Addendum: Response to Criticisms
"It is not surprising that most of the samples found and presented for dating are less than 5,000 years old, and very few are more than 15,000 years old. You need a lot of imagination to read more than that into that graph."
With this you have very well summarized the criticisms of this data. I have heard from two other scientists who share your perspective. However, I'd like to suggest several additional points for your (and their) consideration. Perhaps not as much "imagination" is required as you imagine.
1. Provenance. The provenance of the data reported here plays a crucial role in its interpretation. One could not properly understand the importance of the data showing that the sun bends light waves as they pass it, without knowing that Einstein had explicitly predicted this on the basis of his theory of relativity. Without this prediction the observation would have been important, but its meaning less immediately apparent.
A number of years before this study was done, I postulated that there was a discontinuity in archeological samples, with a sudden drop off in samples somewhere between 5,000 and 10,000 years. This early observation was based on the very limited sampling of what I was seeing in a few of the general science publications such as Science, Nature, Scientific American, Chemical & Engineering News and possibly some others. I first discussed this with Dr. Robert Knocke, a nuclear chemist at the University of Nebraska, in 1964. He was skeptical that it was true, but couldn't suggest any realistic way to test the hypothesis. (I was probably influenced in my thinking by a letter to the editor in C & E News, in which the writer had reviewed the obituaries in the same publication, as well as in a selection of other professional publications, and had found that chemists died several years sooner, on average, than their other professional colleagues - physicists, mathematicians, psychologists, etc.)
I discussed my hypothesis with a number of scientists over the following years, several at the University of California, Davis during the years I was finishing my studies in the Department of Biochemistry and Biophysics. There were only a few rabid evolutionists in the department, several of the professors engaged friendly back and forth on a variety of issues. It was one of them who brought the bristlecone pine chronology to my attention. The initial part of my work in dendrochronology was reported in Pensee, along with a report by Willard Libby, Nobel Laureate for his work in radiocarbon data and Immanuel Velikovsky, the most influential catastrophist in this century: http://www.happythinker.com/stuff/AgesofBristleconePine.html)
My major professor, Dr. Lloyd Ingraham, given my five publications with him from my first year of work there, was very tolerant of my unorthodox thinking. We spent some time brainstorming my hypothesis. (Each of our major brainstorming sessions resulted in a publication at some point, a matter of no small consequence in the university system.) It was his opinion that I had to find the data to allow a tabulation of samples per year if I ever hoped to get anywhere with this. Very fortunately, since my tour at Nebraska, a comprehensive index of all radiocarbon dates from 1950 to 1965 was published (in 1967). Since more than 10,000 specimens were involved, my professor provided the funding to have this data digitized and to produce the chart under discussion.
Neither of us were prepared for the results obtained. However, the discontinuities were so obvious that my professor attempted to write some statistical routines to quantitate them.
To characterize the information presented here as a couple of patent, trivial points suggests an amazing resistance to insight. I'll detail several points:
2. Current to 2000 years B.P. (before present). Critics who dismiss this data tend to do so on the grounds that any relationship between age and number of samples is either non-existent or trivial. And this is pretty much the case during this most recent era. However, this is definitely not the case for the preceding 40,000 years. The regularity of the data for this long period of time (with discontinuities) requires explanation, not dismissal.
3. 2000 to 6000 years B.P. This is the period during which large numbers of samples were analyzed, and is most amenable to statistical analysis. We need to bear in mind that these are carbonaceous samples, including lots of human artifacts: shoes, ropes, baskets, bones and suchlike. We are not considering the decay of radiocarbon, but rather the decay of the samples.
Consideration of the probable type of decay suggested that there would be an exponential relationship between the variables:
N/N0 = e-lt
The validity of this hypothesis is illustrated in Figure I where data
on samples between 20 and 60 centuries of age is
displayed. The number of samples parameter and age parameter exhibit a high degree of correlation (R = -.95).
Figure I. The logarithmic "decay" of recent carbonaceous materials.
The least squares analysis yields a half-life of 18.5
Least squares analysis of this subset of data yields a half life of
18.5 centuries. Factors likely to be represented in this number
1. Rate of formation of archeological samples.
2. Rate of deposition.
3. Internal forces of decay.
4. External forces of decay - climatic, oxidation, etc.
5. Sample selection and collection.
The net result of all these factors (and possibly others) has created a relationship that is very unlikely to have happened by chance. The high degree of correlation (-.95) on over 3000 samples shouldn't be ignored.
4. 6000 to 14000 years B.P. There are actually two regions in this time period, up to 115 centuries and then from there to 140 centuries. The earlier period shows erratic numbers of specimens from century to century, albeit gradually decreasing. The half-life of specimens is 20.8 centuries with a correlation of -.40. An attempt to determine the types of samples represented was made by a review of every specimen reported in Libby's earlier book. (The comprensive index of 1967 gives no information on sample composition. Each is identified only by a number from the corresponding laboratory.) A log buried under a glacier would be typical of these samples, with minimum artefactual material. But it is interesting that in the older portion of the era, the number of samples decreases to an effective baseline that might as well be zero.
5. Greater than 14000 years B.P. It is potentially of great significance that the number of samples per century is a relatively constant 5 to 7 samples (actually, well past 40,000 years). (At the time of this research, the theoretical limit of radiocarbon dating was about 50,000 years, about 10 half-lives of radiocarbon - not to be confused with the half-lives of the archeological samples under discussion.) This is a very anomalous finding, totally inconsistent with steady deposition/decay of samples, in any case. For example, statistically, these samples, though very old, have a very short half-life. This is quite impossible. Hence, another explanation must be found.
I am quite certain that any radiocarbon dated sample of greater than 14,000 years is due to a small amount of contemporary contamination of carboniferous age material. Without a detailed description of the radiocarbon dating method, suffice it to say that the less radiocarbon found in the sample, the older it is assumed to be. So very old samples have no radiocarbon at all. But if you take a very old sample, and it becomes contaminated with more recent carbonaceous material anytime up to and including the present, a very small amount of random contamination would result in a very old random date. This is exactly what is observed. The assigned dates are meaningless.
(As a related matter I began a study of carbon inventories and likely
production rates of formation of C-14 from N-14 in the upper atmosphere
due to cosmic radiation. The carbon inventory studies show that assigning
carboniferous era material very old ages because of the lack of radiocarbon
in them is absurd. If the massive petroleum reservoirs - not to mention
carbonate carbon - were returned to the biosphere pool from which they
obviously came, even with recent elevated rates of production of C-14,
the dilution effect would be so large as to render the carbon pool as detectably
free of radiocarbon.)
For ready reference, the statistics are summarized in this table.
I am always prepared to be wrong, but I believe the analysis described above is closely driven by the data. Not much imagination is required.
the Christian searcher