ABSTRACT
The potential to be a hero is in all of us. This is no mere platitude; it reflects a deep and simple truth: heroism is an extreme form of everyday ethical behavior. Heroes, by our definition, are people who make great personal sacrifices for the benefit of others. From an evolutionary perspective, heroes have a mysterious origin story: how did the “self-interested” process of natural selection give rise to self-sacrifice? Extreme self-sacrificing behavior seems particularly maladaptive: wouldn’t heroes have died out by now? We will argue that this puzzle can be resolved by considering the evolutionary logic supporting milder forms of self-sacrifice, the sort of ethics we see in our everyday lives. We will specifically discuss three types of ethical principles: justice, solidarity, and pacifism. By
justice, we mean behaviors concerned with fairly distributing resources; by solidarity, we mean behaviors concerned with group-beneficial self-sacrifice; and by pacifism, we mean behaviors concerned with the avoidance of harming others. These three ethics are not meant to be an exhaustive list, but rather a sample of key ethical principles with the properties of being (a) cooperative (i.e., individually costly but beneficial to others); (b) adaptive (i.e., individually longrun payoff maximizing) in the context of everyday life; but (c) not in one’s self-interest (and therefore “heroic”) when taken to the extreme. There is something seemingly counterintuitive about the evolutionary nature of our thesis:
how could costly other-benefiting behavior have evolved through a process that is inherently self-interested? The answer has to do with the timescale of evolution. Self-interested behaviors are often thought about in the short term: if I steal money from you today, I will be richer. It is important, however, to consider both the short-and long-term consequences of behavior: if I steal from you today, I will be richer today, but you will avoid me in the future and I will miss out on joint ventures with you that will make me richer in the long-term. Not stealing from you, therefore, may seem virtuous, but really it is in my long-term self-interest not to do so. If I started as a thief and learned the hard way, it may be that I stop stealing from you not because I am concerned about you, but because I am concerned about myself. It is therefore important to consider the consequences of behaviors not only in the short-term but in the long-term as well. Evolution is generally considered in biological terms, as a process that involves environmental
and social pressures selecting for adaptive traits which arise from randomly varying genes. But for humans, evolution can occur in the domain of culture, wherein natural selection acts not on genes, but on “memes,” or units of culture such as rituals, behaviors, symbols, and strategies (Dawkins, 2006; Richerson & Boyd, 2008). Cultural evolution works via “social learning”:
people imitate the actions and beliefs of those whom are seen as successful. Therefore, memes that cause their adopters to succeed (i.e., that increase cultural “fitness”) will spread through the population. We will discuss the evolution of heroic ethics mainly in terms of learning, though we will provide evidence of their biological origins as well. We adopt the dual process model of cognition (Kahneman, 2003; Sloman, 1996), and argue
that social heuristics, or rules of thumb for thinking about social interaction, play a key role in the transition from adaptive everyday ethicality to heroism. Heuristics can increase long-run payoffmaximizing behavior because they avoid the time and cognitive cost of deliberating when in familiar situations (Gigerenzer, Todd, & Group, 1999; Gilovich, Griffin, & Kahneman, 2002). Applying the general idea of heuristics to social interaction leads to the concept of cooperative intuitions that we develop through our everyday social interactions-because cooperation typically benefits our long-term self-interest (Bear & Rand, 2016; Rand et al., 2014). While cooperative intuitions may typically be adaptive-because the (often long-term) benefits generally outweigh the short-term costs-there are situations in which this is not the case. Also, because social heuristics are automatic, non-reflective processes, they are less sensitive to contextual complexity; that is, they respond similarly in situations where cooperation is advantageous as well as those where it is not. When cooperative intuitions are applied in nonadvantageous contexts (i.e., where the short-term costs outweigh the long-term benefits), we often see the resulting behavior as heroic. In other words, everyday ethics are self-sacrificial helping behaviors that benefit the actor in the long run, whereas heroism is self-sacrificial helping behavior that does not benefit the actor in the long run. Heroism, we therefore argue, occurs when cooperative intuitions are over-generalized to
situations where they are net costly (for the individual hero). One might object, however, that not all heroism is intuitive in nature-some heroism is quite deliberate. Indeed, we identify two types of heroism: emergent and sustained. Emergent heroism is an act of self-sacrifice that occurs “without a second thought,” exemplified by many of the people honored as Carnegie Heroes who jump in front of trains or into raging rivers to save strangers (Rand & Epstein, 2014). This type of heroism arises directly from the intuitive over-generalization logic laid out above. Sustained heroism, on the other hand, involves a long-term, often life-long commitment to
self-sacrifice, exemplified by the “moral saints” of history such as Mother Teresa or Mahatma Gandhi. We believe this type of heroism could also arise from the logic above, but via a less direct route: it may be that heroic, individually costly cooperative goals are set via automatic processes, and this goal setting is based on over-generalization, but then deliberative processes are recruited to pursue these goals (Cushman & Morris, 2015). In other words, it could be that automatic processes lead to the establishment of (heroic) extreme goals, and then deliberative processes enact these goals, leading to sustained heroism over time. For example, many sustained heroes have a distinct moment of inspiration, as Mother Teresa did on a retreat in 1946 (Langford, 2008). It could be that her concept of justice changed to a (heroic) extreme in this moment, driven by an intuitive process, and that the rest of her life was effectively spent pursuing this heroic goal. Alternatively, it could be that deliberative processes contribute to sustained heroism by anticipating guilt for not acting in accordance with intuitive responses favoring extreme prosociality. In other words, it could be that when we deliberate and evaluate our cooperative intuitions, we consider not only the material costs and benefits of cooperating, but also the future psychological costs imposed on us by our intuitive cognitive processes: for example, we might anticipate that we would feel guilty if we do not behave cooperatively (Battigalli & Dufwenberg, 2007). Thus, whether heroism is emergent or sustained, i.e., enacted via intuition or deliberation, it may be that over-generalized cooperative intuitions are at its root. Importantly, the claim we make here is descriptive, rather than normative. We do not argue
that heroism is bad because it is individually non-advantageous; on the contrary, we hold that the source of heroism’s virtue-the reason it is so good-is that it is so remarkably selfless (i.e., costly
to the individual). Indeed, exploring the evolutionary roots of moral psychology (Kurzban & DeScioli, 2015) to discover when seemingly altruistic behavior pays off in the long-run (e.g., by benefitting one’s reputation, as in Yoeli, Hoffman, Rand, and Nowak, 2013) allows us to identify when cooperation actually is extraordinary and worthy of being called heroic. In sum: we argue that the everyday ethics of justice, solidarity, and pacifism arise from adaptive
mechanisms. When these cooperative behaviors are typically individually adaptive, they become automatized as social heuristics. The automaticity of social heuristics makes cooperative behavior prone to over-generalization, and when this occurs in individually costly contexts, the resulting behavior is heroic. Having laid this foundation for the origins of heroism, we now provide a more detailed survey of evidence that justice, solidarity, and pacifism are adaptive and automatic.
