Radiocarbon Dates Linked to Tree Rings
THE title of the Twelfth Nobel Symposium was “Radiocarbon Variations and Absolute Chronology.” The title implies that radiocarbon dating is no longer regarded as absolute. The emphasis in the symposium was on the variations in radiocarbon dates and the attempts, only partly successful, to explain them. That which emerged as the absolute chronology was the one based on counting tree rings.
Is that bad news? After all, the method of radiocarbon dating is a specialized technical field for a few highly trained experts, and the theory has been corrected here and adjusted there until it is difficult even for other scientists to understand. On the other hand, everyone knows—doesn’t he?—that a growing tree adds one ring every year around its trunk. And after a tree is cut down you can tell how old it was merely by counting the rings, can’t you? What could be simpler than that? Doubtless many persons will be relieved to learn that the radiocarbon clock, which always smacked a little of scientific magic, is now being kept on time by something as easy and understandable as counting tree rings.
The calibration curve was included in the published report of the symposium (also published in Scientific American, October 1971). It shows, for each year back to about 5200 B.C.E., how many years must be added to or subtracted from the radiocarbon date to make it correspond with the tree-ring date.
At first glance you might mistake it for a chart of stock-market prices. Its lack of any regularity, its random short-term wiggles, and its unpredictable long-term trends all enhance the resemblance. By using this correction curve, the radiocarbon dating laboratories have come to rely fully on the accuracy of tree-ring chronology, also called dendrochronology.
So those who have put their faith in radiocarbon dates must now ask themselves whether that faith is strengthened or weakened by the new linkage to tree-ring dates. The answer, of course, depends on how certain the tree-ring chronology is. Is it a firm anchor for radiocarbon dates, to keep them from floating off into the unknown depths of antiquity?
Not many trees live thousands of years. The magnificent giant sequoias that grow on the mountain slopes of California are famous for their extreme longevity. In recent years, however, it has been found that the bristlecone pine, an unpretentious, scrubby-looking tree that grows on high, rocky slopes in the southwestern United States sometimes lives even longer. One tree in Nevada is reported to be 4,900 years old.
The usefulness of this long-lived tree was first pointed out in 1953, by Edmund Schulman, of the University of Arizona. In the White Mountains of eastern California he found a number of very old trees, some of them still living, others now dead stumps or logs. He collected cores cut from living trees as well as the remains of fallen trees in the grove. He examined them in his laboratory and used them to set up a tree-ring chronology. After his death in 1958, this work was resumed by Professor C. W. Ferguson in the same laboratory. Ferguson reported the present status of the work to the Nobel Symposium. He claims to have established a tree-ring chronology for the bristlecone pine all the way back to 5522 B.C.E. This is a span of almost 7,500 years, a truly impressive accomplishment. Can there be any reason to doubt that it is correct?
Questioned by Some Researchers
Well, we may note that Professor P. E. Damon, of the geology department at the same university as Ferguson, said: “The accuracy of tree-ring dating may be questioned by some researchers.”8a Then let us inquire into the procedure of constructing a tree-ring chronology to see why it may be questionable.
The first thing we should ask about is the basic assumption of tree-ring counting, that one ring equals one year. It may surprise you to learn that this is not always true. Ferguson says on this point: “In some instances, 5 percent or more of the annual rings may be missing along a given radius that spans many centuries. The location of such ‘missing’ rings in a specimen is verified by cross-dating its ring pattern with the ring pattern of other trees in which the ‘missing’ ring is present.”9 Since the investigator adds these “missing rings” to his chronology, it is greater than the actual number of rings counted, by five or more years for each century.
Even more interesting is Ferguson’s comment about the possibility that a tree may produce two or three rings in a single year: “In certain species of conifers, especially those at lower elevations or in southern latitudes, one season’s growth increment may be composed of two or more flushes of growth, each of which may strongly resemble an annual ring. Such multiple growth rings are extremely rare in bristlecone pine, however, and they are especially infrequent at the elevation and latitude of the sites being studied.”9
So, under present climatic conditions, multiple rings are rare. From a uniformitarian point of view, such a statement is reassuring enough. But this viewpoint overlooks the abundant evidence that the climate was much more temperate before the Deluge of 2370 B.C.E. Also, the present-day location of the bristlecone pine groves might then have been at a much lower elevation. Both of these differences, in harmony with the opinion quoted, could have resulted in more multiple rings in trees then living. This would have been true, not only before the Flood, but even for some time afterward, while the earth’s crust was adjusting to new pressures. Who can say how often multiple rings formed under those conditions, or how many extra centuries are included in the chronology on that account?
Piecing the Patterns Together
The next point to note is that no single tree has 7,500 rings. Although it is reported that some standing trees are more than 3,000, and even 4,000 years old, the oldest living tree included in the chronology goes back only to 800 C.E. However, a dead tree was found with some 2,200 rings, and similarities in the pattern of thick and thin rings were found between the outer layers of the dead tree and the inner layers of the living tree. So the ages were considered to overlap from 800 to 1285 C.E., and the older tree was dated back to 957 B.C.E. This process was repeated with seventeen other remnants of fallen trees, ranging from 439 to 3,250 rings, to carry the ring count back a total of 7,484 years.
Now you may ask, How certain is the matching of the overlapping patterns? Ferguson assures us that there is only one possible way to make each of the seventeen fits; as he says: “The master chronology for all specimens involved is unique in its year-by-year pattern; nowhere, throughout time, is precisely the same long-term sequence of wide and narrow rings repeated, because year-to-year variations in climate are never exactly the same.”9 Some persons might be willing to accept this opinion at face value; other researchers might, as Damon says, be among those who question it.
Another question: If it were possible to fit a dead tree segment in more than one place, what considerations would guide the selection of the “correct” fit? This statement by Ferguson may give us a clue: “Occasionally, a sample from a specimen not yet dated is submitted for radiocarbon analysis. The date obtained indicates the general age of the sample, this gives a clue as to what portion of the master chronology should be scanned, and thus the tree-ring date may be identified more readily.”10 And, again: “Radiocarbon analysis of a single, small specimen, that contains a 400-year, high quality ring series indicates that the specimen is approximately 9000 years old. This holds great promise for the extension of the tree-ring chronology farther back in time.”11
Thus it is evident that the carbon-14 dating sometimes serves as a guide in fitting together the pieces of the tree-ring puzzle. Do these admissions give reason to suspect that perhaps the tree-ring chronology is not as well-anchored as it seems to be, but that its proponents look for support to radiocarbon dating? Such a suspicion is not unfounded, for Professor Damon, after assuring us of his personal confidence in tree-ring dates, adds: “Nevertheless, it is reassuring to have some objective comparison, for example, with another method of dating. This is, in fact, provided by carbon-14 dating of historically dated samples.”8
If tree-ring dates need to be bolstered by comparison with radiocarbon dates in the range where they are supported by historical dates, back only 4,000 years, what is to be said of the need 4,000 or 5,000 years before that?
Problems in Dating Wood
The efforts to strengthen the mutual support of the two chronologies are plagued by another problem that occasioned considerable discussion among the experts. Even in radiocarbon analysis of those samples of bristlecone pine that now serve as the basis for all other radiocarbon dates, the possibility of sample alteration must be considered. It is known that inorganic substances, such as the limestone of shellfish and the carbonate in bones, are very susceptible to exchange with dissolved carbonates, either older or younger. For this reason they are almost useless for dating. Organic substances, such as cellulose, are regarded as unlikely to exchange. The live sap in a tree can be washed out of the dead wood, but if it has been circulating through the wood for centuries or millenniums, can we be sure that it has not partly replaced the decaying carbon 14?
Unlike the sap, resin is difficult to remove. Ferguson has referred to “the highly resinous nature” of bristlecone pine wood.12 The experts agreed that resin from younger wood moves into the older wood, where it can cause errors. “The diffusion inward of the resin certainly is a reasonable result.”13 Also, “This resin problem is important, particularly as the correction increases as one goes further into the tree.”13 In one experiment, the extracted resin was apparently 400 years younger than the wood.
However, the experts disagreed as to how effective their chemical treatments are. One said that boiling the wood successively in acid and alkali “removes all of the resin.”14 Another said: “In my opinion, the resins in bristlecone pines cannot be removed completely by treatment with inorganic chemicals.”14 But when they use organic chemical solvents, they have to worry about whether the solvent has been completely removed afterward, because just a little modern carbon from it could apparently rejuvenate a sample of ancient wood. Of course, they work conscientiously to exclude all these errors, but are they completely successful? How sure can we be?
Glacial Varve Counting
A somewhat similar method of counting years into the past was discussed at the meeting, one based on glacial varves. Varves are alternate layers of sand and silt that are supposedly formed annually by a glacier as it melts. It is claimed that these provide a continuous record, one in Sweden going back as far as 12,000 years. This also was proposed as an absolute chronology to which radiocarbon dates might be tied. But how firm a basis is it, really?
The Scandinavian varve chronology is pieced together from sections observed in different places throughout the length of Sweden. The record appears much less useful than a tree-ring chronology, for several reasons.
For one thing, there is no link to the present day, corresponding to the bark ring. Estimates as to the date when the last varve was laid down vary widely. Also, the problem of identifying annual deposits contributes to the uncertainty. So one geologist dated the beginning of the series in Skåne at 12,950 B.C.E., another at only 10,550 B.C.E. Dr. E. Fromm, of the Geological Survey of Sweden, said: “In these cases the geological setting did not a priori limit the possible range of the datings, and the ‘teleconnections’ have obviously given quite unreliable results. Moreover, in these parts of Skåne doubts remain as to whether all varved deposits with sedimentation in small melt-water lakes are really annual varves.”15
Note this admission that varves do not always correspond to annual deposits. In reality, they represent alternate conditions of rapid flow and slow flow, which might occur several times a year under some climatic conditions. “Dr. Hörnsten of the Geological Survey of Sweden pointed out that each varve had to be examined very carefully to avoid counting the varve from one year as two years. One single varve deposited during one year may have one or two pseudo-winter layers, due to variations in the discharge of melt-water (cf. double tree rings).”16 Professor R. F. Flint, a well-known geologist of Yale University, asked for a clear statement of the criteria by which a varve is recognized, but so far as the record of the symposium shows, this was not forthcoming.17
These, then, are the “absolute chronologies” that were offered at the Nobel Symposium. From the articles in popular science magazines it would be easy to get the impression that radiocarbon dating is more firmly established than ever. But a careful reading of the backstage discussion at the Uppsala conference reveals that the uncertainties have multiplied. The radiocarbon theory no longer provides a sound basis for acceptance of its dates. The results of twenty years of study have greatly weakened most of its underlying assumptions.
Now reliance is placed on the work of a single research group on a new method—tree-ring dating. What additional weaknesses in this technique might be revealed by twenty years of intensive study in different laboratories? In its present status, would you be willing to rely upon it, rather than on the Bible, for the vital decisions you must make in the near future?
a References are found on page 20.