Conor Feehly is a science writer from New Zealand.
In Greek mythology, after slaying the Minotaur on Crete, Theseus sailed back to his home in Athens a hero. As a memorial, the Athenians preserved his ship, over time replacing the timbers as they succumbed to decay. Contemplating the staggered replacement of the parts of the famous ship, ancient Greek philosophers wondered: At what point does it become a completely new object?
Like Theseus’s ship, living entities are constantly replacing the components that make them up, engaging in self-sustaining and self-organizing behaviors to maintain their continued existence in relation to an environment that seems, at best, indifferent to their ongoing survival. As the Chilean philosophers Humberto Maturana and Francisco Varela put it in the early 1970s, this continuous self-regenerating behavior, or “autopoiesis,” is the crucial distinction between life and nonlife.
For just about as long as scientists have understood autopoietic behavior as a feature of living systems, it’s helped guide questions about life and its enigmatic beginnings. But what if autopoietic behavior is not exclusive to living systems? What if this tendency for self-preservation could be found in physical substrates usually considered to be inert and lifeless? Could we then make better sense of that inexplicable transition from inanimate matter to animate life?
By investigating analogous examples in minimal physical systems of what life does, a new movement of scientists and philosophers are trying to tell a more plausible story regarding the origin of life’s complex behavioral repertoire, and perhaps the origins of life itself. This is the intuition that has guided a new approach to understanding the puzzle of life’s emergence on our rocky world some 3.5 billion years ago — an approach that applies the tools we use to study the mind to the mystery of abiogenesis.
First Life
For as long as the origin of life (OOL) has been a subject of scientific inquiry, it has most often been framed as a chemistry problem.
In general, two approaches have defined the investigation of the deep, complex riddle of life’s origins. The first is concerned with understanding how geochemical conditions on the newly formed Earth could have created early iterations of the universal biomolecules we see in every living organism today. The second, paleogenetics, starts with sequencing DNA from modern and extinct organisms and working backward to reconstruct ancient genomes based on shared genes. So far, paleogeneticists have been able to trace the tree of life back to a last universal common ancestor — LUCA — that lived some 300-400 million years after the Earth formed.
But just as cosmologists cannot peer all the way back to the Big Bang, biologists cannot see beyond LUCA. LUCA, therefore, can only tell us so much about the OOL since it existed many millions of years after life first emerged on Earth.
As it stands today, researchers in the OOL field tend to disagree about the setting in which life may have gained its initial foothold (among much else). Some think it was surface geothermal ponds, others deep-sea hydrothermal vents. A small minority thinks it’s plausible that life was delivered to Earth via an astronomical source.
The disagreement largely stems from different ideas about what is needed to get life going in the first place. Warm surface ponds are thought to be more hospitable to the synthesis of information molecules like RNA, while the hydrothermal vent scenario is favored by those who see the first living entities as needing a primitive form of energy-giving metabolism.
“The origin of life doesn’t just signal the beginning of complex self-replicating chemistry — it also represents the beginnings of agency, mind and consciousness on our planet.”
Michael Wong, an astrobiologist and planetary scientist at Carnegie Science, and Stuart Bartlett, of Caltech’s Division of Geological and Planetary Sciences, have an analogy for this way of thinking about the OOL: the cellphone. “What’s universal about modern cellphones?” Wong asked rhetorically when I talked to him recently. Well, they all have cameras and touchscreens. Should we then assume that the first cellphones possessed cameras and touchscreens?
“Just because you see something that’s universal in life today doesn’t mean that particular molecule was responsible for the origin of life,” Wong said. You could think of the modern cellphone as the eukaryote of communication technology, with touch screens and cameras starting off as free-living species that eventually became incorporated like organelles into a complex information-processing device.
It’s now generally accepted that presumed conditions on the early Earth could have created lipids, sugars, amino acids, nucleobases and possibly even RNA. The surface pond hypothesis is attractive because this environment would have received a steady source of organic products from the atmosphere, undergone wet-dry cycles that can support chemical combination, and would have been exposed to organic compounds delivered by meteorites.
But whether those conditions actually existed on the early Earth is a topic of intense debate, and even if all the necessary ingredients for life were there, there is no guarantee they could spontaneously assemble into a fully fledged lifeform. Having all the ingredients for a tasty dinner is one thing, but undergoing the processes that combine them into a meal is another. “It’s not just about the material,” Wong said. “It’s about what the material is doing. It’s the processes of life that matter.”
Artificial Life
Matthew Egbert, a computer scientist at Auckland University in New Zealand, has spent the last 15 years building computational models of autopoietic systems in their most basic form. These “cellular automata” help test ideas like autopoiesis outside of the complicated world of biology, where disentangling all the complex chemical machinery of living cells is nearly impossible.
Egbert is fascinated by an idea called “viability-based behavior,” which he described as “something special autopoietic systems can do that non-autopoietic systems cannot do.” Unlike, say, a rock or even a complex machine, autopoietic systems actively behave in ways that promote their own survival. This could be as simple as a bacterium moving toward warmer, more hospitable conditions. In this case, the organism modifies its immediate environment to promote its own health.
This notion echoes a characteristic of living systems called niche construction, whereby organisms actively regulate and modify features of their environment to create conditions that enhance their survivability. Humans build houses, beavers build dams, birds build nests. But even the simple act of movement, of choosing to take one path toward a more favorable location, is a minimal example of viability-based behavior. “It’s not just that the environment is posing a problem that the organism has to solve, but the organism is also affecting the environment, influencing it and selecting it,” Egbert said.
Outside of his computational simulations, Egbert highlights the emerging field of “wet A-Life,” where complex systems scientists study viability-based behavior in basic, chemical systems. Think of oil droplets and waxy, fatty chemical structures floating around on a water surface. Like living systems, these structures maintain a flexible boundary with their environment.
“What if autopoietic behavior is not exclusive to living systems? What if the tendency to preserve continued existence could be found in physical substrates usually considered to be inert and lifeless?”
These chemical entities are far less complex in terms of their molecular structure than even basic biomolecules like RNA, and yet the striking thing is that such incredibly simple systems are still capable of surprisingly complex self-preservative behavior. Behavior that seemingly suggests a capacity for learning, memory and decision making. Behavior that we tend to associate with cognition — with life.
When it comes to typical accounts of the OOL, prebiotic chemicals that were subject to a wide set of contingent environmental factors had to serendipitously find themselves in the same place at the same time. But this is an extremely high bar to reach. Instead, the chemical systems that wet A-life researchers study hint that even before entities we would recognize as life emerged, chemical structures that maintain a boundary with their environment were behaving in ways that we tend to associate with life.
Rather than relying on an astonishing confluence of chemical and environmental factors, it’s possible that life arose out of dissipative chemical systems carrying out self-sustaining behavior, modifying their environment and actively selecting for conditions that were agreeable to them. In this way, the fact that basic “life” eventually emerged may come down to the self-preservative behavior of simple chemicals.
“Could these ideas precede even Darwinian evolution?” Egbert wondered. “Because we have systems that are so simple that they could spontaneously emerge without any genetics … it really changes the way we start to think about the origins of life.”
Persistent Life
For over a century, scientists have been studying the self-propelling behavior of chemical structures called camphor boats. Small camphor pellets, which are made of a waxy chemical substance, are placed on the underside of a boat-like structure and dropped onto a water surface. As the camphor reacts with its environment, the boats spontaneously zip around on the water’s surface in complex patterns.
Initially, chemists like the pioneer of surface science Agnes Pockels were interested in how camphor affected the surface of water to generate Marangoni flows — fluid motion driven by gradients in surface tension. The study of camphor, however, has seen a recent resurgence in complex systems science and active matter research.
Richard Löffler, an OOL researcher at the University of Copenhagen, studies camphor boats precisely because of their self-propelled movement and behavior that mimics what life does. For instance, in some of Löffler’s initial experiments, he found that camphor boats would aggregate into collectives as the chemical dissipation of their waxy surfaces drove them around on a water surface.
In later collaborations with pioneering chemist Martin Hanczyc, Löffler conducted experiments investigating the behavior of motile oil droplets and dye on water surfaces. Similar to camphor, oil droplets maintain a boundary, are capable of movement as they react with their environment and are much simpler than even the most basic biomolecules. Löffler and Hanczyc found that the droplets had the ability to selectively move toward conditions that allow the active surface chemistry of the oil droplet to persist.
It’s easy to consider the behavior of camphor boats and motile oil droplets as akin to a rock rolling down a hill, except the hill in this case is a chemical gradient as opposed to a gravitational one; the boats and droplets, in other words, just go wherever the laws of physics propel them. But Löffler and Hanczyc see it slightly differently. The droplets, they argue, carry their own “fuel” in the form of chemical potential, which they use to move toward conditions that enable subsequent movement. “As it runs out of fuel,” Löffler told me, “it can ‘detect’ where more fuel is and move to absorb it and basically refuel itself, digest it and dissipate for a longer period of time.” As a rock tumbles down a hill, on the other hand, it can’t selectively stop, slow down or change direction.
“Having all the ingredients for a tasty dinner is one thing, but undergoing the processes that combine them into a meal is another.”
There are thus two ways to describe the behavior of the droplet. In purely physical terms: A reaction takes place at the interface of the droplet and its aqueous environment; chemicals and water move along the interface, which drives a convective flow, with environmental conditions influencing the rate of the reaction and therefore the movement. These factors combine to create a surface tension on one side of the droplet, which causes the flow of materials to propel it toward environments that are more conducive to continued motion.
Or in proto-cognitive terms, where the droplet’s actions seem intentional: It runs out of fuel, it becomes slower, it becomes “hungry,” it detects a new source of fuel, it moves and absorbs it, it keeps going. This tendency for the system to actively seek conditions or environments that allow it to “survive” and selectively avoid conditions that hasten its demise represents what Löffler calls a minimal form of goal-directed behavior. In order to persist, then, the droplet senses and acts.
For Löffler and his collaborators, this research is at the precipice of two ontologies — deterministic physics and intentional, cognitive agency. The cognitive description of the droplet’s activity is akin to what philosopher Daniel Dennett described as the “intentional stance,” where a system’s behavior becomes complex enough for outsiders to describe it as having its own goals, internal states and agency. In other words: a minimal mind.
The emergent decision-making of the droplet — behavior that supports its continued existence — represents what the initial fork in the road at the OOL might have looked like. Seemingly deterministic chemical action began to achieve lifelike, goal-directed persistence.
Scaling Up
If we accept that basic nonliving chemical entities are indeed capable of behaviors that faintly echo the cognitive capacities of living systems, how then did they gradually transition into the complex forms we associate with life? Interestingly, complexity is not always an indicator of longevity or successful self-replication for any system: As researchers from the Scripps Research Institute demonstrated in a 1994 experiment, simple synthetic RNA molecules outcompeted longer ones that had a higher chance of breaking, misaligning or mutating.
And yet, life has inexorably continued to grow more complex. As Bartlett’s own experiments have shown, you don’t need to climb the ladder of molecular complexity very far before chemical agents become capable of simple forms of associative learning, such as conflating the presence of a new chemical with a source of food. Such agents, however, are driven to develop more complex internal models of their environment in the presence of competing chemical agents that use the same source of energy. Survival requires the incorporation of variables into extra layers of complexity and more developed responses, a feedback loop that incentivizes more sophisticated learners competing against one another. The result: complexity ratchets up.
Recently, biologists and philosophers have been extending components of cognition — learning, memory, decision making — to entities once considered to be more like mindless machines than mind-bearing lifeforms like fungi, bacteria and plants. Even organs and cells show hints of behavior once reserved only for beings labeled “intelligent.” So, how far down the scale of complexity does cognition actually go?
Wet A-life researchers like Löffler have taken this research agenda one step further, probing for the signatures of mind in chemical systems that do not meet our criteria for what it means to be “alive.” And yet, they exhibit a surprising array of behaviors we usually only attribute to things we think of as living.
As Löffler told me, “I don’t want to rule out something like cognitive aspects of very simple systems, which are clearly nonliving systems, because it might be helpful to wonder if something is there.”
“Incredibly simple systems are still capable of surprisingly complex self-preservative behavior.”
Cognition can usually be evaluated from three different perspectives: What is the physical information-processing circuitry of a system (neuroscience)? How does the system interact with its environment (behaviorism)? And what does it feel like to be that system (phenomenology)? Of course, ascribing cognition to basic chemical entities depends on our definition of cognition itself. If we adopt a more stringent view stringent view that requires some sensory apparatus and internal representation, then these systems may only resemble cognition. Still, in scaled-up lifeforms like ourselves, part of how we determine cognition is through an assessment of capability, which suggests that we shouldn’t rule out the possibility that simple behaviors by simple systems are the precursor elements of more sophisticated intelligence.
The OOL doesn’t just signal the beginning of complex self-replicating chemistry — it also represents the origins of agency, mind and consciousness on our planet, which, as far as we know, are the only such examples in the universe. While understanding the chemical and environmental contexts in which living systems may have emerged on Earth is no doubt important, unleashing new methods when it comes to the study of the OOL may help us make better sense of the features of life that remain the most impenetrable — consciousness and the mind.
If inanimate physical systems, systems that at face value seem dead and dumb, are actually capable of behaviors that require basic cognitive elements, then how does this change the view of the origin of our own conscious capacities and the presence of mind in other entities?
Consciousness is something that we have struggled to square with the physical and chemical laws of the universe. For centuries, philosophers and scientists have attempted to make sense of how the features of the mind emerge from cold, dead, mindless material. The investigation of the ingredients of mind in the most basic physical substrates thus forces us to reconsider our assumptions about where minds reside.
It could also help reframe the origins of our own internal worlds — not as a mysterious feature that “emerged” at some sufficient point of chemical and behavioral complexity in our evolutionary history, but as part of a more ancient, gradual process, one that expands the behavioral repertoire of organisms as they navigate an increasingly complex, uncertain world.
