– by Victor Vikram Odouard

Driving my Fiat 500 convertible through the Swiss countryside surrounding Geneva, a sign caught my eye. In large hand written letters, it read “CAiSSE” – “CHeCKOUT” in French. Behind the sign, a heap of pumpkins lay on a threadbare canvas blanket. But unlike most checkouts, no one was there. And unlike most self-checkouts, there were no cameras in sight. No, this little Hofladen relied on the honor system.

Why wasn’t there a horde of pumpkin bandits raiding our humble farmstand?

Scientists have proposed many mechanisms to explain the evolution of altruistic behavior, such as kin altruism, punishment, and direct & indirect reciprocity. These all share a common characteristic: they provide a second-order benefit to altruists, that compensates for the initial cost.

Most of these mechanisms require knowledge of who is cooperating (I will sometimes call altruistic behavior cooperation in this post). Sometimes that knowledge can come from observation, as is the case with reciprocity (I know who did me a favor in the past). But other mechanisms, like third party punishment and indirect reciprocity, require that knowledge of who has cooperated spread to those who were unlikely to observe the interactions in question. This requires a faithful communication system (read: gossip!) that accurately transmits the relevant information.

And here we arrive at the question we pose in this paper: why should we even have such a communication system, when it might be in people’s interests to lie?

The question echoes the original cooperation question – “why should people benefit society by acting altruistically when it can come at a personal cost?” Here, we simply replace “acting altruistically” with “communicating truthfully”. The parallel points to a solution – perhaps the same mechanism that keeps people cooperative can also keep people communicative.

In our paper, we analyze a model of indirect reciprocity to determine whether such a mechanism can exist. Imagine an agent, Ariel. She meets someone else, Blake, and she can make a choice: she can cooperate or defect. If Ariel cooperates, she pays a cost in order to produce a benefit for Blake. Let’s say she does cooperate. Afterwards, Blake can “gossip” about Ariel’s behavior, in this case, “Ariel good”, since she cooperated. In the next interaction, when Ariel meets Corina, and Corina is deciding whether to cooperate with Ariel, Corina is aware of Blake’s gossip – in this case, that Ariel cooperated. Blake’s gossiping thus produces Ariel’s reputation. Corina can then, for example, decide to cooperate with Ariel since she cooperated with Blake. This is indirect reciprocity.

The twist in our model is that the indirect reciprocity “twists back” on itself: instead of reputations being based solely on how agents act, they can also be based on how agents gossip. How does this work? Returning to our previous scenario, let’s say Ariel cooperates with Blake, but Blake nonetheless says that “Ariel bad”. Perhaps someone, Diana, saw the whole thing go down, and she communicates “Blake bad”, because he lied. In Blake’s next interaction, with Ed, Ed could therefore decide not to cooperate with Blake.

In this scenario, the machinery being used to maintain cooperation (accurate communication) is being co-opted to maintain itself! In our paper, we ask whether such a self-referential system can actually lead to a stable, high-cooperation state, and we found that yes, it could, in the case where (1) agents made occasional (but not too frequent) errors and (2) the norm punishes defection while nonetheless being self-consistent. Both of these conditions will require further explanation.

The first condition reflects an important fact about many complex systems, which is that heterogeneity can enhance stability: for instance, a virus often has a harder time sweeping through a population with many diverse immune responses to its infection, as it must circumvent many different types of defenses. Taking this one step further, if a population has been infected by many different sorts of viruses in the past, it can mount a diversity of different immune responses to any new pathogen – thus, a diversity of past “disturbances” (viruses, in this case), can make a population more robust by diversifying its immune defenses. This is exactly what we see in our model, where error introduces diversity by causing, for instance, some defections, or some untruthful gossip, where they wouldn’t usually occur. We found that this diversity of small-scale disturbances inoculates the system against more serious destabilizing forces.

As for the second condition, a norm in our setting is a determination of whether a given action is good or bad. A gossip system can encode such a norm: for instance, it might call cooperation and truthfulness “good” and defection and untruthfulness “bad”. Agents can then respond to those judgments: for instance, in one of the above examples, Blake was deemed “bad” for having lied about Ariel, and Diana therefore refused to cooperate with him. But a question arises: does the norm judge Diana’s action, which is in effect enforcing the norm (cooperating with those who followed the truthfulness norm, defecting with those who didn’t), as good? In this simple case, no – defection is considered bad. So Diana is judged as “bad” under the very norm that she is enforcing! This norm is therefore not self-consistent – a norm is self-consistent if it does reward the actions enforcing it. A self-consistent norm instead might deem cooperating with “good” agents good, and defecting with “bad” agents good, and everything else “bad”. This norm, called stern judging, is self-consistent, because it rewards the behaviors that enforce it. For instance, there are some societies that consider theft to be permissible if committed against people in bad standing in the community (for instance, people who unjustifiably stole from someone else). This is a self-consistent norm, as the punishment of individuals in bad standing is permissible under the norm’s own rules.

In our analysis, we find that maintenance of cooperation and communication requires such a self-consistent norm. This highlights that for decentralized norm enforcement, self-consistency is critical. If norm enforcement is centralized, self-consistency need not apply, as different rules may exist for judging the centralized body of enforcers and everyone else. For instance, most people judge killing to be bad, but in some countries, this judgment does not apply to a government doling out the death penalty. However, with decentralized norm enforcement, the people who enforce the norms are the people, and therefore no such distinction can exist. Self-consistency is thus needed.

Our study is fundamentally about whether a self-contained norm-enforcement system, not buttressed by anything external, can maintain high levels of cooperation. The external buttress that provoked our study was communication – most indirect reciprocity models assume an already-truthful communication system. But by noticing here that communicating truthfully is not so different from cooperating (in that both provide a public good), we hypothesized that the same indirect reciprocity mechanism enforcing cooperation could enforce truthful communication, without relying on a preexisting communication. The two conditions under which such a state was stable emphasize other important properties of a self-contained norm enforcement system. First, that it can “create its own diversity”, through error, allowing the system to function without being propped up by some external source of diversity. And second, that the norm is able to “reward its own enforcers”, by being self-consistent – this allows enforcement to occur without requiring a separate body of enforcers, operating under different rules.

Read the original article: Odouard, V.V., & Price, M.H. (2023). Tit for tattling: cooperation, communication, and how each could stabilize the other. Evolution & Human Behavior, 44(4), 359-372.

– By Amy Zhao and Brendan Zietsch

Ariana Grande and Ethan Slater have been making headlines due to their alleged affair, which occurred while both parties were married to their respective partners. Another aspect of their relationship that hasn’t gone unnoticed is Ethan Slater’s striking resemblance to Ariana Grande’s brother, Frankie.

This is just one example of a couple who look like siblings. Many of you may even have your own stories of having mistaken a couple for being siblings, a phenomenon so common that it has led to the creation of a popular Instagram account called Siblings or Dating.

Despite all this anecdotal evidence, there has been mixed empirical evidence as to whether couples actually share facial similarities. But even if we assume that couples do have similar faces, it’s unclear why someone would even date another person who looks similar to themselves in the first instance.

Some researchers have proposed that couples might look similar because couples share the same lifestyles, diet, exercise, stressors, etc. and so, couples’ faces become more and more similar over time. However, there is mixed evidence to support this hypothesis.

Instead of couples’ faces converging over time, an alternative theory is that individuals prefer facially similar partners to begin with; that is, individuals find similar faces attractive. However, there is mixed evidence for this association too. The studies from which these mixed findings have been derived are flawed in several ways. Firstly, these studies have exclusively measured facial attractiveness through participants’ ratings of 2D images. These images consist of photographs of real-life participants, or they are computer generated/digitally morphed faces. There are several implications of using images to obtain attractiveness ratings. Firstly, humans have relied (until most recently) on face-to-face interactions to interact and form judgments about another person, especially another potential mate. And secondly, morphed faces can lack realistic features such as facial hair, skin texture, or head hair. Therefore it is unclear whether past findings can validly reflect real-life behaviours and preferences. In addition, facial traits (such as similarity, averageness, and masculinity) are often subjectively rated by participants. These facial trait ratings may be conflated with ratings of facial attractiveness and can be subject to other biases (e.g. the halo effect).

We conducted a large speed-dating experiment with 682 participants and there were 2285 speed-dating interactions. Participants speed-dated members of the opposite sex for 3 minutes at a time. After these interactions, participants rated their speed-dating partners on facial attractiveness, and kindness and understanding. We landmarked the facial photographs of all participants and calculated objective, data-driven scores for facial averageness, facial similarity, and facial masculinity to address the limitations of subjective ratings of such facial traits.

We found that participants rated partners who had geometrically similar faces to their own as more attractive, showing that individuals do in fact have a preference for facially similar opposite-sex individuals.

The question still remains, why do people even find facially similar people more attractive in the first case?

We found that people with similar faces also rated speed-dating partners as more kind and understanding, even when controlling for the ethnicity of the participant and partner. Our findings are in line with an alternative explanation where a preference for similarity is not driven by the attractiveness of similar faces per se, but that similar faces may signal kinship. While cues of kinship can decrease sexual desirability due to the costs of inbreeding, the same cues can increase positive social behaviour by promoting increased chances of survival via altruistic and cooperative behaviour from others (inclusive fitness benefits). There is also evidence that positive assortative mating (mating with someone who is similar) can result in net positive reproductive benefits, as a small degree of genetic relatedness is associated with higher fertility.

Sure, we have shown that individuals do prefer similar faces, but does this preference extend to real-life mate choice?

We’re not sure. Actual mate choice involves the mutual desire to form a romantic partnership, but these decisions are constrained by the mating market. For instance, individuals with high mate value are unlikely to pair with low mate value individuals. Given that not everyone can satisfy their preferences, individuals must make compromises. In addition, the effects we found in our study are small, and therefore these preferences may not carry through to realised mate choice. It would be a worthwhile future endeavour to investigate how the objective facial characteristics measured in our study relate to real couples.

Read the original article: Zhao, A.A.Z., Harrison, K., Holland, A., Wainwright, H.M., Cecatto, J.-M., Sidari, M.J., Lee, A.J., & Zietsch, B.P. (2023). Objectively measured facial traits predict in-person evaluations of facial attractiveness and prosociality in speed-dating partners. Evolution & Human Behavior, 44(4), 315-323.

– by Marcell Székely & John Michael

Imagine that you and your neighbor collaborate to clear away weeds from the common garden where strawberries grow. Assuming you and your neighbor spend about the same amount of time weeding, a reasonable expectation is that you would divide up the strawberries equally.  In contrast, if your neighbor has spent more time weeding than you have, you might think it reasonable that they should receive a greater share of the strawberries. Or if your neighbor spends less time than you weeding, you might think it reasonable to expect more strawberries for yourself.

This everyday example illustrates that we have a sense of fairness, which leads us to divide rewards of joint actions equitably among contributors. Moreover, it shows that this sense of fairness is not just a blind tendency to divide rewards equally but that it takes effort contributions into account. And indeed, recent research has confirmed that our intuitions about this hypothetical scenario are reflected in people’s actual behavior in controlled scenarios. For example, Frohlich and colleagues found that when participants were placed in one room and had to proofread a text to correct spelling errors for joint rewards, they divided their collectively earned rewards in proportional to individual effort costs. Even more strikingly, recent studies in developmental psychology have even shown that three-year-old children take effort costs into account to achieve a fair distribution of rewards. For example, Kanngiesser and Warneken found that three-year-old children kept more stickers after working more than their partner.

But now let’s tweak our hypothetical example: imagine that you and your neighbor partner are tidying up the stairwell together. Notice that in this version of the hypothetical scenario, there is no tangible reward that could be divided up at the end, so fairness cannot be assured through the division of rewards. So what would you do if you perceived that your neighbor was investing a higher level of effort than you? Or a lower level of effort?

In a new article, we describe a series of experiments which we conducted to investigate how people deal with such scenarios. In the experiments, we adapted an effort discounting task developed by Hartmann and colleagues. In the adapted paradigm, people are asked to make decisions about how much effort to invest (namely, by spending time pressing the spacebar repeatedly) for different monetary rewards. This makes it possible to quantify the costs people associate with different levels of effort and to investigate how people trade off those costs against monetary reward.

We turned this paradigm into a social task by letting participants observe their partner performing an effortful cognitive task in order to ‘unlock’ the effort discounting task. Previous research has shown that people spontaneously track others’ ongoing effort investments, and indeed do so accurately. Moreover, we were confident that participants would track the partner’s effort investment using these stimuli because this effort perception manipulation was successfully used in Székely and Michael and Chennells and Michael. Sometimes, participants were rewarded jointly, sometimes they were rewarded separately. This enabled us to test whether participants chose to invest different degrees of effort against monetary reward based on their relationship to the other person. In other words, it enabled us to investigate how social context modulates people’s decisions about how much effort different rewards are ‘worth.’

Our main finding was that people invested more effort when they were rewarded jointly with their partner than when they were rewarded separately. This provides evidence that when rewards cannot be divided to ensure fairness, people instead boost their effort investment as a means of ensuring fairness with a cooperation partner. In other words, the sense of fairness governs not only the distribution of rewards but the allocation of effort in joint actions.

This is important because, while most research on the evolution of the sense of fairness has focused on the division of rewards, a great many cooperative endeavors now, as in deep evolutionary history, do not produce any distributable rewards at all. One reason for this is that the success of cooperative endeavors is always at least to some extent uncertain. Like individual endeavors, they may not succeed because of exogenous factors or lack of competence. In addition, they may not succeed because the collaborative partner may encounter a tempting outside option and so may prematurely withdraw from the joint activity. For example, hunting and foraging in ancestral environments were uncertain endeavors, and they may not have yielded any rewards to distribute at all. Moreover, even if a joint activity yields a reward, this reward may not be divisible. For example, tidying up the stairwell or building a shelter together may not produce any divisible rewards. In such activities, there is nothing to distribute at the end, but it is still possible to distribute the effort costs. This may explain why our participants were willing to invest more effort in a joint action, and thereby to increase the potential rewards for themselves and their partner, when their partners had apparently invested a high degree of effort, even if the partner’s effort investment did not increase the potential rewards.

These findings also open up important new avenues of research on the evolution of cooperation: most of the empirical studies investigating the evolutionary origins and motivational mechanisms to cooperate have focused on the exchange of money, and their results are commonly used to support claims about humans’ willingness and motivation to cooperate more generally. While this focus on monetary rewards is justifiable, people often have to make decisions about how to divide up other resources as well as costs in the context of cooperative endeavors – and had to do so in our deep evolutionary past. It is therefore crucially important to extend the scope of our focus to resources other than monetary rewards. The current findings are an invitation to take up this important challenge.

The findings also provide new impulses for future research about whether and how the sense of fairness may be sensitive to specific features of cooperation partners, such as their competence. While we intentionally bracketed out questions about the relative competence of different actors for our study, parties to cooperative endeavors often contribute with different levels of competence. We look forward to the results of future research investigating how this is reflected in people’s decisions about the distribution of rewards and allocation of effort costs.

Read the original article: Székely, M., & Michael, J. (2023). In it together: evidence of a preference for the fair distribution of effort in joint action. Evolution & Human Behavior, 44(4), 339-348.

– by Megan Bishop & Brian Lerch

Large-scale, advanced cooperative action that transverses familial lines is uniquely, exceptionally human. It lies in the foundation of our economic and political infrastructure, our corporate workflow, the small talk we make with the grocery cashier –evidence of our cooperative, pro-social nature is everywhere. Put simply, the evolution of human cooperation has bound us all to each other in a way that far transcends anything seen elsewhere in the animal kingdom. Such an incredible phenomenon begs a multitude of questions, but upon beginning the work leading up to my recent publication, The influence of language upon the evolution of cooperation, I became particularly enthralled by one: how?

Pursuing this broad question led me to another uniquely, exceptionally human development: advanced spoken language proficiency. The evolution of advanced human cooperation and language have been argued to have a reciprocal relationship, where the evolution of one drives the evolution of the other and vice versa. While the existence of large-scale cooperation assumes an ability to coordinate via complex language systems, few mathematical models examined the ways that language promotes cooperative action, despite the utility of mathematical models for analyzing how evolutionary mechanisms align with argued verbal logic.

Human cooperation has important roots in collaborative foraging, which is well modeled by a stag hunt game. Here, a group of early human foragers are presented with two options: individually hunt a hare with a guaranteed fitness benefit or risk the opportunity cost of the hare to collaborate with others in pursuit of hunting a stag. Upon the successful hunt of the stag, the fitness payoff would be evenly split amongst the group, with a much higher payoff than the hare. However, this depends on successful cooperation. Without successful cooperation, the stag hunt fails, leaving cooperators without the fitness benefit from the hare or the stag. Given the important evolutionary origins of collective human foraging, we build upon an existing stag-hunt model and include language in three various mechanisms to analyze its implications on early human cooperation.

First, we analyze an “enhancement” mechanism of language, where we assume better language systems allow for more complex coordination among cooperators, allowing them to pursue larger game with a greater fitness payoff. Here, our findings align with verbal arguments: more proficient language always makes conditions more favorable for cooperation. In our model, cooperation steadily evolves when language proficiency is better. However, this is limited by the proportion of cooperators in the population. Regardless of language proficiency, language cannot promote the evolution of cooperation with too few cooperators present.

Next, we studied an “assortment” mechanism of language, where we assume better language systems allow cooperators to communicate with, and therefore identify, other cooperators. In other words, with better language proficiencies, cooperators are more likely to join groups with other cooperators. Like our “enhancement” mechanism, our findings align with verbal arguments that better languages always favor the evolution of cooperation when enough cooperators are present in the population. However, this “assortment” mechanism is especially beneficial for cooperation when there is a high frequency of cooperators, leading to larger benefits of cooperation than is seen in the “enhancement” mechanism.

Finally, and most interestingly, is our “threshold” mechanism. Here, we assume better language proficiencies allow humans to achieve successful cooperation with fewer cooperators participating. This mechanism challenges the idea that better language proficiencies always favor the evolution of cooperation. We find that while better languages allow cooperation to evolve in populations with fewer cooperators, the overall fitness of individual cooperators decreases with better language. Upon reflection, this makes sense. Even when a smaller group can establish advanced complex coordination with improved communication, more defectors are able to “free-ride” off of the cooperators’ payoffs.

So, I answer my question of how? with an answer that frustrates answer-seeking scientists like myself, yet is universally true about our world: it’s complicated. It is likely a combination of these mechanisms that has influenced the evolution of human cooperation. For instance, perhaps the benefits of our assortment mechanism were slowed by the effects of our “threshold” mechanism. However, certain environments might not allow for the existence of one or more mechanisms, such as an environment with tightly constrained group membership would not enable positive assortment. The nuance between language and cooperation depends greatly on the mechanism of language. This model pokes and probes at our existing understanding and how language influences cooperation, unveiling that there is greater nuance than what is commonly assumed. My hope is that by challenging such a long-standing assumption, the scientific community continues to grow in our study and understanding of the fascinating, unique phenomenon we call human cooperation.

Read the original article: Bishop, M.E., & Lerch, B.A. (2023). The influence of language on the evolution of cooperation. Evolution & Human Behavior, 44(4), 349-358.