Creationists aren’t just operating on a misunderstanding and ignorance of the science (often wilful); they are also operating on broken epistemologies. The case of biogenesis affords us an illustration. I’ve written many articles on this. For example, in Oh No! Biogenesis Is Impossible?? A Case Study in Creationist Lies I document Jerry Bergman’s catastrophic ignorance and recourse to outright lying in his attempt to dissuade the public from believing actual scientific facts. But I also included one section on his broken methodology. Although given his documented dishonesty, he might well know it’s broken and is just hoping readers don’t notice; but recently I encountered a creationist who seems to sincerely not know this methodology is broken. Time to educate.

In response to Bergman I made the crucial point that “having more than one evidence-based theory of how something could have come to be increases the probability that it came to be by one of those methods; it does not decrease that probability.” And accordingly, “the more natural-cause theories we have that have been shown to be plausible on a basis of evidence, the more likely it is that a natural cause is responsible,” whereas “in any competition between a theory supported by evidence and precedent and a theory supported by none, the former is always more likely.” And “these two principles are not arbitrary assertions, but mathematical certitudes.” I demonstrate that there, so I won’t repeat it here, but merely summarize: vast evidence establishes naturalist theories tend to be true and theist theories never are (Naturalism Is Not an Axiom of the Sciences but a Conclusion of Them), therefore all else being equal, “god did it” is always the least likely explanation of anything (especially given The Argument from Specified Complexity against Supernaturalism). And when we have numerous theories each of which fully explain an observation, and all of which fully follow from and cohere with all empirically established scientific facts to date, there can be no logically valid argument that god must have done it instead. The evidence establishes that that is almost certainly not the case. Even when we don’t know which of those many competing theories is true. Just as with UFO’s: we don’t need to be able to prove which specific explanation of them happened in any given case; because we already know, on a basis of accumulated evidence, that any number of explanations are vastly more likely than “aliens did it” (UFOs Are Not That Remarkable).

If there are a dozen theories, all cohering with and following from the empirical facts, each of which rendering the observation that biogenesis occurred on the early Earth probable, then it is an actual fact of the matter that we know from empirical evidence that a natural biogenesis probably occurred on Earth. We therefore have no need of any other explanation. The question of which specific way biogenesis historically occurred on Earth remains an interesting question; but we no longer need look for any other theory than natural ones. If there are, let’s say, twelve competing naturalist hypotheses, all plausible, then the fact that biogenesis is a natural occurrence is simply P(H1) + P(H2) + P(H3) + P(H4) + P(H5) + P(H6) + P(H7) + P(H8) + P(H9) + P(H10) + P(H11) + P(H12) + P(H*) = ~100% (where the ~ symbol here indicates approximation and H* indicates “any other possible hypothesis”). Consequently, having so many theories substantially increases, not decreases, the probability some natural explanation is true. And we know that sum probability is very near 100% because the historical success of naturalistic explanation (in contrast with the exactly converse historical failure of supernatural explanation), and given that each of those theories make the appearance of life an expected outcome, and all match extant evidence (none are ruled out by it). There is therefore simply no logical case to make that “god did it” needs to be included in the theories we are exploring. And that’s that.

Q Doesn’t Get It

An anonymous creationist using the moniker “Q” attempted to deny this in response to my article The Latest Proposal for a Probability of Abiogenesis (in a comments thread beginning here). I do not know their gender but unless corrected I will here operate from the statistical assumption they’re a man (in my experience, creationists who dispute with people online with this particular kind of sustained stubbornness tend to be). He begins with the opening line, “The lack of an empirically verified naturalistic explanation for the origin of life remains problematic for metaphysical naturalism.” In that single sentence he betrays his broken epistemology. The conclusion is a non sequitur. He falsely presumes we have to know exactly what happened on the early Earth to give rise to life (which of a dozen or so different scenarios), in order to know that, whatever it was, it was probably a natural event (one of those dozen or so different scenarios, or any other comparable). That’s illogical. Just as we can empirically prove someone was murdered without knowing exactly who murdered them (or even how), we can empirically prove biogenesis was a natural event without knowing exactly which pathway nature took to get there. As long as we have several explanations in total agreement with science which posit nothing we have not already empirically proved possible or likely on the early Earth, and all those explanations are natural-cause explanations, then we know for an empirical fact that biogenesis was probably a natural event. End of story.

Hence there is no “problem” for metaphysical naturalism here. As I explained to Q, we have several naturalistic explanations for the origin of life, all match known science and available evidence. Q is confusing that, with knowing which one is true. But we don’t need to know which one is true. Because we already know there are a dozen, any one of which can be true. Those explanations all completely fit the evidence and known science, and do so vastly better than any made-up supernatural nonsense: as I demonstrated in Bayesian Counter-Apologetics, “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,” therefore naturalism “predicts essentially that” whereas “theism does not.” The evidence matches that prediction, and thus confirms that biogenesis was far more likely a natural event rather than a supernatural one.

Creationists like Q will want to claim that we should go against all this evidence “because” we lack any explanation for how biogenesis could be a natural event. But that’s false. Not only do we not lack an explanation, we have a dozen or more explanations. That’s exactly the opposite of having no explanation. So that argument is blocked. So then they try to say that because we only have models—we haven’t yet reproduced any in the lab (because we can’t; the durations of time needed are not available to us)—this means they “can’t” be true. But that’s a non sequitur. As long as those models are fully in conformity with the evidence (and all the findings of the sciences to date) and have some specific evidence in their support, and match peculiar predictions (like the bizarre age and size and material scope of the universe and the oddly long and meandering history of life on Earth from molecules to plants and animals), they have a far higher probability of being true than theories for which we have no evidence whatever and that go against all past findings of the sciences and fail to match any predictions about life and the cosmos. Having a lot of evidence always makes a theory more probable by far than having none; so it does not matter whether we could have a great deal more evidence still. The conclusion is already reached by that point: whatever caused life to emerge on the early Earth, it was almost certainly some natural cause.

Desperate and on the ropes now, a creationist like Q will try then to move the goal posts, and admit, okay, yes, we do have explanations and that we “lack” one was a mistaken assertions, “but” our explanations are all just speculations, and therefore not “evidence” of the point. But now Q has fallen into a pit as rife with fallacies as the Well of the Souls was rife with asps. First, none of these theories is “just speculation.” Every single one rests on a vast body of empirical scientific evidence establishing they would work given the conditions: with the right environments plus enough time, a natural biogenesis is ~100% expected. Not a speculation. Fact. No such case can be made for the “god did it” hypothesis. And that leaves only the question of how probable it is that there are enough environs and time in the universe to ensure at least one of those methods would produce life somewhere in the universe to a probability near enough to 100% to be assured. And vast empirical and scientific evidence establishes the answer is that there is: there are so many planets and moons, and the conditions required on them are so readily produced in nature, that that probability is indeed as expected. Which is itself a confirmation: why else would the universe and the history of life be so unusually like that, unless life arose just as we suspect?

Hence as I said to Q, “In the same way, we don’t need to know which horse Caesar rode when he crossed the Rubicon. We have enough evidence to know he crossed the Rubicon, and probably on a horse. The fact that we don’t know which horse does not warrant concluding he therefore was carried across on a flight of angels.” It was probably a horse. End of story. Indeed, that it was even a llama rather than a horse is still vastly more probable than a flight of angels. For llamas and the means to transport them to another continent have been empirically proven to exist and function as required. So the fact that it almost certainly wasn’t a llama tells us the probability that it was angels is as near to zero as makes all odds. We don’t even have to consider it. By the same reasoning, the probability life originated on Earth by natural process equals the total probability of natural biogenesis occurring anywhere in the universe in conditions relevantly similar to the early Earth’s. Just as, if I won the lottery, you can’t claim I didn’t win because winning lotteries is improbable, because lotteries are routinely won, and I’m as likely to be the winner as anyone else who played; therefore you would need evidence against my likely being the winner (evidence, for example, that I’m lying about that, or mistaken somehow) before you could rationally conclude otherwise. And there is no evidence against natural biogenesis on Earth. So that rhetorical strategy is forestalled.

So when Q ignored all this and insisted instead that “proponents of abiogenesis have yet to even demonstrate empirically that abiogenesis can happen” (we have abundantly proved empirically that it can) “and that it has happened” (there is abundant empirical evidence that it did) and therefore it’s only “an empirically unverified working assumption,” he appears not to understand what “evidence” is or how it works to increase the epistemic probability of a conclusion. So what we have here is a broken epistemology. The conclusion that any of a dozen or so ways life can arise by natural chance accident “can” have occurred on the early Earth is not a “working assumption,” as if we just insist upon it on a basis of no evidence. We have considerable empirical scientific evidence that these pathways could work; this is a conclusion based on evidence, which is the opposite of an assumption. If it were an assumption, we wouldn’t have all that evidence these theories are argued from. But we do. And that’s why they are declared plausible theories. This is the opposite of a “working assumption.” Meanwhile, the evidence that one of these pathways did occur on Earth is similarly strong: vast evidence establishes life did originate with a single molecule, and had to evolve over billions of years into multi-celled life-forms; and vast evidence establishes that the conditions required for each of the dozen or so possible ways it could have so originated actually did exist on the early Earth, or too easily could have to rule them out. We do not “assume” any of this. These conclusions all derive from a considerable amount of evidence.

By contrast, “god did it,” rests on no evidence whatever. That is what a mere “working assumption” looks like; indeed, worse, as it contradicts all this evidence. “God did it” does not of itself predict the observation that life would begin with a single molecule and take billions of years to even figure out how to assemble into more complex life (and still take hundreds of millions of years more to figure out how to make people). Whereas natural biogenesis predicts exactly that. And does so by pathways empirically proven to exist in the known universe, and empirically proven to exist on the early Earth. Only some theories depend on elements proven only to be possible on the early Earth, but even that evidentially exceeds a mere assumption, and many theories don’t even require this step. We therefore have many well-evidenced theories; the theist, does not. The evidence looks exactly like it would have to look if God was not involved. The laws of evidence entail the conclusion must then be that God was not involved. The only way to reverse that conclusion is to produce evidence that (a) God exists and (b) desired to create life in precisely exactly the way that looks like it wasn’t created by God and (c) did so. Theists can present no evidence for any one of those assumptions. And the cumulative evidence against (a) is, honestly, already insurmountable (see the whole of Bayesian Counter-Apologetics). It is all this evidence for natural biogenesis, and the complete lack of evidence for any god as a cause of anything (much less one who’d act so strangely as this, much less did), that remains problematic for theism. Not the other way around.

Q Is Bad at Math

Understanding the significance (the empirical and logical consequence) of evidence requires understanding mathematics, at least to a sixth grade level (in the American system). For example, Q kept insisting that “the 1 in 10^41 odds you give is for (‘toy model’) assembly of a single, self-replicating molecule, NOT for the origin of life,” even after having been told why that’s not the case. Only two things need occur for biogenesis: (1) the right molecular linking of amino acids, (2) in the right environment (an environment conducive to rather than hostile to that chain forming, metabolizing, and reproducing). Of course this is assuming that’s the only way life can arise; others might be possible, but I’m setting that aside in the present case because I think it would be too complicated for Q to grasp. He has already shown he does not understand how evidence works or how probability works; he therefore is incapable of understanding how the potential availability of other bases of evolved life than amino acids is not an assumption but, once again, based on evidence, and the consequence of which is an increase (not a decrease) in the probability of natural biogenesis on Earth. He also does not appear to understand degrees of evidence. The evidence for natural amino-acid-based molecule-chaining being a likely event on the early Earth is enormously greater than the evidence for the potentiality of life arising in this universe from other molecules; yet there is evidence for the latter. Degrees of evidence alter only degrees of certainty. And yet the evidence for other molecular bases being available is quite strong—strong enough to be confident it’s true. Ergo the evidence for natural amino-acid-based molecule-chaining being a likely event on the early Earth, being so much greater, produces an equivalently greater confidence. Whereas nothing the like exists for “god did it.” That’s how evidence works. Anyone who does not understand any of this, cannot comprehend anything about this matter. But let’s pretend that’s not the case and assume only amino-acid chaining can generate life, and that any creationist we are talking to actually understands how evidence and probability work (I know that’s a tall order).

As I replied to Q, “I have said over and over again that the remaining components” required for biogenesis (which all relate to defining the environment the right molecule chain must arise in) “have a probability of coordination in the cosmos of essentially 100% and therefore do not affect” the probability of biogenesis. In other words, when we “add in” the need to account for the second component (environment), even if there was only one chance at it in the entire history and expanse of the cosmos, “the probability of biogenesis is [still] (1 in 10^41) x (~1), which is as near to (1 in 10^41) as makes all odds.” Therefore none of the other components required for biogenesis have any significant effect on that probability. That’s why the assembly of the chain is the only relevant probability scientists bother looking at (see Why Life Must Be Complex (and Thus Probably Won’t Be on Mars)). All the other components (the conducive environs) have been established (and established on a basis of evidence) to already happen with abundant expected regularity across the universe (as even demonstrated in Rare Earth despite that book being obsolete now; empirical discoveries since have even further shored up its point). To get a better estimate than just that 1 in 10^41, all we need estimate is how much total “environment” there is in the universe (in both chemical volume and duration of existence), and how many other ways life can arise in such an environment, accounting for both of which vastly increases the probability above 1 in 10^41. Not the other way around. And conservative, evidence-based estimates exist for these things (see Could Be a 38% Chance We Are the Only Civilization in the Known Universe and, again, The Latest Proposal for a Probability of Abiogenesis). Not assumptions; conclusions from evidence. An assumption would be to make up numbers on a basis of no evidence. Instead, estimates based on what conservatively follows from actual evidence (the quantities that have to at least exist given present evidence, lest we have to invent some new arbitrary facts to explain otherwise) are not “assumptions.” Yes, they aren’t necessarily a conclusively proven fact (like the existence of, say, the Higgs boson, proved to a probability around 99.999%). But we don’t need that scale of certainty to be assured of something. Anything approximating 100% is good enough a basis on which to be confident (many sciences settle on 95%; I prefer at least 99%; but philosophically even just 90% is enough to rule out less probable alternatives like gods).

This is why Q doesn’t understand the irrelevance of his claim that “a single self-replicating molecule is neither living, nor sufficient in itself to originate (or evolve) life.” That statement is also false. In the context of biogenesis, self-replication is the definition of life (as that entails metabolism, and every other component associated with life—like “growth and functional activity”—is an evolved, not originating, feature of life). But it’s also a non sequitur. Because no one is claiming the chain itself is “sufficient,” as if the requisite molecular chain arising in the vacuum of intergalactic space is something we are counting as biogenesis. It is not. We full well know (and all our theories and calculations of probability assume and take into account) that the molecule must arise in an appropriate environment. But the probability that such environments exist is already known on abundant evidence to be (for all intents and purposes) 100%. We therefore don’t have to account for it’s existing. That’s why the only probability we talk about is of the chain occurring. That’s the only thing left to explain. And even then, in actual fact, when we calculate that probability we are taking into account the need of conducive environments, and thus include conservative evidence-based estimates of how many opportunities there have been for it to occur: we know as a matter of empirical fact that there was not “only one chance at it in the entire history and expanse of the cosmos,” but quite a lot of chances; because the universe is vast. So the total probability must take that into account. But we are not including instances of this occurring on the Moon or the Sun or in the vacuum of space or even almost anywhere on Earth; we only count potential places (limited volumes of space) even on Earth (as well as every other moon and planet possible) as relevant for the summation of odds.

Confusing Biogenesis with Evolution

A very common creationist mistake is to keep falling into confusion between biogenesis and evolution. Even after the difference is explained to them and they seem for a while to get it, they keep falling back into the same confusion. And this is a crucial mistake. Because the two scenarios involve completely different math. The arising of the first self-replicating molecule—the protobiont—in a godless universe is necessarily a product of chance accident, a straight probability calculation of the right molecules linking together in the right order from random mixing in a chemically active volume. That’s why it can be figured from a straight odds of a sequence (e.g. the self-replicating Lee peptide, which scientifically could have evolved into the entire present biome, has a one-off probability of random assembly of 1 in 10^41) measured against a number of trials (how many times the “parts” that already exist in nature have been mixed in ways that could produce the requisite molecule chain). For example, if the required naturally-occurring molecules (glycine-backed peptides in the case of PNA-first scenarios; nucleotides in the case of the RNA-first scenarios, per Michael Marshall’s summary for New Scientist) only ever existed in a suitable environment in a quantity, across the entire universe and billions of years, to make one Lee peptide, and only chained together at random a single time, then the probability of natural biogenesis via the Lee peptide would be, simply, 1 in 10^41. And of natural biogenesis generally, it would be that probability, P(Lee), + P(any other PNA origin) + P(any TNA origin) + P(any ANA origin) + P(any RNA origin) + P(any other origin), which is logically necessarily a sum more probable than P(Lee) alone, and vastly more than P(God). Creationist William Dembski has shown that events of an improbability as high as 1 in 10^150 can be expected to occur by accident in the known universe (that’s how huge and old and full of stuff this universe is; which is itself evidence that just such a natural biogenesis is what happened). This means events as unlikely as 1 in 10^41 will have occurred by random accident in this universe over 10^109 times. That’s a one followed by one hundred and nine zeroes. Yeah. That’s what we call effectively 100% certain to have occurred.

So we’re done, really. I could drop the mic here. Even if a tiny collection of peptides came together in a suitable environ only once ever, across all trillions of galaxies after billions of years, and chained only once, the probability that something this unlikely has happened somewhere in the universe by now is essentially 100%. Something of the like is guaranteed to have happened. Why, then, do we need any other explanation? But this isn’t the going argument for natural biogenesis. Because this is still assuming we’re quite lucky, that of all the events that improbable (many of which will not be this event, and thus won’t have originated life), one of them happened to be this one (thereby originating life). There is already enough evidence to support the conclusion that that happened (it can readily happen by chance accident in a universe this large and old, and so the fact that we observe ourselves to be in just such an unusually large and old universe is evidence we are indeed that chance accident). But when scientists calculate the probability of biogenesis, they are making a far stronger point: that, on existing evidence, we can be confident that the natural occurrence of PNA or RNA components is probably so large as to alone render an event as improbable as 1 in 10^41 to have occurred many times, and thus we’re not even lucky. Life is an inevitable product of this universe (albeit rare, hence Rare Earth).

To illustrate the distinction, think of poker. While the odds of drawing a royal flush if poker is only ever played one time in the entirety of history may be remarkably low, it will still remain well within the range of random accidents that can be expected to occur, even just on Earth, much less across all known space and time. With six billion people living lives of over half a century apiece, events in people’s lives the odds against which are nearly three million to one will happen thousands of times across the planet, even if each one of them is an entirely unique event. Therefore, such an event would be impressively lucky, but would not signify any intelligent design had it happened. Whereas, the odds of drawing a royal flush once across ten million games of poker is ~100% all on its own and is therefore not even remarkable. So if we observe a royal flush appearing and ten million games of poker played, we definitely need no further explanation but natural chance. That observation is simply ~100% guaranteed. So…do we have evidence that there were enough “games played” for life to be like that? Yes.

Totani conservatively estimated on existing evidence that we can expect there to have been formed in the early history of the Earth at least 10^25 nucleotides in the requisite environs (which we know, on a basis of evidence, can happen: Biscans 2018; Cafferty et al. 2016), which is one thousandth the number of nucleobases, 10^27, that we can already expect there were (a fact of which we are even more certain). He’s thus assuming a very rare formation of nucleotides even in the presence of abundant nucleobases. And this is over the whole Earth over the course of hundreds of millions of years (and then, iterated across the entire universe and billions of years). From this he estimates, counting all comparable planets and moons with suitable environs abundant evidence entails for the known universe, the probability of a self-replicating chain of RNA (which we know exist: Tews & Meyers 2017, Robertson & Joyce 2014, Lincoln et al. 2009, etc.) arising in those environs somewhere in the universe. In actual fact it’s more likely RNA is an evolved structure, and that life originated with a simpler molecule like PNA, which is more robust and far easier to assemble in nature (Singhal et al. 2014). Simple self-replicating PNA strands are an established fact (the Lee peptide being just one of them; we now have the Plöger-Kiedrowski peptide, the Singhal-Nielsen peptide, and so on). Indeed PNA automatically assembles in the expected conditions (Leman et al., “Dynamic Chemical Assembly of Peptide Nucleic Acids” in XVIIIth International Conference on Origin of Life 2017; and Rodriguez et al., “Nitrogen Heterocycles in Miller-Urey Spark-Discharge Mixtures: Using Chemical Trends to Elucidate Plausible pre-RNAs on the Early Earth”; and now Liu et al. 2020, Frenkel-Pinter et al. 2020 and Scognamiglio et al. 2021). So the probability of natural biogenesis is very much higher than even Totani calculates. All based on evidence.

Mathematically, all one needs for the probability of even just a Lee petide biogenesis happening even just once in the known universe is for there to have been 10^43 PNA molecules across the cosmos mixing randomly in the right environments. Not all in one place, but across all space and time. Scattered experimental pools of chemical reactions can span trillions of galaxies and billions of years; the outcome probability remains the same. Just as lotteries are routinely won, even though millions of people are playing and almost all of them lose. A factor of 10^43 is 100-times more than the 10^41 odds against a spontaneous Lee peptide formation, thus accounting for the dozens of PNA molecules needed to form that specific chain (which is less than forty monomers long) while ensuring a near 100% occurrence (in the same way millions of poker games ensure someone somewhere will draw a royal flush). If we scattered this requisite material among environs spanning one trillion galaxies housing just one million suitable environs apiece (despite hundreds of billions of moons and planets per galaxy) and over only one billion years (there actually have been many billions of years), and assume only one chance assembly per year (an enormous under-estimate of the way chemical reactions work), the average supply of PNA we’d need to find per select environ is 10^16, vastly less than Totani’s estimate of 10^25 nucleotides, despite nucleotides necessarily being much rarer than nucleopeptides. That’s a difference of 10^9, or one billion. In other words, we can assume Totani overcounted by at least a billion times, and still get an inevitable Lee peptide, in a suitable environ for its forming and evolving, somewhere in the universe. And evidence establishes such environs unquestionably existed on the early Earth (an unlikely coincidence on any other theory but this, or others like it). And that’s just one self-replicating PNA molecule; we know there are others, as well as countless self-replicating RNA molecules; and likely even other molecular bases for life entirely. And we also know, on a basis of abundant evidence, that more material and environs are likely to exist than these conservative estimates have it. So the actual probability will be many orders of magnitude higher still. We have no such evidence for “god did it.”

Understanding the Mathematics

Q attempted to take me to task for saying Totani’s “only [Class] 5 error is ignoring pre-RNA worlds and over-estimating even the minimum RNA size based on some faulty logic.” A Class 5 error, I established under peer review, is “Begging the size of the protobiont,” as in, “not deriving a sound evidence-based estimate for how small (i.e. how structurally simple) a self-replicating molecule can be.” Q insists “Totani’s article is very speculative and rife with uncertainties, as many theoretical works are” and “there’s nothing wrong with that.” But I didn’t say there was. To the contrary, I support Totani’s article and work and results. My critique was actually reinforcing his conclusions by showing how his inputs are too conservative. Q seems not to understand this. The only problem I find in Totani’s study is that his results are calculated for a larger (and thus less probable) self-replicator than we already know exists. Know exists; not speculate exists. If we were to redo Totani’s analysis using the actual empirical evidence we have of the simplest (and thus most probable) protobiont, his calculation would indicate a far higher probability of biogenesis than even he estimates. And this is a fact. Not a speculation.

Q wants to disingenuously defend Totani against that point with the fallacious argument that “you can’t fault research for being no more or less than what it purports to be,” but actually, you can. If a study purports to determine the probability of biogenesis, and leaves out (even by admission) facts affecting that probability, that is a fault—and indeed, one that must be pointed out to anyone who does want to know what the probability of biogenesis is. I then reincorporate what Totani left out and show the effect of that on his conclusion: it increases the probability of biogenesis. His conclusion is therefore even more evidentially supported a fortiori. Q seems instead to have mistaken me saying Totani “speculated” that the smallest self-replicating RNA molecule “could not be smaller” than 40-60 nucleotides, as my saying Totani speculated that the smallest self-replicating RNA “could be as small as” 40-60 nucleotides. To the contrary, what I said was “speculating” was that that was the minimum; i.e. that there can’t be smaller ones. That there can be self-replicating molecules as simple as 40-60 nucleotides was actually empirically demonstrated in the papers Totani cites (Szostak 1993; Robertson & Joyce 2012); they show that countless functional RNA sequences exist in that range and that it is statistically inevitable many will be recursive (once RNA is performing functional operations, that’s just one kind of basic molecular function among them; those cited studies even discuss empirically the structures required). Rather, Totani’s “mistake” is in assuming the studies he cites prove the minimum size of self-replicating molecules. In fact they only prove that there can be self-replicating molecules that small; not that there aren’t smaller ones. The Lee peptide already empirically proves there are, and if that’s the case for PNA, it’s more likely the case for RNA, too, because it has a more reliable and efficient functionality (that’s why all life is now based on using it). Thus evidence establishes Totani’s estimated probability is far too low. Q is therefore missing the evidence for the trees here: we actually have evidence of smaller self-replicators—not a speculation; fact. Ergo Totani is grossly under-estimating the probability of natural biogenesis.

Another error Totani makes having the same effect relates to a different problem. Totani mentions that we have already discovered actual self-replicating RNA sequences between 100 and 150 nucleotides (Wachowiusa & Holliger 2019; Horning & Joyce 2016); but these have been shown to self-assemble in natural conducive environments from smaller sequences. So they did not have to arrive purely at random as Totani assumes; a common mistake I identified under peer review as a Class VII error, and discuss in my analysis of Totani: evidence (not speculation) proves that some aspects of the first protobiont don’t require purely random assembly, but assemble automatically, with a probability in the described environments of effectively 100%. No one has yet run the complete calculation to determine their total probability of random assembly; and these were single molecules essentially picked at random; which means countless other self-replicating RNA molecules must exist in the same range, so even their probability of spontaneous assembly is not the probability that any such molecule will thus assemble. The latter probability is far higher; a mathematical point Q never seems to grasp. It’s that same math he keeps stumbling on: the probability of a lottery being won is not the probability that one specific person will win, but the probability any person will win, which approaches 100% as the number of players increases. And evidence establishes the number of players here (the number of possible “winning numbers” in the lottery) is high.

Even if we were to estimate the probability of a single specific one, we still end up in the realm of accessible probability. For example, Wachowiusa & Holliger 2019 took a self-replicator of 150 nucleotides and broke it up into shorter strands of 20 or 30 nucleotides each and showed a self-replicating 150 nucleotide strand forms inevitably from them in the right context at a measurable rate (countless copies spontaneously formed). Across all volumes in the universe of randomly chaining nucleotides, there will form 10^120 more chains of 30 nucleotide length than of 150 nucleotide length. That’s a one followed by one hundred and twenty zeroes. The probability that several such chains of the right sequence will form in the same volume and then link together, by that inevitable process, into a 150-nucleotide self-replicator is vastly greater than the probability of a spontaneous random sequence of 150 nucleotides forming by itself. As the probability of their so chaining once available is effectively 100% in the specified environ, the total probability of such an assembly approaches that of the spontaneous assembly of molecules 40-60 nucleotides long. Which is likely one of the reasons Totani conservatively chose that as his range. But as easy as this is to hit upon in a universe so vastly old and large as ours, randomly hitting upon something like the Lee peptide will still be vastly more frequent. So we have no need of Totani’s reliance on RNA self-replicators. We already know life is many orders of magnitude more likely to arise from PNA, because PNA self-replicators will have arisen in this universe many orders of magnitude more frequently. Q does not understand the mathematical implications of this; or how it is based on actual evidence, not “assumptions.”

Q misunderstood several other things, such as assuming the structural shape of a PNA or RNA molecule must be determined separately from the sequencing of nucleotides; in fact the sequencing of nucleotides causes the resulting structure and shape. Ergo, we only have to calculate the odds of the sequence. Which is why Totani only bothers with that. Q also did not understand how mathematics entails the conclusion: in attempting to deny PNA self-replicators will be the more likely originators of life on worlds across the cosmos, Q insisted “experiments simply demonstrate that PNA-based replicators as short as 32aa can be produced; they don’t prove (and can’t prove) that that’s what actually happened historically.” But they don’t have to. Because that evidence already proves the mathematical conclusion that it will be vastly more frequent. This is not an assumption. It is a logical entailment from the evidence established. There is actually no way to avoid it: it is an empirical fact that PNA replicases will spontaneously arise on worlds with vastly greater frequency than RNA replicases. Therefore, it is an empirical fact that it is always more probable that life on any moon or planet arose in a PNA-first scenario; ergo it is more probable that’s how life originated on Earth. Yes, “maybe” Earth is one of the comparably rare worlds where it did begin with a far-less-commonly-formed RNA self-replicator. Sure. But it’s very unlikely, given what we empirically know about PNA replicases. And in any event, one cannot argue biogenesis is improbable “because” RNA replicases are improbable, when we know for a fact PNA replicases exist. So Q can’t get to his desired conclusion here no matter which way he turns. He simply doesn’t understand the conseqences of known facts.

There are actually more reasons than that to expect a PNA biogenesis: as I explained in the article Q was commenting on, PNA is “an even simpler and stronger peptide” and thus is for that reason also a more probable outcome of random chemical activity, and being more robust in harsh environments it can form and survive in more environments and therefore there are far more opportunities for a PNA biogenesis. We can therefore expect, by empirical fact, that far more biomes in the universe will have arisen from a PNA biogenesis than an RNA biogenesis. And since RNA can evolve from PNA—the sequence of steps is calculable and well within the probability range for a hundred million years of molecular evolution—we actually don’t need RNA-first theories in the first place. RNA is already an explicable outcome of PNA in applicable natural environments; and PNA protobionts are already known for a fact to be a far more expected occurrence in nature. But regardless, by the laws of probability, that biogenesis on Earth was by “either a PNA or RNA protobiont” remains on present evidence ~100%.

It therefore does not even matter to what creationists are trying to argue that, factually, P(PNA) > P(RNA). What matters is the sum total probability: P(PNA) + P(RNA), plus the probability of every other pathway known to be possible on current physics. So we don’t need to know “what historically happened” here on Earth. All we need to know is what most probably did. And evidence firmly establishes that’s a natural biogenesis, by either a PNA-first or RNA-first process; and indeed, by any available process for either. Because we know the options were quite abundant. Because factual evidence establishes numerous available physical pathways and environs on the early Earth for forming PNA or RNA. Yes, we don’t know which one happened. But we don’t need to. Evidence already establishes many suitable pathways and environs existed on the early Earth, and that the necessary initiating event is essentially 100% likely to happen somewhere in the known universe; and like any lottery, whoever the winner is is as likely to have been the winner as anyone else playing. So there is nothing more to explain as to why it happened here. The specific how of it we will continue to investigate; but the probability of it being some other unscientific, unevidenced mechanism is effectively zero. So as explanations go, theism simply isn’t even in the running here. All the evidence we have already establishes it had nothing to do with it, to a certainty exceeding all else.

Understanding the Science

Q attempted to claim Totani “neglected” many variables in his math that should lower his final odds. But he did not. That’s why his “math” passed peer review in a science journal. Totani takes into account, for example, the rarity of naturally-forming nucleotides (as opposed to nucleobases, which have been proven wildly abundant in nature). Just as he takes into account the rarity of conducive environments (even on the early Earth; likewise in the whole universe), not hostile or destructive ones; the probability specifically of stable self-replicators, not unstable ones; and the need of replicases to still multiply even in noisy, destructive environments (thus limiting the viable replicases and environs for him to count). The papers Totani cites show reproduction can exceed destruction in quite a large range of environments with numerous different replicases. Even just a 0.5% net gain per unit-time (in this case, a matter of mere seconds or minutes) is all that is needed for evolution by natural selection to go on to generate an entire tree of life over the ensuing millions and billions of years. That’s how powerful evolution is as a process. Replicases that can’t achieve that, environs that won’t allow that, Totani is not counting. Thus his math is unaffected by them. That’s the whole point of his study.

Other factors simply aren’t relevant, which is why Totani didn’t bother with them. For example, Q insisted that biogenesis “requires a reducing atmosphere which didn’t exist on the early earth,” but everyone in protobiology knows that that simply isn’t relevantly true anymore. It’s only an ignorant talking point you’ll find on creationist websites now. People should not be so easily duped. They should make an effort to understand the actual science instead. The particular kind of anoxic environments needed are in fact abundant even on present-day Earth, much more so before our biome generated most of the oxygen now present. One need only account for the anoxic volumes so-enclosed (how many, how large, how much stuff is in them), and Totani’s estimate for that, even for the early Earth, is conservative. He thus in no way assumes the entire Earth was anoxic. To the contrary, he assumes very little of it was. As do all current theories of biogenesis (e.g. clay-surface, ocean-vent, zinc sulfide, etc.). And Totani is indeed being conservative here. Disputes actually continue as to whether if or for how long a reducing atmosphere enveloped Earth. For example, a study published in 2011 only confirmed it wasn’t reducing as of 500 million years after Earth formed—not whether it hadn’t been hundreds of millions of years earlier; whereas there remains some evidence the earlier Earth was in fact anoxic for its first two hundred million years (example; example; example), and life may have originated in that earlier window (in fact, probably did). And this would greatly increase the probability of biogenesis that Totani calculated. But Totani conservatively assumes Earth’s early atmosphere was not reducing; the opposite of “not taking it into account.”

It’s also not as true anymore that reducing atmospheres are needed. They assist early life-formation. But the original Miller–Urey experiments assumed that amino acids had to be formed on Earth, which does require anoxic conditions. But we now know for an empirical fact that amino acids regularly form in, and arrived on Earth from outer space. Totani discusses this and cites a mere sample of the confirming studies; and he assumes this was the primary source of nucleobases available. In other words, none of Totani’s calculations rely on the atmospheric premise in Miller–Urey. Totani also mentions that suitable environments nevertheless would still exist on Earth, and just be fewer and thus less productive, so he does incorporate a conservative estimate of a terrestrial contribution to the nucleobase supply; but for this, he still does not assume a reducing atmosphere enveloped Earth. Once biogenesis has occurred, self-replication chemistry does not require a reducing atmosphere; hence several models of biogenesis, matching known scientific facts, omit that requirement. Indeed, all current leading theories do (see the summary by Gerald Bergtrom as well as the TalkOrigins summary and its associated section CB035 (and for arguments based on confusing oxygenated environs with non-reducing ones, CB035.1, CB035.2, and CB035.3). You can also peruse the TalkOrigins Abiogenesis FAQ.

Similarly, contrary to Q’s repetition of bogus creationist website rhetoric, Totani well knows, as do all protobiologists (including I’m sure his peer reviewers), “compartmentalization or encapsulation” is not a necessary component of the first life. Keeping the chemistry inside a cell or even just naturally occurring lipid nodules is certainly extremely helpful; but it is not required. It is now believed to be an evolved, not an original, feature of life. But even if we assume life began that way (and few now do), we already know such naturally-occuring places abundantly exist for life to form and grow in (which are now believed to be habitats that early life moved into and occupied rather than originated in). All the ability to repair and modify that enclosure is then just another avenue of an organisms’s evolved ability to control its environment (see The Role of Lipid Membranes in Life’s Origin and The Origin and Evolution of Cells and Uncovering the Genomic Origins of Life). So Totani has no need of accounting for this. No model he is calculating from requires life to originate in lipid enclosures, and lipid enclosures are so abundantly ocurring in naure that the probability of their presence to occupy is ~100%. So this does not have to be accounted for. Evidence already accounts for it.

Likewise, Q further reveals he does not understand the science when he complains that “what the empirical and experimental evidence indicates is that cells with minimal genomes of only a few hundred genes and proteins can exist.” These are all evolved features, not original features, of life. Genes, cells, protein metabolism: none of these exist in any biogenesis model. The self-replicating RNA and PNA we have empirically proved exist have none of these things. They do not possess genes. They do not metabolize proteins. They do not reside in cells. They do not even possess codon coding. So Q really does not understand here the difference between biogenesis and evolution. He clearly has not read any of the pertinent science on biogenesis. For if he had, he would not make this mistake. Confusing evolved for spontaneous features is a Class VI error. Hence you cannot understand biogenesis, if you keep confusing those two things. It is evident that Q, in ignorance, is just flailing around for any excuse he can come up with for rejecting the mathematically inevitable consequences of abundant empirical evidence. And that’s a broken epistemology on full display.

Another example of this “error mode” is when Q asks—as if no one has ever thought of this before—that “it is further difficult to explain how a single self-replicator that was so successful (in terms of fitness), would then be replaced, so as to leave no vestige of its prior existence.” This is about as scientifically illiterate in this context as one can get. All protobiologists know this requires no explanation at all, because it is literally self-explaining: the highly adapted powerhouses of DNA-based organisms vastly out-compete relatively frail naked RNA and PNA organisms. This is why RNA life only survives today in the form of viruses, which get by now by coopting DNA-based reproductive machinery: that machinery is so much more effective and advanced, it simply isn’t possible for any virus to out-compete another that learned to coopt that. Thus viruses that reproduced themselves simply went extinct; thriving in their place, are the viruses that learned to do one better. So if you want to know what happened to the entire original pre-DNA biome, quite simply, it was all eaten. It was outdone by the far more advanced organisms that evolved from it. This is why it no longer exists. This is also why PNA organisms no longer exist: they weren’t even capable of the adaptation RNA viruses exploited that kept them around, because they didn’t even have a physical structure capable of it. So their extinction was assured, once a biome of more evolved organisms filled every niche on Earth. Which cannot be stopped from happening in relatively short time. Every expert knows this. Which is why they no longer bother discussing it, confusing inattentive creationists, who never actually learn how evolution works.

Or how evidence works. Q insisted the claim that there is more than one possible self-replicating molecule is an assumption. But it’s not. It’s actually the other way around: that there is only one is an assumption; worse, an assumption now empirically refuted. The Lee peptide is not the only self-replicating molecule we’ve already discovered; and the evidence establishes it is essentially impossible that the few random examples we have found “are” the only ones, as there are countless self-replicators on Earth already, proving that countless ways to assemble one exist. It is therefore impossible that we have found them all, even in the shortest sequence range. We have already empirically proved countless functional RNA chains exist in the 40-60 nucleotide range, and RNA in that range can mutually catalyze into self-replicating chains of over 100 nucleotides. We could never explore every possible sequence in that set to test it for self-replication ability, but since we struck on the ones we do know by chance, statistically it is trillions to one against that what we’ve found are the only ones in that range, or indeed that the number there are is even small. It is empirically necessarily the case that there must be quite a lot more of them. And Totani’s estimate of how many is by his own admission absurdly conservative. And he didn’t even include known PNA replicases in his analysis, which prove there is an even larger range of possible self-replicators. So that there are far more than only one is an empirical fact, not an assumption.

Similarly, Q complains that “it is not just a simple matter of ‘rolling the dice’ enough times” because “that is not how mass action chemistry actually works.” But it is—as much as is needed for the math to hold. Totani’s calculation is accounting for the fact that the requisite chemical events, and the requisite chemical environments, are rare. The question is simply what the probability is of the right sequence of chance events, producing the needed chain of molecules, in the required chemical conditions. Totani’s math thus accounts for all the countless greater numbers of failures there inevitably will be, owing to the messiness of chemistry and its environs, and thus all the countless ways a biogenesis process can go wrong or be undermined. It’s just an empirical fact (to Q’s chagrin) that the conditions that will be instead conducive rarther than destructive to such random chaining are ubiquitous in the known universe, owing to the universe being so absurdly huge and various, and that only a protobiont with a minimal amount of advantage is required: just enough to allow reproduction to cause a population to grow—even by the slightest amount, and thus even if almost all reproductions fail or are destroyed. This is not assumption. It is all an empirical fact or a logically valid inference from empirical facts. As has been proved by all the scientists Totani cites, and by all the known science underlying his work and theirs (from physics to biochemistry), establishing how random assembly of monomers in the right environments leads to exactly this kind of “natural experimentation” in chaining sequences, even to quite long chains; and enough environs remain across the cosmos that would permit it to then grow by replication, and thus evolve, to ensure that somewhere, it did.

Factoring in Evolution by Natural Selection

You have to ask yourself: what are the odds life would indeed have originated in the only way it naturally can (God has no need of naturally chaining molecules, or of such vast scales of space, matter, and time to ensure enough random mixing of them), precisely where and when a natural biogenesis would be expected to occur (the various suitable environs known to exist on the prebiotic Earth)? Natural theories explain all these oddities (why start with a single naturally-occurring chemical, why take billions of years to build it into more, why have trillions of galaxies around and billions of years to burn?); while theism does not (without gerrymandering excuses for it, no evidence for which exists). There is simply no case left for “god did it.” Odds are, nature did it. All the evidence establishes no other conclusion. Once any self-replicating molecule exists in any suitable environment, mutation and natural selection will inevitably produce a biome of advancing speciation. A tree of life; inevitable outcome. Which is how all the “other” functions and features and capabilities of life come about: they are not a part of biogenesis, but of evolution by natural selection; a completely different phenomenon. And this is because the mathematics of evolution are inescapably much more favorable than of mere biogenesis. Evolution is by its very nature a probability sieve: it skips the need to arrive at complex structures “spontaneously” by arriving at them through stepwise advances, each of which is highly probable (in fact, given enough time, inevitable; which is why evolution accomplishes so much with such extraordinary amounts of time).

Like most Christians who have a hard time grasping the significance of vast spans of time, Q incorrectly thinks a hundred million years is too “short” an amount of time. It is not. When replications occur in mere seconds, a hundred million years is equivalent to a billion or more years in evolutionary time. For example: over the last half billion years on average mammals could double their number every 15 years or so (losses preventing that then drive evolution over that period); if they could have done that in 15 minutes, as protobionts more than can, then they would be 4 x 24 x 365 x 15 = 525,600 times more evolved in that same period, ergo one year for a protobiont is equivalent to 35,040 years in animal-scale evolution. A hundred million years is then equivalent to 100,000,000 x 35,040 = 3,504,000,000,000 or three trillion years of equivalent evolution. That’s the math. And that’s why scientists don’t need to explain why evolution from simple protobionts to our familiar single-celled life could occur within that window. That’s damn well plenty of time. And no “straightforward trajectories” are needed. Evolution by natural selection finds pathways readily no matter how meandering, for which scales of time this vast are ample, as I have demonstrated before: even with extremely conservative assumptions, the T4 bacteriophage is guaranteed to arise within a billion years of evolution; with more realistic assumptions, simple bacteria, such as even still exist, would take but mere millions of years—and the early Earth had over a hundred million to work with. This follows if we adopt something closer to an empirically confirmed rate of additive mutation and a small average population size of subcellular life-forms, which would get us one relevant progressive mutation per year (still absurdly conservative). In those conditions recognizable cellular life will populate the earth in less than half a million years after the first protobiont. Given the greater simplicity of life in PNA and RNA worlds over codon-based DNA worlds, the transition from each to the next can be expected to have occurred in similar timeframes. Ergo, a hundred million years is ample. In fact, that there was so much time, is evidence this was indeed nature taking its course. God just makes stuff; instantly. Only nature requires millions and millions of years. So that life required the latter to get to even just single-celled DNA-codon models, proves nature, not gods, produced life on Earth.

Hence the power of the evolutionary process is one of the reasons we now know it is what happened. It’s too incredible a coincidence that such a simple, blind, iterative process can produce such scales of complexity on such scales of time. Just as we see. It therefore must have done it. And all the evidence we have confirms that.

Conclusion

All I did in the article Q couldn’t handle was show that Totani’s estimated probability of natural biogenesis in the universe (which was still ~100%) is even still too low. Once we correct his Class VII error (his not considering nonrandom natural processes in contributing to the formation of protobionts; many of which are empirically verified) and his Class V error (his grossly over-estimating the smallest possible self-replicating molecule; smaller self-replicators are empirically verified), the probability of natural biogenesis can be shown to be enormously greater than even he finds it to be. So on a basis of evidence we get a conclusion even more certain than Totani’s that, as I said, “the universe is so unfathomably large that abiogenesis is inevitable even on Totani’s overly conservative assumptions.” That this conclusion follows with even more certainty when Totani’s oversights are corrected is what follows from actual empirical facts—not speculations or “mere assumptions.”

That’s why naturalism is not “in trouble” over biogenesis. Naturalism has already more successfully explained biogenesis and all its surrounding facts than any other worldview: it predicts and thus explains why life started that way (a single molecule of naturally-chaining chemicals), why it took so long to get to making even plants and animals much less people, why the universe has to be so large and old and deadly and full of wildly varying worlds, why the early Earth looked the way it did, and why its history, geo-climatically and biologically, proceeded in the particular way that it did. None of this is predicted by theism. All of it is predicted by naturalism. A naturalist theory of biogenesis is therefore far more probable on the evidence. And you can’t object to this result with a “but you don’t have any theories of how the first protobiont could have formed,” because we do; and not just theories, but theories abundantly based empirically on evidence—such as of the physical processes proposed and their expected outcomes and their availability on the early Earth. Theism comes nowhere near to this scale of empirical status. Not even by a thousand miles. So that leaves naturalism quite comfortable on present facts.

Throughout this conversation we find Q committing a lot of scientific errors—ignorance of facts, getting facts wrong—which, yes, does partly explain why he is trapped in his delusion. But mere ignorance and error would readily be corrected in the mind of anyone operating with an effective epistemology. Case in point, Q caught one error in my paper: I was wrong to say that “nowhere in [Totani’s] paper does he even mention, much less account for, PNA-first models of biogenesis.” Upon learning this, I immediately corrected the error. Totani does not explicitly mention such models, but he has one line inclusive of them, with an admission that he is leaving them out of account; my article is now worded accordingly. You won’t see Q do anything like this very often. But in that respect and many others, it is not mere ignorance and error, but his broken epistemology that keeps Q trapped in this hall of mirrors. He does not understand—and I would wager, owing to the resulting cognitive dissonance he’d encounter, does not want to understand—the difference between an actual assumption and an inference from evidence. He also does not understand degrees in the latter category; that some inferences from evidence are weak, some are strong, some are far stronger still. Nor does he understand how mathematical consequences follow from observed empirical facts. And thus, for example, he cannot understand that evidence proving something is possible can well be strong enough to be confident it’s likely, without having additional evidence confirming the result actually does or has happened; while conversely, lacking that same evidence for even the possibility of a process, like divine creation, much less its actuality, puts you in a vastly worse epistemic position, not a better one. It is the theist who is in trouble here. Not the naturalist.

This is the same error we observe when Q refuses to grasp or admit the difference between the expected features of a protobiont arrived at by random natural biogeneses, and the more elaborate and impressive features of subsequent life that then evolve by natural selection. Yet no one can understand biogenesis who does not grasp this distinction. Spontaneous biogenesis requires an extremely improbable initiating event; evolution by natural selection never does, but to the contrary, automatically builds incredible specified complexity in very short orders of time. So when it comes to any natural biogenesis, all we need to have occurred is the conjunction of two things: (a) the right chained sequence of chemicals producing any self-replicating molecule and (b) a suitable environment to allow its continued reproduction and thus evolution to take hold. The probability of (b) occurring in this cosmos has been scientifically proved with abundant evidence to be ~100%. So neither I nor Totani nor any naturalist has to account for it. That leaves only the probability of (a). Which we know, for all the reasons I lay out (which are references to evidenced facts) is substantially higher than Totani calculates. And that’s that. There are no assumptions here. Just empirical facts, and what logically necessarily follows from those empirical facts. Theism, by contrast, in this case has zero empirical facts to call upon in its defense.

Thus naturalism wins here, by an overwhelming preponderance of evidence. And that evidence is already so overwhelming that if you ever do get to look into the “black box” containing the answer here, what actually historically happened on Earth, whether it was “a natural or unnatural biogenesis,” we can already be quite confident that what we’ll see in that box is “natural.” Only a straight up fool would bet against that now.