On the Origin of Life and Falsifiability

Dr A. beat me to the point I was going to make, but I'll reiterate it in my own words anyway.

To me, the evidence you presented against the RNA World and the evidence you presented against Panspermia are analogous. Remarkably so, in fact.

In grade school, I remember being taught to use this format for analogies:

x:y = a:b

It's read "x is to y as a is to b."

(actually, I think the "=" might have been two colons "::", but it's close enough)

Now, insert the following values for the variables:

x = "RNA is inherently unstable"
y = RNA World Hypothesis

a = "basal prokaryotes are vulnerable to cosmic rays"
b = Panspermia

To me, this analogy holds perfectly true with these values inserted. The two evidences (x and a) address very similar questions about their respective proposed protobionts, don't they?

They both rely on a reasonable modern surrogate to examine the chemical/physiological shortcomings of their putative biotic progenitors.

Where they differ is in where their explanatory power comes from. I would argue that the phylogenetic falsification of the Panspermia hypothesis has more power from a historical perspective, but less power from a mechanistic perspective, than the biochemical falsification of the RNA World hypothesis.

The core of your argument is that only the historical approach grants the power of Popperian falsification. I would argue that you are just showing a little phylogenetic chauvinism.

Can you explain that a little more? Are you basically saying, if you have one phylogenetic tree, it's hard to make inferences about anything that doesn't fit inside that tree?

That's true, but since the "data analysis" part is the part Ringo identified as differing for science and ID, what exactly is the problem with that?

Since we're in the game of philosophical nitpicking, science doesn't start with theories: it starts with hypotheses (or propositions), and the iterative process of the scientific method is what eventually turns a subset of those hypotheses into theories, while discarding the remainder.

For creationism/ID, the process doesn't really begin with a hypothesis or proposition: it begins with an axiom, and the data is evaluated or rejected based on its conformity to that axiom. It's "conclusion first, even at the expense of data."

To bring this back toward the topic, one of the major failings of Intelligent Design and creationism is the falsifiability criterion: objections or divergences from the predictions of the creation hypothesis are only used as a basis for obscuring or interpreting away shortcomings in the hypothesis, and not for evaluating the accuracy of the overall Design concept, which is what makes it an "axiom."

Genomicus is proposing that several mainstream OoL hypotheses could fall into this same trap. The RNA World Hypothesis seems particularly popular these days, and Genomicus's objections to it are one way of trying to "keep the science honest," so to speak. This is part of the evaluation process: Genomicus is trying to make sure we only consider valid, testable propositions, and he's challenging our hypotheses on those grounds. Ultimately, I think he's just got some bits of the philosophy wrong, but he still deserves a lot of credit for showcasing how science is different from creationism: we consider and reconsider our propositions fastidiously, just as Ringo said.

I have three objections to your comment here:

1. I don't really understand why this evokes such incredulity from you. The validity of this data as a falsification hinges on the premise that there is no other means of tolerating cosmic radiation aside from certain patterns in the abundance/diversity of proteases, ABC transporters and nucelases. Accepting that premise requires a fair helping of hubris, doesn't it?

2. I don't really understand why you think LUCA and FUCA should be similar. After all, in the panspermia model, FUCA is supposed to have come from an alien planet, so it seems quite reasonable to anticipate a whole suite of very novel selection pressures immediately after arrival on Earth. So, there seems plenty of reason to anticipate massive genomic changes occurring between FUCA and LUCA; and it seems a bit hasty to dismiss any possibilities the way you have.

3. And besides, isn't this the exact form of the argument you used to demonstrate the unfalsifiability of the RNA World hypothesis? That is, what if someone published a phylogenetic tree claiming to provide evidence about the existence of self-catalytic RNA's coded in the genome of LUCA? Wouldn't that beg a couple of big questions about whether or not those RNA's were the actual prebiotic RNA catalysts, and about why/how those prebiotic catalysts became encoded into LUCA's DNA? How is that substantively different from the questions Dr A is asking now about Panspermia?

Edited by Blue Jay, : No reason given.

This is well outside my area of particular expertise, but one pattern really jumps out to me. See all the words above that I highlighted in yellow? Those words all describe active processes; as in, they only work when the cell is actively metabolizing. And every one of those mechanisms you describe includes these yellow-highlighted words. So, unless a microbe is actively metabolizing during its long space voyage, these are not mechanisms that would enhance an organism's survival during panspermia.

In fact, none of these mechanisms actually protects a microbe from ionizing radiation: rather, they repair a microbe after ionizing radiation. They would need some other means of actually surviving the radiation in order to get the chance to repair themselves afterwards (unless, again, they have some means of maintaining an active metabolism in deep space).

So, perhaps the real "trick" to panspermia isn't about being able to repair yourself: it's about having enough of yourself left to repair once you arrive. And, if it's possible to survive a panspermia voyage without activating any genomic mechanisms, wouldn't it be hypothetically possible to find a panspermia scenario that doesn't actually require any particular genomic characteristics at all?

---

There are a lot of weasel words in there. How different is "starkly different"? What degree of overlap is "not inconsiderable"? And, what does that imply for the mutability of a planet-jumping genome? You can't falsify a hypothesis by applying assumptions that are not strictly essential to the hypothesis. Any putative falsification must work equally well for any range of possible tolerances, or it isn't a true falsification.

I agree with this. But you've slipped into an argument about the plausibility of what I'm going to call the "LUCA~FUCA Hypothesis." As you said in Message 1, "...no amount of evidence for the plausibility of {a hypothesis} will be able to establish the historical accuracy of that hypothesis."

It is entirely possible that the genome of FUCA was radically different from the genome of LUCA, with different types of proteins serving the functions of ATP transport, breakdown of oxidizing agents, etc.; and that entire suites of FUCA's molecular machinery, optimized for function in an alien environment of unknown character, had been completely replaced in LUCA by entirely new machinery that operates more efficiently in Earth's environment.

Is that possibility plausible? I don't know, but your own conditions were that we don't incorporate evidence about plausibility, so I raised my objections and proposed alternative hypotheses.

It seems to me that any falsification attempt will inevitably require components addressing both the plausibility and the historicity of the proposed hypothesis. Much of the historicity is only available to us via extrapolation, no matter which route we take, and we will ultimately rely on arguments about plausibility to fill in those gaps.

Edited by Blue Jay, : No reason given.

Edited by Blue Jay, : No reason given.

Yeah, that's what I was trying to get at.

I also think Genomicus was too dismissive of the possibility that an icy comet may in fact be sufficient protection against ionizing radiation.

I was going to hold off on the details until I saw Genomicus's rebuttal, though: give him the chance to steer the conversation, since it's his topic.

I don't think those papers are as relevant as you think they are. Here's an excerpt from the methodology on the first paper:

quote:

---

All plant cultures were grown for 20 days in closed microfuge tubes in standard tissue culture growth chambers under illumination by cool white fluorescent lights with an incident light flux on the samples of 80 μEm−2 s−1 using 16-h light–8-h dark
photocycles. Algal cultures were grown in similar growth chambers in vials with a punctured cap that allowed air exchange. The algal cultures were contained in glass jars saturated with water vapor to allow full light exposure but prevent evaporation.

---

In a nutshell, they grew plants and algae on meteorite-based soil under Earth-like* conditions. I don't think that does very much to advance a panspermia argument.

*I had to google "μEm−2 s−1": it's micro-Einsteins per square meter per second. An Einstein is a mole of photons. So, using this conversion table, 80 μEm−2 s−1 for a cool white fluorescent light is about 6000 lux. Full daylight is 10,000 lux, and interior lighting is usually around 50-100 lux. They don't mention temperature, so I assume they used ambient temperatures, with this light as the only additional heat source. That's probably going to be a rather warm growth chamber.

I'm not very familiar with the panspermia literature, so I guess I can't argue with you. But, I was always under the (apparently unjustified) assumption that panspermia generally assumes microbial spores, rather than active organisms. Having the microbes active during transit really complicates the issue of evolutionary adaptation before, during and after transit, though.

Well, then I think you need recognize that your falsification would only apply to that specific type of panspermia, and not to the overall Panspermia Hypothesis as a whole. The main reason why I disagree with your claim that Panspermia is falsifiable is because your proposed falsification is much narrower in scope the overall concept. And, I believe that this will be true of any proposed falsification. There will always be possibilities for panspermia that are not easily falsifiable.

A comet would not be any more frozen than a meteorite or any other space object, so these same restrictions should apply to any panspermia model.

Unfortunately, there are a lot of anti-freeze proteins (AFPs) in life on Earth. Freeze tolerance is apparently possible with a very small suite of semi-specialized AFPs, and many different clades of organisms have independently evolved (and lost) novel types of AFP, so it's much harder to wield the power of phylogenetics against this one.

What about a meteorite at deep-space temperatures?

And I completely agree with you on this. My whole argument is that Panspermia is no different from the other models in this regard: there are always going to be a whole lot of messy contingencies that can't be easily dismissed, so we'll have to continue considering alternative panspermia models for a very long time after these phylogenetic have ostensibly "falsified" panspermia. To me, that means it isn't really a falsification at all.

I think I made a mistake somewhere. I latched onto ABC transporters as my example because they were one of the few proteins you mentioned that I actually knew a little about. But, they don't seem to be on the list of proteins Cavalier-Smith suggested were absent from LUCA, so I guess they aren't an example of what I thought. Sorry about that.

My general point is that you're trying to have it both ways: you're trying to claim that a gene can be both essential to an organism's survival, and possibly absent from LUCA.

If a certain genomic feature is legitimately essential, then all organisms should have it, and the LUCA~FUCA assumption is appropriate. But, such a genomic feature would be useless for your phylogenetic falsification, because it can't be missing from any organism.

On the other hand, if the genomic feature is not essential, then it is useful for your falsification, because it could hypothetically be lacking in some organisms. But the LUCA~FUCA assumption is no longer appropriate, because such a genomic feature could believably be lost between FUCA and LUCA.