Searching For Eden
In my last post, I reviewed the standard theory that life on Earth got very lucky to complete a series of hard try-try steps to get to our human level before its window for life closes. I said that this theory has had only mixed success in predicting Earth history timings, and does noticeably badly in predicting that Earth should score well on the key figure of merit of (V*M*W)N, for number of hard try-try steps N, volume V, metabolism M, and time window W not used for easy steps. This seems a pretty big deal, as this is a pretty basic theory on a pretty important process. If we are very wrong about it, that’s important to know. (Just as its important to get to the bottom of credible UFO sightings.)
This evidence conflict might be tolerably weak if one only estimated N=1, one very hard step. But, I said, life designs look so complex and well-integrated that I estimate at least ten hard steps, much more than the few perhaps seen in Earth’s fossil record. This graph conveys a similar intuition, suggesting that those added hard steps happened either as many try-once steps very early in Earth history, or as many try-try steps before Earth.
In my last post, I suggested that this conflict could at least be cut by positing many hard try-once steps, instead of the usual hard try-try steps, early in Earth’s history. But now I’ll admit I don’t think that is enough. I find it hard to believe that more than half of (the integrated magnitude of) hard steps are of the try-once sort, and yet even N=5 gives a quite strong evidence conflict. So I’m forced to take seriously panspermia, the hypothesis that life had another oasis before Earth, and was transferred from that oasis to Earth very early in Earth’s history. Call that prior oasis “Eden”.
Yes, interstellar panspermia seems hard and risky, in effect adding another try-once great filter step which life must compete. Here’s a recent estimate:
But if this scenario allows an Eden with a volume times metabolism as large as Earth’s, and if there are R times as may oases that could support the more robust early sort of life, relative to the more fragile multicellular sort that comes later, then the relative chance of this Eden scenario, compared to just-Earth, is a factor of R*2N times the chance that Eden could successfully transfer life to another (any other) suitable oasis. If Eden’s volume times metabolism were larger that Earth’s, this factor gets larger. So even a low chance to transfer to another oasis might be more than compensated by such a factor.
If there really was an Eden, then finding it is likely to become one of the great historical quests of our descendants. So what can we say about Eden now, to advise this quest? Here are some clues:
Until we replace our usual theory, we still probably want to use the usual figure of merit to predict Eden. Except N becomes the number of hard try-try steps that happen at Eden, and the time deadline for delivering life to Earth says that the time window W can only be extended by having an oasis that starts earlier. So Eden likely meets the usual constraints (like temperature), has a large volume V and metabolism M, and started early.
A set of planets or moons that are close to each other in the same solar system may have a high enough rate of life traveling between them to count as one larger planet, thereby gaining a big advantage. The same may apply to stars that are close and have much moving stuff nearby to induce high rates of transfers. For example, if seeding one star in a stellar nursery effectively seeds S starts in that nursery, then the panspermia theory gets a factor of S boost relative to alternate theories.
The more try-try hard steps that happened on Eden, as opposed to on Earth, the weaker is the evidence conflict re Earth. So maybe most such steps happened on Eden.
Eden mainly needs to give rise to single-cell extremophiles that could travel well enough. So it needn’t support fragile multicellular life, and may work better if it has high (but not overly high) variability to encourage the evolution of robustness. So there seems to be a substantial R factor, and Eden may have been far from a gentle protective “garden”.
Eventually, Eden would have been home to the sort of life that gave rise to Earth’s sort of life. With carbon, water, and DNA. So that rejects exotic hypothesized life such as made of silicon, or in plasmas or neutron stars.
This robust single-celled life seems much less vulnerable to the gamma ray bursts, supernova, and asteroids that tend to kill off fragile life like us close to the galactic center. Or close to the large solar flares common near red dwarf stars. So while Earth was not allowed to be in such places, Eden is allowed there.
Panspermia gets easier in places where stars are spaced closer together, as toward the galactic center. So all else equal, expect Eden in such places.
However, if dormant cells only survive between stars for a million years, then if its dust or rock travel host moved at a typical relative velocity of 30km/s, it could only travel 100 light years in that time, which doesn’t get you far in the galaxy. Thus either Eden was quite near to Earth when Earth acquired life, or dormant life can last much longer, or hosts could fly much faster.
The usual analysis of interstellar panspermia gets pretty low rates. But the chance of panspermia should be increased by the density of stuff flying around near the travel origin and destination locations. Such stuff can kick up life from Eden, and help grab stuff traveling past Earth. Earth had lots of stuff flying about when its solar system formed, and that was embedded in larger complex turbulent dynamic molecular clouds which had more stuff flying about. So if Eden was close to Earth then, Eden was plausibly in a similar cloud area then, which helped induce the travel origin. Seems worth analyzing how molecular clouds change panspermia rates.
Life could have continued on Eden long after it sent life to Earth, but the selection effect of seeing our existence doesn’t enhance the chance of that. The higher is our estimate of the number of oases to which Eden life would have spread, the easier it will be to find such life out there. But unless that chance is enormous, or R is enormous, we expect Eden and any of its other descendants to be quite hard to find. Stars that are siblings of our Sun, born in the same nursery, seem good candidates.
The hypothesis of two prior oases in sequence, instead of one, would also be penalized by a low chance of transfer between them, but might allow a larger total time window W, and a boost in R via more possible oases. Furthermore, this scenario might allow a split wherein try-try steps happen in the oasis with a large figure of merit, but try-once steps happen in multiple parallel small oases, giving them a larger chance of success.
Life in the atmosphere of a brown drawf seems an interesting possibility, but it seems harder for passing stuff to kick out or grab life from such a reservoir. Those things seem easier for life on a planet near a red dwarf, but those may suffer too much variability.
Life may have been possible in a few places 10-17 million years after the Big Bang, from heavy elements formed by supernovae in rare star-forming fluctuation regions that constitute ~10-17 of all matter.
(I’ll add more here as I or others suggest them.)
Added 17Dec: Note that a prediction of the Eden scenario is that the earliest and simplest form of life on Earth is likely a form that enabled panspermia, staying alive but dormant deep in rock for long periods. So life now deep in rock on Earth is predicted to be early and simple, instead of being variations on surface life that migrated down and colonized deep rock.
Very insightful. A series of nearby planets would also reduce risk of life dying of.
Yes, one plausible idea is that Eden seeded the sun's nursery, and thus many of the sun's siblings at the same time.