12  What Is Life?

In 1944, a physicist who had never cultured a bacterium or cored a sediment sat down in Dublin and asked a question that biologists had been circling for centuries: What is life?

Erwin Schrödinger’s little book did not answer its own title. What it did was more lasting. It reframed the question in the language of physics – order, entropy, free energy, information – and insisted that the answer, whatever it turned out to be, must not violate the laws that govern everything else.1

We opened this book with his question. Now, having traveled from the first chemical reactions on a cooling planet to microbial communities persisting 2.8 kilometers underground on energy budgets thinner than a candle flame, we can attempt something Schrödinger could not: an answer grounded in evidence, calibrated in \(\Delta G\), and tested against the porewater profiles of real sediments.

12.1 Three definitions, all correct

Throughout this journey, three definitions of life have kept surfacing. Each one captures something essential. None is complete alone.

Life is the process of maintaining non-equilibrium conditions by extracting energy from the environment.2

In Chapter 1, this was a bacterium dividing once per century and a formula: \(\Delta G = \Delta G^\circ + RT \ln Q\). By the time we reached the deep biosphere, it was something you could measure with an electrode and a gas chromatograph. The Gibbs budget that seemed abstract in the opening pages became the accounting system for every metabolism we encountered – from aerobic respiration’s generous payoff to the razor-thin margins of anaerobic methane oxidation.3 We learned that “maintaining non-equilibrium” is not a poetic description. It is a quantitative claim. It means spending maintenance energy, continuously, just to keep enzymes folded, membranes intact, and gradients from collapsing. Maintenance energy is not an abstraction. It is the first law of microbial existence.4

Life is inseparable from the concept of the organism – discrete units that reproduce.5

This definition is the one that keeps biology from dissolving into pure chemistry. A fire also maintains a non-equilibrium state, consumes fuel, and exports entropy. But a fire does not reproduce with variation. It does not evolve. The discreteness of organisms – bounded, self-replicating, subject to selection – is what makes microbial communities more than reaction networks. We saw this most clearly when we traced how organisms emerged from communities and communities from chemistry: the transition from an abiotic world of mineral-catalyzed reactions to cells with membranes, genomes, and metabolic strategies. That transition remains one of the great unsolved problems in science, but the direction is clear. Somewhere between geochemistry and biology, information began to matter.

Life is the ability of a replicator to copy itself using resources from the environment.6

This is the information definition, and it is the one Schrödinger anticipated most clearly with his “aperiodic crystal.” We traced it from the RNA world hypothesis through the emergence of DNA-based heredity to the full metabolic machinery of modern cells. The replicator definition captures something the other two miss: the arrow of evolution. Non-equilibrium maintenance is necessary but not sufficient. Discrete organisms are necessary but not sufficient. What makes life life – what distinguishes it from a cleverly maintained chemical gradient – is that the system carries instructions for its own reproduction, and those instructions can change.

12.2 Two irreducible properties

Strip away the details – the metabolisms, the redox ladders, the transport equations – and you are left with two things that cannot be removed without losing the phenomenon entirely:7

  1. Genetic information (DNA, or whatever preceded it). A stable, copyable record of how to build and maintain the system.
  2. Active implementation of self-maintenance and reproduction, powered by energy extracted from the environment (proteins, or whatever preceded them).

Information without work is a library with no readers. Work without information is a fire. Life is both: a system that reads its own instructions and pays the energy cost of following them.

12.3 The arc of the book

This book began with a bacterium in the dark and the energy accounting it forces on us, and it ended with planetary engineering. The same physics operates at every scale.

The same \(\Delta G\) that governs electron transfer in a hydrogen atom governs whether a bacterial community 2.8 kilometers underground will thrive or slowly starve. The Michaelis-Menten kinetics that describe a single purified enzyme reappear, in effective form, as the rate laws of entire ecosystems – not because the enzyme and the ecosystem are the same system, but because saturation logic applies at both levels. The same conservation law that tracks a solute diffusing through a sediment column applies to carbon moving through the global ocean – though the transport operators and boundary conditions change substantially.

The underlying physics does not change. The models built on it must.

The claim is empirical, and it is the reason that reaction-transport models are portable at all. If the physics changed with scale – if microbial communities invented new thermodynamics – then every model would be an ad hoc curve fit. The fact that a model calibrated on porewater sulfate in one fjord can make useful predictions in another – even if recalibration is needed – is evidence that the non-equilibrium framework is not just a metaphor. It is the actual mechanism.

We moved through that framework in four stages:

Part I: The Rules of the Game established the minimum accounting that the rest of the story needs: free energy, kinetics, and the planetary boundary conditions that kept Earth wet, reactive, and open long enough for chemistry to become biology.

Part II: The First Society introduced the organisms, but in their most ancient and minimal forms: the first metabolisms, the first communities, the first catastrophic success (oxygen). Here the emphasis was on how life does not merely inhabit environments but reshapes them – sometimes constructively, sometimes catastrophically.

Part III: The Great Mergers traced how competition, cooperation, and endosymbiosis produced the cellular architectures we see today. Syntrophy turned out to be not an exotic curiosity but a dominant strategy: organisms that cannot survive alone thriving in partnerships where one’s waste is another’s fuel.

Part IV: The Hidden World and the Future opened by turning that machinery into a way of reading real systems: the conservation equation, the transport operators, the rate expressions, and the practical art of explaining a profile rather than merely admiring it. It then carried the story into the deep biosphere, groundwater redox, water treatment, climate feedbacks, and the open questions that define the frontier of the field.

12.4 What we still do not know

A book that pretends to have all the answers is advertising. Here are questions that remain open, drawn from the same source literature that informed every chapter:

How did the transition from inorganic catalysts to protein enzymes occur? We know that mineral surfaces can catalyze many of the reactions that enzymes perform today.8 We know that ribozymes can catalyze a subset of reactions using RNA alone.9 But the mechanistic path from mineral-catalyzed chemistry to the protein-dominated metabolism of modern cells remains sketchy. The gap is not just historical curiosity – it determines how we think about the likelihood of life elsewhere.

What would count as convincing evidence of life elsewhere now that ocean worlds and Mars are back on the table? Enceladus has phosphates, organics, and hydrogen. Europa is finally being surveyed with instruments designed for habitability rather than mythology. Mars has begun yielding redox-sensitive minerals and organics in settings that look, to a geomicrobiologist, uncomfortably familiar. But none of these observations is self-interpreting. We still do not know what combination of isotopes, minerals, textures, and disequilibria would amount to a persuasive biosignature rather than an intriguing chemical story.

How far does satisficing take us? This book has argued that microbial communities satisfice rather than optimize. I have used Simon’s term as an extension to microbial systems, not as established geomicrobiological doctrine.101112 The framework explains fuzzy redox boundaries and the coexistence of “losing” metabolisms. Critics of satisficing argue that the concept can collapse into optimization once you price search or information costs into the objective function.13 Microbial communities may resist that reduction because no agent surveys the whole reaction landscape and no shared objective joins unrelated lineages into one maximization problem. That defense is not enough by itself. The framework has to earn its place by predicting field behavior that optimization stories miss.

How do we model the lag phase? Every microbiologist knows that cells do not respond instantaneously to a new substrate or a new environment. There is a lag – sometimes minutes, sometimes weeks – during which gene expression shifts, enzymes are synthesized, and the population adjusts.14 Most reaction-transport models ignore this entirely. Whether that matters depends on the timescale you care about, but for predicting transient responses to environmental change, the lag phase may be the largest unmodeled source of error.

What is the mechanistic basis of organic matter degradation? We model organic matter breakdown with rate constants and reactivity distributions, and those models work surprisingly well.15 But the actual mechanism – the sequence of enzymatic attacks, the role of mineral protection, the feedback between microbial community composition and degradation rate – remains poorly understood at a mechanistic level. The “reactive continuum” is a powerful abstraction, but it is an abstraction.

Can we build predictive reaction-transport models that work across all environmental conditions? Current RTMs are good at reproducing observations in the environments where they were calibrated.16 Transferring them to new settings – different temperatures, different organic matter sources, different mineral assemblages – often requires recalibration. A truly predictive RTM would derive its parameters from first principles: thermodynamics, enzyme kinetics, and transport physics. We are not there yet, but the framework in this book is designed to move in that direction.

What is the relationship between energy supply, energy demand, and microbially catalyzed processes in the deep subsurface? The organisms living kilometers below the surface exist on maintenance energy budgets so thin that the distinction between “alive” and “dormant” becomes blurred. Understanding how these communities persist – and whether they are truly at steady state or slowly running down – is a question that connects microbiology to geology on timescales of millions of years.

These are not footnotes. They are the frontier.

12.5 The invisible architects, still at work

The title of this book is a statement of fact. Microbes built the oxygen atmosphere. They regulate the carbon cycle. They mediate the transformation of minerals, the cycling of nitrogen, the fate of sulfur. They created the chemical conditions that made complex life possible, and they continue to maintain those conditions today.

But “invisible architects” is also an invitation to shift perspective. We tend to think of the living world as the part we can see: forests, animals, coral reefs. The actual biological engine of the planet is invisible. It lives in sediment pores, in deep aquifers, in the fractures of basalt kilometers below the seafloor, in the thin films of water coating soil grains. By some estimates, the continental subsurface alone harbors on the order of 10\(^{29}\) cells – a substantial fraction of all prokaryotic life on Earth.17 These organisms are not waiting to be discovered as curiosities. They are running the geochemistry that makes the surface habitable.18

12.6 Three claims

This book has made three claims, nested from the microscopic to the planetary.

First: microbial communities appear to satisfice. I use Simon’s term as a lens for microbial systems, not as settled geomicrobiological doctrine. The claim is that communities cover maintenance costs under local thermodynamic and kinetic constraints rather than maximizing growth rate, energy yield, or community structure. This framework explains fuzzy redox boundaries, the coexistence of “losing” metabolisms, and the gap between laboratory rate constants and field rates. How far the lens extends remains an open question. It earns a place in the book only if it keeps explaining field behavior that optimization stories miss.

Second: the conservation framework carries across scales. The same accounting principle – accumulation equals net flux plus net reaction – applies to a sediment pore, a treatment wetland, a regional aquifer, and the global ocean. The transport operators, boundary conditions, and closure terms change with each setting, but the underlying logic does not. Reaction-transport models work because the physics does not reinvent itself at new scales. There is no “ecosystem thermodynamics” separate from “molecular thermodynamics.” There is only thermodynamics.

Third: water treatment underperforms when it ignores the geomicrobiology. The same broad classes of metabolism and the same thermodynamic and kinetic constraints that shaped the Archean atmosphere, that cycle sulfur through sediments, that maintain the deep biosphere on maintenance energy alone – these are the processes that clean contaminated groundwater. Better data will always matter. But data alone cannot fix the problem if the biology is modeled as a black box with fixed parameters rather than as an adaptive community under thermodynamic constraints. The theory in this book is one piece of what is needed.

These three claims are not separate arguments. They are one argument at three scales. Satisficing describes the biology. The conservation framework explains the math. And the water crisis is sharpened when we ignore both.

The invisible architects have not retired. They are still at work – in every grain of sediment, every drop of groundwater, every square centimeter of your skin. They preceded oxygen, eukaryotes, animals, and thought – and they will outlast us.

The question is whether we will make the effort to understand them.

  1. Erwin Schrödinger, What Is Life? The Physical Aspect of the Living Cell (Cambridge University Press, 1944). (Schrödinger 1944)↩︎

  2. Pier Luigi Luisi, “About Various Definitions of Life,” Origins of Life and Evolution of the Biosphere 28 (1998): 613–622. (Luisi 1998)↩︎

  3. Katrin Knittel and Antje Boetius, “Anaerobic Oxidation of Methane: Progress with an Unknown Process,” Annual Review of Microbiology 63 (2009): 311–334. The energetics of AOM at the sulfate-methane transition demonstrate life operating at the thermodynamic edge. (Knittel and Boetius 2009)↩︎

  4. Douglas E. LaRowe and Jan P. Amend, “Catabolic Rates, Population Sizes and Doubling/Replacement Times of Microorganisms in Natural Settings,” American Journal of Science 315 (2015): 167–203. Catabolic rates vary over twelve orders of magnitude across Earth’s biosphere. (LaRowe and Amend 2015)↩︎

  5. Carol Cleland and Christopher Chyba, “Defining ‘Life’,” Origins of Life and Evolution of the Biosphere 32 (2002): 387–393. (Cleland and Chyba 2002)↩︎

  6. Gerald Joyce, “Foreword,” in Origins of Life: The Central Concepts, ed. David Deamer and Gail Fleischaker (Jones and Bartlett, 1994). This formulation became the NASA working definition of life. (Joyce 1994)↩︎

  7. Eors Szathmary and John Maynard Smith, The Major Transitions in Evolution (Freeman, 1995). (Szathmáry and Maynard Smith 1995)↩︎

  8. Günter Wächtershäuser, “Before Enzymes and Templates: Theory of Surface Metabolism,” Microbiological Reviews 52 (1988): 452–484. The iron-sulfur world hypothesis posits that mineral surfaces catalyzed the first metabolic cycles. (Wächtershäuser 1988)↩︎

  9. Thomas R. Cech, “A Model for the RNA-Catalyzed Replication of RNA,” Proceedings of the National Academy of Sciences 83 (1986): 4360–4363. Discovery that RNA can catalyze reactions without protein enzymes. (Cech 1986)↩︎

  10. Herbert A. Simon, “Rational Choice and the Structure of the Environment,” Psychological Review 63 (1956): 129–138. The satisficing framework: organisms find strategies that are good enough, not optimal. (Simon 1956)↩︎

  11. Yohay Carmel and Yakov Ben-Haim, “Info-gap robust-satisficing model of foraging behavior: do foragers optimize or satisfice?,” The American Naturalist 166 (2005). Carmel and Ben-Haim provide one of the clearest empirical defenses of satisficing against optimal-foraging predictions. (Carmel and Ben-Haim 2005)↩︎

  12. Thomas J. Brennan and Andrew W. Lo, “An evolutionary model of bounded rationality and intelligence,” PLOS ONE 7 (2012): e50310. Brennan and Lo show how bounded strategies can persist under selection when cognition and information carry costs. (Brennan and Lo 2012)↩︎

  13. Werner Callebaut, “Herbert Simon’s silent revolution,” Biological Theory 2 (2007): 76–86. Callebaut reviews the attempt to fold bounded rationality back into optimization and argues that the reduction misses Simon’s target. (Callebaut 2007)↩︎

  14. Harvey W. Blanch, “Invited Review: Microbial Growth Kinetics,” Biotechnology and Bioengineering 23 (1981): 1691–1722. Lag phase dynamics and the limitations of unstructured models in microbial kinetics. (Blanch 1981)↩︎

  15. Sandra Arndt et al., “Quantifying the Degradation of Organic Matter in Marine Sediments: A Review and Synthesis,” Earth-Science Reviews 123 (2013): 53–86. The mechanistic understanding of organic matter degradation remains a major challenge for RTMs. (Arndt et al. 2013)↩︎

  16. Carl I. Steefel, Donald J. DePaolo, and Peter C. Lichtner, “Reactive Transport Modeling: An Essential Tool and a New Research Approach for the Earth Sciences,” Earth and Planetary Science Letters 240 (2005): 539–558. (Steefel et al. 2005)↩︎

  17. Cara Magnabosco et al., “The Biomass and Biodiversity of the Continental Subsurface,” Nature Geoscience 11 (2018): 707–717. Estimated 2–6 × 10\(^{29}\) cells in the continental subsurface; combined with other estimates, total global prokaryotic biomass of ~23–31 Pg C. (Magnabosco et al. 2018)↩︎

  18. Paul G. Falkowski, Tom Fenchel, and Edward F. Delong, “The Microbial Engines That Drive Earth’s Biogeochemical Cycles,” Science 320 (2008): 1034–1039. Review of microbial contributions to global biogeochemical cycling. (Falkowski et al. 2008)↩︎