Chapter 5


[Part 3]

Of the dating methods which have been discussed, potassium/ argon dating of basalts is slightly in favor of a short chronology, rubidium/strontium dating is against a long chronology, and the others are not much help in either direction, with the exception of uranium/lead dating, in which the data reported by Gentry et al. strongly support a creationist position. This leaves us with thermoluminescence dating and its relatives (optically stimulated luminescence and electron spin resonance dating), fission-track dating, pleochroic haloes, amino acid dating, obsidian hydration, and the cosmic ray nuclides to consider.

Thermoluminescence dating depends on the fact that natural crystals have tiny imperfections which can hold electrons. When [163] a crystal is irradiated, the radiation sometimes knocks electrons into these “holes”. They then remain trapped until the material is heated, whereupon they jump back where they belong with the emission of a photon per electron. The method depends on knowing the dose of radiation (uranium, cosmic ray, potassium, etc.), the characteristics of the crystal, and the association of heating with the event to be dated (and no heating since). It is also technically demanding. Finally, there seems to be a plateau effect (when all the defects are filled with electrons, no further “aging” can occur), which would tend to blur the difference between an evolutionary and a creationist time scale. This method is primarily used in archaeology and is not much help with our problem. Interestingly, from an evolutionary standpoint, “the interrelation of the TL signals from meteorites in terms of radiation ages or terrestrial ages has not yet been solved.”113
113Geyh and Schleicher, p. 270, citing Sears DWG, Hasan FA: “Thermoluminescence and Antarctic meteorites.” In Annextad JO, Schultz L, Wanke H (eds): International Workshop on Antarctic Meteorites. LPI Tech Rep 86-01. Houston: Lunar and Planetary Institute, 1986, pp. 83-100. Sears and Hasan give Antarctic meteorite thermoluminescence values which are scattered over two orders of magnitude and are all within the range of values for modern (witnessed) meteorites. It is fascinating that the heat of entry into the earth’s atmosphere is not supposed to reset the thermolumincence that the meteorites acquired from cosmic rays in space.
114It is of interest that electron spin resonance dating of the flowstone reported by Geyh and Hennig in note 103 was too high by up to an order of magnitude, especially in the recent material.
A creationist explanation of these data is perfectly straightforward.

Optically stimulated luminescence dating and electron spin resonance dating are simply two more ways to measure the displaced electrons that are measured by thermoluminescence dating. They give us no additional information at this time.114

Fission track dating is based on the spontaneous fission of 238U. The constant for this decay is around 7 to 8 10–17/year (there is some uncertainty in the measurements and more uncertainty in the method of measurement), which is much less than the alpha decay constant (most 238U atoms decay by alpha decay). When a 238U atom fissions, the two major fragments fly in opposite directions with enough force to disrupt the crystalline or polymeric structure over a track of about 10 to 20 microns. These tracks can be seen if the mineral is ground and polished, then etched with a chemical solution such as hydrofluoric acid or [164] potassium hydroxide. It should then be simple to measure the 238U concentration, count the tracks, and relate the two by a formula. But there are several complications.

First, it seems that 238U is not always (in fact, is usually not) distributed uniformly in the crystal. In some cases starburst tracks can be seen apparently originating from 238U nodules. Secondly, the volume in which a given number of fission tracks is located is hard to determine (it is easy to measure the area but not the effective depth). So a surrogate for the 238U concentration is used which gets around both of these problems. With 238U, 235U is found in the standard ratio. The 235U can be caused to fission by irradiation with neutrons (this neutron-induced fission is the principle behind the atomic bomb). This reaction produces independent fission tracks which can be compared with the 238U fission tracks. With a relatively simple formula, the age can be determined from the two counts.115
115The formula is , where is the density of fission tracks, sf is spontaneous fission, if is induced fission, k238 is the alpha decay constant of 238U, ksf 238 is the decay constant for the spontaneous fission of 238U, 235 is the cross-section of 238U for thermal neutrons (5.802 10-22 cm2), and is the neutron flux in neutrons/cm2.
116 Where is the track density, m is the monitor, sf is spontaneous fission, and if is induced fission. Strictly speaking, the formula given in Faure, p. 346, col. 1 (20.13 and 20.14) is more accurate.

But this approach is not without problems. The decay constant for the spontaneous fission of 238U is uncertain, and makes the derived time imprecise. More importantly, the neutron flux is hard to determine, and varies from place to place in a reactor. So the preferred procedure at present is to compare the spontaneous and induced fission track densities on the sample rocks to those on several reference (monitor) rocks whose age is known. The age of the sample is found by t = tm (sf/if) / (msf/mif).116 Of course, if the zeta correlation, as it is called, is used, it makes the age of the sample depend entirely on the age of the monitor. This means that fission tracks when interpreted this way give only relative dates.

In addition, there are problems with annealing (track fading). Often there are not enough visible tracks to match the presumed age. To correct for track fading, samples are heated and [165] the spontaneous to induced fission track ratio is measured for various temperatures (or the same temperature for various times). The induced tracks, and the spontaneous tracks that have not faded, begin fading promptly at a critical temperature. Tracks that have been partially annealed do not begin increased fading until a higher temperature is reached, so the spontaneous tracks reach a higher ratio to the induced tracks before starting to fade. This gives a plateau age which can be correlated with other radiometric ages (for example, potassium/argon ages).117
117Geyh and Schleicher, p. 294, figure 6.96, citing data from Naeser CW Izett GA, Obradovich JD: “Fission-track and K-Ar ages of natural glasses.” US Geol Surv Bull 1980;1489:31.

At first glance the agreement that was reported by Naeser et al. appears to be a powerful confirmation of both fission track dating and the dating method with which it was compared, and a good argument for plateau ages. This is particularly true when it is realized that fission tracks in some minerals can be demonstrated to be reset under geologically reasonable conditions.118
118For example, see Storzer D, Wagner GA: “Correction of thermally lowered fission-track ages of tektites.” Earth Plan Sci Lett 1969;5:463-8.

However, several considerations appear to have been overlooked. First, one may remember that the use of glasses in potassium/argon dating is discouraged. To use them here because the data obtained fit a particular theory is opportunistic unless such use is further justified. Second, Naeser et al. used a value for the 238U spontaneous fission decay constant that was lower than the value found using determinations made on modern standards,119
119Thiel K, Herr W: “The 238U spontaneous fission decay constant redetermined by fission tracks.” Earth Plan Sci Lett 1976;30:50-6. The error is about 17%.
thus throwing off that beautiful correlation line by about 17%. And third, the precision of the method is only ± 10%, and even greater if the uranium distribution is not as uniform as expected. Frankly, the data in Geyh and Schleicher (following Naeser et al.) look too good for the precision of the method, raising the question of whether the data have been filtered. And it turns out that they have been. Naeser et al. in the original paper do not themselves believe that the data are as good as appears in Geyh and Schleicher’s treatment.120 [166]
120See Naeser et al., note 117, p. 13: “It should be noted that this procedure does not always yield concordant results (McDougall [sic], 1976, and this study).” (citing MacDougall JD: “Fission-track annealing and correction procedures for oceanic basaltic glasses.” Earth Plan Sci Lett 1976;30:19-26. Italics mine). I can find two glasses which did not give concordant dates that were left out of the graph. In addition, many more were not reported. See p. 3: “These natural glasses were chosen from a large group in our collection . . .” Other papers, such as Naeser CW, Fleischer RL: “Age of the apatite at Cerro Del Mercado, Mexico: A problem for fission track annealing corrections.” Geophys Res Lett 1975;2:67-70, suggest that annealing does not always provide the answers expected.

Perhaps it would be worth quoting the abstract from Naeser and Fleischer (written 5 years after Naeser et al., note 117):

Fission-track dating and K-Ar dating indicate that the age of apatite from Cerro de Mercado, Mexico, is 30 m.y., in contradiction to previous corrected fission track ages of 40 and 57 my. by other works. Annealing data for the “plateau method” correction of fission-track ages for the Cerro de Mercado apatite lead to corrections by a factor of sixty or more, which give geologically unreasonable ages. In addition, published data concerning the length of fission tracks and the annealing of minerals imply that the basic assumptions used in an alternative procedure, the length-reduction-correction method, are also invalid for many crystal types and must be approached with caution unless individually justified for a particular mineral.
121Geyh and Schleicher, p. 287.
122Geyh and Schleicher, p. 293.

Fission-track dating is not an easy way to date fossiliferous formations. Problems in the method are such that “the quality of the data is very dependent on the skillfulness and experience of the laboratory staff.”121 But even with a good laboratory, “The main disadvantage of the FT method is that the results are often difficult to interpret in terms of actual ages.”122

Why does it take a skillful and experienced laboratory staff? The procedure itself is not hard. One simply irradiates half of a sample with neutrons (expensive but not technically difficult), mounts the samples on epoxy, then puts them in the appropriate reagent for a specified time. There can be up to 50% variation in the time needed in the reagent without significantly altering the results.

The problem is in identifying the tracks. Theoretically, if all the tracks were the same size and shape and there were no confusing structures, the job should be easy. But there are such structures.123
123For example, to finish the quote from Naeser et al. in note 120, “These natural glasses were chosen from a large group in our collection because they appear very fresh and because they are essentially free of microlites and crystallines, which can make fission-track dating impossible. Many obsidians are crowded with microlites and crystallines (gobulites and trichites), and these form fission-track-like etch pits following etching with hydrofluoric acid. The etch pits of the microlites and crystallines are difficult to separate from real fission tracks formed from the spontaneous decay of 238U, and accordingly, calculated ages based on counts including the microlite and crystalline etch pits are not reliable.”
124I have not run across any photographs of partially annealed tracks in geological materials in the literature. I am assuming that the annealed fission tracks are not all uniformly decreased in size, like the photograph in Storzer and Wagner (see note 118, p. 465). If they are 99+% the same size, the only model that will fit the data is that of accumulation of fission tracks without annealing followed by a relatively sudden heating episode <1% of the putative age ago. That is not usually a reasonable evolutionary model, and the fission tracks would actually be evidence against the evolutionary time scale in that case. Or perhaps these pits are not fission tracks at all, in which case they are totally irrelevant to our question. Perhaps only the normal-sized tracks should be counted, or perhaps the difficulties of separating fission tracks from other artifacts in glasses are so great that the method cannot give meaningful data.
125See MacDougall in note 120.
126See Gale in note 84.
And the very fact that some minerals have partially [167] annealed tracks says that the fission tracks themselves vary in length and width (older tracks should be more annealed than younger tracks).124 The fission tracks do not come labeled as such. So it takes an experienced eye to pick them out. How do we know that the eye is experienced? Because it gives us the “right” dates. We again have circular reasoning, unless the data can be obtained by individuals who know nothing about the presumed age of the rock.

Even with the plateau correction technique, most fission track dates are too young, even for specimens that have remained (as far as we know) below room temperature over their entire history.125 Inherited tracks are a concern in some minerals.126 There is great need for more review, and for more data, particularly blinded data; with all observations fully reported.

Even if fission-track dating presents a major problem to evolutionary theory, a creationist should not feel comfortable about the results at this time. Fission-track dating is theoretically able to disprove a creationist time scale. And there are enough old dates to make a creationist uncomfortable. So we will look at the difficulties for a creationist in more detail.

In order to disprove a short chronology, several conditions must be fulfilled. First, the etching pits must be shown to be due to fission tracks and not to other imperfections (this will be difficult in glasses). Second, the mineral must be shown to have been either formed, or heated to a specified minimum temperature for a specified minimum time (so there are no inherited tracks), at or following the time of the dated event. Third, the deposition should be clearly correlated with the existence of life. Fourth, the dating [168] must be absolute and not relative. The zeta correction technique is not adequate without a securely dated standard. (In addition, fission tracks from neutron-induced fission of 235U must be negligible, although in the average granite this works out to ª0.01% of the 238U tracks.) I have not had time to review all the literature, but I haven’t yet seen the “smoking gun” proving a creationist time scale wrong. The literature should have a thorough review from this standpoint. Incidentally, it would be fascinating to date secondary minerals, perhaps such as aragonite in corals or calcite in brachiopods, or perhaps petrified trees, with the fission-track method.

Pleochroic haloes are fascinating from the standpoint of the age of the earth. They can be found in multiple minerals, but the easiest to use is biotite mica, because it can be easily split into thin flakes. Pleochroic haloes are formed by alpha particle damage, similar to fission fragment track damage. Since alpha particles have higher initial speeds and are less positively charged than fission fragments, they produce minimal track damage until they slow down near the end of their range. At this point they create maximum damage. Thus if one splits a crystal at the point of a uranium inclusion one finds rings around the inclusion which are lighter at the center and darker at the edge. These rings are darkened in proportion to the radiation damage. The diameter of a ring is proportional to the energy of the alpha particles producing the ring. Generally, more stable isotopes emit alpha particles of lower energy and thus make smaller rings, whereas more unstable isotopes emit alpha particles of higher energy and make larger rings. A mineral inclusion of uranium which has all the daughter products present will create a series of concentric rings, with a size characteristic of the mineral.

It would be easy to measure the discoloration of the rings around an inclusion, measure the discoloration of a known alpha dose,127
127Or according to Faure (p.355) one can calibrate by “studying halos in other specimens of that mineral of known age.” This would immediately involve us in circular reasoning unless the true age of the “known age” mineral was indeed known.
measure the uranium and thorium content of the inclusion, and calculate the age. However, this method “is hardly used anymore because of the thermal instability of the haloes and the fact that very old samples often reach a saturation point . . . and [169] thus yield apparent ages that are too small. The intensity of the coloring may even decrease when saturation is exceeded.” “. . . many geologically unacceptable results have been obtained with this method.”128
128Geyh and Schleicher, p. 299.
129Faure, p. 355.
“This method has been abandoned in favor of the isotopic dating methods.”129

But is it really that unreliable? Notice that the above results suggest a younger age than the “geologically” acceptable one. One straightforward interpretation would be that at least some pleochroic haloes support nearly contemporaneous formation of some igneous and metamorphic rock formations, much younger than usually assumed. Perhaps the data should be reviewed. It just might be compatible with a creationist time scale.130
130The literature on the subject is apparently not large, but mostly old and published in obscure journals. In addition, many of the dates are Precambrian and do not bear directly on our question, such as those in Deutsch S, Kipfer P, Picciotto E: “Pleochroic haloes and the artificial coloration of biotites by a particles.” Nuov Cim, series 10, 1957;6:796-S10. Some of the literature does not give good geologic characterizations of the specimens (for example, Joly J: “Pleiochroic halos of various geological ages.” Proc Roy Soc London A 1923;102:682-705, where the Devonian halo-containing rocks are identified simply as “Co. Carlow (Ballyellen) mica”, without a word as to the geological environment, and similarly “Tertiary mica of the Mourne granite” on p. 695).

To falsify a creationist time scale, one would have to show that pleochroic haloes were found in rocks which had either formed or been heated enough to erase the previous haloes at or after their association with fossils, and that the halos could be reliably dated to well beyond a reasonable creationist time frame. This is indeed theoretically possible. It appears that pleochroic haloes can be erased in mica at 500-705° C (Holmes A: “The age of the earth.” Bull Nat Res Council (U.S.) 1931;80:159-96, esp. p. 188). In addition, fossils such as petrified trees might produce haloes which would falsify the creationist time scale (the only evidence of this kind that I know of is in favor of the creationist time scale. See Gentry et al., note 93).

Of course, any evolutionist who tries this will also have to deal with the published work on polonium haloes (conveniently collected in Gentry RV: Creation’s Tiny Mystery. 3rd ed. Knoxville, TN: Earth Science Associates, 1992). There are no good mechanistic evolutionary explanations of polonium haloes. The argument that they must be from uranium decay products ignores the evidence from the “uranium-poor White Mountain (New Hampshire) granites” cited on p. 332-3 (reprinted from Gentry RV: “Response to ‘Radioactive halos: Geologic concerns.”’ Creat Res Soc Quart 1989;25:176-80). I don’t imagine that any evolutionist relishes the job of making coherent sense of all the important data within standard evolutionary theory.

Amino acid dating is based on the fact that all amino acids except glycine have an asymmetric carbon (threonine and isoleu-[170]cine and the secondary amino acids hydroxyproline and hydroxylysine have two), which can come in either right-handed (D) or left-handed (L) forms (see chapter 2). These amino acids are all found in only one form (the L form) in living organisms (with rare exceptions like D-alanine in some microbial cell walls). These amino acids slowly transform from L to D forms (and back again) randomly. If the transformation constant is known, amino acids may be used like radioactive isotopes for dating.

The obvious disadvantage of this dating method is that several enviornmental influences such as acidity (pH) and temperature influence the transformation “constant”. The influence of temperature is particularly striking. A 1° increase in temperature increases the “constant” by 25%.131
131Geyh and Schleicher, p. 350.
Thus “constants” must be calibrated by other methods. “Ages that are not based on such site-specific calibrations can deviate by several orders of magnitude from the actual ones (Dungworth 1976).”132
132Geyh and Schleicher, p. 350, citing Dungworth G: “Optical configuration and the racemisation of amino acids in sediments and in fossils—a review. Chem Geol 1976;17:135-53.
This inaccuracy is enough to blur the difference between the creationist and evolutionary time scales.

One of the apparently repeating phenomena noted in amino acid dating is the decreasing of the racemization “constant” in older specimens.133
133For example, see Geyh and Schleicher, p. 347, fig. 8.2. Also see Dungworth in note 132. On p. 149 he states, “The results for modern bone and that of Aztec origin are in excellent agreement; 700 years age disparity disclose no noticeable difference in the magnitude of the rate constants. . . . In Mammoth bone there is a distinct decrease in the magnitude of the rate constant, while the much older deer and walrus bones, of Pleistocene age, display rate constants which are about one order of magnitude less than those in modern bone. The implication is that the rate of the racemisation reaction is decreasing with time.” In fact, on p 140 he reports Jurassic (conventional age 180 million years) material which still has a considerable excess of L-form amino acids. This is quite surprising from an evolutionary perspective.
This is particularly interesting in view of the fact that a straightforward interpretation of the creationist time scale largely eliminates the crookedness of this curve. Several very “old” samples still have significant amounts of L-amino acids and are not in equilibrium. In fact, this trend toward smaller “constants” in “older” samples is true for the entire literature.134 [171]
134See Brown RH: “Amino acid dating.” Origins 1988;12:8-25, which gave an exhaustive survey of the available literature.
135The problem has to do with the way of averaging. We usually average by adding the items and dividing by their number. If there are 2 objects, one being 3 meters high and one 5, their average is (3+5)/2, or 4 meters. But if you go 30 kilometers/hour for 90 kilometers and 90 kilometers/hour for 90 kilometers, your average speed is not 60 kilometers/hour but [(303)+90]/4 or 45 kilometers/hour. This is because you spend 3 hours traversing the 30 km/h stretch and only 1 hour on the 90 km/h stretch. In fact, there is a puzzle which says, “If you want to average 60 miles per hour over a 60 mile road and you spend the first 30 miles going 30 miles per hour, how fast do you have to travel the remaining 30 miles?” The answer is not 90 or 120 miles an hour, but at infinite speed. That is, you can’t make it. You have already used up the hour during which you should have finished the 60 mile trip.

In the same way, if a racemization “constant” for a 3,000 year old sample is 6.93 10-4/y, so that its “half life” is 1,000 years, and a racemization constant for a 4,000,000 year old sample is 6.93 10-7/y so that its average “half life” is 1,000,000 years, the discrepancy is even greater than it appears. For if the environment of the 6 million year old sample was the same as that of the 3 thousand year old sample for the last 3000 years, then it had at least a two-stage history. In the last stage it lost 7/8 of its leftover unmatched L-amino acids over 3000 years, and in the first stage it lost 1/2 of its unmatched l-amino acids over an approximately 4,000,000 year period (3,997,000 to be exact), for a constant of 1.73 10-7/y. So if one calculates the early “constants” for the older samples, they are even smaller, and in some cases vanish entirely! (a 20,000,000 year old sample and a 20,000 year old sample with the same D/L amino acid ratio in the same environment would imply a racemization constant of zero for all but the last 20,000 years, or else an erroneous time scale.)

136One of the problems with using amino acid racemization as a dating technique is that the racemization “constant” not only varies with temperature, acidity, etc., but also with the position of the amino acid residue within the peptide. In general, the residues on the end of peptides are easiest to racemize (have the highest “constants”). However, there are strange cases such as isoleucine which racemizes easiest inside the protein (Kriausakul N, Mitterer RM: “Isoleucine epimerization in peptides and proteins: Kinetic factors and application to fossil proteins.” Science 1978;201:1011-4). Thus instead of a simple exponential relationship between racemization and time one has a complex curve, depending on how much of the protein has hydrolyzed. There are even cases of the apparent reversal of racemization (Kimber RWL, Griffin CV, Milnes AR: “Amino acid racemization dating: Evidence of apparent reversal in aspartic acid racemization with time in shells of Ostrea.” Geochim Cosmochim Acta 1986;50:1159-61; Kimber RWL, Griffin CV: “Further evidence of the complexity of the racemization process in fossil shells with implications for amino acid racemization dating.” Geochim Cosmochim Acta 1987;51:839-46), so that older specimens may look younger than younger specimens subjected to the same conditions. Specimens may also appear to age suddenly. Obviously, this makes the method as presently done incapable of proving anything, and even suggestions must be tentative.

In addition, it is possible that amino acids have a preferred chirality (handedness) at equilibrium when in the middle of a protein. In that case the equilibrium mixture may not be a 1:1 D to L ratio, but something greater or (more likely) less than this. Thus a significant excess of L-amino acids may not necessarily prove that the protein is less than some number of years old. The Jurassic amino acids cited in note 133 may not be younger than the 2 million year limit given by Kimber and Griffin (ibid.). This is why I have not given great weight to the above analysis.

137The equation used is d2 = Ate–E/RT, where d is the thickness of the layer A is a frequency factor, E is the activation energy, R is the universal gas constant, and T is the “effective hydration” temperature.
138Contrary to the claim that “Moisture content and pH of the surrounding environment seem to have no influence.” (Geyh and Schleicher, p. 362).
139Geyh and Schleicher, p. 366.

The decrease is almost linear. In fact, the decrease approaches asymptotic if the effect of more recent racemization is considered.135 The chemistry behind this phenomenon is not easily understood when viewed from an evolutionary perspective. Deep (older) sediments would be expected to be hotter, not colder, than surface ones. However, a creationist explanation eliminates this problem. This would seem to indicate that a creationist time scale is more in accord with the amino acid data than an evolutionary one. But the uncertainties inherent in the method undermine its validity as an argument for anything.136 [172]

The obsidian hydration method is based on the idea that freshly broken obsidian hydrates at its surface at a rate that is completely dependent on the type of glass, the temperature, and the time.137 Because the thickness of the hydration layer varies with the square root of time, older dates are inherently subject to more inaccuracy than younger dates. Also, pressure, water, and solutes must have something to do with the rate138 because the experimental hydration rates are determined in a pressurized reaction vessel with deionized water. In actual application, “dating errors may occur when the obsidian artifact has been subjected to heat because this changes the rate of hydration. Moreover, erosion of the surface, which is difficult to detect, also leads to incorrect ages.”139 Note that erosion would produce too young ages (since it is difficult to detect, one can simply throw out young ages that one does not like simply by claiming that there was erosion). In fact, erosion turns into peeling at 40-50 microns, or less with mechanical or heat stress. So there is an absolute limit to the dating accuracy.

With these problems in the method, I do not know of any data that cannot be explained by either a creationist or an evolution-[173]ist time scale. There are too many fudge factors. Obsidian hydration is not much help in determining the most likely time scale.

Now we come to our last group of dating methods, those involving radioactive isotopes produced by cosmic rays. They are 53Mn, 36C1, 81Kr, 129I, 26A1/10Be, and of course, 14C. Of these, all but the last are of limited application. Carbon-14 dating deserves its own section, so we will start with 53Mn.

The 53Mn method is dependent on the constant production of Manganese-53 from iron by cosmic ray bombardment, almost exclusively in meteorites. After the meteorites reach the earth, the production of 53Mn essentially ceases. It then decays with a half life of 3.7 ± 0.6 106 years. Thus if one knows the initial activity and the activity now, one can calculate the time required. However, the initial activity is difficult to determine. If we knew that all meteorites were saturated with respect to the radioactivity induced by cosmic rays when they entered the atmosphere, we could guess at the terrestrial age of a given meteorite. But “cosmic ray ages” are obtained from meteorites also, which implies that at least some meteorites are not saturated with respect to 53Mn. Thus a meteorite that appears to have an old terrestrial age may simply have a young cosmic ray age instead.

Another complication is the fact that what is measured is not absolute 53Mn concentration, but the 53Mn/55Mn ratio. Suppose a meteorite started out with an inhomogeniety in the distribution of iron and manganese. Since the 53Mn is mostly produced from iron, the 53Mn/55Mn ratio would vary with location. The theoretical uncertainties in the method are such that dating either meteorites or meteorite dust in ice cannot be done with confidence. And for dust dates the possible interference from 53Mn-free manganese from volcanoes would have to be taken into account.

From a creationist point of view it would be interesting to try dating Paleozoic and Mesozoic meteorite fragments. This will never be done in a scientific community dominated by evolutionary theory, but could provide evidence for a young earth. But it would not provide incontrovertible evidence for creationism, because of the uncertainties noted above. I also do not know if the appropriate meteorites have been found.

The 36Cl method is dependent on the production of 36Cl from 36Ar by a neutron-proton reaction, and to a lesser extent from 40Ar by spallation, from 40Ca, and from various potassium species. It is also produced underground by neutrons from 35Cl. The half-[174]life of 36C1 is 3.01 ± 0.04 105 years. Unlike 14C, which is widely distributed, 36C1 is concentrated at latitude 45° N and S. The 36Cl/Cl ratio can vary by 6 orders of magnitude. 140
140Geyh and Schleicher p. 197.
The uncertainties in initial levels of 36Cl and the probability that such levels would have been disturbed by a Flood make the method unsuitable as evidence for or against a long chronology.

The 81Kr method is dependent on the production of 81Kr from Rb, Sr, and Zr in meteorites by spallation. Apparently, the supply of 81Kr on earth is largely from meteorites. Its half life is 210,000 years. It is used primarily to date meteorites. Like 53Mn, it is used to measure both cosmic ray ages and terrestrial ages, and the two are mutually exclusive (it is interesting that it is apparently not driven off by the heat produced by entering the earth’s atmosphere). The method can be safely ignored in our discussion.

The 129I method is dependent on the production of 129I from 129Xe by cosmic rays, and by uranium fission and muon bombardment of tellurium ores underground. Its half life is 15,700,000 years. It is used to find cosmic ray ages of meteorites and to date tellurium ores. The variables behind this clock and the difficulty being sure the clock is reset make the 129I method of little use deciding our question.

The 26Al/10Be method is dependent on the production of the respective isotopes by cosmic rays, the former apparently from argon and to a lesser extent from silicon and stable aluminum, and the latter from nitrogen, oxygen, and carbon. The production rates of the two isotopes are assumed to be proportional to each other so that the difficulties caused by their uneven production around the world can be ignored (neither the 10Be method nor the 26A1 method were considered more than experimental by Geyh and Schleicher). 141
141The possible accretion of 26Al from cosmic dust, of concern to some earlier investigators (for example, Amin BS, Kharkar DP, Lal D: “Cosmogenic 10Be and 26Al in marine sediments.” Deep-Sea Res 1966;13:805-24) is apparently ignored.
The method is used for dating ice, sediments, coral, manganese nodules, and “oceanic particulate matter”, although the assumptions behind the latter dates seem staggering to me. The assumption of equal deposition of 10Be and 26Al during and immediately after a Flood seems strained, so from a creationist standpoint the method would seem to lack validity. This method again seems not to offer much help in answering our question. [175]

To summarize, potassium/argon dating of basalts is in favor of a short chronology. Other potassium/argon dates are easily explainable from a creationist perspective, except for evaporites where there are problems for both interpretations. On rubidium/ strontium (and potassium/calcium) dating the evidence strongly favors the creationist time frame. The data on uranium/thorium/ lead dating is moderately in favor of the creationist position; if one trusts the Gentry data it nearly excludes the evolutionary time scale. On the other hand, the evidence on fission track dating is slightly in favor of the evolutionary time scale, although not coercive. The other methods are simply not helpful enough, although some of them, such as amino acid dating, lean toward the creationist position. All this is true without altering the radioactive time constants.

This will come as a surprise to many. Many have not even considered creationism to be a valid scientific option, let alone the most scientifically defensible one. But there is an even bigger shock in store. The next dating method we will examine, carbon-14 dating, almost mathematically eliminates the evolutionary time scale and almost mandates some kind of creationist time scale. We will examine that evidence now.

Carbon-14 Dating

Carbon- 14 dating is based on the production of 14C in the atmosphere by cosmic rays interacting with 14N (nitrogen).142
142Cosmic rays actually produce little or no 14C directly. Rather they release neutrons which react with 14N (nitrogen) to produce 14C and 1H. To a much lesser extent the neutrons react with 17O (oxygen) to produce 14C and 4He, and with 13C directly to produce 14C. The only other natural process that produces 14C (outside of meteorites) that I have seen considered in the literature is the reaction of high energy 4He nuclei (alpha particles) with 11B (boron) to produce 14C and 1H. This process is unusual even underground, and is practically nonexistent in the atmosphere because of the extremely small amount of boron there.
143 For decay counting the ratio is usually given in decays per minute per gram of carbon, which relates directly to the 14C/C ratio. This is why I have chosen to refer to the 14C/C ratio in the text, even though this is not the precise ratio measured in all methods of radiocarbon dating. For decay counting, the 13C/12C ratio is also measured by mass spectrometry to correct for isotopic enrichment effects. For accelerator dating the amounts of 14C, 12C, and 13C are measured and the 14C/12C and 13C/12C ratios are calculated. These measurements can be used directly to calculate the radiocarbon age. The 14C/C ratio determined by decay counting can be compared with the ratios obtained by accelerator dating straightforwardly: It essentially equals the 14C/(12C + 13C) ratio, which is only about 1% lower than the 14C/12C ratio and is proportional to it to within the limits of the measurements.
144Levin I, Münnich KO, Weiss W: “The effects of anthropogenic CO2 and 14C sources on the distribution of 14C in the atmosphere.” Radiocarbon 1980;22:379-91.
145When reading the literature one has to keep in mind that many references use a half life of 5568 years because of an old inaccurate measurement of the half life. To keep the literature consistent, “radiocarbon years” are usually given in terms of the older (shorter and less accurate) half life. We will follow this procedure here. There is only a 3% difference between the two half lives.
The production rate is nearly constant at the present time. The 14C produced is rapidly turned into 14CO2, which mixes in with regular CO2 to form (before modern industrial society) a 14C/C ratio143 of 1.2 10-12. The mixing is very efficient; within 10 years [176] of when atomic bomb tests doubled the amount of 14C in the northern hemispheric atmosphere, it was thoroughly mixed throughout both hemispheric atmospheres.144 The 14C decays back into 14N by beta decay, with a k of 1.21 10-4/year, corresponding to a half life of 5730 ± 40 years.145 The 14C/C ratio in the biosphere (excluding the deep ocean regions) has remained nearly constant through the last few thousand years, thus providing the basis for measurement of the age of various carbon-containing substances. Since living things are made largely of carbon compounds and water, this method has the advantage of directly dating plant and animal remains.

Like the other methods we have considered, the 14C method depends on assumptions. For 14C dating, the assumptions are:
1. The decay constant of 14C is invariant.
2. The 14C/C ratio in the biosphere has remained constant.
3. The dated object was in equilibrium with the biosphere at time t0.
4. The dated object has not gained any carbon since time t0.
5. We can measure the present 14C/C ratio in the object.
If these assumptions are correct, then

14C/C = (14C/C)0 e–kt,
and with a little calculus we get
t = ln [(14C/C)0 / (14C/C)] / k.
Graphically the dating method can be represented by the following:


To find a radiocarbon age, one measures the 14C/C ratio in a sample, finds it on the right edge of the graph, and follows an exponential curve to the left until it intersects the “present” 14C/C ratio in the biosphere. That point gives the radiocarbon age.

(If assumption 3 is valid, the specimen was in equilibrium with the biosphere, and if assumption 2 is valid, the biosphere had the same 14C/C ratio that we find today when the specimen was last in equilibrium with the biosphere. Combining the two assumptions, the point at which we would expect to equal the present 14C/C ratio is the point at which the specimen was last in equilibrium with the biosphere.)

Again we will assume (in fact, in this case insist on) the invariance of assumption 1. Assumptions 3 and 4 can be violated, but for our purposes we will assume that we have carefully chosen samples which were in equilibrium with the biosphere146
146There is a small isotope fractionation effect. One can compensate for this by measuring the 13C/12C ratio. The difference in 13C/12C ratio between ones sample and a standard is almost exactly half the correction needed for the 14C/C ratio. This effect is rarely larger than 3% and is insignificant for our purposes.

Meteorite dating using 14C is not directly comparable to conventional 14C dating. In meteorite dating, the assumption is made that 14C is made by cosmic rays at a constant rate in a given type of meteorite, and that this production essentially ceases when the meteorite lands on the earth. Thereafter the 14C decays exponentially as does the 14C used in conventional radiocarbon dating (see Sears and Hasan in note 113, and Kigoshi K, Matsuda E: “Radiocarbon dating of Yamato Meteorites.” In Annexstad et al., see note 113, pp. 58-60).

The 14C found in meteorites is apparently largely produced by the interaction of fast (>10 Mev) neutrons with 16O [16O (n,2pn) 14C], so stony meteorites, which have more oxygen, have higher concentrations of 14C than iron meteorites. The amount of carbon in meteorites is variable, as is its ratio to oxygen, and so the 14C/C ratio is useless in dating meteorites. What is used instead is the amount of 14C per gram of meteorite.

Unfortunately, not all meteorites, even of the same general type, have the same concentration of 14C. Some reports in the literature give the impression that the variation in 14C concentration in recent falls is narrow. For example, Suess and Wänke (Suess HE, Wänke H: “Radiocarbon content and terrestrial age of twelve stony meteorites and one iron meteorite.” Geochim Cosmochim Acta 1962;26:475-80) give a range of 37 to 56 decays per minute (dpm) per Kg of meteorite for their stony meteorites and 5.5 dpm/Kg for their iron meteorite. However, a different range is given in Goel and Kohman (God PS, Kohman TP: “Cosmogenic carbon-14 in meteorites and terrestrial ages of “finds” and craters.” Science 1962;136:875-6), namely 47-78 dpm/Kg and 1.64-1.80 dpm/Kg respectively Boeckl may describe the situation more accurately (Boeckl R: “Terrestrial age of nineteen stony meteorites derived from their radiocarbon content.” Nature 1972;236:25-6). His range is 36-108 dpm/Kg for stony meteorites. This is quite a wide range and would give an uncertainty of approximately 9000 years in the calculated age of a find (a meteorite whose fall was not witnessed). It appears that meteorites, and even stony meteorites, have a wide range of 14C concentrations when they fall. The range may be even wider than that given by Boeckl.

One is tempted to make the data more precise by measuring the 14C/O ratio. However according to Boeckl (p.25), “Finds usually show substantially higher oxygen values than falls, a fact which can be attributed to weathering.” This would tend to decrease the 14C/weight ratio, and to decrease the 14C/O ratio even more. The 14C dating of meteorites needs more study before it is helpful in our discussion. Its precision is not comparable to that of conventional 14C dating.

and [178] have since not gained (loss does not matter) any carbon from their environment. It is usually fairly easy to satisfy these requirements.

Assumption 5 deserves some discussion. It turns out that initially it was difficult to measure the 14C/C ratio in various samples. If carbon was measured as a solid, the total C was easy to measure; one simply weighed the carbon. But the 14C was hard to measure; the beta decay of 14C in solid carbon could occur in any direction, including into the rest of the sample or into the support, which meant that an uncertain and untestable factor had to be added into the equation. So decay counting is now done by using a carbon-containing gas like carbon dioxide, methane, or acetylene, or sometimes by liquid scintillation counting.

But gas decay counting and liquid scintillation counting have one major drawback; they detect not only 14C decays, but also background radiation. And the background radiation is high, swamping the 14C decays. Some of this background is from radon. The radon can be eliminated by allowing the sample to stand until essentially all the radon has decayed. Some background comes from neutrons, which can largely be absorbed by surrounding the chamber with paraffin and boric acid. But most of the background is produced by cosmic rays. One can shield against these by using steel and/or lead shielding. One can also ignore them by the use of anticoincidence detectors. [179]

In an anticoincidence system the sample counter is surrounded by other counters. If a cosmic ray hits the sample counter, chances are very good that it will hit one or more of the other counters at the same time. A computer can be told to ignore discharges that happen simultaneously in the sample counter and one or more of the other counters, and to count only discharges that occur in the sample counter alone.

Using these methods, one can get back to around 30,000 radiocarbon years, or about 1/40 of the present 14C/C ratio (2.5 percent modern carbon or pmc). With special shielding deep underground and long counting periods it is possible to extend the range to 50,000 radiocarbon years (0.2 pmc).147
147Some laboratories (for example, Grootes PM, Mook WG, Vogel JC, de Vries AE, Haring A, Kistemaker J: “Enrichment of radiocarbon for dating samples up to 75,000 years.” Z Naturforsch 1975;30a:1-14, Grootes PM: “Carbon-14 time scale extended: Comparison of chronologies.” Science 1978;200:115, and Stuiver M, Heusser CJ, Yang IC: “North American glacial history extended to 75,000 years ago.” Science 1978;200:16-21) are apparently able to obtain dates in the neighborhood of 60,000 radiocarbon years without enrichment, and 75,000 radiocarbon years with isotopic enrichment techniques.
Decay counting also requires about 5 to 10 grams of carbon, which may call for bone samples of as large as half a kilogram. Thus there was considerable interest when it was discovered that 14C ions could be separated from all confounding ions by a device called a tandem accelerator mass spectrometer, or AMS for short. The figure illustrates the operation of an AMS.

A beam of negative carbon ions is formed by negative cesium ions striking a carbon target. These ions may be partially separated according to mass by use of an analyzing magnet (not shown [180] in the figure). They are then attracted to a positive electrode, where they are stripped of part (or all) of their electrons by either a foil or high-pressure gas. As positive ions they are then repelled from the electrode to very high velocities. These high energy ions are formed into a beam and sent through a magnetic field which separates them by their charge-to-mass ratio. A very specialized target is used for 14C, which measures the amount of energy a given particle gives up traveling a definite distance through a semiconductor, and also measures the total energy of the particle. Sometimes the time of flight (and therefore the speed) of the particle is also measured. This gives a unique identification for 14C. Cosmic rays, 14N atoms, or other background factors should not be able to mimic 14C atoms in this detection process. The prediction was repeatedly made that the machine background would be zero.148
148To be precise, the prediction was greater than 100,000 radiocarbon years range in Muller RA: “Radioisotope dating with a cyclotron.” Science 1977;196:489-94; less than 1 count per run (50,000-60,000 radiocarbon years) in Nelson DE, Korteling RG, Scott WR: “Carbon-14: Direct detection at natural concentrations.” Science 1977;198:507-8; less than 1 count per day in Doucas G, Garman EF, Hyder HRMcK, Sinclair D, Hedges REM, White NR: “Detection of 14C using a small van de Graaff accelerator.” Nature 1978;276:253-5; and greater than 70,000 years in Bennett CL, Beukens MR, Clover MR, Gove HE, Liebert RB, Litherland AE, Purser KH, Sondheim WE: “Radiocarbon dating using electrostatic accelerators: Negative ions provide the key” Science 1977;198:508-10. There were some objections, but these tended to be centered on the difficulty of preventing contamination of the samples.
This made it theoretically possible to date very old samples, of the order of 100,000 radiocarbon years (0.0004 pmc). And at the same time it meant that the machine should be able to give dates on milligram-sized samples, as every decay per minute represents some 400 billion carbon atoms. The AMS method is also much faster (minutes versus hours of counting time) than the decay counting method.

The AMS development was particularly interesting from the creationist point of view. It made possible the testing of a creationist prediction that was incompatible with any evolutionist prediction, but which seemed mandatory from any creationist view except that of a gradually decreasing decay constant for radioactivity (which, as we have noted above, is nearly completely parasitic on evolutionary predictions). That prediction is that there should be measurable 14C in all fossil carbon.

To understand the importance of this prediction we should [181] first note that 14C dating is arguably the most important dating method in establishing the evolutionary time scale. First, using a rather straightforward interpretation of 14C ages, it gives dates that are compatible with the evolutionary time scale in the vast majority of cases. Second, it can be quantitatively tested on recent material and has passed that test repeatedly. So a creationist cannot simply disregard the method entirely (as is sometimes done for the potassium/argon method, for example). He or she has to explain why it works well for recent samples but not for older material.

A good creationist model for radiocarbon dating would seem to have to start by acknowledging that our assumptions 3-5 can be reasonably fulfilled in many kinds of organic material. Altering assumption 1 without a reason would seem to be an ad hoc solution and thus should be discouraged, at least at present. So we are left with alterations in assumption 2.

What could disturb the 14C/C ratio in the biosphere? If one interferes with the transport of 14C from where it is produced to the earth’s surface, one will only decrease the amount of 14C by the amount that decays on the way down, which in 100 years (a long time by meteorological standards) would be only about 10% with the most favorable assumptions (and probably closer to 1%), not nearly enough to account for the difference between the creationist and evolutionary time scales. One cannot vary the nitrogen content of the atmosphere much. Cosmic ray flux could conceivably be decreased by a stronger magnetic field on the earth, but the maximum reasonable effect would be only to drop the 14C concentration by a factor of four,149
149 Brown RH: “The interpretation of C-14 dates.” Origins 1979;6:30-44.
and its actual effect would probably be less.

But increasing total carbon, the denominator of our ratio, has been demonstrated to give lowered 14C/C ratios and falsely elevated 14C dates. In the late 1800’s, with the increasing use of fossil fuels, particularly coal, there was a marked (~5%) decrease in the 14C/C ratio, which is called the Suess effect after its discoverer.150
150Suess RE: “Radiocarbon concentration in modern wood.” Science 1955;122:415-7. This is the earliest reference I can find specifically measuring the effect we call the Suess effect (The effect is hinted at in Seuss HE: “Natural radiocarbon measurements by acetyline counting.” Science 1954:120:5-7 There is a reference in the Science articles to a “[H. E. Suess], Proc. Conf. on Nuclear Processes in Geologic Set, Williams Bay, 1953” or “H. E. Suess, paper presented at the NSF Conference, Oct. 1953, Williams Bay, Wis.”, which presumably predicts the existence of the effect. I have been unable to locate this paper yet). Because of the Suess effect, wood formed in 1850 was determined to be the reference standard for the 14C/C ratio, rather than wood formed just before the first atomic bomb. This is true even though 14C dates are customarily given in years BP (before present, present being defined as 1950).
So an increase in the total carbon in the biosphere would [182]entail a corresponding decrease in the 14C/C ratio.

And that is precisely what a creationist theory would postulate for the antediluvian (before the Flood) world. All the coal and oil, and some of the limestone, in the fossil record should have been in rough equilibrium with the carbon dioxide in the atmosphere. A standard estimate of this mass would indicate that the 14C/C ratio, from this effect alone, would have been about [183]
1/200th of its present value.151
151 Brown RH, see note 149.
So a model has been proposed in which the 14C/C ratio was initially at about 0.1 to 0.5 pmc. According to this model, at the end of the Flood 14C continued to be formed, but was now diluted in a much smaller pool of ordinary carbon, so that the 14C/C ratio of the biosphere rapidly (and perhaps somewhat irregularly) rose, leveling off to near its present concentration within a few hundred years. Thus radiocarbon dates are not discarded, but are reinterpreted as shown in the above figure.

To find the age using this flood model, find the 14C/C ratio at the right of the graph, follow the (exponential) curve to the left until it intersects the flood model 14C/C ratio curve, then read the time from the scale on the bottom. But this model implied that there should be residual 14C in all antediluvian material. Specifically, if the Flood happened around 1 half life of 14C ago,152
152A Septuagint date—the Masoretic Text would be slightly shorter (4,300 to 4,500 years or 0.8 half lives ago) and creationist theories which believe the Genesis 11 record is incomplete might have a Flood date as much as 2 to 3 half lives ago.
153The correction factor would be 1/200 for the ratio of antediluvian to postdiluvian biomass, 1 to 1/4 for the effect of the antediluvian magnetic field, and 1/2 for the passage of time. In point of fact, there is some uncertainty in the estimate of fossil carbon, so the factor of 1/200 might be better estimated at 1/ 100 to 1/400. There is also the theoretical possibility that the earth started out at the time of creation with no 14C whatever, which would give an additional correction factor of down to 1/5 because of non-equilibrium conditions. This is doubtful even on creationist assumptions, since other radioactive minerals which naturally occur in living organisms (for example, 40K) seem to have either been created or have maintained their identity through creation.
the antediluvian 14C/C ratio should be approximately 1/400 to 1/1600 the present ratio (0.25-0.0625 pmc).153

These measurements are out of range for all but the most careful, time-consuming, and expensive experiments using conventional decay counting. More importantly, with decay counting, it is necessary to measure background counts using a counter identical to the sample except for the absence of 14C. This has usually been done by using fossil carbon, which is “known” to have had all its 14C transformed to nitrogen. But if the question is whether this material still has 14C, no amount of measuring could find 14C by comparing fossil carbon with fossil carbon. Both [184] measurements would come out the same, almost by definition. But when the AMS method was developed, there were repeated and theoretically persuasive arguments that the background could be essentially eliminated. So if there is 14C in antediluvian material, it should be detectable with the AMS method. Thus we could have a clear-cut method to decide which time scale most accurately reflects the correct time scale for the history of life on the earth.

The earliest reports of measurements on “infinitely old” material were mixed. Several experiments on AMS gave backgrounds of 48,000 to 70,000 radiocarbon years.154
154Some examples follow. Bennett CL, Beukens MR, Clover MR, ElmoreD, Gove HE, Kilius L, Litherland AE: “Radiocarbon dating with electrostatic accelerators: Dating of milligram samples of graphite.” Science 1978;201:345-7, 48,000 years. Andrews HR, Ball GC, Brown RM, Burn N, Davies WG, Inahori Y, Milton JCD: “Radiocarbon dating experiments with the Chalk River MP tandem accelerator.” In Gove HE (ed): Proc 1st Conf on Radiocarbon Dating with Accelerators. Rochester, NY: University of Rochester, 1978, pp. 114-26, 58,000 years (on “graphite”., with 2 counts. Dolomite had S counts, which would give it a calculated age of about 50,600 years). Litherland AE: “Radiocarbon dating with accelerators: Results from Rochester-Toronto-General Ionex Corporation.” In Gove, op. cit., pp.70-113, 65,000 years (on “graphite”). Bennett CL et al., note 148, 70,000 years (on “petroleum-based graphite”).

It is tempting to use the “petroleum-based graphite” of Bennett et al. as a measure of carbon from the antediluvian biosphere (from a flood model perspective) or from quite old carbon (from the evolutionary perspective). In this case one could argue that breaking the 50,000 radiocarbon year barrier mentioned below is a function of the form of carbon. However, while I have not been able to find out the geologic history of this particular carbon, I have been told by someone in the field that often graphite that is dated is simply bought from a supplier and the geologic history is not known by the experimenter. It is possible that this graphite was not in equilibrium with the biosphere at any time. Further observations and experiments could clarify this issue.

But as time has continued, it has become general knowledge that there is a wall at about 50,000 radiocarbon years (about 0.2 pmc) that is not breached in practice. This is well within the limits of the creationist prediction, and outside the evolutionary prediction, or even of evolutionary theory.

The first evolutionist reaction to the data was to say that the machines were somehow giving background counts. This was unlikely theoretically, but possible. One should be able to test this possibility by dating carbon that had had practically all its 14C removed, say, by mass spectrometry, and possibly also by [185] dating material which had not been in equilibrium with the biosphere near the time of the Flood. Two possible examples that came to mind were carbon from igneous rocks and Precambrian carbon.

I had intended to test this possibility (and in fact had written to one AMS lab asking to arrange for experiments along that line) when I became aware that the most critical experiments had already been done by Schmidt et al.155
155Schmidt FR, Balsley DR, Leach DD: “Early expectations of AMS: Greater ages and tiny fractions. One failure? - One success.” Nuci Instr and Meth 1987;B29:97-9.
They dated “geological graphite” to 69,030 radiocarbon years (0.0185 pmc). Prepared slightly less carefully, it dated at 58,590 to 65,840 radiocarbon years (0.028 to 0.068 pmc). Carbon-12 from the Faraday cup of the accelerator dated at 61,000 radiocarbon years (0.050 pmc).

At the same time, their anthracite coal dated “up to 52,000” radiocarbon years (0.154 pmc), and their marble ran up to 49,690 radiocarbon years (0.206 pmc). Thus, for antediluvian carbon they hit the same wall as other investigators, but they were able to go through this wall with graphite which may have represented carbon not in equilibrium with the antediluvian biosphere. Similar results were obtained with carbon from which 14C had been mostly removed by isotope separation. Thus machine background is not an adequate explanation for more than 0.0185 pmc (radiocarbon age 69,000 years), and probably not for even that much (with the sample holder completely empty, their machine produced “≥ 90,000 years”, or no counts in a 30 minute run).

With machine background eliminated as a reasonable explanation, there are only four ways I can think of to explain the background in anthracite coal and marble (and oil). It could be contamination during sample preparation, source contamination with modern carbon, in situ formation of 14C, or residual activity.

Contamination during sample preparation seems unlikely to explain all the difference between most fossil carbon and the geologic graphite noted above. It should have affected the geologic graphite and the purified 12C implant as well. And in order to explain the difference the samples would have to be consistently contaminated with a known contaminant (modern post-bomb carbon) at about 1 part in 1000. Anyone who did that would have flunked an analyti-[186]cal chemistry class. However, it is still possible that the experiments haven’t been done carefully enough,156
156There are two articles I know of that give data that suggest this as a possibility First, in Grootes et al., see note 147, it is noted that oxygen and nitrogen carrier gas apparently have enough carbon dioxide contamination to invalidate very old dates (compare their GrN 6553 with their GrN 6808). Their data are also difficult to interpret if one assumes that anthracite coal has a ratio of approximately 0.2 pmc; the data suggest rather that the ratio is >0.014 pmc. A repeat experiment to confirm these findings would be helpful. For some reason the method is little used nowadays.

Second, Beukins reports (Beukins RP: “Radiocarbon accelerator mass spectrometry: Background, precision and accuracy.” In Taylor RE, Long A, Kra RS (eds): Radiocarbon After Four Decades: An Interdisciplinary Perspective. New York: Springer-Verlag, 1992) that the contamination in multiple samples from different sources was 0.076 to 0.081 pmc in the IsoTrace AMS. This, if reproducible, would seem to indicate that the major source of background in most AMS apparati is contamination during sample preparation. It would also suggest that the major source of background in the IsoTrace experiments was either residual activity or contamination at the reduction step. This differentiation should be subject to experimental determination.

so one could simply take the “geological graphite” and subject it to the same process that the anthracite coal undergoes. If the dates are still different, then sample contamination may be ruled out.

Contamination of the source implies a worldwide exchange of carbon in all the world’s known carbon deposits that is relatively even, in spite of the differences in the physical state of coal, oil, and natural gas. It also requires a roughly 50% exchange of the entire biosphere with fossil carbon, within the last 6000 years or so (the longer ago it happened, the greater the required degree of contamination). And again, the exchange rate is 0.2% of ancient carbon, which in a 1,000,000 barrel oil field is approximately 2,000 barrels, again within the past 6000 years. This amount of contamination is hard enough to believe with oil, and frankly incredible with coal.

In situ formation of 14C would involve, in the easiest case, neutrons in massive quantities. There are strong arguments that “Subsurface production of radiocarbon is negligible (Zito et al. 1980; Florkowski et al. 1988).”157
157Geyh and Schleicher, p. 165, citing Zito R, Donahue DJ, Davis SN, Bentley HW, Fritz F: “Possible subsurface production of carbon-14.” Geophys Res Lett 1980;7(4):235-8, and Florkowski T, Morawska L, Rozanski K: “Natural production of radionuclides in geological formations.” Nucl Geophys 1988;2:1-14. Zito et al. calculated the production of 14C in groundwater from neutrons. Using their best case (granite), the 14C concentration would be 0.00266 pmc for an apparent age of 87,100 years (5730 year half life). Florkowski et al. seconded their calculations. Redoing the calculations with oil (assuming no nitrogen and a density of 0.9 g/ml which will cancel eventually) and assuming a 13C neutron cross section of 0.0014 barns (using the terminology of Zito et al., P = 3.89 10-7 and N = 14,400 atoms of 14C per liter of oil), yields 2.7 10-8 pmc, which is ridiculously low. It would almost take a neutron bomb to produce enough 14C from neutrons to give the contamination presently found in phanerozoic carbon, and the neutron irradiation would have to be within the last 6000 years or so, or else most of the 14C formed would decay to nitrogen before the measurement took place. Nitrogen is 110,000 times more efficient at producing 14C from neutrons than carbon, so any production of 14C from neutrons would be heavily influenced by the nitrogen content of the fossil material. I know of no such effect reported in ancient carbon. Perhaps it should be sought.
And if this were the case, one [187] might expect the effects to be variable on antediluvian material and to also affect “geological graphite” to a similar extent. In situ formation seems highly unlikely.

That leaves us with residual activity. But this mathematically eliminates the evolutionary time scale. For if we started with the entire earth’s mass being 14C, within 1 million years all of the 14C would have decayed to 14N except for 1 atom, and that one atom would have a roughly 99% chance of decaying. Each 5,730 years further back doubles the number of earths we would need to have that one atom, and we would need to have filled the universe before we get to 2 million years. The universe is demonstrably not made of 14N. There is no way that 60-600 million-year-old material should have any residual 14C, and thus if there is 14C in this material which is not contamination, it is simply not that old. This evidence is in almost complete conflict with the evolutionary time scale. The phanerozoic is almost certainly less than 60,000 years old, and very probably 4 to 8 thousand years old.158
158I say “in almost complete conflict with” instead of “incompatible with” because it is still just possible that one of the other explanations for the presence of 14C in Paleozoic and Mesozoic carbon is correct. Further experiments, as suggested above, to rule out or confirm these explanations should be done (and are being planned).

To summarize the experiments, they are: 1. to repeat the experiments of Schmidt et al. (see note 155), 2. to take carbon that dates older than fossil carbon (determined from experiment set 1) and run it through the same preparation procedures that are required for fossil carbon, 3. to determine the dates of fossil carbon from many different sources, and 4. to date fossil carbon of varying nitrogen content. If the fossil carbon from experiment sets 2-4 all has similar 14C contents, which are significantly higher than that of our material found in experiment 1, then the evidence for a short history of life on the earth would be overwhelming. All the strata in the geologic column would be of roughly the same date, and it would be entirely reasonable (since they were almost all, if not all, formed under water) to attribute them to the Flood. If fossil carbon can be prepared with sufficient care (without significant isotope separation) that its 14C content is <0.005 pmc, then this creationist model would be eliminated, leaving only the evolutionary model or a parasitic creationist model which assumes a change in the decay constant at the Flood (which at this point is not convincing). It is also possible that neither of these eventualities will happen, in which case one will simply have to take the best available data and make one’s best guess. Given the relative ease with which the experiments should be able to be done, we should not have to settle for this, but it does appear that the evolutionary explanation for the data we have at present is by far the more strained.

The most likely date for the Flood, based on the present 14C data, is one which would account for a present age of fifty thousand radiocarbon years in a biosphere with 200+ times the modern amount of carbon, which would require a Flood roughly 1 radiocarbon half life ago. I have given reasons why I do not trust the standard interpretation of the uranium series disequilibrium dating of corals, the only other consistent physical dating method to match (roughly) 14C dates. Historical dates extend only 5,000 years back at best, and so are not a major problem for a short chronology. The only other major objection to a short chronology is tree ring dating. I have serious reservations about the accuracy of this method as usually applied. I plan to deal with tree ring dating more fully when I take up the subject of the Exodus, and I hope that currently unpublished material regarding the bristlecone pine chronology will be available at that time.

However, even if tree ring dating does turn out to be accurate as usually applied, it still would not solve the 14C problem for an evolutionist. It would simply mean that Genesis 11 did not contain a complete chronology. The earth would still have to be much too young for any kind of evolutionary explanation to be adequate, and Genesis 1-9 would still contain the best available explanation for the fossil record.

159 See the discussion in Faure, pp. 410-2. See esp. Inoue T, Tanaka 5: “10Be in marine sediments.” Earth Plan Sci Lett 1976;29:155-60. On p. 155 Inoue and Tanaka stated, “The scarcity of cores having a uniform “10Be concentration at different depths suggests that the sedimentation at the ocean floor has not been uniform but disturbed by some geophysical events in the past.” It is of interest that wherever the data roughly matches an exponential curve, it is assumed that production and deposition of 10Be and sedimentation rates have all been constant (Amin BS et al., see note 141, although they also noted on p. 823 a “high frequency of cores with erratic 10Be concentrations at different depths, even within depths of only one meter, suggest[ing] that the ocean floor is physically disturbed in many regions.”), whereas later evidence indicates that given the evolutionary time scale, production rates at least had varied (for example, Somayajulu, see note 109).
160 Faure, p. 415, citing Brown L, Klein J, Middleton R, Sacks IS, Tera F: “10Be in island-arc volcanoes and implications for subduction.” Nature 1982;299:718-20. On p. 718 they state that the phenomenon “certainly gives one pause.” They try to explain this by the activity of cosmic rays. It would be interesting to test this theory by dating basalts from different depths. About 10 meters should be sufficient for testing purposes.
161Gentry RV: “Forum: Time: Measured responses.” EOS; Trans Am Geophys Union 1979;60(22):474. He also predicted a 14C/C ratio of approximately 0.01 pmc in Paleozoic and Mesozoic fossil material.

This conclusion concerning 14C dating is compatible with the weight of evidence for the other dating methods mentioned above, as well as a straightforward interpretation of the data on 10Be (a “failed” dating method, that is, one which is not easily interpretable using the evolutionary time scale). Ocean floors show no consistent gradients of 10Be concentrations,159 which is more easily [189] explained by a Flood than by very slow deposition of ocean sediments.

In addition, lava apparently formed from subducted ocean floors (which contain 10Be from cosmic ray production) contains up to 7 106 atoms/g of 10Be, whereas isolated volcanoes like Kilauea in Hawaii have 0.1 106 atoms/g of 10Be. However, a basalt from the Columbia river plateau with an evolutionary age of 14 million years contained 1.0 106 atoms/g of 10Be.160 The half-life of 10Be is 1.5 million years. If we assume that this lava started out with the highest known modern concentration of 10Be, then only 0.01 106 atoms/g of 10Be should have been left from the initial lava flow. Where the extra 10Be came from is hard to say assuming the evolutionary time scale, but quite easy if the lava flow in question happened only a few thousand years ago. The existence of this 10Be was predicted by Gentry in 1979 on the basis of a creationist model.161

If we are committed to following the weight of evidence we are led to discount theistic evolution and multiple creations as explanations of life on earth. We may also discount the (creationist) theory that decay constants have varied significantly with time, at least back to the Flood. And if it needed any further demonstration, mechanistic evolution is thoroughly discredited. Creationism may not have solved all its problems, but it has solved the major ones, and it is not unreasonable to believe that the rest will be solved with further study, whereas theories requiring millions of years for life (including theistic evolution and multiple creations) appear incapable in principle of solving the 14C problem, and there is solid evidence that they are wrong in their interpretation of other dating methods.

This also means that we should give credence to the early Genesis record, and that we should seriously consider the claims [190] of Mosaic authorship, and also the accuracy of the rest of the Pentateuch and Joshua. I will not argue these points at this time. I hope to be able to do so later. For now I will assume that the entire Bible is reliable in the sense noted in chapter 3. I have now outlined, as a scientist might say, my materials and methods. Next, we will deal with some preliminary results.

Home Page
Go to Chapter 6: God, Freedom, and Time