Chapter 9: Is a Living Planet Necessarily a Firing One?

This is the chapter where the map runs out, and I want to say so at the top rather than let you discover it halfway down.

Register: this chapter is speculative from beginning to end. Where I lean on established results I will say which, but the question being asked has no settled answer and I am not going to manufacture one. If you have been reading for the parts of this book that stand on measurement, Chapters 7 and 8 were those parts. This one is reaching, and it is offered as reaching.

Part III has been building an argument: that a planet's interior heat engine supplied the air and the ocean, manufactured the continents, and runs the thermostat that has held this world in the liquid-water range for four billion years. Chapter 8 tested it against four neighbours and the argument survived — indeed it sharpened, because Venus forced the correction that it is the circuit that matters and not the eruption.

And then Chapter 8 handed you the weapon, which I would rather do myself than have a reviewer do for me. If there is life in the ocean of Enceladus, then a biosphere exists on a body with no atmosphere, no weathering, no subduction, no continents and no carbon cycle — none of the apparatus I have spent two chapters arguing is necessary.

So the question has to be asked properly.

Three questions wearing one coat

Most discussions of this topic run aground because they conflate claims of very different strength. Let me separate them, because once separated they have different answers and the differences are the interesting part.

Question one: is a geologically active interior necessary for life to begin?

Question two: is it necessary for a biosphere to persist over billions of years?

Question three: is it necessary for a biosphere to be detectable from another star?

These are not the same question. They are not even closely related. And I think the honest answers are, respectively: probably not, probably yes for one kind of biosphere and probably not for another, and — this is where the chapter is going — almost certainly yes.

Question one, briefly

Chapter 6 did most of this work and I will not repeat it.

The strongest-developed origin scenario places the event at an alkaline hydrothermal vent, and such vents are driven by serpentinization rather than by magma. The volcanic contribution is one link further back: the tectonic system that exposes mantle rock to seawater. The rival scenario, at subaerial hot springs, is magmatic, and has real arguments behind it, and is less well developed.

What every scenario except panspermia requires is a wet rocky body maintaining chemical disequilibrium. That is a requirement about the availability of gradients, not about eruption specifically. And Chapter 8 established that tidal flexing can supply the energy to maintain those gradients on a body with no radiogenic engine worth speaking of.

So: probably not necessary, in the strict sense. Necessary that something keeps the chemistry from equilibrating. Volcanism is one way. It is not the only way.

Question two: the case for necessity

Here the argument is much stronger, and I want to put it at full strength before I attack it.

The climate argument. Stars change. Over billions of years a main-sequence star brightens substantially — ours by something like a third. A planet with a fixed atmosphere and no compensating mechanism will therefore not stay habitable; it will pass through a habitable window and out the other side. The carbonate–silicate thermostat is the only mechanism anyone has identified that can hold a surface climate steady across that kind of change, and it requires volcanic outgassing as its source term and tectonic recycling as its return. Remove either and you get Mars or you get Venus, both of which we have.

The disequilibrium argument. Life runs on gradients, and gradients run down. A geologically dead planet equilibrates: its reduced species get oxidized, its reactive minerals get weathered to completion, and the free energy available at the surface goes to zero. Interior activity is what continuously reintroduces reduced material — hydrogen, methane, sulfide, ferrous iron — into an oxidizing environment, and thereby keeps the battery charged. A planet with no interior activity is a planet whose chemistry has finished happening.

The nutrient argument, which is the one most often left out and which I think is among the strongest. Life needs phosphorus, and phosphorus is not abundant. Its main source is the weathering of apatite in igneous rock. On a tectonically active planet, uplift continuously exposes fresh rock, weathering liberates phosphorus, rivers carry it to the sea, and the biosphere is resupplied — while burial in marine sediment removes it, and subduction and uplift eventually return it. Without that circulation, phosphorus is a one-way flux into deep-sea sediment, and a surface biosphere on a tectonically dead planet slowly starves. The same argument applies with variations to the trace metals — iron, molybdenum, nickel, cobalt — that sit in the active sites of ancient enzymes.

The habitat argument. Continents make dry land, varied topography, shallow seas, isolated basins. That is where most of the diversity and most of the evolutionary novelty on this planet has happened.

Taken together, that is a serious case. A surface biosphere on a rocky planet, persisting for billions of years, appears to need a mechanism for regulating climate, regenerating chemical gradients, and recycling limiting nutrients — and the only mechanism we know of that does all three is a working interior heat engine coupled to a tectonic circuit.

Question two: the case against

Now the objections, and they are not weak.

The sample size is one. This is the fundamental problem and everything else is a variation on it. We have exactly one example of an inhabited planet, and every single feature of that planet is perfectly correlated with the presence of life on it. Earth has plate tectonics and life; it also has a large moon, a particular atmospheric mass, a specific rotation rate, a Jupiter in the outer system, and an oxygen-rich atmosphere. From n equals one you cannot distinguish a necessary condition from an incidental one. Any argument of the form "life needs X because Earth has X and Earth has life" is unfalsifiable in its current state, and I have just spent two pages making four such arguments.

I do think the four arguments above are better than that, because each proposes a mechanism rather than merely noting a correlation. But mechanisms can be real and still not be the only route.

The subsurface objection. This is the one Chapter 8 set up. A biosphere in a subsurface ocean, heated tidally, feeding on hydrogen from water–rock reaction, requires essentially none of the apparatus described above. No atmosphere to regulate. No weathering. No continents. No nutrient delivery by rivers — the rock is right there at the bottom of the ocean. No thermostat, because there is no surface climate to stabilize; the ice shell does the insulating and the tides do the heating, and neither cares what the star is doing.

If Enceladus is inhabited, question two is answered in the negative, decisively, and Part III's claim has to be narrowed to surface biospheres specifically.

The stagnant-lid objection. There is modelling work suggesting that planets without plate tectonics may nevertheless outgas enough carbon dioxide, over long enough, to maintain habitable surface conditions for billions of years — that a stagnant lid with volcanism can partly substitute for a full tectonic circuit, depending on the planet's size, its volatile inventory, and its radiogenic budget. Venus is not a proof that stagnant-lid worlds fail; Venus is a proof that a stagnant-lid world at Venus's distance from this star failed. Move it out and the outcome might differ.

The alternative-regulation objection. Seafloor weathering — basalt reacting with seawater — also consumes carbon dioxide, and it does not require continents. It may be capable of providing a weaker thermostat on a waterworld with no exposed land at all. If so, the continent-making part of the argument is doing less work than Chapter 7 implied.

The duration objection. How long does a biosphere have to last to count? I have been assuming billions of years, because that is what happened here and because complex life took most of that time. But if the question is whether life exists elsewhere, rather than whether it becomes complex, then a habitable window of a few hundred million years might be plenty, and over that span the regulation requirements are considerably looser.

The analogy I am not going to make

I need to stop and refuse something explicitly, because this chapter's material sits adjacent to a set of ideas that it would be very easy to slide into, and sliding is exactly the failure mode this book promised to avoid.

The planet is not an organism. Volcanism is not a metabolism, except as a figure of speech, and I have used it as a figure of speech in this book and I want to withdraw any weight it might have accumulated. A metabolism is a coordinated set of catalysed reactions maintained by an entity that reproduces and is subject to selection. A planet losing heat is a heat engine. The resemblance is a resemblance.

It is worth distinguishing two claims that often travel under the same name. The weak version — that life substantially alters its planetary environment, and that biology and geology are coupled systems that must be studied together — is true, well evidenced, and not controversial. Photosynthetic organisms oxygenated the atmosphere. Land plants measurably accelerated silicate weathering, which means biology is a term in the very thermostat I described in Chapter 7. That coupling is real and important.

The strong version — that the biosphere regulates planetary conditions homeostatically, as an organism regulates its body temperature, in a way that suggests planetary-scale purpose — is a different claim and I do not think it is supported. The standard objection is the decisive one: homeostatic regulation of that kind is the sort of thing that arises by natural selection, and natural selection requires a population of competing, reproducing entities. There has only ever been one Earth. It has no siblings, it does not reproduce, and there is no mechanism by which planetary-scale self-regulation could have been selected for.

The thermostat in Chapter 7 is not evidence for the strong claim, and I want to be emphatic about this because it is the most likely misreading of this book. The carbonate–silicate feedback is abiotic in its core and is not for anything. It is a negative feedback of the same species as the one in a refrigerator. It stabilizes because of the temperature dependence of a reaction rate, not because anything wants it to. It has, several times in this planet's history, run in directions that killed most of what was alive — which Part IV is about — and it will, in the end, be part of what ends the biosphere entirely, which the coda is about.

An indifferent mechanism that happens to have kept the surface habitable is not less remarkable than a purposeful one. It is more remarkable. But it is not the same thing, and conflating them is how a scientific argument becomes a consoling story.

What we could actually detect

Suppose we wanted to test any of this. What would we look for?

The proposal that gets most attention is sulfur dioxide. Its virtue is its short atmospheric lifetime: it is destroyed on timescales far shorter than a planet's age, so detecting it in an exoplanet atmosphere implies something is actively replenishing it. Volcanism is the obvious candidate. A persistent SO₂ signature would be, in effect, a geosignature — evidence not of life but of a planet that is still running.

Other candidates have been proposed: particular ratios of carbon species, transient signals from individual large eruptions, thermal variability.

Now the honest assessment, which is considerably less exciting than the proposals.

We have not yet confidently characterized the atmosphere of a temperate rocky exoplanet at all. The most accessible targets are small rocky planets around small cool stars, because the planet-to-star size ratio is favourable, and the best-studied of those systems has so far mostly produced results consistent with the inner planets having thin atmospheres or none. Detecting a rocky planet's atmosphere is at the edge of what current instruments can do. Detecting a specific minor constituent within such an atmosphere, and attributing it to volcanism rather than to something else, is beyond it.

And there is a deeper problem, which I think is underappreciated and which follows directly from Chapter 8.

The thing we might be able to detect is not the thing that matters. Chapter 8's whole lesson was that volcanism alone is insufficient — Venus has volcanism in abundance and is a furnace — and that what matters is the complete circuit including the subduction return path. But an SO₂ signature indicates outgassing. It says nothing whatever about whether the carbon is coming back. There is no proposed observable that directly indicates subduction, and I am not aware of any credible route to one.

So we might, eventually, learn to identify planets that are volcanically active. That would be a real achievement. It would not tell us the thing this book has spent Part III arguing is the important thing.

Lava worlds

Briefly, because it is the furthest reach in the book and I do not want to indulge it.

At the extreme end are rocky planets orbiting so close to their stars that their daysides are molten — surface magma oceans, tidally locked, with atmospheres composed of vaporized rock: sodium, silicon monoxide, oxygen boiled off a liquid surface. Several are known. Some have had contested atmospheric detections, and the interpretations have been revised more than once.

These are not habitable and nobody suggests they are. They are interesting for a different reason: they are the only place we can currently observe silicate magma at planetary scale from outside, and they may eventually teach us about magma ocean chemistry — the state every rocky planet, including this one, passed through in its first few million years.

That is genuinely the entire case, and I am going to leave it there rather than dress it up.

What would break this

A hypothesis without a failure condition is a slogan, so here are three things that would seriously damage the necessity claim, in rough order of how soon we might get them.

One: a confirmed biosphere in a subsurface ocean. If Enceladus or Europa turns out to be inhabited, then life persists without atmosphere, weathering, subduction or thermostat, and the necessity claim collapses to a claim about surface biospheres only. This is the one that might actually be settled — not soon, but plausibly within a few decades, which is remarkable fortune for a question about planetary history.

Two: a well-characterized stagnant-lid planet with a stable long-term climate and biosignatures. This would sever the link between the return path and habitability. It is a long way off.

Three: a demonstration that seafloor weathering alone provides adequate regulation. This is a modelling result, not an observation, and it could arrive at any time. It would not kill the argument but it would substantially weaken the continental part of it.

I should note what would support the claim, since it is easier to list threats than confirmations: a statistical pattern, across a large sample of characterized planets, in which atmospheric biosignatures correlate with indicators of geological activity. That requires a sample we are nowhere near having.

Where I land

Let me give you my actual reading, flagged as opinion, since I have asked you to follow me this far.

On origin: volcanism is probably not necessary. Sustained chemical disequilibrium is. Volcanism is one way to get it and tidal heating is another.

On persistence: for a surface biosphere lasting billions of years on a rocky planet, I think the case for necessity is strong — climate regulation, disequilibrium, and nutrient recycling all point the same way, and each proposes a mechanism rather than merely a correlation. For a subsurface biosphere, I think the case fails, and Enceladus is the reason.

And on detection — this is where I have ended up after writing the chapter, and it was not where I expected to end up.

If most habitable environments in the galaxy are subsurface oceans on icy bodies, then most habitable environments are effectively invisible. They have no atmosphere to take a spectrum of, no surface to image, no signal that propagates. A biosphere under twenty kilometres of ice is undetectable at interstellar distance by any method I can imagine anyone proposing.

Which means that the biospheres we could find are the ones with atmospheres — and an atmosphere on a rocky planet, sustained over geological time against escape and chemical loss, needs continuous resupply from the interior. A planet that stops outgassing loses its air, as Mars did.

So the answer to the question in the title may be: no, a living planet need not be a firing one — but a living planet that we can find almost certainly is. The engine is not a requirement for life. It may be a requirement for life that is legible from outside.

That is a stranger conclusion than the one I set out to argue, and I think it is more defensible. It also means that every search for life beyond this solar system is, whether or not it is framed this way, a search for planets whose interiors are still hot.

Part III is done. The case is made, with its limits stated: the engine built this world's air, ocean and land, it runs the thermostat, and it is very probably the condition for any biosphere we will ever detect.

Now the other half of the ledger. The same machinery, at rates nothing can buffer.

Draft notes — verification status

Standing convention. This chapter is explicitly speculative and makes fewer hard factual claims than any other, but the empirical scaffolding it leans on still needs checking — particularly the observational-status claims, which are time-sensitive and which the chapter uses to deflate its own proposals.

Pending verification — high priority:

The current state of temperate rocky exoplanet atmospheric characterization. The draft claims no temperate rocky exoplanet atmosphere has been confidently characterized, and that TRAPPIST-1's inner planets have yielded results consistent with thin or absent atmospheres. This is the chapter's main deflationary claim and it is time-sensitive — verify against the latest results and update before publication.
SO₂ as a proposed volcanic geosignature: atmospheric lifetime, the resupply argument, and current detectability assessments.
The claim that no proposed observable directly indicates subduction or tectonic recycling. This is a strong negative claim and carries the chapter's most original point. Check the literature for proposed tectonic biosignatures/geosignatures before asserting it.

Pending verification — standard:

Solar luminosity increase over Earth history — already flagged in Chapter 7 as inconsistent between passages there ("a quarter to a third" vs "thirty per cent"); reconcile across all three chapters.
The phosphorus argument: apatite weathering as the dominant P source, burial in marine sediment as the sink, and the claim that a tectonically dead planet's surface biosphere would become P-limited. Verify this is a real position in the literature rather than an extrapolation.
Trace metal requirements (Fe, Mo, Ni, Co) in ancient enzyme active sites and their resupply pathways.
Stagnant-lid habitability modelling — the claim that stagnant-lid planets may sustain habitable surface conditions for billions of years under some parameter ranges (Foley, Smye, Tosi and others). Confirm attribution and the strength of the result.
Seafloor weathering as a potential carbon sink capable of climate regulation without continents; its estimated strength relative to continental weathering.
Land plants accelerating silicate weathering — the magnitude of the effect and whether "measurably" understates or overstates it.
The Gaia critique: the argument that planetary-scale homeostasis lacks a unit of selection (Doolittle, Dawkins and successors). Confirm the standard formulation before relying on it.
Lava worlds: named examples (55 Cnc e, K2-141b, Kepler-10b), rock-vapour atmosphere compositions, and the contested status of 55 Cnc e's atmospheric detections. Recent and revised more than once — check current status.
Europa Clipper and any Enceladus mission timelines relevant to the "few decades" falsification claim. Time-sensitive; see also Chapter 8 notes.

The same machinery, at rates nothing can buffer