Spacek vs. Benner: Thanksgiving showdown about life on Mars

Cite as: Spacek, J., Benner, S. A. (2023) “Spacek vs. Benner: Thanksgiving showdown about life on Mars”. Primordial Scoop, e20231123. https://doi.org/10.52400/ZSJS2155

JAN: Steven generated the BARSOOM (bacterial autotrophs respiring with oxygen stored for use in their overnight metabolism) hypothesis (1, 2, 3) using already observed Viking data. In my rebuttal post I tried to argue why such a post hoc hypothesis using alien biology as an explanation of the observed results is problematic. In fact, I generated a few more post hoc hypotheses, a.k.a. stories, while fishing yesterday (e.g. O₂ generation might be Martian distress signal analogous to cis-3-hexenal – generated by freshly cut grass). It is easy to do so using ad hoc alien biology, while disregarding Earth’s biology. Certainly, other astrobiologists already generated uncountable post hoc explanations for all of the observed Viking results. They were merely not bothered to publish them.

The problem is not that Steven’s hypothesis is not possible, the problem is that it might be true. Without any additional data we cannot estimate how probable Steven’s hypothesis is. That’s why I concluded my previous post by saying “An explanation without an experiment is doomed to remain merely a story.”

Zach Weinersmith from SMBC comics read this discussion, got inspired by it, returned to 2013, made this comic strip and then covered up all his tracks. This figure caption also might be true.

STEVEN: Jan has deployed the “But it is a Just So Story!” line to criticize the BARSOOM model for life on Mars. That model interprets the Viking 1976 results as the expected outcomes of Martian near-surface autotrophs that fix carbon in daytime, perhaps photosynthetically, and respire at night using stored O₂.

For those unfamiliar with the “just so” epithet, it comes from Rudyard Kipling, who wrote a book with this title to “explain” things to children. As an example, the camel got its hump as punishment for not working hard enough when the world began.

Jan dismisses as “merely a story” the BARSOOM model because he thinks that it cannot be immediately subjected to direct experimental test. However, as I explained in my book, Life, the Universe, and the Scientific Method, the most interesting questions are exactly those. Jan’s dismissal would seem to limit science to only 8^th grade science fair topics.

As Life explains, useful models are transparently different from “Just So Stories” because they appeal to universal rules. They do not rely on “one time only” contingent events, like those that afflicted Kipling’s camel.

And so it is for BARSOOM. Carbon fixation, O₂ storage, respiration, and CO₂ pre-organization are all known in Terran biology when needed for fitness in the relevant environment. Evolution is expected to evolve analogous processes whenever the environment demands. I have suggested that the Viking environments demand them.

Further, only 8^th grade science fairs seek to “estimate how probable” a hypothesis is. As Life details, real science is a back-and-forth where models are set opposite contrasting models, with data used to assign burdens of proof as they move back and forth.

For Mars, both sides have failed to follow this process. This failure is reflected throughout NASA policy, including what Terran astrobiology NASA funds and what missions it launches.

Thus, those arguing that Viking found life never completed a model that might shift the burden away from them. They did not, for example, suggest a biological oxidant for the label release experiment. They did not propose a systems biology model explaining why Viking microbes might be prepared to respire food that Viking dropped on them. 

And, if Jan is correct that “astrobiologists [have] already generated uncountable post hoc explanations for all of the observed Viking results”, they have committed scientific malpractice by not “bother[ing] to publish them.” Had they done so, they may have arrived at the BARSOOM model decades ago. Perhaps they were spooked by the misinterpretation of the GC-MS data.

The BARSOOM model is important because it shifts the burden of proof to those who wish to argue that data from Mars indicate that Viking sites were sterile. Importantly, experiments are immediately possible to develop this non-life side of the argument. These too must appeal to universals. But since no Martian autotroph is part of their “story”, those experiments can begin today.

JAN:
1) BARSOOM hypothesis passed a test against data it was generated on but failed a test against Terran life.

2) Steven did not suggest a way to test/falsify the BARSOOM hypothesis.

3) Steven said that BARSOOM shifts the burden of proof away from him.

From this loaded question “How might clever evolution adapt bio-systems to store “O₂” in a way most fit for the Viking Mars environments?”, we can see that Steven assumes that Martians follow Bennerian biology: unlike microbes living under anaerobic conditions on Earth, Martians store O₂. There goes Steven’s “appeal to universal rules” presented above.

A more plausible, yet still unlikely, explanation of the observed O₂ release is that it comes from dying Martians as they release perchlorate reductase and chlorite dismutase (analogs) from their cells.

My position is that to this day no presented hypothesis satisfactorily explains Viking results. If anything, the hypothesis privately disclosed to Steven and me by Chris McKay explaining the Viking results abiotically is more plausible than Steven’s (although it would benefit from a quantitative work).

BARSOOM, just like Barsoom, is fiction.

The burden of proof is still on Steven.

STEVEN: Let us review the basics of process in science, covered in Life, the Universe, and the Scientific Method.

Models are constructed to “explain” all data available at the moment. Epistemologists write chapters to relate “explanation” to prediction, but in this context, an observation is “explained” when the explainee says “Ah! I see! That observation is exactly what I would have predicted.” 

The reaction may be stronger if the explanation is consistent with generally accepted principles unrelated to the observation at hand. It may be still stronger if its pieces are exemplified, here, in some known biological systems.

BARSOOM does all of this. It explains the Viking results. It is consistent with the presence of organics found later at other Martian sites. It is consistent with chemical and thermodynamic principles. 

And Terran biology holds examples of its pieces. Humans themselves store O₂ in readily releasable forms (in hemoglobin). Respiration by biocatalysis is common. C4 plants concentrate CO₂ to manage their ecology. 

Jan objects because the Viking data were known before the BARSOOM model was constructed. He calls the BARSOOM model “post hoc”, as an epithet. Even though that is what all models are. 

Of course, we all might have preferred had BARSOOM been constructed before Viking flew. Then, Jan would have seen the Viking results as BARSOOM passing “tests”. Predictions, not “retrodictions”. But the chronology does not work. I was not scientifically active in 1976. Jan was not yet born. We do what we can do.

So why do I suggest that the BARSOOM model shifts the burden of proof from those who argue that Viking encountered a Martian biosphere to those who argue that the Viking sites were sterile? This is more complicated, since where such burdens are placed depends on the culture of those placing them. 

Sara Seager, William Bains, and I wrote a piece a decade ago analyzing this in a wonderful case where the arsenic DNA was proposed as the genetic molecule in microbes living in environments containing high arsenic levels. Biologists and chemists placed the burden on those proposing arsenic DNA. Geologists and physicists placed the burden on those who doubted arsenic DNA. The scientists from all four disciplines were reacting reasonably, given their culture. Only because geologists were the peers who refereed the Science paper was it published with the evidence that it presented.

In any culture, in the abstract, if a model explains all of the available data, is consistent with first principles, and has its pieces exemplified in our experience, it is no longer susceptible to the complaint: “But you now need to test it”. It has met the burden. Other models must now be developed until they meet these standards, and then some. In courts of law, this is known as the “preponderance of evidence” criterion. Then the burden shifts back.

Possible tests are easily described for BARSOOM. But most of them require going to Mars. And so they are not readily done. But that is not important, since the burden has shifted to models that have not explained all available data consistent with first principles.

Now, humans being human, nothing is more difficult than accepting such a shift. Confirmation bias, and so on. Jan at several points reveals how his past determines what he accepts as “plausible”. 

Jan even writes “My position is that to this day no presented hypothesis satisfactorily explains Viking results”. Even when presented with the BARSOOM model, which does. Here, the trigger word is “satisfactorily”. What is not satisfactory to Jan? Well, he does not say.

We train scientists to have the intellectual discipline to understand how their past blinds their perspective. One way to inculcate this discipline is to have them play “games”. Thus, in Life, we challenge the student to react appropriately to the claim that water is H₃O, not H₂O. 

The correct reaction is not to dismiss the claim as uneducated. That reaction would cause us to overlook every major advance in science. Rather, the student must be disciplined enough to react: “How interesting”. And then to ask: “What else among things that I think are true must be false if water really were H₃O, not H₂O?

Jan has not yet achieved this Zen level of intellectual discipline. He argues that because he knows of no Terran life that has the BARSOOM life style in toto, the BARSOOM model still must bear a burden of proof. Indeed, he argues that this constitutes the failure of BARSOOM to pass a test on Earth. He adumbrates a model where the Viking O₂ release came from perchlorate reductase and chlorite dismutase activities from organisms being murdered by Viking.

No problem by me. Elaborate the perchlorate reductase model to the point where it explains all of the observations on Mars in a way consistent with general principles. Then it too would not bear a burden of proof. 

Of course, Jan’s perchlorate model is also a biological model. BARSOOM fills a hole in the culture created by the absence of any previously proposed model to explain all of the Viking results in terms of a Martian systems biology. And so the burden is not shifted to another biological model, no matter how interesting they would be if developed. Rather, it should shift the culture to examine non-biological explanations of the Viking results.

I can conceive of such non-biological models. Some of them strike me as potentially compelling. For example, irradiation over millions of years of the soil might generate kinetically strong oxidants that do not dissipate, even over millions of years, but do emerge when the soil is wetted.

What is nice now is that the burden can be met without traveling to Mars.

JAN: From Steven’s reaction and from our private conversation, it seems that up to this point, I failed to fully express and convey my ideas. Let me try to tackle the cultural and language barrier again to attempt to explain my position and then what I learned from this discussion so far.

The reason why I am dissatisfied with Steven’s BARSOOM explanation comes in two different flavors. (1) I find post hoc explanations based on alien biology unsatisfactorily in general and (2) I do not like Steven’s BARSOOM explanation in particular.

The general dissatisfaction arises from the notion that “extraordinary claims require extraordinary evidence”. Living Martians are culturally perceived as an extraordinary claim par excellence, yet evidence we can generate to support it will remain weak in the foreseeable future. Viking missions were the only experiments to seek extant life on Mars and therefore the only evidence we can bring up is post hoc evidence that is culturally and statistically perceived as weak.

The particular dissatisfaction with the BARSOOM explanation stems from my intuition based on my understanding of extremophile life on Earth. Steven dismissed this point because an intuition is not a valid test. I must concede this point to Steven. Intuition is not a test, and BARSOOM cannot be tested against life as we know it. But sometimes, when dealing with complex problems, intuition is the only thing to guide us.

Chemistry is like a game of chess, where strength of any position can be precisely calculated. Biology is more like a game of go, where strength of a position is judged by intuition gained through playing the game.

Hence the chemist will dismiss biologist intuition: Just because no anaerobic life on Earth generates and stores oxygen, it does not mean that Martians would not do that. And just because arctic soil bacteria prefer trace gas metabolism over photosynthesis to generate its biomass, it does not mean that Martians would not have the opposite preference (even though they have to contend with harsh UV and less sunlight). And just because most of the species of photosynthesizing anaerobic organisms on Earth are anoxygenic (not producing oxygen), that does not mean that Martians would be in this group. (see Evolution is cleverer than you are for more examples)

…and still biologists will dismiss chemists’ dismissal, because we learned to trust our intuitions.

Despite these differences in language and culture, we both try to advocate that we should seek life on Mars. And here lies our other contention.

For me the best position to convince the world that we should seek life on Mars is to point towards the intuition and analogy: The more we understand about extremophile life forms on Earth, the more we are convinced that there might be some kind of life on Mars, even though we are unlikely to guess which kind of life that would be.

The second best point to make is to say: Despite 5 decades of trying to do so, humanity failed to fully explain Viking results using abiotic chemistry.

These two arguments (analogy with life as we know it, and absence of explanation of the Viking results) are sufficient to generate excitement to seek life on Mars.

On the other hand, Steven believes that the best move to make is to generate a specific model of Martian metabolism, regardless how biologists feel about it.

Trying to guess the exact Martian metabolism is not just unlikely to be correct (based on my intuition), but also detrimental to our effort to convince the world to support the search for life on Mars before we send humans there. When the world is dissatisfied with the model Steven generated, Steven will be perceived as unreliable and so will be our call to search for life on Mars.

Now for what I learned from this discussion so far: I believe that we should be able to disregard the notion of “extraordinary claims require extraordinary evidence,” when generating hypotheses, just like Steven does. This notion clearly stifles creativity. But this can be done only if everyone, from peers to funding agencies or private mission funders, can see the difference between brainstorming done through hypothesis generation and claims that actually require extraordinary evidence.

From NfOLD astrobiology meetings I learned that we should not even publicly speculate about alien life, because being wrong can be harmful.

Hence there are two battles to fight: (1) to advocate for change in the culture – it is permissible to generate hypotheses about alien life and (2) we should seek alien life on Mars.

The question is which battle you want to fight first. Conducting one is detrimental to the other and I do not believe there is enough time to win both before we have people on Mars.

STEVEN: A fascinating discussion. Jan, at last, “gets it”.

The question is: Can those steeped in NASA culture also “get it”? Or, if not, can private individuals be induced to place contrarian bets in the “Mars Life Finding” business?

First, let us lay aside a linguistic issue. All explanations are “post hoc”. That is what distinguishes explanations from predictions. And, as reviewed in Life, the Universe, and the Scientific Method, a century of discussion has explored the relation between prediction and explanation.

As also discussed in Life, explanations are demanded only when observations differ from intuitions. Carol Cleland speaks of these as “anomalies”. And Carol sees anomalies as events that drive discovery and paradigm change in science.

Or, at least, they should.

However, as discussed in Life, science does not work “as it should”. This is, of course, the conclusion of many historians of science, notably Thomas Kuhn.

In general, when observations come along that contradict “intuitions”, people do not change their intuitions. Rather, they deny the observation. This behavior is part of the human propensity for “confirmation bias”.

Kuhn and others have long struggled to understand how science advances despite this. Kuhn argued in his Structure of Scientific Revolutions that if enough anomalies accumulate, scientists will eventually question their intuitions. Others have argued, perhaps more realistically, that science advances one funeral at a time.

Examples are numerous. 

• Ernst Mayr, the doyen of evolution speaking from Harvard, refused for years to accept Woese’s interpretation of ribosomal RNA as indicating a third kingdom, the Archaea. 
• Geologists’ “intuition” refused to believe that solid rock continents can move above the Earth’s mantle. After all, “what force could move continents?”
• The chemiosmotic theory for ATP synthesis was developed by Peter Mitchell only because he was rich enough to fund his own research. Even his Nobel Prize was insufficient to pay him back.

One can compliment Jan’s forthrightness here. He admits that he does not like the BARSOOM model because … well … he just does not like it. He admits that he has “learned to trust [his] intuitions”, fully aware of one consequence of that trust: He must stand aside whenever science advances past his intuition.

However, Jan understands the pragmatic problems for those of us who want to seek life on Mars before our funerals.

• The search requires enormous resources. Even someone as wealthy as Peter Mitchell could not swing it. 
• Most people make decisions based on how they “perceive” things, a word that Jan uses three times, each time appropriately. These are those goddamn “intuitions”.
• To get a NASA committees to agree to commit those resources will be a “hard slog”.

Nothing illustrates the last point better than the 2022 “Community Report from the Biosignature Standards of Evidence Workshop”. That Workshop had 82 attendees carefully selected from 355 applicants; it was no collection of rabble. On the contrary, the Report represents the “important” people in the NASA community, the people who peer review grant applications, decide what research is funded (and what is not), and advise NASA on what missions to fly.

Indeed, NASA considered the Workshop Report so important that NASA brought it to the National Academy of Sciences for its endorsement.

The Academy declined, Instead, it added to, and in many respects mirrored, criticisms of the Workshop Report made on these pages by biologists, geologists, and philosophers.

The Academy began by cautioning that “consensus standards” might “become a barrier to entry into the field as a form of gatekeeper“. It noted that NASA was seeking “[a] formal process of additional internal vetting … by committee [italics added]” that is “antithetical to the established peer-review process. It is also likely to stifle innovation.”

It did not help that the Report suggested that its committee “to coordinate communication and information exchange” would be supported by “intergovernmental organization, e.g. the United Nations.” Nor did it help that the Workshop “canceled” one of its 82 members who argued too strongly against the “framework for discussion” presented by the Workshop leads.

Entrepreneurs like to be contrarian, to go against intuitions. That involves risk, because most of the time, intuitions are right.

However, every now and then, an opportunity comes along where the “consensus” is so bad that a contrarian wager seems to be a sure bet. The archetypal example from recent history was the 2008 financial meltdown. As noted in The Big Short, it was possible for “a few outsiders and weirdos” to see “the giant lie at the heart of the economy … by doing something the rest of the suckers never thought to do. They looked.”

The situation with Mars life detection may offer a similar opportunity.

But what should our strategy be to recruit the outsiders and weirdos? Here, I disagree with Jan on the impact of the BARSOOM model.

BARSOOM does something that 50 years of post-Viking analyses failed to do: It offers a coherent biological model that can explain all of the Viking results. The absence of such a model has long been a valid reason for the Viking data to be “culturally and statistically perceived as weak.”

Thus, I see BARSOOM as necessary to get past this deficiency in the discourse. Even if it turns out to be wrong. Three months ago, my “intuition” told me that such a model was not possible, After all, no one had done so. Because I had to give a plenary at the Mars Society, I had to think about the problem. Writing it out has changed … um … my intuition.

So what will I be doing next? Well, I am the author of Life, the Universe, and the Scientific Method. I practice what I preach.

So I will next work to build a non-biological explanation for the Viking results. As Jan said, “despite five decades of trying to do so, humanity [has] failed to fully explain Viking results using abiotic chemistry.” That explanation is needed. I might even submit an application to a NASA committee to fund the effort.

Happy Thanksgiving.

JAN: Steven’s eloquent speech leaves me no option but to find life on Mars, to demonstrate that its metabolism does not match his model.

Happy Thanksgiving Steven.