When scientists are not able to answer a certain question, sometimes they move to an adjacent question that is more tractable. This happened with the free will/determinism debate. Do humans (or other animals) have freedom to act, or is every decision predetermined by every other event that happened before? This question is hard to answer. Meanwhile, we may talk about agency. The notion of agency seems less burdened and simply means the capacity to act. While free will is hard to measure, many aspects of agency can be tested in experiments. The book in today's post looks at agency in the evolutionary context. It's “An Evolutionary Story of Agency: How Life Evolved to Act on Its Own” by Álvaro Moreno and Juli Peretó.

The authors define agency as controlled and purposeful actions carried out by agents. Agents have a capacity to pursue goals and causally change their environment with their actions. The book covers all different entities (organisms, beings) that possibly had or currently have those capacities.

The authors propose criteria for agents: the organism must be able to act on the environment and to regulate its own actions. Modern humans clearly satisfy the criteria, but what's with primordial entities?

It all began some 3.8 to 4 billion years ago. One hypothesis is that, at the dawn of life, protocells emerged, which were just water-filled lipid vesicles, and RNA molecules were probably getting trapped inside these vesicles. (This is only one hypothesis, and the dawn of life remains an active area of research.) The major shift happened when those cells began to reproduce. Reproduction with heritability made natural selection possible. It is hypothesized that evolution favored genetic instructions, so RNA became the genetic code for protocells. However, even when those protocells had an impact on the environment, it was rather by chance and not regulated by the protocell itself, meaning that protocells were not agentic.

Prokaryotic organisms appeared some 3.5 billion years ago. The two types of prokaryotes are bacteria and archaea. Although both are predominantly unicellular, bacteria and archaea are two different biological domains, as they substantially differ structurally and genetically. Archaea are less known to the general public than bacteria, probably because there are no known archaeal human pathogens. Actually, archaea, just like bacteria, live in the human gut. Many bacteria and archaea are capable of interaction with the environment via movement, which is why they are considered agentic according to the authors’ criteria. The types of movement that various species are capable of include swimming and surface motility, such as swarming, twitching, gliding, and sliding. Prokaryotes are also capable of chemotaxis, which is a primitive sensing of and responding to chemical gradients. Furthermore, some bacteria form a temporary coliving residence, called a biofilm. In a biofilm, a colony of bacteria sticks to the surface and covers itself with a slimy matrix made of sugars and proteins. A biofilm is protected from adverse effects better than a single bacterium floating around. An example of biofilm is a dental plaque. Although already agentic, the actions of prokaryotes are limited mostly due to limitations in the ways that they can move. This was solved in the next step.

Eukaryotes emerged some 1.8 billion years ago. Presumably, they evolved from archaea and also incorporated bacteria inside the cell to become mitochondria. Eukaryotic cells evolved to be much larger than prokaryotic cells. Another key evolutionary change was the emergence of actin-myosin networks. Actin and myosin are proteins that work together to generate contractions (movements). Larger cell bodies, together with actin-myosin networks, helped some eukaryotes to generate innovative and diverse behaviors. Moreover, larger cell bodies allowed them to detect chemical gradients in a novel way, relating gradients spatially, for instance, to the side of the body (on the left or on the right). Furthermore, some eukaryotes developed subcellular sensing structures. For example, certain species of unicellular marine plankton possess an eye-like organelle called ocelloid, which consists of a lens surrounded by a layer of mitochondria functioning like a cornea, and specialized photosynthetic plastids functioning like a retina. This “eye” probably helps plankton to determine the direction of light. Clearly, unicellular eukaryotes meet criteria for agency. However, an increase in sophisticated behavior required multicellularity.

There are known organisms that oscillate between the single-cell stage and the multicellular stage. For example, when Dictyostelium discoideum has enough food, it is single-celled, or amoebae, and it feeds and divides like usual single-celled organisms. However, when food is not available, individual cells come together to form a multicellular organism, a slug, which can migrate towards more favorable conditions. Once the environment with light, heat, and humidity is reached, a slug transforms into a fruiting body that releases spores, which later become single-celled amoebae. However, aggregate multicellularity is not true multicellularity. In a true multicellular organism, some cells lose their ability to move and are surrounded by other cells (without direct access to oxygen) and thus rely on other cells for nutrients. True multicellular organisms require a mechanism for coordination between cells.

Ancestors of modern-day animals reached the multicellular coordinated stage some 600 million years ago. The key developments were integrated machinery of movement and cell-to-cell communication mechanisms. Later on, the nervous system had developed, probably, to link sensory organs to motor organs, and to regulate essential functions of the body. Animals, including humans, act on their environments, pursue their goals, and thus satisfy the criteria for being called an agent.

The authors start and end by reiterating the argument for the agency of living beings. The environment did not fully determine the fate of evolution. Living creatures changed the environment through their actions and therefore determined their own future.

The book is short and packed with interesting and fascinating facts about living beings: self-organized bacteria, fungi and zombie ants, and moving and carnivorous plants. I agree that even bacteria have agency. However, the argument about agents acting on their environments doesn’t remove determinism: perhaps the environment is acting on the animal in such a way that it can only act on the environment such and such. Yet, causation in a feedback system is a topic in and of itself, and the book is about something different. The book is about life, and its different manifestations. It was beautiful and inspiring.

Favorite quote:
”There is no agency below the level of the cell, regardless of how intricate a chemical system may be. The cell is both the minimal unit of life and the minimal unit of agency.”

October, 2026