The following article from Approfondimento Sindone isreprinted with the permission of the publisher. Copyright © 1998by Centro Studi Medievali (Pontremoli MS, Italy). Nothing may becopied or reproduced in any form without the written permission ofthe publisher.
The results of the work of three radiocarbon laboratories to datesamples from the Shroud of Turin by the radiocarbon method1, whileaccepted by many, have not convinced everyone. In this Note Idescribe how the principles and techniques of the method have beenbrought to bear on this particular problem, and give a scientificperspective, as an expert, of the result.
I consider it beyond my scientific perspective to commentauthoritatively on non-scientific aspects of the dating operation.For example, it has been asserted that the sub-samples of the shroudcloth (but presumably not of the control samples) were deceptivelysubstituted by ones not from the Shroud of Turin. Having witnessedthe sampling operation, I find this assertion incredible, but I canproduce no scientific evidence to refute it (although thecharacteristic weave of the shroud cloth is certainly unusual andeasily recognised). Another assertion, that the sample was from arepair of the cloth, is also contrary to my own perception, but thereare others involved in the sampling, more expert than I, who arebetter able to refute this argument. Here I am concerned withexplaining how the measurements obtained leave no room for doubt thatthe textile sample, shared between the participating laboratories,was made from plants growing in the early part of the secondmillennium after Christ.
The basic principle was formulated, and tested on known agematerials, by Libby (who obtained the Nobel prize in 1960 for thiswork) fifty years ago. Two isotopes of carbon, 13C and 12C, are knownto be stable and were formed when the other atoms of this planet werealso formed, and they constitute almost all the carbon component inorganic and inorganic material on earth. A third isotope, 14C or«radiocarbon» is radioactively unstable, so that any formedwith the planet will have long since disappeared. However, 14C isbeing continuously formed in small quantities as cosmic rays impactthe earth's upper atmosphere, and the amount found on earthcorresponds to the balance between cosmic ray formation and thenatural disappearance of the isotope due to its instability. Thisbalance is extremely small, 14C being approximately one part in amillion million of the overall carbon content. The important point,for dating purposes, is that the 14C formed in the atmosphere is ableto be taken up, along with 12C and 13C, in atmospheric carbon dioxideby photosynthesizing plants, and thereby spreads throughout thebiosphere, so that all living things in fact have a similar ratio of14C/12C. This ratio is maintained throughout because the carbonisotopes all have the same chemical behaviour. However, once livingorganisms die, their carbon content is no longer maintained by theirmetabolism, and, the level of 14C falls according to its radioactivehalf-life. Therefore a measure of the 14C/12C ratio provides ameasure of the time elapsed since the organism died. This principle,as simply given here, requires a number of qualifications if accuratedating is to be sought, and a great deal of work, in many radiocarbondating laboratories, has been extended over decades in order to showexactly what modifications are required.
The main modification has been to show that the«balance» level of radiocarbon in the atmosphere hasfluctuated. This can be seen by dating known age material (such astree rings) and comparing the result with the known age. There is asmall difference at 0 AD and 1000 AD of about 60 years, but thedifference can be quite large (e.g. 500 years at 3000 BC). It is,however, universal for (almost) all living systems and in allregions, at least to an accuracy of better than 30 years. Thequalification above, of almost all living systems, is due to a secondmodification, which recognises that not all life obtains its carbondirectly from the atmosphere. Actually atmospheric carbon dioxide isin equilibrium with a far larger quantity dissolved in the oceans -this huge reservoir is in fact buffering the atmosphere against theexcessive carbon dioxide released from industrial fossil fuelburning. The ocean reservoir lags behind the atmosphere, andradiocarbon dates from, for example, sea shells, tend to be somecenturies older than terrestrial animals and plants, and so needtheir own calibration curve. A third qualification is essentially anelaboration of this point, namely that the origins of all the sourcesof carbon in a sample need to be considered and separately accounted.For example, in archaeological potsherds, carbon may be present fromcarbonaceous material in the original clay, from the addition ofanimal or vegetable temper (e.g. chopped straw), from fire smoke,from surviving fats from cooked food, or from mobile carbonaceousmaterial in the soil in which the potsherd was deposited after use.All these sources can have quite different times since their carbonatoms were part of living systems, and so the dating of the potterywill depend on which source is selected and measured. I havedescribed these qualifications in some detail to show that there is avery rigorous attitude to understanding the ways in which radiocarbondating requires a more detailed approach than the simple basicprinciples imply.
There are additional further considerations, but usually with onlya restricted application (such as effects of local volcanic gaseousemissions, or of hardwater chemistry - i.e. containing dissolvedlimestone), but I do not have the space to cover them all.
The very small relative amount of radiocarbon present (a maximumof only fifty million atoms per milligram of carbon) means specialtechniques are needed to measure it. Furthermore, to be useful, themeasurement must be rather precise (an error of forty yearscorresponds to a measurement error of 0.5% in the 14C/12C value). Asystem of international calibration and checking of blind results hasbeen recently set up, to which more than 50 radiocarbon labs submitresults2. This enables, at least for the samples measured and thetechniques and routines used, individual laboratories to make anunbiassed comparison of their abilities, and to apply quality controlmethods to their work. All three laboratories engaged in dating theshroud have performed well in these intercomparisons (which wereinstituted after the shroud measurements). The fact that the shroudand control results, taken as a whole, compared well between allthree laboratories, despite their somewhat different approaches,argues very strongly for the validity of the measurements. It is anessential part of science that results can be replicated by otherscientists, and, without question, that was established in this case.The measurements were performed «blind» in order to preventclaims that the results had been biassed by wishful thinking orcollusion.
Two aspects for measurement need further mention. They are theassessment of error in the result, and the avoidance of contaminationin the sample which is being measured. Essentially, the error termthat the laboratory gives with its result is an estimate of thelikely reproducibility of that result if it were to be measuredrepeatedly. There are various terms which go into estimating theerror - often the largest is the fact that the number of 14C atomsdetected during measurement is subject to a statistical samplingerror (as in taking an opinion poll), and this contribution can beaccurately estimated. Other errors include various measuredcharacteristics of laboratory processes, such as the subtraction ofvariable background signals. Whether the error is generally beingcorrectly estimated or not can be checked, over time, by thelaboratory measuring known age or replicate material and looking atthe statistical distribution of results (for example, checking theextent to which they are Gaussian). The question of samplecontamination is more difficult to quantify. One tries to extract apure chemical compound for dating, but archaeological materials areoften complex, with degraded chemistry, and it is rarely possible tobe certain that all extraneous material has been removed.
Radiocarbon laboratories have built up a great deal of experiencein developing extractive chemistries appropriate to many individualtypes of material3, and the results can often be compared (e.g.dating bone and charcoal from the same stratum should give the sameresult, when each has been environmentally contaminated, andsubsequently de-contaminated in the laboratory by quite differentchemical methods.
This sample was provided, unidentified, together with severalother control samples of linen textile whose age was known (thoughnot divulged to the participating laboratories). The main questionfacing the laboratories was the best way to decontaminate the cloth.Most methods used were similar, although Oxford used an additionalstep of a lipid extraction in order to remove any grease or candlewax, as well as oxidatively removing lignins from the basic cellulosestructure of the cloth. In any case, essentially the same result wasobtained by all three laboratories for the whole group of samples,although taking only the shroud result, there was a juststatistically significant difference between Oxford's result and theother two laboratories (this is most likely to be due to anunderestimate - of 5-10 years - of the errors by the laboratories; inany case, in the context of the question whether the Shroud datecould be in error by centuries, the difference is negligible).
The method of measurement used, accelerator mass spectrometry(AMS), was necessary because of the small sample size available.Although at that time AMS was a less well-developed technique thanthat used for the majority of radiocarbon dates, the participatinglaboratories had already measured several thousand dates by the AMSmethod, and its accuracy, both then and subsequently, has been shownto be comparable to the best of the laboratories using conventionalmethods.
From a scientific point of view, the laboratories were in goodagreement overall, and in adequate agreement for the shroud datingitself. The fact that, when calibrated, the most probable date wasclose to the historical time when the shroud was definitely known toexist, while not strengthening the scientific case for the date, isnevertheless extremely powerful circumstantial support for it.
However, the date has been disputed as being wrong on variouscounts. Firstly, it is worth noting that small systematic errorscannot be completely ruled out; while unlikely, some contaminationwhich evaded the chemical purification methods is conceivable.However, if the main consideration is whether the shroud is 1stcentury AD or 13th century AD, the degree of contamination requiredto shift a 13th century date by 1300 years is very large (such ashift would require the addition of about 50% more material of«modern» carbon), and this quantity, or indeed any amountabove a few per cent, can be totally ruled out. A more subtle form ofcontamination, whereby the carbon atoms of the cellulose areexchanged or carboxylated with those from a hot carbon dioxideatmosphere, has been proposed as possibly having taken place duringthe fire to which the shroud was historically exposed when kept atChambéry. The scientific grounds for this proposal are, in myview, quite dubious, but nevertheless have been published in ascientific journal4, together with experimental evidence that such aneffect is possible, and so do require careful consideration. At leastthree experiments have been carried out to test the possibility, withone been published so far5, and none has been able to find any sucheffect. Experiments we carried out at Oxford put an upper limit toany possible change of the date of about 5 years. The issue isimportant, not just for the Shroud dating (although it is mostunlikely to affect the authenticity question, since a massive amountof exchange would be required to change the date by 1300 years), butbecause if such a process did occur during combustion events, itcould spoil the accuracy of many other archaeological dates. The carewith which radiocarbon laboratories have responded to even theoutside chance that some additional process which might feasiblyaffect the date, demonstrates the level of vigilance which is feltnecessary to ensure that radiocarbon dates are accurate.
Most other possibilities are less scientifically testable. Forexample, whether the shroud samples were irradiated by neutrons(which certainly could make a date appear younger than the truedate). This idea is difficult to disprove, but also difficult to takeseriously. In any case, to manage an irradiation so finely calculatedas to give a very credible historical date is a remarkablecoincidence.
Therefore, short of deception in the selection of the sample, orsome parapsychological influence, or some unique scientificphenomenon totally unknown and unguessed at by the scientificcommunity of radiocarbon dating at large, the date of the shroudcannot be significantly different from that of the publishedcalibrated radiocarbon results.
1 Damon, P. et al., Radiocarbon dating of the Shroud of Turin, in«Nature» 337 (1989), pp. 611-615.
2 Gulliksen, S. and Scott, M., Report of the TIRI workshop,Saturday 13 August 1994. Proceedings of the 15th international 14Cconference (eds. Cook, G. T., Harkness, D. D., Miller, B. F. andScott, E. M.), in «Radiocarbon», 2 (1996), pp. 820-821.
3 Hedges, R. E. M., 1992, «Sample treatment strategies inradiocarbon dating», in Radiocarbon After four decades: aninterdisciplinary perspective (eds. Taylor, R. E., Long, A. and Kra,R. S.). New York: Springer-Verlag, 1992, pp. 165-183.
4 Kouznetsov, D. A., Ivanov, A. A. and Veletsky, P. R., Effects offires and biofractionation of carbon isotopes on results ofradiocarbon dating of old textiles: the Shroud of Turin, in«Journal of Archaeological Science», 1 (1996), pp.109-122.
5 Jull, A. J. T., Donahue, D. J. and Damon, P. E., Factorsaffecting the apparent radiocarbon age of textiles: a comment on«Effects of fires and biofractionation of carbon isotopes onresults of radiocarbon dating of old textiles: the Shroud ofTurin» by D. A. Kouznetsov et al., in «Journal ofArchaeological Science», 1 (1996), pp. 157-160.