As we continue going through Christoph Adami’s book, The Evolution of Biological Information, we come to a chapter on the origin of life, entitled “The Informational Origin of Life.” He briefly discusses the RNA world hypothesis but he focuses on the real issue, information. The fundamental challenge in explaining the origin of life is explaining the origin of the encoded information necessary for a self-replicating system.
Adami runs through an example. In his example, 100 bits of information are required. This would take about 1030 attempts to produce by random chance. If we generously grant a billion attempts for every second in the half-billion years it took life to arise on earth that would give us 1025 attempts. That is still five orders of magnitude short of what would be necessary.
Adami asks “Is the spontaneous generation of information thus impossible?” “No,” he says, “because the equation relies on an assumption that is unrealistic and also leads to a dramatic underestimate of the actual likelihood.” What might this assumption be that leads to a dramatic underestimate?
Those Typing Monkeys
The assumption in question is that each monomer is equally likely. To understand, consider the case of monkeys attempting to type out Shakespeare. It is practically impossible that monkeys would, by pure random chance, hit the right keys to type out even a single line of Shakespeare. But what if instead of keyboards, the monkeys were drawing Scrabble tiles? Each letter appears only once on a keyboard. But for Scrabble, there are more tiles for common letters than rare ones. This makes it much easier to assemble words and phrases randomly.
Take for example the line “all the world is a stage.” If we randomly pushed letters on the keyboard, the odds of that phrase are approximately one in 1027. However, if we choose letters from the Scrabble tiles, the odds are approximately one in 1024. That is approximately 1,000 times more likely. But given how much improbability remains, one might rightly be unimpressed and question if the effect is appropriately described as “dramatic.”
In a Biological Scenario
But Adami seeks to argue that the effect might be dramatic in a biological scenario. He considers a protein from HIV and computes a probability of approximately 1 in 1099 for it. But he argues that it could be only 1 in 1085 “if the distribution of amino acids was just somewhat skewed in the direction of the probability distribution of amino acids in typical proteins” and that it could be 1 in 1035 given a more extreme skew.
But there is a fundamental problem with his analysis. This approach works only when the typical distribution of amino acids is strongly non-uniform with some amino acids being far more common than others. It works relatively well in English because some letters are far more common than others. But in the case of amino acids or nucleotides, while there are differences in the frequencies, they are not that dramatic. The cases that Adami considers “just somewhat skewed” are mathematically impossible.
But there is a deeper issue. On what basis would we think that the production of monomers for the origin of life would be skewed in a helpful direction? Adami says:
Yes, the environmental conditions to pull this off are probably very rare, but finding such conditions is still far more probable than the likelihood of finding a large piece of information by chance.
Adami does not defend this assertion, apparently thinking that it is self-evidently true. However, it is not true as work by William A. Dembski on conservation of information shows. His recent publication shows the probability of an environment that greatly improves the odds of producing the origin of life must itself be highly improbable.
The problem is that for every environment that Adami can imagine that favors the origin of life, there are myriad environments that make the origin of life less likely. The more we imagine that an environment favors the origin of life, the more alternative environments can be imagined that favor other non-living products. The problem of finding a suitable environment is no easier than trying to produce the sequence randomly in the first place.
Assembling a Self-Replicating Sequence
But the situation is worse than that for Adami. If we grant the naturalistic production of these monomers to be assembled into a primitive thing, it is most likely that the production would be highly skewed. Some monomers would be very common while others would be rare. Since living sequences are not so highly skewed, taking this into account makes the assembling of a self-replicating sequence less likely.
Adami is incorrect, taking this into account does not dramatically increase the likelihood of an origin-of-life scenario. The scenarios he proposes are mathematically impossible because the frequency of monomers in biological usages are not so far from uniform. Even if they weren’t it would be extremely unlikely to find such a favorable environment. In practice, typical environments are more likely to be inimical to the production of functional sequences. In truth, taking this consideration into account dramatically decreases the likelihood of the origin of life.