The Case for a Creator, Chapter 9
How likely is the spontaneous origin of life? In chapter 9, Stephen Meyer likens it to one of those tornado-in-a-junkyard scenarios that creationists love so much:
“Imagine trying to generate even a simple book by throwing Scrabble letters onto the floor. Or imagine closing your eyes and picking Scrabble letters out of a bag. Are you going to produce Hamlet in anything like the time of the known universe?” [p.229]
Obviously, the answer is no. Almost as obviously, however, this is not a question that bears on the origin of life.
Let’s see how Meyer’s facile comparison holds up if we put some actual numbers on it. I downloaded Hamlet from Project Gutenberg and did a character count on the text file. Not counting spaces, punctuation or the copyright notice, I came up with a total of 129,839 characters. Since the alphabet has 26 letters, it takes a minimum of 5 bits to specify any single letter, which means that Hamlet has (129,839 x 5) = 649,195 bits of information.
To contrast to this, consider the smallest known genome: Carsonella ruddii, a bacterium that lives in the guts of leaf-eating insects called psyllids. It has only about 160,000 base pairs of DNA, coding for 182 proteins. But since there are only 4 base pairs in DNA, it takes only 2 bits to specify each one, which means that Carsonella‘s genome contains (160,000 x 2) = 320,000 bits of information: less than half of Hamlet! And Carsonella is the smallest modern genome. The very first life, which was probably little more than a self-replicating hypercycle of molecules, would have been smaller still.
Obviously, this analogy is still rigged in Meyer’s favor: neither evolution nor the laws of chemistry are very much like picking Scrabble tiles out of a bag. The laws of English are such that the vast majority of possible arrangements of letters are meaningless gibberish, but this is not true of proteins and DNA. Because a protein’s function is defined by its shape, virtually every possible string of amino acids potentially “means something” in a way that random combinations of English letters don’t. In the primordial sea, there would have been billions of different molecules drifting around, bumping up against each other, interacting in countless ways. Until we know the smallest possible interacting set of molecules that could be called alive – and we don’t know that, at least not yet – there’s no basis for any claim about how likely it would have been for such a thing to arise by chance.
“There’s a minimal complexity threshold… There’s a certain level of folding that a protein has to have, called tertiary structure, that is necessary for it to perform a function. You don’t get tertiary structure in a protein unless you have at least seventy-five amino acids or so. That may be conservative. Now consider what you’d need for a protein molecule to form by chance.
First, you need the right bonds between the amino acids. Second, amino acids come in right-handed and left-handed versions, and you’ve got to get only left-handed ones.
Creationists are fond of invoking this “handedness” problem (the technical term is “chirality”). It refers to the fact that certain organic molecules like sugars and amino acids naturally come in two stable configurations that are mirror images of each other, like your left and right hand. Most living things use only left-handed amino acids and right-handed sugars, and creationists often suggest that no natural force could produce this bias.
But, in fact, there’s a wide variety of natural mechanisms that can sort molecules by chirality. Some common crystals, such as calcite, selectively absorb molecules of one handedness on one crystal face and the other handedness on the opposing face. (The chirality of all modern life may simply be because that vital first set of chemical reactions occurred on one side of a rock rather than another.) Circularly polarized ultraviolet light also selectively destroys molecules of one handedness. (The Murchison meteorite, which contains amino acids produced in the early solar system, has an imbalance of left-handed amino acids, and some scientists feel it’s for precisely this reason. See also.) There’s also a chemistry principle called “majority rule” in which certain reactions that begin with a weakly chiral mixture can produce products that are strongly chiral. Some scientists even believe that the laws of physics are not completely symmetric and one chirality is energetically favored over the other. Any of these mechanisms, or several of them in combination, could plausibly be why life has one chirality and not the other. We don’t know the true cause for certain – but it’s not that we have no idea how it could have happened; it’s that we have too many candidates and can’t choose among them!
“Third, the amino acids must link up in a specified sequence, like letters in a sentence.
Run the odds of these things falling into place on their own and you find that the probabilities of forming a rather short functional protein at random would be one chance in a hundred thousand trillion trillion trillion trillion trillion trillion trillion trillion trillion trillion. That’s a ten with 125 zeroes after it!” [p.229]
There’s no explanation of how Meyer got these numbers (again, footnotes are absent just where they’d be most helpful). But I strongly suspect he’s committing the poker player’s fallacy again: assuming that only one amino acid sequence can provide the function he wants. His mention of “a specified sequence” implies as much. But in any plausible origin-of-life scenario, there wouldn’t be one miracle sequence, but a large number of functionally equivalent sequences. We already know this to be true in modern life: a large percentage of mutations are neither positive nor negative but neutral, having no effect on the overall shape or functioning of the protein.
But if Meyer’s numbers are right, it should be easy to prove in an experiment: just generate some organic molecules at random and see what happens. If he’s correct, nothing interesting or useful will ever emerge. Well, unlike creationists comfortably ensconced in their armchairs, real scientists do run experiments like this. From a Usenet post by the biologist Howard Hershey:
Random syntheses of 50 nucleotide long RNAs generates certain specific selectable functional ribozyme (RNA enzyme) activities relevant to biological functions that would be needed for an ur-organism (RNA ligases, terminal transferases, etc.) in the range of once every 1014-1017 molecules (a mole of molecules is about 1023, so it is a virtual certainty that you will have a number of molecules with the needed activities in a millimole of such randomly generated RNA (you would certainly be able to hold this in a thimble). Moreover, ALL these activities would be present in the SAME millimole of RNA.
For technical details, see this similar paper from the journal Science: “Structurally Complex and Highly Active RNA Ligases Derived from Random RNA Sequences”. The authors say, “The fact that such a large and complex ligase emerged from a very limited sampling of sequence space implies the existence of a large number of distinct RNA structures of equivalent complexity and activity.”
To put this in layman’s terms: a thimbleful of randomly generated RNA sequences contains numerous enzymes with an interesting variety of biologically relevant abilities. This is a far cry from Meyer’s “ten with 125 zeroes after it”. Either nature is pulling a prank on us by defying the odds every single time, or else the ID advocates’ calculations are based on unrealistic assumptions. Note that these RNA enzymes were only 50 nucleotides long – shorter than Meyer’s “minimal complexity threshold” of 75 amino acids or more – and yet were still able to perform biologically interesting functions. And for any plausible origin-of-life scenario, we’re not talking about a thimbleful of molecules, but a whole planet’s worth.
Other posts in this series: