The Evolution of Agency

Part I: The complexities of agency

Photo by Rishi Ragunathan

This week, we’ll begin a two-part series that explores one of the enduring mysteries of philosophy: how much agency do we have in our own lives? How free are we to make our own decisions, control our own actions, shape our own character?

Our guide for this exploration is evolutionary psychologist Michael Tomasello and his 2022 book, The Evolution of Agency: Behavioral Organization from Lizards to Humans. Tomasello is currently James F. Bonk Distinguished Professor of Psychology and Neuroscience at Duke University, and he also holds professorships in evolutionary anthropology and philosophy (as well as being emeritus director of the Max Planck Institute for Evolutionary Anthropology)—so he is especially well-placed to theorize about the distinctions between humans and other animals. He has spent his career conducting studies on the social and intellectual capacities of great apes and human children, and his work has profound implications for a question asked by the ancient Stoics and philosophers throughout history: What makes us human? What sets humans apart from other creatures?

In his 2019 book, Becoming Human: A Theory of Ontogeny, Tomasello proposed that while great apes are highly intelligent as individuals, only humans have the capacity to be highly intelligent as a group. Our evolved capacities for sharing attention and intention with other humans, working toward a joint or collective goal, is what enables us to accomplish great things. This ability is biologically costly, however, and only happens because of our extremely long childhood (during which time we are cared for by our relatives). A child’s developing brain matures in the context of a specific social and physical setting, where she learns language, social expectations, culture, and other survival skills for her particular context. This is why humans have the longest childhood of any animal. It’s a time for us to, quite literally, become human.

We can’t learn our group’s social and cultural expectations before birth, since it’s too complicated and there’s too much to learn. But we’re hardwired for this learning to occur after birth. There is a period of intense social learning throughout childhood and adolescence, which slows down (but never completely stops) in adulthood. So you might say that we are all born with the capacity to become good group members—acquiring language, developing a sense of fairness, learning the rules of social interaction—but we have to learn these skills through long-term interactions with people around us. Genes and environment work together to create a cognitively flexible, context-responsive, socially cooperative human.

The result is that humans have a great deal of agency in our own lives. But it remains a matter of contention just how much agency we have. We are obviously subject to environmental influences, such as what family and culture we grow up in. If we grow up in a social environment that values sharing resources with others, then that’s what we’ll learn to do. If we grow up in a social environment that values material acquisition, then that’s what we’ll learn to do. Evolutionarily speaking, this is a feature of the system, not a flaw. We are prepared to interact in the specific environment in which we find ourselves.

This social adaptability can lead to problems, though, when our social environment is less than optimal. Particularly when children are raised by abusive parents, or in dysfunctional families or communities that unintentionally teach dysfunctional behavior. Even if there’s no abuse, families may teach the wrong values (like selfishness) or emotional habits (like getting angry over the slightest things). Everyone is already familiar with these problems, so I don’t think I need to go into detail about what they are and the negative impact they have on people.

What we want to focus on today is how much agency we have to change the values or emotional habits we acquire from people around us. A great many people grow up with less-than-optimal social influences (perhaps this even applies to you). But does that mean we’re doomed to repeat the mistakes of our family members? Are we at the mercy of the social forces around us, or do we have a say in our own actions and character?

We will explore some Stoic answers to this question in Part II. But right now, I’d like to start from a different direction. Sometimes approaching an old problem from a new angle can create new insights, or at least provide a fresh way of thinking about them. So that’s what we’re going to do today. Guided by Tomasello’s work, we’re going to look at human agency from the perspective of evolutionary psychology. But we’re not going to start with humans…we’re going all the way back to single-celled organisms. Let’s see what The Evolution of Agency can tell us about agency throughout the animal world.

What is Agency?

We’ll start with the basics. Forgetting humans for a moment, let’s think about agency in general. What is it, and what purpose does it serve in nature? Tomasello emphasizes that agency is not the same as complexity of behavior. Complex behavior is present in the delicate efficiency of a spider building a web or a honeybee communicating a prime pollen location to hive-mates through its dance. Many organisms are hard-wired to perform complex tasks, but that’s just it—they are hard-wired. They don’t seem to have much choice in the matter. They can’t suddenly decide to stop doing it or to do it in a different way. As Tomasello puts it,

Even when they are doing something highly complex, the behavior of ants and spiders and bees does not seem to be under the individual’s control. Their evolved biology is in control. In contrast, even when they are doing something relatively simple, primates and mammals seem to be making active and informed decisions that are at least somewhat under the individual’s control. In addition to their evolved biology, they are cooperating with a psychology of individual agency.

p. 1

So agency, according to this theory, is not so much about what organisms do as how they do it. In other words, it’s about the organization of behavior. Agency is a special type of behavioral organization in which individuals “direct and control their own actions” (p. 2). More specifically, agency is “feedback control organization in which the individual directs its behavior toward goals—many or most of which are biologically evolved—controlling or even self-regulating the process through informed decision-making and behavioral self-monitoring” (p. 20).

So here we have some crucial points about agency. It is

• A feedback loop (informed by feedback from the environment)

• Controlled by the agent

• Goal-directed

• Continuously monitored

• Dynamic and flexible

A squirrel searching for nuts is exhibiting agency because, although it is fulfilling a biological imperative to gather food, the individual determines the exact time, place, and route for securing nuts based on its previous experiences in a given area. Its actions meet all the criteria above for agency. This is in contrast to nonagentive actors, such as unicellular creatures that float around and passively consume whatever food bumps into them. Tomasello describes unicellular organisms (along with simple multicellular animals such as the famous C. elegans) as “stimulus-driven, not goal-directed” because they cannot exert behavioral control.

Why would some creatures have agency and others not? Tomasello suggests agency is an adaptive response to unpredictable environments. An amoeba floating in a primordial sea might do very well without agency because its environment is stable and predictable, allowing it to consistently find food and not get eaten by predators. But in more complex and variable natural settings, a degree of flexibility is an advantage:

A plausible hypothesis is that in some cases the environmental niche of a species is too unpredictable across time and space for hardwired perception-behavior pairings to be effective. In the face of such unpredictability, Nature—if we may personify the process of evolution by means of natural selection for ease of exposition—needs someone “on the ground,” so to speak, to assess local conditions in the moment and decide on the best course of action. What thus evolves is an underlying psychology of agency that empowers the individual—in some key subset of situations—to decide for itself what to do according to its own best judgment.

p. 2

Squirrels are biologically programmed to cache nuts, so they have certain instincts and abilities such as jumping across tree branches, searching for nuts, locating a good caching site, being wary of predators, etc. But any specific squirrel must apply these abilities in its specific situation if it’s going to stay alive and reproduce. So it must have the freedom search for that particular nut in that particular location while looking out for that particular predator. It uses its individual agency to identify where, when, and how to go after the best nuts in its particular environment. However, it can’t just do whatever it wants; its capacity for individual decision-making is “always exercised in the context of an organism’s evolved capacities” (p. 2).

Types of Agency

As you can see, this definition of agency could encompass a wide range of species, and that is exactly how Tomasello applies it. According to his theory, members of different animal species have agency to greater or lesser degrees. Primates, and especially humans, have more agency than most other animals, but all mammals have some agency, and even many vertebrates such as reptiles have a limited agency within their ecological niche. Here is his proposed hierarchy of agency in the natural world, from less to more:

• Goal-directed agents (lizards)

• Intentional agents (squirrels)

• Rational agents (great apes)

• Socially normative agents (humans)

Goal-directed agents include lizards and other reptiles, amphibians, and perhaps fish. They show “flexible, context-sensitive behavior” (p. 31) in finding food and escaping from predators, and in experiments they are able to learn which of their existing behaviors are most rewarding (although they are not able to learn new behaviors). These animals are thought to make go-no-go decisions in which they determine whether or not to make a specific action from within their behavioral repertoire.

Mammals have even more agency—they are intentional agents: “Mammals direct their actions toward goals not just flexibly but intentionally, as they cognitively simulate possible action plans toward their goal before actually acting” (p. 43). Animals such as mice and squirrels are capable of more complex decision-making because they are capable of more complex cognition, including emotions, learning, and an awareness of their own actions and perceptions (which we might call consciousness). Mammals can flexibly learn new skills, and they understand how their actions impact their environment.

Great apes (chimpanzees, bonobos, orangutans, and gorillas) approach human-like rationality with their understanding of “the underlying causal structure of events in the physical world and underlying intentional structure of actions in the social world” (p. 67). I think Tomasello is being deliberately provocative here by calling apes rational agents, since the term rationality is generally reserved for humans’ special cognitive abilities (which, by any measure, are quite different from apes). However, he argues that “it is clear empirically that great apes are agentive in ways that are very similar to humans” (p. 90):

The criteria for rationality I have used here … [are] often used by philosophers in addition to the use of socially normative standards. In particular, I have used as criteria (i) thinking about the external world using logically structured causal and intentional inferences, providing rational coherence to experience; and (ii) adopting a reflective and self-critical stance to one’s own thinking and decision-making, including adjudicating between conflicting goals before acting by reflecting on their relative merits, providing rational coherence to one’s psychological functioning in general.

p. 89

Whereas all mammals are able to form intentions and then act them, they cannot reflect on or control those processes—they can’t think about their thinking. Great apes can. They have an additional tier of executive control that enables them to reflect on their intentions and control them in ways that other mammals can’t. They even perform as well as human children (3-5 year-olds) at some of these tasks. In Tomasello’s words (p. 88), apes are able to

• “Plan for future goals”

• “Make logically organized inferences based on an understanding of external causal and intentional relations explaining why things happen”

• “Are self-critical of their own decision-making processes, causally analyzing problems or conflicts and intervening to resolve them”

• “Display impressive skills of inhibitory control and the resolution of goal conflicts at the stage of action execution”

All together, these remarkable skills enable apes to experience the world differently from other creatures. Tomasello suggests that apes have a reflective tier of executive function, distinct from the executive function of other mammals. This special cognitive equipment allows them to reflect on their own decision-making and align their own internal experience and with external events. In effect, great apes have a very sophisticated understanding of themselves as agents in the world.

On to Humans

Clearly, apes have impressive cognitive abilities and are similar to humans in many ways, even if they are not “socially normative” in the way we are. But what does all this have to do with the questions we asked at the beginning of the chapter? Why are we talking about squirrels and apes when we started by asking about humans? As Tomasello puts it,

An obvious but nevertheless profound conclusion is that these processes cannot have as their evolutionary or ontogenetic origin in anything deriving from uniquely human forms of culture, intentional instruction, or language. Rather, they constitute an evolved system common to all great apes, enabling individuals to make effective and efficient—indeed, reflective and rational—behavioral decisions.

p. 90

I think you can see where we’re headed with this. If agency isn’t a unique characteristic of humans, but is a widespread feature of the animal kingdom, that has important implications for our understanding of human agency. At a minimum, humans have at least as much agency as great apes. Now we can ask questions like: What are our evolved capacities for agency? How far does this agency extend into uniquely human, socially-normative behaviors? What can a species-comparative approach to agency tell us about ourselves?

When we start thinking about agency for humans, we introduce a new dimension: joint (with other individuals) and collective (within a group) agency. What sets humans apart from apes more than anything else is our ability to join forces with other people, to set a shared intention, to work together toward a shared goal. Apes have some limited capacity to do this, but nothing close to human capability in this domain. It is humans’ ability to cooperate, collaborate, coordinate, and communicate that makes us as intelligent as we are:

Most of the unique psychological capacities of the human species result, in one way or another, from adaptations geared for participation in either a joint or a collective agency. Through participation in such agencies, humans evolved special skills for (i) mentally coordinating with others in the context of shared activities, leading to perspectival recursive, and ultimately objective, cognitive representations; and (ii) relating to others cooperatively within those same activities, leading to normative values of the objectively right and wrong ways to do things. Individuals who self-regulate their thoughts and actions using “objective” normative standards are thereby normative agents.

p. 117

Our need to cooperate with others in order to survive in unpredictable and highly competitive environments led to a new layer of executive functions (on top of great apes’ already impressive metacognitive abilities as agents.) Not only do we have extraordinary decision-making abilities as individuals, we are able to form intentions with other people and act on those intentions as a group. We can then reflect on those actions, noting the impacts of other people’s behavior as well as our own, and taking this information into account in our future group actions.

That means humans have a type of agency—the ability to make decisions collectively—that is not available to other animals. It also means that each individual human has multiple and simultaneous commitments as an agent: individual and collective (either to other specific individuals or to the group as a whole). Or, as Tomasello puts it,

Modern humans are individual rational agents who sometimes (though not always) subordinate their individual agencies to various shared agencies when doing so is either instrumentally or normatively appropriate, with ontogeny in a culture helping to shape those judgments of appropriateness. p. 118

In other words, we can choose which agentic relationship we want to prioritize at any given time—self, partner, group. Tomasello provides the example of someone in a hunter-gatherer culture deciding whether to take meat back from the hunt to share with the whole tribe, or whether to give most of it to a favored partner (or keep it for their own family). There could be various reasons for choosing each (you owe your partner, since he gave you a large share last time; your tribe is very hungry and depending on you for their food). But Tomasello’s point is that these overlapping commitments give rise to genuine moral dilemmas for which there may never be one right answer. These “conflicts within an individual among different agencies with different goals and values” (p. 118) are simply part of human social complexity.

Clearly, this is an issue that we all face at various times throughout our lives. We’re not hunter-gatherers anymore, but we may encounter competing and mutually exclusive demands from, for example, our job (get this project done) and our family (be at your child’s dance recital). Or we may face competing obligations from within our own family (needing to take care of an aging parent and a young child at the same time), or we might confront demands from society (pay taxes) that require personal sacrifices (less money for our family). According to Tomasello’s framework, we learn these expectations from our culture (or specific social influences within that culture) through our years of interpersonal interactions through collective agency.

Next Time…

In the next post we’ll continue this discussion, examining how this evolutionary theory can inform a Stoic approach to agency. We’ll return to our practical questions about agency and social influence, and we’ll look at what steps we can take to realize our own agency in everyday life. See you then!