.
An interesting facet of this is haplodiploid theory, which states that the clade hymenopterans (which includes bees, ants, etc.) is more likely to exhibit eusociality due to its mating habits and the principle stated above. In a colony of this clade, all males are haploid, since they are born through the unfertilized meiosis of females. Therefore, they share half of their DNA with their mothers, and none with their siblings. However, when these haploid males mate with diploid females, the resulting children both have all of their father’s genetic information, and half of their mothers’. Therefore, they are guaranteed to share half of their genomes, derived from their fathers, and on average share a quarter of their genomes deriving from their mothers. Hence, sisters typically share ¾ of their genetic information with each other, while only sharing half of it with their mothers, meaning they are more highly related to their siblings than their parents, promoting eusociality. However, by the 1990s, this hypothesis had all but been abandoned. As new species exhibiting eusociality were found, the correlation between haplodiploid reproduction and cooperation dipped below statistical significance. Unfortunately, it had too many assumptions to ever be realistic. It assumes that all interactions are pairwise and additive, and hence excludes many synergistic, complex games which are closer to simulations of real human interaction. Though there has been some debate on the validity of kin selection theory [7], this is the widely adopted conclusion. When the fitness (the so-called inclusive fitness) an individual gains is simply seen as the additive sum of the fitness it gains and the fitness its partner gains times their relatedness, larger-scale, more involved interactions cannot be modeled.
Instead of it being because of kin being close to kin, the unity of certain groups may instead be caused by common cultural background. To see why, suppose that we have a large group of people, and that they intermix and move around rather slowly, all while picking up local cultural traits rather quickly. Say they span two environments with low cultural exchange. Through rapid local adaptation, those living in one particular environment will develop a culture centering around the maximization of fitness. Thus, they will initiate cultural selection, where both groups quickly become distinct due to the systematic passing down of knowledge from generation to generation. Every single subsequent generation will take what the last one left behind, tweaking and improving it, leaving just a slightly better final product behind. Take, for example, Northern Alaska, where the kayak is crucial for the whaling of its indigenous tribes, and yet no single human intellect could come up with the perfect kayak through pure power of reasoning because of its enormous spatial complexity. Over centuries of tinkering, however, where the best “adaptation” to the kayak is the one that lasts, the kayak at least approaches its ideal form through its own form of selection. As Charles Darwin said in his book The Descent of Man, “...if some one man in a tribe, more sagacious than the others, invented a new snare or weapon, or other means of attack or defense, the plainest self-interest, without the assistance of much reasoning power, would prompt the other members to imitate him; and all would thus profit” [8]. As such, cultures can evolve and separate themselves from one another, leading to many small groups. This is shown in Figure 2.
In these groups, any behavior can be kept stable. For example, if one improves the current design of kayak, their reputation benefits. If they steal, it worsens. Thus, any behavior can be kept stable when employing the three Rs: reciprocation, reputation, and retribution. ‘Good’ acts come back through reciprocity and retribution, and improve reputation, all the while ‘bad’ acts decrease reputation, thus causing negative reciprocation as the result of retribution. Depending on a group’s set of values, cooperation may be favored more or less. Finally, the mechanism by which cooperation evolved in human cultures becomes clear. More cooperative groups tend to outlive less cooperative ones: after all, a society may live long if a soldier is willing to sacrifice themselves for their peers and their country.