I think it’s extremely improbable we’ll find life on Mars. Yet NASA is “well on its way” to finding life there. Or so headlines told us. The NASA head, Jim Bridenstin, who said that actually meant we were close to being able to test Martian soil for “biosignatures.” He did not mean the tests would be positive; although he did express hope they would be.

I can understand his naive enthusiasm. We’ve found all the basic ingredients there: organic chemicals, methane signatures, and liquid water (extensively on the surface in the Martian past; and maybe still today in some underground locales). And we know life can arise and thrive in extremely alien and inhospitable environments. Ward & Brownlee’s book Rare Earth long ago made this point already, leading them to propose that simple life may be cosmically common, but complex life must surely be rare, as it requires more hospitable and stable environments than places like Mars have to offer. Which is why we find it on earth.

So when we talk about “life” we do have to distinguish what we mean: there is simple life (e.g. microorganisms), complex life (e.g. multicellular organisms), and advanced life (e.g. conscious observers). Each is considerably more complex than the previous. And though we know evolution by natural selection causes relatively rapid rises in organism complexity, there are caps. Mars could harbor microorganisms. But it probably lacks conditions sufficient now to sustain multicellular organisms; and certainly isn’t compatible with conscious observers—which is why, if we ever go there, we will have to bring hospitable environments with us, just to survive (and even that is dicey).

But that Mars “could harbor” microorganisms does not mean it’s at all likely any arose there. As even Ward & Brownlee point out, that’s actually extremely unlikely. By local standards. By cosmic standards, they expect protobiology has spontaneously arisen in many star systems across the Milky Way and beyond. But getting life simultaneously on two planets in the same system is less likely than winning two poker hands with royal flushes back to back. And this has implications for science, philosophy, theology, and atheism.

The Theological Problem

Theists rightly note that even simple life requires such dizzying complexity that it’s appearance by accident seems unbelievable. Therefore, God must’a done it. Their mistake is in the math. Not the premise. Life does require the spontaneous formation of dizzying complexity. And locally, that is effectively impossible without intelligent design. But we aren’t an isolated locality. There are countless trillions of localities. Trillions of galaxies even. Which are just the ones we can see. And “the spontaneous formation of dizzying complexity” is actually probable on such scales.

In fact, this is really good evidence there is no designer god responsible for life at all. Because a god supposedly can just create life. He does not need all the things that are required for a godless universe to do it without him. Indeed it’s extremely weird that a god would make the world look exactly like a world without a god in it; and then expect us to conclude it was designed. In fact, I’d say that would be positively insane; fundamentally irrational. Hardly the behavior of anyone smart and capable, who was intent on being recognized or listened to.

As I’ve noted before, in my book Why I Am Not a Christian and my blog Bayesian Counter-Apologetics and most of all in my most detailed chapter on design arguments in The End of Christianity, the evidence pertaining to biogenesis is actually all evidence against divine causation. Not the other way around.

Indeed, just looking at the Historical Development of Origins Research, it’s obvious mechanistic explanations of the origins of life have been abundantly demonstrated, even if we don’t know the exact mechanism that operated in our case. Because mechanistic protobiological theories have born far more fruit predictively and empirically than any alternative. By contrast, no “God did it” theory has ever made a useful scientific prediction or led to any remarkable empirical discovery, much less generated an amazing database of pertinent knowledge.

But it’s worse than that. To summarize my blog’s summary:

The only way we could exist without a God is by an extremely improbable chemical accident, and the only way an extremely improbable chemical accident is likely to occur is in a universe that’s vastly old and vastly large. So atheism predicts a vastly old and large universe; theism does not. Observation confirms atheism’s prediction, not theism’s.

Similarly, the only way we could exist without a God is by an extremely long process of evolution by natural selection, beginning from a single molecule, through hundreds of millions of years of single cells, through hundreds of millions of years of cooperating cells, to hundreds of millions of years of multicellular organisms. So atheism predicts essentially that; theism does not. Observation confirms atheism’s predictions, not theism’s.

Likewise, if chance produced this universe, we should expect it to be only barely conducive to life, indeed almost entirely lethal to it, since there are vastly more ways to get “just barely hospitable” universes by chance selection, than to get a universe perfectly suited to life throughout. But the universe we observe is almost entirely lethal to life. Almost all of it is a lethal radiation filled vacuum. Of what isn’t that, almost all of it is stars and black holes on which no life can live or arise. Of what tiny, minuscule proportion remains after subtracting all of that, almost all of it is lifeless dust and ice and asteroids that can never harbor or produce life. Life-friendly locales are extraordinarily rare in our universe. Just as atheism predicts; not theism.

So theology gets no comfort from the fact that life requires improbable accidents. That fact actually refutes theism. It proves we indeed arose by improbable accident. Leaving no role for any god.

The Metaphysical Problem

But what if even an atheist challenges the premise? What if someone were to insist we actually don’t know that life requires an extremely improbable accident to arise? Because, after all, we only have a sample size of one to judge by.

I’ve already somewhat answered this in my 2004 paper on biogenesis in Biology & Philosophy (on which see now The Latest Proposal for a Probability of Abiogenesis). But there I was mostly concerned with theists trying to fabricate lower probabilities for spontaneous biogenesis than science, math, and logic allow any credibility to—rather than dealing with the contrary attempt to fabricate higher probabilities than science, math, and logic allow.

But even there, I noted that the simplest known self-replicating molecule has a complexity of random assembly on the order of 10^41. Which sounds low. But given that the size and age of the universe is such that we can fully expect events to occur by random chance that even have a probability as low as 1 in 10^150 (and that’s as admitted by creationist William Dembski himself), events with a probability of 1 in 10^41 could actually have occurred by random chance literally billions of trillions of times—in fact, more than even 10^100 times. So that kind of complexity is peanuts for our universe. It will be an extremely common event on that scale.

However, the probability of two such events occurring in the same solar system remains extraordinarily unlikely. Though it will happen. A lot. In fact, the odds are 1 in 10^82. Which means it could have happened well more than 10^40 times across all the trillions of known galaxies. But it would be extremely rare to find ourselves in one of those solar systems. Even at best, fewer than 1 in 10^40 systems with life, will be dual-life systems.

Of course, all this just means “events that improbable,” not the formation of life specifically. So it is not the case that there are 10^40 solar systems with two spontaneous biospheres. Rather, there are 10^40 solar systems in which events as improbable as 1 in 10^82 will have occurred—which events may have nothing to do with the forming of life. They could be solar systems with an asteroid shaped like Darth Vader’s helmet or something. The only thing we can deduce here is that the accidental assembly of a self-replicating molecule is vastly within the realm of things that definitely will happen in a universe such as ours. So we get no evidence from its improbability for intelligent design being involved. To the contrary, these facts are ample evidence against that.

Of course, once the first life forms, more complex life requires no extraordinary improbabilities. The environment rapidly selects increased complexity, as information bleeds from the environment into the organism’s reproduced code. Hence, evolution. Which in our case went from first life to complex life in just a few billion years. And it is a tautology that absent evidence to the contrary, you are always more likely to be typical than exceptional. Ergo, that’s probably a typical time scale. As long as the environment life begins in remains hospitable enough for a few billion years, complex life is probably inevitable without any intelligent meddling. So it’s really just a question of how long the average hospitability window is, that determines how common complex life will be. In all those zones that are hospitable enough for complex life at all; simple life will find a home in many more places than that.

But it’s getting that first spark that’s the hard part.

So can’t we bypass that? Do we really need peptide or nucleic molecule assembly? Do we really even need replication? Just because that’s how we happened to get life here, doesn’t mean that’s the only way to get it. Right?

There are two reasons why that reasoning won’t fly.

• The first is that we have a fairly thorough inventory now of all the “things that exist” at the fundamental level capable of materializing any structure we’d call life—and basically, that means atoms. Nothing else making up the contents of the universe lasts beyond fractions of a second or can form stable structures at all. So life has to be some assembled molecule of atoms. This is not because it’s the only way we’ve seen it done. It’s because we can determine already it’s the only way it can be done, anywhere.
• The second is that even in some hypothetical other universe, a comparable statement will always be true. Life is by definition a complex. And not merely so, but a complex structure that is stable enough over time to store and transmit enough information to even count as life at all. And that requires some relatively stable material, arranged in some particular complex order. Maybe in supernatural universes you can have life without this; but exclude the supernatural, and all you have left is this. There has to be some complex structure. And something must hold that structure over time, to keep it from dissolving.

Either way, what we need is a complex arrangement of something. Which by definition requires either intelligent design or random accident. Even a process that is driven by physical forces, e.g. the spontaneous ordering of crystals on the surfaces of clays, which is neither intelligent nor random, and which could be one of the ways the first self-replicating molecules arose on earth (and, potentially, on Mars), requires a random accident. Not merely the accident of the presence of those forces, and those crystals, and those clays; because that would only explain how a long enough molecule can get assembled. It does not cause that assembly to conveniently be in exactly the right order needed for the resulting molecule to be self-replicating. Indeed, if there were physical laws that did that, we should have in that very fact evidence of divine design of the universe itself. And in any event, if such remarkable “life-making” forces existed, particularly on earth, we should have observed them by now.

But do we even need self-replication?

Certainly, we know that requires highly complex molecular order; if it didn’t, we’d have made simple self-replicators by now. But even when we control every aspect of the mechanism, as in self-replicating computer programs, the minimum complexity required is huge, from a “random assembly” perspective. And in chemistry, we’ve been trying, countless ways for a hundred years now, and the smallest we’ve come up with is the Lee peptide, weighing in at 32 amino acids—themselves already each a rather complex molecule, but at least one that we know arises spontaneously in nature. But getting them assembled, and all in the right order, is another matter. You can randomly rearrange those amino acids, and what you get is sterile. No self-replication. Only when in that one order, do they start spontaneously creating copies of themselves (and only in a conducive medium) as a consequence of known physical laws. And we have found nothing simpler capable of doing that—even though, again, we’ve inventoried the basic materials of the universe pretty thoroughly by now.

But maybe you can have life without replication? What would that look like? NASA’s page on defining “life” has something to say about this. Many features are considered definitive of life, but when you strip away the ones that are found in things we do not count as life (e.g. replicating crystals), the one feature that remains inseparable is learning. Which does not mean only learning through a nervous system sensing the environment and storing information about it, but learning through natural selection.

For example, the DNA of a bacterium or RNA of a virus both “learn” over generations, by acquiring and storing information in their nucleic “code,” which information is then used to perform tasks, which are fundamentally describable as intelligent. Not intelligent in the sense of consciously aware. But intelligent in the same sense as an everyday computer: able to make discriminating decisions that obtain particular goals, such as metabolism, defense, and reproduction. Crystals don’t really do that. They don’t have a memory. They don’t make intelligent decisions. With every generation of them, information rapidly randomizes and degrades over time; it is not stored and preserved; nor intelligently used.

But one should ask, what would it take for crystals to be able to do those things?

This is where we get to the fundamental requirements of life. What you always end up with is analytically the same in every possible universe, and thus every possible planet and habitat in our universe: you need the ability of an organism to acquire information, store that information, and use that information to make decisions that ensure (among other things) the continuing preservation and use of that information—decisions without which the environment (or even just the laws of thermodynamics) would simply dissolve the organism back into the background chaos whence it came, before it would even be around long enough for us to notice it, much less care. And certainly, such an inevitably dissolving system, can never evolve, and thus can never produce advanced life. The only kind of life that can contemplate where it came from. Life like us.

But an organism that can do those minimal things requires a baseline set of complex information already stored: the minimum amount needed to start acquiring, storing, preserving, and acting on that information in the first place. This is why life is always by definition complex, and in fact extremely complex, by the standards of random assembly it requires. It is complex by definition. As in, life is analytically, necessarily complex. In a physical world, at any rate, it would be a logical contradiction to have a living organism that wasn’t complex in at least this minimal sense. And this complexity is high enough that random assembly is extremely improbable. And thus it must be extremely improbable to observe it happening twice in one locality—like one solar system.

This explains why we haven’t found simpler self-replicators than the Lee peptide. Even the Lee peptide is a poor replicator, as peptides lose information per generation far more quickly than most living organisms do; but so far as we know, not fast enough to prevent its evolving better information preservation, given enough time and an environ of adequate size—neither of which we can replicate in a laboratory. Protobiologists don’t have whole lakes and millions of years to work with, to wait around and see what would come about. So we can’t test it. At least not yet; computer simulations may be a key way to “create” the volumes and times required.

If there isn’t replication, then there has to be immortality. Without either, what you have left dissolves, rather than evolves. And replication is observably the most robust mechanism of preserving and building information, because it allows redundancies, and learning at the code level (through mutation and selection). It’s thus essential for evolution. And thus essential for any kind of life worth discovering. If all we found was a bunch of Lee peptides under the surface of Mars, that had never evolved into anything since first ever forming billions of years ago, that would be interesting, but not very impressive. We’d even debate whether that should be called “discovering life” at all.

Immortality seems a much less likely route for life to evolve by. How would it learn? How would it evolve? And immortality is hard to maintain; that’s why replication is so successful: most shit gets destroyed by the chaos of real world environs; so unless you are making tons of copies, survival becomes progressively unlikely. Perhaps there is some pathway, however extremely rare, by which an organism can grow and preserve its information database without replication, and long enough to even be observed before going extinct (dare we even say long enough to evolve consciousness, although that’s betting on some seriously long odds).

But even there we are still back to the same thing: a starting block of highly complex stored information. Without which we just have a bunch of rocks or goo. Chemistry. But not life. So getting even that unusual organism started, requires a spontaneous event of extraordinary improbability: the baseline starting arrangement of information minimally needed to keep it alive and continually adding information rather than losing it. Because things randomly ordered, or inadequately complex in order, cannot do either.

So there’s no getting around the extreme improbability of spontaneous biogenesis; and hence the need of an extremely large and old universe to ever have life, absent gods or the functional equivalent.

The Scientific Problem

On finding life on Mars, Wikipedia has a decent entry. It all focuses on only one side of the equation though: available resources (materials, conditions, etc.). But that’s not enough. Merely having the right set of resources still requires an average window of time for random interactions to assemble the minimal molecule needed to finally start storing and adding rather than losing information, and acting on that information to continue storing and adding rather than losing information. Once you have that, evolution by natural selection does the rest. And then it’s just the average window of time needed for that process to leverage its way to advanced cognitive systems. I discussed these kinds of calculations recently.

In practice, for life to have survived on Mars long enough even to be detectable today (much less still around today), it had to have an information system complex enough to preserve information (which requires intelligent repair mechanisms), to add information (which requires at least mutation and selection and thus reproduction or some equivalent operation), and to act on that information (at the very least for survival, of itself if immortal or its descendants if replicating). Much of which would have evolved by natural selection over time. But there had to be a minimal amount from the start—without evolution, which means by random and hence improbable accident—or else it cannot have ever begun doing any of these three things.

And it’s just extremely unlikely that will have happened on two planets in the same solar system independently. Mars may have all the right conditions. But it’s still very unlikely ever to have produced life. Indeed, if we do find life there, I put good odds it will be the same life as on earth. And I mean literally: we’ll determine Martian life evolved from microbes transmitted from Earth by an asteroid impact; or vice versa. Even more likely, it will be life we recently accidentally took there on the very probes we’ve been sending—though NASA has been taking extreme precautions to ensure against that, when do human efforts always work out as planned?

It’s of course even more problematic that Martian probes are only looking for amino-acid based life just like earth’s. They don’t even have instruments capable of detecting life based on some other structural substrate, if such is even possible. So if you’re of a mind to say that alien life is not likely to be just like earth life owing to our single-sample bias, then you should be even more pessimistic than I am that we’ll ever find life on Mars. Because frankly, in that scenario, we aren’t even actually looking for it.

There are many valuable reasons to explore Mars. Chiefly because we need to know all we can about it to move there, as inevitably we may have to do. And secondarily because it’s one of the easiest ways to do a bunch of extra-planetary science that has countless side applications, including improved understanding of earth science by increasing our sample size beyond one. But of all the reasons to be exploring Mars, the prospect of finding life is not one of them. Other than to run a falsification test on my very thesis: to confirm life isn’t there despite the conditions being suitable. Negative findings are still useful findings. They just aren’t perhaps the findings people want to spend money on.

Suitable conditions are certainly going to be commonplace; I’m not even surprised to see that confirmed. Ward & Brownlee already predicted it. But “suitable conditions” aren’t remotely sufficient for life to arise. A rare accident must also occur. One that can start what I just defined as a cascade of growing intelligence. And of course even that’s not enough to get all the way to a species capable of thinking about the universe (and thus pondering how it got there). The latter requires many more stages, of much longer windows of time, in environments far more suitable than Mars has proven to be.

The probability is pretty high, I think, that the first life on Mars will be us. And if aliens ever come to dig up our dead civilization, they’ll only find life on multiple planets here, because we put it there.

This is all hypothesis, of course. I’m placing bets. Not asserting it a fact. And I’m just explaining what the basis is for where I’m placing those bets. Currently, I don’t see any evidence warranting moving those bets.