One of the cornerstones in the crumbling foundation of creationist "science" is the notion that evolution contradicts the second law of thermodynamics. The classical version of this law may be stated as follows: The entropy of an isolated system can never decrease. (An isolated system is one that does not exchange energy or matter with its surroundings.) Creationists originally argued that a decrease in entropy is exactly what evolution requires, hence the conflict with the second law. This argument was used in an article by Dr. Morris of the Institute for Creation Research (ICR) as late as 1973. As is the usual practice among creationists, he tried to support it with out-of-context quotations from the writings of respected scientists.
Actually, it is not difficult to find inaccurate statements regarding entropy in popular science literature. Ever since the time it was first defined, entropy has been recognized as a most elusive quantity as far as understanding its physical significance is concerned. Defining it mathematically in terms of other quantities is no problem; however, this cannot be done to advantage in popular debates, a situation that creationists have been quick to capitalize on. Entropy has been defined nonmathematically as a measure of disorder, equilibrium, uncertainty, and unavailability of energy. Actually, to consider only the entropy content of a system is not enough; a system can gain entropy and, at the same time, become more organized, unbalanced, and richer in information and available energy. (A few examples will be considered later on.) What is important is the entropy deficiency of the system. We define this as the difference between the system's entropy capacity (the maximum amount of entropy the system is capable of holding with its present energy content) and the amount of entropy it is actually holding. This deficiency may also be referred to as negentropy (short for negative entropy)—a concept which, had it been generally adopted, might have been less confusing than entropy. Negentropy, then, has been defined as a measure of order, information, lack of equilibrium, and the availability of energy for doing work.
But most fundamentally, negentropy—or entropy deficiency—is a measure of the improbability of a system being in a given state. For this reason, when we discuss such things as the improbability of a certain nucleotide sequence, for example, we are also discussing entropy and the second law of thermodynamics.
A final warning: the word order in popular usage is highly ambiguous and should be scrupulously avoided in explanations of entropy for the benefit of anyone not already familiar with scientific jargon, lest it cause a great deal of confusion. (The mathematically inclined reader can refer to such works as Sears in 1959 and Brillouin in 1962 for more detail.)
The creationist argument given in the first paragraph contains a gaping flaw, and evolutionist debaters wasted no time in pointing it out: While the classical version of the second law does indeed state that the entropy of an isolated system cannot decrease, evolving systems are not isolated! One might expect that at this point the issue would be considered settled and everyone would pack up and go home. However, such an expectation would never be entertained by anyone familiar with the peculiar tenacity of creationists.
Let us see how Morris responds after he has been confronted with the clear evidence that evolving systems are open. In 1976, he said: "The second law really applies only to open systems, since there is no such thing as a truly isolated system." This statement suggests that he lacks the ability to distinguish between theoretical and practical concepts—an ability which is absolutely essential for the understanding of much of physics. It is certainly true that the second law applies to all thermodynamical systems; it wouldn't be much of a law otherwise. But the particular statement of the second law that Morris has in mind—namely, that the entropy cannot decrease—applies only to isolated systems. It is a purely theoretical statement, and in theory, any desired system can be postulated whether or not it can exist in practice. Let me mention another example: The concept of an ideal gas is utilized throughout thermodynamics and is extremely useful, even though no such substance actually exists. Just as real gases approximate an ideal gas, some better than others, there are real thermodynamical systems that are very nearly isolated. In these systems we do not expect the entropy to decrease. On the other hand, in a wide open system the entropy can either increase, decrease, or remain constant. The second law does not in any way prevent entropy decreases and the generation of entropy deficiencies in local systems so long as there is an equal or larger increase in entropy outside the system. This concept is easily grasped by most college and even high school students of science but not, apparently, by creationists, including those boasting Ph.D.s in the sciences.
It might now seem that all we have to do is give some examples of open: systems in which the entropy decreases and then we can pack up and go home. But alas, no such luck. In an attempt to counter this, creationists have introduced a new device, which one creationist, Mr. Elmendorf, calls "The Creative Trinity," a properly descriptive phrase with an appropriate ring that I will therefore adopt.
The Creative Trinity
According to this creationist concept, a system can become entropy deficient only if three conditions are satisfied (Morris, 1976). (1) Free energy must be supplied to the system. This is actually incorrect, since a loss of energy can also generate an entropy deficiency; however, the need for the system to be open is universally recognized, so further discussion is unnecessary. (2) The system must contain an energy conversion mechanism. When creationists are pressed, we find that just about anything qualifies as having a "mechanism," including matter itself, so the statement becomes quite meaningless. (3) The system must contain a directing program. This is variously referred to as intelligence, information, control system, and so forth by creationists. The idea is that this directing program did not arise through natural processes but was created by God. The Creative Trinity can also be interpreted as a statement to the effect that there are different kinds of entropy which are not interchangeable.
We must take careful note of an elementary fact which is often missed in debates on evolution and the second law: In spite of what they claim, creationists are no longer talking about the second law. They wish to give the impression that science, in this case thermodynamics, is on their side in their opposition to evolution. But the fact is there is nothing in thermodynamics that contradicts the phenomenon of an entropy deficiency being produced in a system when energy flows through it. On the contrary, this is what thermodynamics leads us to expect, and nothing else is needed, such as a directing program, etc. It is interesting to note that, in his resolution of the long-standing paradox of Maxwell's demon, Brillouin showed that, to enable the demon to distinguish between fast and slow molecules, energy has to be supplied to the system, thus producing an entropy increase elsewhere in just the amount required by thermodynamics (Ehrenberg, 1967). And it doesn't matter whether the demon is an intelligent being or a simple mechanism.
Creationists are not showing that evolution contradicts the second law of thermodynamics; instead, they are saying that the second law, as accepted by conventional science, is incorrect and insufficient to explain natural phenomena. They insist that something else of their own making must be added—namely, a divinely created directing program or a distinction between different kinds of entropy. Let us now look at several examples to see how creationists attempt to support their claims and to show that their notions are wrong and unnecessary.
The example of crystal growth is particularly interesting, because it has been misunderstood and misused by evolutionist debaters as well as by creationists. While the growing crystal is certainly an example of an open system in which entropy is decreasing, there is an important thermodynamical difference between it and a living system. In the crystal, the entropy is always at a maximum. In other words, while it is true that the entropy decreases as the liquid changes into a solid, this happens because the entropy capacity of the system decreases. The living system, on the other hand, contains an entropy deficiency, and this deficiency increases as the system grows or evolves. It should now be obvious that a debater who tries to draw too close a parallel between crystals and living systems will be in trouble.
Nevertheless, creationists have expended a great deal of effort attempting to explain the entropy decrease inherent in crystal growth. Elmendorf claims that there is no decrease in entropy, because liquids are more orderly than crystals (1978). When I pointed out to him in an exchange of letters that gases turn into liquids by a similar removal of heat, he decided that gases are the most orderly of all. I might have asked him why we observe changes of state in nature which proceed in the opposite direction by means of the simple addition of heat, such as snowflakes melting, however, I did not pursue the matter any further.
It is more interesting to examine the claim by both Elmendorf and Morris that crystals grow because of the divinely created directing program built into matter. Elmendorf simply tells us that "the molecules are pre-programmed," while Morris, with somewhat greater sophistication, explains that crystals are able to form only because of "the electrochemical properties of the molecules in the crystal" (1976). This quotation from Morris may sound perfectly reasonable (or should I say conventional?), but only because it is out of context. He subsequently informs us that these properties "could never arise by chance" or "within the constraints imposed by the second law," and finally concludes that they must be the work of "an omniscient programmer."
Two points should be noted here. First, Morris confuses the origin of matter and its properties with the process of evolution. This undoubtedly is done intentionally, since it is a common obscuring tactic among creationist debaters. Second, the divine programs built into matter are claimed to be capable of bringing about such entropy-reducing processes as crystal growth, development of a seed or egg into a mature organism, growth of populations, evolution of complex technologies, and so forth, but not capable of bringing about biological or even comparatively simple astronomical evolution. Creationists have nothing but contempt and ridicule for theistic evolutionists, an attitude made possible only by this severe inconsistency in their own belief system.
In their attempts to prove their version of the second law, creationists often use the example of a pile of bricks lying in the sun. This is supposed to represent an open system that, although it is receiving an abundance of high-grade energy, is not exhibiting any reduction in entropy. Creationists gloatingly draw our attention to the fact that such bricks have never been observed to organize themselves spontaneously into a building. What they apparently fail to understand is that under the given conditions, an entropy deficiency is in fact generated in the pile. After several hours of exposure to the sun, the temperature will be higher at the top than at the bottom. If we were to measure the temperatures throughout the pile, it would be a fairly simple matter to calculate the entropy deficiency. Useful energy could actually be extracted from the pile by means of a thermocouple, for example. Creationists should tell us where in this mundane pile of bricks we find the divine directing program and conversion mechanism, supposedly necessary for an entropy deficiency to be generated in the system.
Incidentally, this pile of bricks, absorbing heat at the top only, is an example of a system that becomes entropy deficient even though the entropy in the pile actually increases. This seeming paradox results from the fact that, as heat is added, the entropy capacity of the pile increases faster than the amount of entropy contained in it. If we began again with a uniform temperature throughout the pile and then allowed heat to be removed from the top, as when cooling at night, the entropy would in fact decrease in addition to an entropy deficiency again being generated. We may also note that in this case the cause is a loss of energy. When discussing crystal growth, we saw that a loss of energy produced a decrease in entropy, but not a deficiency. Almost any combination is possible and we have to be extremely careful in making general statements concerning entropy.
Other, more impressive convective systems, in which large entropy deficiencies develop spontaneously as a result of the simple influx of solar energy, are meteorological systems such as hurricanes, tornados, and lightning storms. And consider the water cycle: Heat from the sun evaporates water from the ocean; the vapor is carried over the land by winds, which are also generated by solar heat, and is forced up by mountains, where it precipitates; the water eventually forms rivers with waterfalls and finally flows back into the ocean to close the cycle. The waterfalls, of course, constitute a well-known source of available energy. Where, creationists, are the directing programs in these highly organized, entropy deficient systems?
Mutations and the Genetic Code
The growth of a seed or egg into a mature organism constitutes an observable process involving a large and spontaneous increase in the entropy deficiency of a localized system. Creationists naturally claim that the genetic code making this possible is just the directing program included in their Creative Trinity. It is certainly true that the genetic program determines just what the egg will grow into. But it is not true that this program is what enables the system to develop an entropy deficiency. In the course of a year, the earth receives 1.6 x 1021 watt-hours of energy from the sun and reradiates almost the same amount into space. But, because the incoming radiation originates on a high-temperature source (the sun) and the outgoing radiation on a low-temperature one (the earth), the whole process results in an outflow of entropy or inflow of negentropy. This negentropy flux can be calculated to be 3.2 x 1022 joule/ °K per year (Tribus and Mclrvine, 1971). A significant portion of this negentropy is used in biological processes directed by genetic programs, but a considerably larger portion is used to generate entropy deficient meteorological systems without the benefit of directing programs. Thus, the genetic program only insures that a small portion of the negentropy is used to develop a particular type of entropy deficient system. The only legitimate question left is whether the first bit of replicating genetic material could have come about naturally without violating the second law.
We may first note that all the information stored in a fertilized mammalian egg-cell is equivalent to only about 4 x 10-12 joule/ °K of negentropy. Ordinary everyday processes that we observe all around us spontaneously develop entropy deficiencies that easily amount to billions of times this amount. Thus, it is not the generation of the entropy deficiency that constitutes the problem, although this is what creationists imply when they say that a natural origin of the genetic code would violate the second law.
Experiments of the type first performed by Stanley Miller have shown that the basic building blocks of life—amino acids and nucleotides—are generated spontaneously in a reducing atmosphere, consisting of compounds of carbon, hydrogen, oxygen, and nitrogen, when energy in the form of electrical discharges or high-energy radiation is supplied. We are unable to choose at this time the particular mechanism whereby these units assembled themselves into proteins and DNA (or RNA) respectively; there are several possibilities. A more important question is the probability of the spontaneous formation of such a chain with sufficient autocatalytic properties so that, once formed, it would promote its own duplication. Once this hurdle has been overcome, evolution can be expected to proceed through the combination of mutations and natural selection, as discussed later. For years creationists have been indulging in calculations intended to prove that the formation of the original functional chain is statistically impossible. Let us examine one such attempt by Dr. Gish, also of ICR (1978).
Gish begins by assuming that a functional chain would need to consist of 100 amino acids of the 20 different kinds found in living organisms. He then states that there are 10130 different varieties of such a sequence, which is correct.
He then assumes arbitrarily and, he thinks, generously that 1011 of these variations might be functional. Stated more directly, he has assumed, entirely without justification, that only 1 out of 10119 combinations is useful. But, to show what an extremely generous man he is, Gish then assumes that 1021 varieties are formed every second during a period of 5 billion years. He is still perfectly safe, of course; with his assumption of 1 in 10119, the useful chain would never form. Gish doesn't mention whether anyone has systematically examined the properties of any significant number of such sequences. But even if thousands had been investigated, this would be nowhere near 10119, and it would be just as reasonable to assume that 1 in a trillion (1012), 1 in a billion (109), or even 1 in a million (106) has the desired characteristics. Actually, the evidence we have points in this direction. For example, examination of hemoglobins of different species shows that only 7 out of a total of 140 sites always have the same amino acid (Perutz, 1968). The probability of these 7 sites being correctly occupied, assuming again 20 different amino acids, is 1 in a little over a billion (1.3 x 109).
Now, if we go by what little evidence we have and make the far more reasonable assumption that 1 in 109 is functional, and assume further that only one sequence forms each second (anywhere on earth), a functional one could be expected to form in about 32 years! On the time scales we are dealing with, even 32 million years is nothing, so we too can be generous and assume that only 1 out of 1015 randomly generated 100-member sequences is sufficiently autocatalytic. Let us see Gish or anyone else prove this impossible!
Perhaps the greatest unanswered question in biological evolution concerns the manner in which proteins and DNA (or RNA) became associated with each other. Creationists maintain that because we don't now know how this happened naturally, it could only have happened through divine design, and it is useless to investigate it further. We are fortunate that such attitudes have not prevailed universally at all times or science would never have evolved out of the Dark Ages.
We may speculate on whether evolution could at one time have proceeded through mutations and natural selection involving chains of amino acids only, but in the present discussion we will leave aside these early developments, of which enough is not yet known. Let us look, instead, at the evolution of the genetic program from that of primitive organisms even simpler than (and different from) modern viruses, to that of complex ones such as mammals. Although we recognize the enormous amount of variation possible in the normal genetic mixing associated with sexual reproduction, the only way in which something entirely new can be introduced is through mutations, including such phenomena as gene duplication. Creationists contend that, because of the second law, only detrimental mutations are possible. An examination of the mechanism involved will show that this contention is absurd.
Four nucleotides constitute the characters in the genetic code, and, for convenience, they are designated A, C, G, and T in the case of DNA. They are read in groups of three called codons, each of which codes for an amino acid. A simple type of mutation is one in which one nucleotide is replaced by a different one, and, as a result, a different amino acid is coded for. (Because of a redundancy in the code, this does not always happen.) Since the genetic program has already been brought to near perfection through natural selection, a mutation is usually detrimental to the organism. It therefore tends to be weeded out of a population or, if it gives rise to a recessive gene, is limited in its spread. But there is, of course, no natural law which prevents an occasional mutation from benefiting the organism, especially if the latter exists in a changing environment. Such a mutation would tend to become more common and spread throughout the population. (An example is the acquisition of drug resistance on the part of asexually reproducing organisms, where variability due to genetic mixing does not play a part.) The important point here is that, as far as the second law is concerned, it makes no difference which nucleotide substitution occurs. The entropy content of the genetic message does not depend on whether the substitution turns out to be beneficial or detrimental to the organism.
We might profit from an examination of the fallacy that an accumulation of beneficial mutations would contradict the second law. It undoubtedly derives from the fact that, if such an accumulation were the result of a totally random process, it would indeed be contrary to the predictions of the second law. However, if each beneficial mutation is favored over an indifferent one, which in turn is favored over a detrimental one, then the process is by no means random, and we cannot invoke the second law to predict its outcome. The selective process just described is, of course, what we commonly refer to as natural selection.
In order for the complexity of the code to increase, a simple nucleotide substitution is not enough; instead, nucleotides need to be added to the existing sequence, perhaps through the process of gene duplication. Such an addition does constitute a minute negentropy increase, but, as we have seen, this does not at all violate the second law, since there will be a corresponding entropy increase elsewhere. In other respects, the addition is like the simple substitution discussed earlier; in particular, the entropy change in the genetic material is in no way dependent on whether the organism is helped or harmed, and the few beneficial mutations will be favored and accumulate, here adding complexity.
In their first and crudest attempt at creating the illusion of a contradiction between evolution and the second law of thermodynamics, creationists simply ignored the fact that evolving systems are not isolated. Their next endeavor consisted of altering the second law by maintaining that it precludes entropy decreases in all systems, not just isolated ones.
Although they still occasionally make either or both of these claims in debates, they apparently realized at some point, presumably after having been confronted with examples proving them wrong, that a new device was needed. So, they invented the "Creative Trinity." This actually replaces the second law, but they still refer to it as the second law of thermodynamics in order to maintain the air of scientific respectability.
There is a virtually unlimited number of examples of natural systems in which entropy deficiencies develop spontaneously, provided only that energy is allowed to flow across their boundaries, thus disproving the creationist requirement for a divine directing program or different kinds of entropy. We are awaiting coherent responses from creationists dealing with these examples.
This leaves only the task of examining the validity of the claim by creationists that genetic programs could not have developed naturally and must therefore have been intelligently created. A simple calculation of the probability of formation of a sufficiently autocatalytic chain of amino acids and an elementary examination of the process of evolution through mutations and natural selection from simple organisms to complex ones show that, whatever difficulties occur in the natural origin of life, they do not involve any violations of the second law of thermodynamics.