SETI Guy's Errant Estimation

For so long astronomer Seth Shostak has been the face of the Search for Extraterrestrial Intelligence that you might call him “Mr. SETI,” just as we might call Tom Brady “Mr. Comeback.” A few days ago Shostak published an article entitled “Simple math shows how many space aliens may be out there.” The article contains both misinformation and poor logic.

First, Shostak commits a big factual error. He links us to a NASA page entitled “At Least One in Six Stars Has an Earth-Sized Planet.” But Shostak describes this as “recent research showing that one in six stars hosts a planet hospitable to life.” There's a huge difference between those two claims. Most of those Earth-sized planets referred to by NASA are not in the habitable zone, and are either too far from the star they orbit or too close to the star they orbit. And even many of the planets in the habitable zones of stars would not be hospitable to life, because they might lack water or suitable atmospheres.

An Earth-sized planet may not be Earth-like (credit: NASA)

This page lists the number of discovered exoplanets as 3588, but this page says the number of potentially habitable exoplanets is only 53, which includes "super-Earths" much larger than Earth. So the truth is that the number of stars hosting a planet that might be hospitable to life is less than about 1 in 60, a fraction ten times smaller than the 1 in 6 number Shostak states.

Then Shostak gives a very common argument for thinking that life should be common in the universe. He tells us this:

Life on our planet began quickly: random chemical activity in 350 million trillion gallons of ocean water spawned a reproducing molecule within a few hundred million years. So maybe biology doesn't need much of a goad to get started. I don't think it's unreasonable to figure that at least half of all planets suitable for life actually produce it.

Our planet is 4.6 billion years old, and claims are made that there are geological signs of life dating back to 3.5 billion years. But such claims are doubtful, as they rely on what are called stromatolites, unusual-looking geological features which some claim were formed by bacteria. We see no cells or biological structures in the oldest stromatolites. The claim that very old stromatolites (older than 3 billions years) are signs of ancient life relies on a rather complicated and debatable line of reasoning. It's quite possible that they are not signs of early life, and that there are alternate geological explanations. This scientific paper says the evidence for life older than 2.5 billion years is “meager and difficult to read.”

Moreover, as discussed here, many scientists think that the earth's oceans are almost as old as the earth itself, having been brought here by comet bombardments. If that assumption is true, there may have been as much as a billion years between the time when life first had a chance to arise on our planet, and the time that it first did arise. If the shaky claims about the oldest stromatolites are in error, there may have been as much as 1.5 billion years between the time when life first had a chance to arise on our planet, and the time that it first did arise. So Shostak's claim that life “began quickly” may be very wrong, even if we define “quickly” as within a few hundred million years.

Even if it were true that life on Earth arose 100 million years after it first had the opportunity to arise, this would not be a strong reason for thinking that life in the universe is common. The article here states the following by MIT professor Joshua Winn (referring to this scientific paper):

There is a commonly heard argument that life must be common or else it would not have arisen so quickly after the surface of the Earth cooled," Winn said. "This argument seems persuasive on its face, but Spiegel and Turner have shown it doesn't stand up to a rigorous statistical examination — with a sample of only one life-bearing planet, one cannot even get a ballpark estimate of the abundance of life in the universe."

It is easy to show the fallacy involved in this type of “if if happened relatively soon, it must been relatively likely” type of reasoning. Consider the assassination of President John Kennedy. On November 22, 1963 Kennedy arrived in Dallas at 11:38 AM. Within an hour he was assassinated. It would not at all make sense for us to reason like this:

1. President Kennedy was killed within an hour after entering Dallas.
2. So if a US president enters Dallas again, there will be a fairly high chance that he will be assassinated – at least 1 in 10.

Of course, such reasoning does not make sense. The length of time between when Kennedy arrived in Dallas and the time of his assassination is just a random data item that in no way suggests that there is a fairly high chance that a US president will be assassinated if he visits in Dallas.

Similarly, it would make no sense to use this reasoning about the tragic assassination of Robert F. Kennedy, the brother of John Kennedy.

1. In 1968 Senator Robert F. Kennedy received a fatal wound very shortly after entering a hotel kitchen.
2. So if a US senator enters a hotel kitchen, there will be a fairly high chance of him being assassinated.

Just as such logic is not valid, it is not valid to reason that there is a relatively high chance of extraterrestrial life appearing on a planet because earthly life supposedly arose relatively soon after it had an opportunity to arise.

Even the simplest living organism is an item of incredible complexity compared to nonliving chemicals. Based on the improbability of getting by blind chance the “organization explosion” necessary for life to get started, we should conclude that if nothing special is going on, the likelihood of life appearing on any particular extraterrestrial planet is incredibly small, less than a millionth of the “one chance in two” that Shostak estimates.

Let's imagine you were to go to a casino in Las Vegas, and you see a red gambling machine which has a slot where you can put in dollars. The machine is simply labeled, “Put in one dollar – you may get 1000 back.” You put in one dollar, and another dollar; but nothing happens. When you put in your third dollar, suddenly a thousand dollars pour out of the machine. You now have to judge: how likely is that this machine will give you a thousand dollars back if you put in a hundred dollars of your money?

If you only consider timing considerations, you might conclude that there is a strong likelihood of winning another thousand dollars after gambling a hundred dollars. After all, you won a thousand dollars relatively soon. But then you suddenly realize: there's a very strong reason for thinking the chance of winning another thousand dollars must be very low on each opportunity. The reason is: you are in a casino, which is a profit-oriented business; and the odds always favor the casino. So you sadly realize that the chance of getting another thousand dollar win is very low each time you try.

This situation with the red gambling machine is quite analogous to estimating the chance of life appearing on another planet. In both cases:

1. There either was or may have been a “relatively early success” that might make you think the chance of success was fairly high on each try.
2. Such a consideration is overwhelmed by a very strong reason that exists for thinking that the chance of success on each try is very low. In the case of life appearing on other planets, that consideration is the extreme unlikelihood of the occurrence of the “organization explosion” needed for any living thing to appear from lifeless chemicals.

Just as the user of this red gambling machine should actually judge that the chance of success is very low on each try (despite one case of an early success), based on blind chance considerations we should conclude that the chance of life appearing on some randomly chosen planet is very low (even if there was one case of an early success on our planet). Shostak has mentioned some minor timing consideration in regard to the origin of life, while making no mention at all of the vastly weightier consideration that should have been considered: the incredibly low mathematical probability of a transition from non-life to life, requiring a sudden leap in organization like a thousand logs transforming into a hotel made of logs.

Exactly the same error is committed on the FAQ of Shostak's SETI organization seti.org, which refers only to the timing consideration, making the extremely dubious claim that earthly life arose “100 million years after life was even possible” (for the reasons given above, the actual gap could easily have been a billion years or more). The same SETI FAQ tells us that “scientists have developed a theory of cosmic evolution that predicts that life is a natural phenomenon likely to develop on planets with suitable environmental conditions.” This is not at all correct, and Darwinian evolution theory makes no prediction at all about the likelihood of life appearing.

Let's consider only one of the innumerable difficulties in explaining the origin of life. In modern cells, proteins are synthesized by little units called ribosomes. But those ribosomes are themselves largely constructed from proteins. So it seems we have a “which came first, the chicken or the egg” problem in regard to the origin of proteins. We cannot simply say that first there were ribosomes, and later there were proteins.  Even the simplest life form requires many functional proteins, but we have no natural explanation for how any protein could arise from the "random chemical activity" Shostak evokes. Ditto for a genetic code and self-replicating nucleic acids.  None of these things has ever been produced by scientists simulating early earth conditions.  Shostak's claim that such things arose from "random chemical activity" is therefore a metaphysical article of faith without any scientific basis.  

In the rest of his article, Shostak continues to just arbitrarily pull numbers out a hat when he estimates that there is 1 chance in 100 that a life-bearing planet would bear intelligent life, and that extraterrestrial civilizations “continue to hang out for 10,000 years before self-destructing.” Given the reality that natural selection is a very poor explanation for the characteristics of minds like ours, the first number could easily be a million times smaller. Shostak fails to include the following relevant things that should be considered when calculating the probability of intelligent life on another planet:

1. The probability of primitive prokaryotic cells ever evolving into the much more elaborate eukaryotic cells (there are reasons for thinking this probability is very low, and the current theory to explain it relies on a far-fetched kind of “sudden transition by ingestion” tall tale that is not at all Darwinian).
2. The probability of unicellular life ever becoming organized into multicellular life, something hard to explain.

Shostak then says, “Do the arithmetic, and you'll find that one in 100 million star systems has technically adept inhabitants.” The arithmetic? There's a thousand different ways to do calculations of the likelihood of extraterrestrial life, and the more realistic ones (assuming only blind chance) lead to estimates thousands, millions, or billions of times smaller than the estimate Shostak has made. It could be that intelligent life is common, but only if some special metaphysical or teleological factors are at play. Calculating only from physics, chemistry, and known biology, the odds are extremely poor. By suggesting that there is some clear mathematical rationale for some arbitrary number that he has picked out of a hat, Shostak resembles someone dogmatically declaring that the chance of you having a blissful afterlife is 57%.

Whenever we hear scientists make assertions about any debatable topic, we should always ask: what type of sociological factors, ideological factors or economic factors may be influencing these assertions? In Shostak's case, we know of a strong economic factor that may be influencing his calculations. Shostak is a leader of an organization (the SETI Institute) looking for extraterrestrials by using radio searches. If you think the chance of nearby extraterrestrial civilizations is high, you will be more likely to donate to Shostak's organization, which solicits donations.

Postscript: There are about 250 billion stars in our galaxy, one of billions of galaxies. Shostak's estimate of one technical civilization per 100 million stars amounts to an estimate of about 2500 technical civilizations in our galaxy.  But in this estimate, astronomer Carl Sagan said, "When we do the arithmetic, the number that my colleagues and I come up with is around a million technical civilizations in our Galaxy alone." Obviously Sagan's math was very arbitrary indeed, and Shostak's coming up with a number about 500 times smaller is just as arbitrary.  When we consider the organization miracle needed for abiogenesis (life appearing from chemicals), the likelihood of any extraterrestrials in our galaxy seems low. The only hopeful hint involves sightings of UFOs.  You would think Shostak would mention these to beef up his case for galactic extraterrestrials, but he always seems to refer to such sightings in a disparaging tone.  This seems like a case of not using your best ammunition.