First posted on January 14, 2000. Last update January 30, 2002.
Spetner calculates probabilities
Spetner discusses randomness
Spetner discusses information
In the book titled Not by Chance. Shattering the Modern Theory of Evolution, Dr. Lee M. Spetner offers its readers interesting pieces of information from molecular biology and adjacent fields of science. Unfortunately, rather than limit his discourse to scientific arguments, Spetner's book contains passages which introduce the extraneous question of the compatibility of science with religious dogma.
Spetner introduces the above mentioned question at the very beginning of his book. In the Preface (page ix) Spetner writes: "...I met the evolutionary theory in a serious way, and I found it hard to believe. It clashed not only with my religious views, but also with my intuition about how the information in living organisms could have developed."
This is a damning admission for a scientist. The latter is supposed to base his/her views only on facts established via a verifiable and reproducible procedure of an unbiased exploration. Religious views (i.e. beliefs) to which every scientist is, of course, fully entitled, are supposed to be kept aside from his/her scientific search for truth. As to intuition (often alternatively referred to as common sense) it has its legitimate place in research, but the problem with it is that it often leads different people into different directions, and as often as not leads nowhere. (I have described elsewhere situations where relying on common sense, i.e. on intuition, would lead to disaster – e.g. in the case of orbital flights .)
It can be assumed from the above passage that the motivation behind Spetner’s effort had been his desire to reconcile his religious beliefs with scientific evidence. In other words, Spetner seems to have had an agenda rooted in his religious beliefs, i.e., beyond the boundaries of science. A non-scientific agenda chosen a priori is something a scientist is supposed to avoid (although I am not trying to assert that scientists always succeed in completely avoiding ideological presuppositions).
The book in question leaves the impression that Spetner the believer was watching Spetner the scientist thus limiting his freedom of following only the facts wherever they might lead him.
Spetner the believer is in full force as the book starts, but becomes less and less visible as his story unfolds. Toward the end of the book, Spetner the scientist seems to have won the battle, as he writes, on page 212, about his theory, which he calls NREH: "The NREH, on the other hand, is agnostic..."
Of course, Spetner the believer hates to admit his weakness, so this other Spetner surfaces in the continuation of the quoted segment, as follows: "...and poses no contradiction to creation. The NREH, as an explanation of evolution, is in fact derivable from Talmudic sources."
While readers have no reason to doubt that Spetner himself may sincerely believe in what he says about the alleged derivation of his theory of evolution from the Talmud, the quoted statement is irrelevant as far as the theory itself goes. In the space of two hundred pages, Spetner offered a variety of seemingly scientific arguments against the Neo-Darwinian Theory (NDT) and finally suggested his own hypothesis without using any notions, originating either in the Talmud or in any other religious sources or beliefs. Hence his sudden reference to the Talmudic sources sounds like nothing short of a ransom paid to his agenda. Moreover, in such an immense compendium of commentaries and interpretations as the Talmud, one can certainly find some commentaries of this or that rabbi seemingly compatible with any views.
In view of the above, the subtitle of the book in question Shattering the Modern Theory of Evolution seems to be misleading.
In fact, Spetner's theory itself is that of evolution. Contrary to what some readers could conclude from the title and subtitle, Spetner's argumentation is not against the theory of evolution per se. Spetner actually only argues against some aspects of the Neo-Darwinian theory of evolution, namely against the idea that evolution necessarily included random variations, suggesting instead what he calls "Non-Random Evolutionary Hypothesis" (NREH).
However, the term used by Spetner for his hypothesis, Non-Random Evolutionary Hypothesis, is itself misleading. The Neo-Darwinism does not maintain that evolution occurred in a purely random way. On the contrary, the Darwinian theories in all of their modifications maintain that evolution included both random and non-random elements. According to the neo-Darwinian approach, the random element of the evolution entails mutations in the genome. Mutations are believed to be random - this is indeed the assumption of the Neo-Darwinism (although even this assumption must be qualified – see a further discussion). However, the evolution theory is rather far from being limited to mutations. Its second, equally crucial part is the concept of natural selection. The latter is by no means random. The natural selection is directed by the environment. Using Spetner's terminology, natural selection is led by the signals from the environment. In that sense, evolution according to Neo-Darwinism is governed both by random and non-random factors.
Spetner seems to be blissfully unaware of the actual basic principles of the Darwinian evolution theory. If he paid attention to what actually the Darwinist theory of evolution entails, he would have realized that the the concept of randomness is only a part of that theory which entail non-random elements as well.
Moreover, whereas the adherents of Neo-Darwinism who are the target of Spetner's assault, do indeed view mutations as occurring randomly, they do qualify that concept in a quite substantial way. For example, one of the most distinguished defenders of Darwinism, Richard Dawkins, who, unlike Spetner, is a professional biologist, in his very popular book  provides five aspects in which the mutations are not completely random (pages 305-307 in Dawkins's book). Here is just one brief quotation from Dawkins' book (page 306): "There are, in truth, many respects in which mutation is not random." On page 308 Dawkins continues: "The Darwinian says that variation is random in the sense that it is not directed toward improvement, and that the tendency toward improvement in evolution comes from selection."
In view of the above, the very name Spetner chose for his hypothesis sounds as an obvious misnomer. As the matter stands now, Spetner seems to fight against a straw man.
What is then the real difference between the Neo-Darwinism which Spetner rejects and his NREH?
Spetner's theory shifts the non-random step of the evolution from natural selection to mutations. According to Spetner, random mutations followed by natural selection could not ensure the rate of evolution rapid enough for the appearance of the enormous multitude of the existing species. Hence, insists Spetner, to explain the actual rate of the evolution of the species, we must assume that mutations are non-random but rather caused by signals from the environment.
What Spetner seems not to be aware of, is that his hypothesis is by no means a novel one. Similar ideas had been suggested before Spetner and rejected by the overwhelming majority of biologists for a number of reasons, mainly because they are not supported by evidence. Spetner's predecessors sometimes are referred to as "mutationists." For example, in the above quoted book by Dawkins, we find a rather detailed criticism of the views of such "mutationists," (pages 305-309).
Maybe, unlike his predecessors within the ranks of "mutationists," Spetner came up with some hitherto unknown arguments in favor of non-random mutations, triggered by a "signal" from environment and directed toward improvement? Unfortunately, this is not the case. There is not a shred of positive evidence in Spetner's book which would support his hypothesis. All his argumentation is of negative character, wherein, rather than offering evidence in favor of his hypothesis, Spetner tries to show that the Neo-Darwinian theory entails serious faults. Of course, even if Spetner's critique of the Neo-Darwinism were well substantiated, this in itself would not signify the validity of his alternative hypothesis.
Whereas Spetner's is entitled to suggest whatever hypothesis he chooses, in order to be accepted in science, his mislabeled Non-Random Evolutionary Hypothesis would require, first, an explanation of what is the nature and mechanism of the supposed signals from the environment which cause the necessary useful mutations. And, second, his theory would require empirical evidence to support his hypothesis of directed rather than random mutations. Unfortunately, Spetner's discourse does not meet either of the two mentioned requirements. His hypothesis is an unsubstantiated assumption which lacks both empirical foundation and explanatory details which are necessary to lead from his ideas to a real scientific theory. No wonder Spetner published his work as a popular book for a broad audience rather than in a peer-reviewed scientific journal.
Spetner's critical discussion of various aspects of the Neo-Darwinism comprises three parts - probabilities, information and randomness. I will discuss all the three elements of Spetner's discourse in subsequent sections of this article.
It is hard to disagree with Spetner when he says that his hypothesis is agnostic. Every scientific hypothesis is. This statement by Spetner, made at the book's end, contradicts his agenda as it could be understood from his statements at the book's beginning.
The questions of faith cannot and must not be treated by means of scientific research, which has no tools to either confirm or reject religious claims, except for the cases of a deliberate fraud that can be debunked by an impartial test. Spetner's hypothesis provides no arguments either in favor or against any religious beliefs, however strongly some defenders of religious views, including Spetner himself, may wish to derive from that hypothesis such conclusions.
It is interesting to mention that Spetner's book, which its author intended to serve as a weapon defending religious beliefs, was subjected to rather strong criticism  in the Jewish Action magazine which serves the Orthodox Jewish community. The author of the review of Spetner's book in that magazine, Dr. C. Feit, concentrates on the very heart of Spetner's discourse, namely on Spetner's rejection of randomness in the process of evolution. Feit is mainly interested to show that randomness is not contrary to the biblical story. While the religious reasons to disagree with Spetner are beyond the scope of my discussion, there are certain points in Feit's review which could very well be accepted in a rational analysis of Spetner's writing. As Feit indicates, Spetner's premise can be characterized as the "God-in-the-Gaps" approach to the conflict between religion and science. The adherents of that approach look for "gaps" in scientific theories, i.e., for points either insufficiently explained or not explained at all by the existing theories and insist that the only explanation of such points is in invoking the idea of a supernatural deity. Feit, who is a believer, rejects such an approach because the ongoing development of science can eventually lead to a scientific explanation of the yet obscure points, thus forcing religion to retreat with humiliation. Feit also indicates that Spetner has left out many important areas of modern biological research such as (I am quoting) "complexity theory, neutral mutations, self-organization in complex systems, artificial life, common attractors, the modular domain structure of modern proteins." Since Feit is an expert in biology, this part of his critical remarks casts a shadow on the very aspects of Spetner's book which purportedly constitute its strongest points. Since I am not a biologist, I leave it to Dr. Feit and other experts in that field to discuss the biological aspects of Spetner's book, discussing instead some points which are closer to my own expertise.
A detailed critical discussion of Spetner's work from a biological standpoint was given, for example, by biologists Ian Musgrave  and Gert Korthof. These two critics, using very specific arguments from biology, have shown very convincingly the absence of substance in Spetner's hypothesis.
In chapters 4 and 6 of his book, Spetner devotes much space to the discussion of probabilities. Chapter 4 is titled Is the deck stacked? In that chapter Spetner discusses probabilities of random mutations and of other events which, according to the Darwinian approach, were the steps of evolution. Spetner repeats here the same well known calculations of probability which had been copied with slight variations from book to book written by opponents of the theory of evolution but which are actually irrelevant for the problem of the origin of life.
I discuss those calculations of probabilities at length at Improbable probabilities. I will not repeat here all those arguments, but will only discuss briefly a few points in regard to Spetner's calculations.
As I point out in the above posting, from the cognitive viewpoint, the concept of probability first and most reflects the level of ignorance about a situation. If we possessed the full knowledge of the latter, we would deal with certainties rather than probabilities. The extremely low values of probabilities of various events, such as the spontaneous emergence of life and the like, which are often quoted by the opponents of evolution, first and foremost reflect the paucity of information about the events which would lead, for example, to useful surviving mutations or to the emergence of life. This lack of knowledge forces us to assume an enormous number of possible competing events thus resulting in the very low probabilities cited by Spetner, Aviezer , and others.
How is the probability of an event estimated to be, say, 1/N where N is a very large number? This estimate is tantamount to the assumption that there could be N possible events, of which, say, the emergence of life is just one. Assuming further that all those N possible events are equally probable, we divide 1 by N and arrive at the very low probability 1/N, which turns out to be, say, one in ten to the twenty-seventh power (as in Spetner's example). The conclusion derived by the anti-evolutionists is that the cited probability is so small that the event in question is all but impossible.
This way of thought, common to many discussions by the opponents of evolution, is flawed as it misinterprets the meaning of the mathematical quantity called probability. Indeed, if the probability of an event (say, of a useful mutation) is calculated as 1/N, it means that we assumed the competition of N equally probable events. One of those N equally probable events must necessarily happen (even though we don't know in advance which one). If any of these events is viewed as impossible, then by the same token each of them must be viewed as impossible, as all those events have exactly the same chance of happening or not happening. This conclusion, based on the assumption of impossibility of any one of N events is absurd. Hence the premise was wrong.
If a certain event, A (for example a certain mutation) whose probability was calculated as 1/N did not actually happen, it only means that some other event B, whose probability was equally small, happened instead. Why could event B happen but the equally probable event A could not? From the standpoint of probability, there is no difference between all those N events, even if one of them is very special from some non-mathematical viewpoint (for example, being the spontaneous emergence of life).
Moreover, the Neo-Darwinism is not based on the premise of a purely random process. The combination of random mutations with a suitable law (for example, the non-random natural selection) can accelerate evolution by many orders of magnitude (as, for example, has been demonstrated by Dawkins ) and this makes Spetner's probabilistic exercise immaterial.
Add one more comment. Discussing the analogy with a lottery, Spetner considers a favorite example of the opponents of evolution, that of multiple wins, and states correctly that the probability of cheating seems to be much larger than the probability that a person wins a lottery twice in a row. Similarly, he states that in a poker game the simultaneous occurrence of two straight flushes indicates fraud with a much larger probability than its happening by chance. He writes (page 94): "As we have seen in the story of the poker-playing cowboy, too much luck might be not good for you." (In the story in question, the cowboy who had the straight flush was shot to death for the alleged fraud). Then Spetner continues: "So too in nature, if we see the occurrence of the event with exceedingly low probability, we must suspect the event was not random..." This statement is unsubstantiated. Indeed, fraud can be more probable than two simultaneous straight flushes in a poker game, or a double win in a lottery. However, this example is irrelevant in regard to nature where there is no analogy of fraud. In nature an event is perfectly possible despite its exceedingly small calculated probability, because, first, the very small value of the calculated probability is mainly due to the extreme paucity of information about the situation, and, second, because the event whose probability was calculated as very small is actually as probable as any other of N events assumed to be possible.
The chapter in question contains many other calculations of probabilities, but all of them are fraught with similar fallacies, discussed and dismissed many times before. Therefore the argumentation by Spetner against the theory of random evolution, based on his calculations of probabilities in his chapter 4, is not convincing.
In chapter 6, titled The Watchmaker's Blindness, Spetner returns again to probabilities estimation, this time using it to refute the ideas suggested by a prominent adherent of the NDT, Richard Dawkins . Here, again, Spetner offers calculations of extremely small probabilities of such events as encountering a "perfect bridge hand." According to that calculation, the probability of getting a perfect hand in one deal is one in ten to the twenty eighth power. This number, although correct from the formal viewpoint, is irrelevant. Spetner accuses Dawkins of not understanding the meaning of probability. Actually such an accusation could be directed toward Spetner himself, who obviously knows how to calculate probabilities but ascribes to them properties those quantities do not possess. The value of probability does not predict the outcome of any particular event. Despite the extremely low calculated probability of an event, it can well happen on the very first trial, whereas an event whose calculated probability is much higher may not happen even in hundreds of thousands of tests. When such problems as the possibility of the spontaneous emergence of life are discussed, their calculated probability is rather irrelevant and cannot be used as a proof of any opinions on that subject.
The hypothesis Spetner suggests instead of the NDT is, as we mentioned, what he calls NREH, which stands for Non-Random Evolutionary Hypothesis. While NDT is based on the assumption that evolutionary changes occurred via random mutations, followed by non-random natural selection, Spetner's idea is of non-random mutations triggered by the demands the environment imposes on species. Obviously, since the difference between NDT and NREH essentially boils down to that between random and non-random chains of events, the concept of randomness becomes germane for the discussion of those two hypotheses. Consequently, we could expect that Spetner would provide some definition of randomness, as the fundamental concept of his hypothesis. Strangely, Spetner seems not to be worried about the precise meaning of a term he uses so frequently in his discourse. As a result, in different contexts, he uses the terms random and randomness in different ways, seemingly not noticing the vagueness this sloppiness of usage imparts to his discourse.
Here are a few examples.
On page 44 Spetner writes: "The motion of these genetic elements about to produce the above mutations has been found to be a complex process and we probably haven't yet discovered all the complexity. But because no one knows why they occur, many geneticists have assumed they occur only by chance. I find it hard to believe that a process as precise and as well controlled as the transposition of genetic elements happens only by chance. Some scientists tend to call a mechanism random before we learn what it really does. If the source of the variation for evolution were point mutations, we could say the variation is random. But if the source of variation is the complex process of transposition, then there is no justification for saying that evolution is based on random events."
We see that Spetner used in the quoted segment the term random twice, once in relation to a "mechanism," and once to "events." He, apparently, assumes that the term in question is universally understandable without an explanation, in both contexts he uses it. Let us guess what Spetner means when he uses the word "random." From the quoted text it seems to follow that, according to Spetner, random is such event that, first, occurs just by chance, and, second, is simple, comprising only one step. As to the "random mechanism," this term seems to mean a combination of consecutive random events.
If my interpretation of the meaning Spetner implies when saying "random" differs from that by Spetner himself, it only means that he had to define the usage of this term in an unambiguous way.
On page 46, Spetner writes: "But Darwinian theory asks that the mutations be both spontaneous and random." It is not clear whether Spetner intended the quoted sentence as a quasi-direct quotation from some writings by adherents of NDT, or it is given as his own formula, but in any case he uses the expression "spontaneous and random" without any indication that he might view it as imprecise or misleading. Does that expression mean that events (and mechanisms?) can be random but not spontaneous, or spontaneous but not random, or both spontaneous and random? What exactly is meant by these terms?
The necessity for a stringent definition of the above terms and for following that definition consistently is crucial because the concept of randomness (and related to it concepts of complexity and spontaneity) are fundamental for his hypothesis.
Spetner gives no indication that he is familiar with the mathematical definitions of randomness and complexity.
At the core of Spetner's hypothesis is the suggestion that the variations leading to evolution are triggered by forces of environment and are directional rather than random. To substantiate that suggestion, he must first clearly understand in what way the variations in question are not random. To this end he must clearly define what is random and what distinguishes non-random from random. Without first building the foundation in terms of random vs non-random, Spetner's idea remains too vague for a scientific hypothesis.
Spetner may believe that the variations he admits to be the steps in evolution are non-random, but he actually has no knowledge which allows him to clearly establish that the variations in question are indeed non-random. So far his classification of events as random or non-random remains a matter of his personal preference rather than a clearly established fact.
The concepts of randomness and of the closely connected complexity are strictly defined in the algorithmic theory of probability (ATP). Let us recall these definitions. The "complexity" of a system (or of a process) which is often referred to as Kolmogorov's complexity, is defined in ATP as the minimal size of an algorithm (or of a program) which can encode the system (or the process) in question. From this definition follows the definition of randomness as follows: A system (or a process) is random if its complexity approximately equals the size of the system (or process) itself (in bits).
The term "approximately" appears in the above definition because, as ATP shows, randomness is a matter of degree. The closer the value of a system's (or of process's) complexity to the system's/process's own size, the closer that system (process) is to perfect randomness. This shows that the concept of randomness is more complex than it may seem at first glance. The demarcation between random and non-random events or mechanisms is diffuse. Events or mechanisms can be more random and less random. This feature of events is absent in Spetner's discussion.
To base any hypothesis on the concept of random vs. non-random requires a quantitative approach. There are no signs of such an approach in Spetner's book. Therefore when Spetner tells us that certain events are random while some other are non-random, we really cannot verify his statements and with them to judge his line of thought rationally.
The above mathematical definitions are universal and applicable to all systems and processes. Since Spetner's treatise lacks reference to the strict definition of randomness, his whole discourse in relation to random vs. non-random events and mechanisms has little meaning.
Spetner, according to his biographical data, has extensive experience in research related to communication systems, and also, according to his book, must be well familiar with information theory. Therefore he is expected to also be familiar with the mathematical concept of randomness, even more so because he has based his theory of evolution on the distinction between random and non-random events. Strangely, there is not a single sentence in Spetner's book which would reveal his knowledge of the pertinent mathematical concepts and this gives his suggestion of the supposedly non-random sources of evolution the flavor of speculations by a dilettante.
Chapter 5 in Spetner's book is titled "Can random variation build information?"
One of Spetner's main arguments is that random variations never or almost never lead to the build-up of information. Since this concept is one of the foundations of his NREH, one would expect that the fundamental concept of that hypothesis, information, would be presented in an unambiguous, rigorous manner.
Information theory is the scientific basis of modern communication technology. Spetner has been introduced to the readers as an expert in signal processing, a field largely based on the information theory. Hence the readers are entitled to a good professional discussion by Spetner of matters related to information in its application to NREH. Unfortunately, this is not the case.
Like with probabilities and randomness, Spetner uses the term "information" vaguely, as if this concept has no quantitative measure. Therefore, when he maintains that certain variations do or do not add information to a system, these statements remain unsubstantiated. Having accused some other authors (e.g. Dawkins) of lacking calculations which would support their theories or hypotheses, Spetner himself does not offer a single quantitative estimate of the change in the amount of information caused by the variations he discussed. Without such calculations his assertions that this or that variation did or did not add information to a system remain ambiguous and non-verifiable. His statements in regard to particular variations allegedly adding no information are based not on factual evidence but only on his intuition, which may or may not lead to the right conclusion.
One example of such an ambiguous situation is Spetner's example with a streptomycin molecule fitting into a specific location in a ribosome (figs. 5.3 and 5.4 in Spetner's book). There is a steric match between the molecule of streptomycin and the bacterium's ribosome. The streptomycin molecule fits a specific site in the ribosome like a key fits into the proper lock. This inhibits the bacterium's normal functioning thus eliminating its detrimental effect on the human body. A random mutation can cause a change in the shape of the pertinent site in the ribosome making it unsuitable for the attachment of a streptomycin molecule. Such a mutated bacterium acquires immunity against streptomycin thus improving its chance for survival. Spetner maintains that the described mutation leads to a loss rather than increase of information. This assertion is not substantiated.
The described change in the ribosome's shape may either increase or decrease the amount of information associated with that ribosome. To decide whether it is a loss or a gain in information, a detailed calculation is necessary based on the detailed knowledge of a particular shape's change. As long as such knowledge is not available, my guess as to whether the information increased or dropped, is as good as yours or as Spetner's.
Spetner's attempt to substantiate his assertion that the amount of information is decreased by the described mutation because this mutation makes the ribosome less specific is itself unsubstantiated. The ribosome may become less specific in relation to streptomycin, but may become instead specific in relation to some other substance. Since information about such a possibility is absent, there is no reason to assert that the specificity in Spetner's sense has indeed dropped. Therefore Spetner's assertion that the mutation in question resulted in a decrease of information is pure speculation with no evidentiary value.
In his effort to substantiate his hypothesis, Spetner used, as the three foundation stones, calculations of probabilities, the randomness vs. non-randomness comparison, and the consideration of information's build-up. After we have dismissed the calculations of probabilities as irrelevant, the notion of "random events not capable of causing evolution" as uncertain, and the alleged absence of information build-up as not proven, the natural question to be asked is, "what then remains of the book in question to be taken seriously?"
My answer to that question is "almost nothing." There is nary a chance Spetner's book may be taken seriously by biologists. Of course, adherents of Intelligent Design may praise Spetner for allegedly showing that evolution cannot generate information, but this thesis has been promoted by other ID creationists on a much higher level of sophistication, and all of their arguments were rejected by the overwhelming majority of both biologists and information theorists (see, for example, the discussions of books and articles by Dembski, Behe and Johnson on this site).
Anyway, regardless of Spetner's hypothesis being correct or wrong, it entails no religious consequences, even if that might be contrary to Spetner's own intention. Spetner's attempts to interpret his hypothesis as proving certain religious concepts were, first, not useful, and, second, unsuccessful, since his hypothesis can be equally viewed as either confirming or contradicting those concepts which have no relation to that hypothesis's veracity. Another weakness of Spetner's book is his attempts to substantiate his hypothesis by invoking probability, randomness, and information, and having done this in a way not meeting the requirements of scientific discourse, even on the level of a popular narrative
Spetner had a chance to write a serious discussion of an important problem, suggesting his view of that problem, in a non-sensational way. He chose another way. Therefore the title of his book, Not by Chance would become more proper if changed to A Missed Chance.
(1) Lee M. Spetner, Not by Chance, Shattering the Modern Theory of Evolution, (New York: The Judaica Press, 1998).
(2) Mark Perakh, Improbable probabilities, accessed on April 4, 2002.
(3) Richard Dawkins, The Blind Watchmaker, (New York: W.W. Norton & Company,1986).
(4) C. Feit, "Review of Not By Chance by Dr. Lee Spetner," Jewish Action, Spring 1999.
(5) Ian Musgrave, Spetner and Biological Information. Accessed on June 23, 2003.
(6) Gert Korthof, Could It Work?, 2001. Accessed on June 23, 2003.
(7) Nathan Aviezer, In the Beginning (Hoboken, NJ: KTAV Publishing House, 1990).
Mark Perakh's home page.