Group Selection

What we can learn from Monopoly, psychopathic chickens, and cancer

Giulia May

We have previously introduced a selectionist view of the world. To briefly recap, traits vary, adaptive traits are selected and replicated and non-adaptive traits are not. In an organism’s lifetime, behaviours that lead to good outcomes replicate, and behaviours that do not are suppressed or undergo extinction. Many of our behaviours can be explained by these processes.

Often, each behaviour is the product of multiple selection pressures. Sometimes, these selection pressures come into conflict with one another. We have discussed one process by which this occurs in delay discounting. Sometimes, we eat too many brownies because they are too damn delicious, or we find ourselves still awake at 3 AM watching our 5th consecutive episode of The Last of Us the night before an important meeting. In both instances, the present self makes choices that the future self regrets. Analysing behaviour solely at the individual level typically suffices.

Yet, there are cases where we must go beyond analysing behaviour at the individual level. A passer-by leaps into a freezing river to save a drowning child. Mother Teresa forsook her comforts to serve the poor. Aragorn led a suicide charge so Frodo has a shot at destroying the One Ring.

Maudit

Less heroically, a colleague of mine regularly donates blood. I restrict my meat intake to one meal per day to lower my carbon footprint. In each instance, the individual makes choices at their own expense for the benefit of others. That’s strange! If selection favours adaptive traits, why would these seemingly maladaptive traits persist? To answer, we must expand our unit of selection from the individual to the group.

Today, we will unpack group selection through three analogies—Monopoly, psychopathic chickens, and cancer—borrowed from the evolutionary biologist David Sloan Wilson. I hope I articulate them at least half as eloquently as he did. Needless to say, I hold Wilson in the highest regard.

Monopoly

Imagine you are playing a game of Monopoly with 3 others. The goal is simple—bankrupt them before you yourself are bankrupt. You acquire properties, manage your cash wisely, plot, scheme, and do anything to build and hold an advantage. You may form temporary alliances to take down an opponent who is ahead, but you will just as quickly turn on your allies if it advances your own position. Ultimately, your actions are completely self-serving.

Imagine now that the rules are slightly tweaked to create a 2 vs. 2 game—bankrupt the opposing pair before both you and your partner are bankrupt. Pause for a moment to consider the kinds of behaviours you will engage in now. If you are anything like me, your goals and actions will radically differ from the single-player version. In place of me goals, you now have us goals. How do I get ahead is replaced by how do we get ahead. If you cared to notice, other behaviours tangential to the game itself also emerge. You find yourself smiling at your partner, giving them high fives, then fist bumps, then secret handshakes. A warm feeling of fellowship spreads through your body. At the right moment, a blasphemous thought might even enter your head—I’ll sacrifice myself and go bankrupt if it means my partner crushes them.

We have arrived at our first axiom:

Different behaviours are selected for at individual and group levels.

When group selection pressures are sufficiently strong, some behaviours that appear maladaptive at the individual level surface. Conversely, other behaviours that appear adaptive at the individual level are suppressed (the plotting and scheming continues, but not at my homie). Taken together, we have the rudimentary beginnings of complex group phenomena such as cooperation, altruism, and morality.

Psychopathic Chickens

Imagine you are a poultry farmer, and you want to increase egg production in your brood. What is your strategy? Conventional wisdom has it that you will pick the most productive hens, use them to breed the next generation, and so forth. After a few generations, you should expect to possess a brood of egg laying machines. In the 1990s, biologist William Muir and his colleagues did just that. In the span of a few generations, the conventional wisdom hypothesis was thoroughly refuted. Instead of laying more eggs, each subsequent generation laid fewer eggs. Whatever the cluck happened here?!

Turns out, the most productive hens were also the biggest bullies. They achieved their productivity by aggressing towards other hens. When the bullies were put together, they went to town on each other, leading to a steep decline in egg laying. Each subsequent generation grew more psychopathic than the last. You can see an image of one of the cages after five generations of breeding below. Of the 9 hens the cage originally housed, 6 were murdered, and the remaining 3 had stripped each other of their feathers.

Not the kind of hens you want near your children. Muir et al. (2014)

In a parallel experiment, Muir and his colleagues monitored the productivity of entire groups of hens and selected the groups with the highest productivity. As you can by now guess, productivity increased greatly after a few generations, the hens did not murder each other, and they sang Kumbaya in Henglish. Here is an image of one of the cages after five generations, with all hens alive and fully feathered.

Muir et al. (2014)

With that, here’s our second axiom:

The similarities between our tale of hens and our own organisations and network systems might be immediately apparent to some of you. Some of you may be in groups similar to the psychopathic chickens, groups that select for individual achievement and selfish traits. Often, there is intense within-group competition, resulting in a few winners (our three unfeathered hens), many losers (our six unfeathered and dead hens), and a severely dysfunctional group. Other groups are like our Kumbaya chickens, selecting for group-level achievement. We don’t see many winners or losers in these groups, but we see a strong, cooperate group.

In reality, most groups fall somewhere in between—individual achievement is rewarded, as is group success. No one wants their own achievements to go unnoticed, and the fate of a group is determined by outcomes of the collective efforts of its members. Further, placed in an environment promoting individual achievement, not everyone turns into a psychopath. Some folks absolutely have good intentions, and individual success does not always have to come at the expense of others. The interplay between individual and group selection is complex. It is important for groups to critically evaluate and evaluate often their processes and the corresponding selected behaviours.

One final point before we move on. A while back, I was having a discussion with some friends on diversity in government. One of them believed that competency should be prioritised over all else, including voter representation. In short, governance should be based on competence, not populism. I fully agree that our government should comprise the most competent individuals. The challenge lies in defining competency. To some, we should select high status males. I suspect, for my friend, the selected group will be mostly male, Chinese (the majority ethnic group in Singapore), likely to have attended the same elite schools, worked in the same places, and have very similar lived experiences. I am inclined to disagree with this selection, but he may be right. Either way, our axiom holds true—be careful what you select for.

Cancer

First, a short detour before we discuss cancer. In our Monopoly and psychopathic chicken examples, we examined selection at two levels—the individual and the group. Life is far more complicated than that. Groups can be nested within bigger groups, themselves nested within even bigger groups, so forth. When I served in the army, I was in a team, within a platoon, within a company, within a battalion, within a brigade, within a division, within the army, within the Singapore Armed Forces. Zooming in, each of us consist of smaller groups of cells (including many not our own) that comprise yet smaller groups down to the most basic self-replicating unit of the gene.

At each level, new selection pressures are introduced that compete with selection pressures at lower and higher levels. In short, group selection occurs recursively across multiple levels. For our third axiom:

Group selection is really multilevel selection.

Onward to cancer.

Each of us consists of trillions of cells that work together to sustain us. Our cells routinely divide and differentiate to replace other cells that die. Occasionally, mutations occur that allows a cell to grow rapidly, avoid programmed cell death, and even increase the rate of subsequent mutations. Cancer provides a fascinating insight into the tug-of-war between selection pressures at different levels. These adaptations are obviously very good for the mutated cell and its lineage, and bad for us. Our bodies react to nullify these cells, failing which, we die.

Acute myelocytic leukaemia. National Cancer Institute

Our cancer analogy applies for across multiple level-pairs, including cell-individual, individual-small group, and small group-larger group. At the lower level, the goal of the unit is to outcompete other units by behaving selfishly. At the higher level, the goal of the group is to reduce selfishness and ensure cooperation among its constituent units. After all, a group in shambles has scant hope of outcompeting other groups.

We have arrived at our final axiom. In the words of David Sloan Wilson and E O Wilson (no relation):

Selfishness beats altruism within groups. Altruistic groups beat selfish groups.

Once we understand this, we begin to see it everywhere. We have seen an example of this with our psychopathic chickens. Within group, the three remaining hens produced more eggs than their dead counterparts (and they didn’t die), but the Kumbaya group beat the psychopath group handily. An interesting example is unfolding today in the United States. Like mutated cells, fringe groups with their own agendas can be found in both major political parties. In the case of one, a fringe group has seized control over the party. It remains to be seen if this party can still effectively compete for votes.

In Summary

We have journeyed through Monopoly, psychopathic chickens, and cancer. In the process, we have learnt how different behaviours are selected at individual and group levels, why we must be careful what we select for, and how selection occurs recursively across multiple nested levels. Perhaps most crucially, we've learned that while selfishness may triumph within groups, altruistic groups ultimately outcompete selfish ones.

These insights offer more than just explanations for observed behaviors—they provide a powerful framework for shaping our future. As we design our institutions, craft our policies, and build our communities, we have the opportunity to consciously consider the selection pressures we're creating, and how selection pressures at different levels interact with each other.