“16th Century Weavers were Magicians.”

A recent update at academia.edu details Joe Marino’s latest research into his “invisible mending” hypothesis. ‘Further Empirical Data Indicating Repairs in the C-14 Sample Area of the Shroud of Turin,’ drags in the Bayeux Tapestry, Roman tunics, and Medieval tapestries, Joe’s pet peeves (that everybody who disagrees with him is either a lunatic or a liar, and that peer-reviewed papers are valueless if they disagree with him but the stamp of invincible authority if they agree), and a couple of outdated and long disproven “evidences” (the Quad Mosaics and William Bogolub’s Frenway Invisible Weaving), together with some less questionable evidence, to create a dense smokescreen behind which the real weakness of the invisible mending hypothesis – that it’s invisible – hopes to escape undetected. No chance.

The paper has been the subject of considerable discussion at academia.edu, so although I have contributed there, I think a detailed analysis is merited. As so few people will read either I don’t think the occasional duplication will matter!

Perhaps I would not dwell quite so comprehensively on Joe’s publication if I were not alluded to several times in it, never by name, of course, as one for whom “psychological factors” are the real obstacle to my accepting Joe’s thesis, rather than the fact that there is insufficient evidence for it. In this I join the vast majority of authenticity-doubters, from Bishop Pierre d’Arcis through Ulysse Chevalier and Walter McCrone to Michael Tite, all of whom, it seems, are mad, bad and dangerous to know, rather than simply unconvinced.

There is some evidence for the hypothesis, of course, so it is a pity that it is buried among obfuscation and accusation, which will deter genuine students of the Shroud from taking it seriously. An early paragraph is devoted to the failings of the peer-review system of editorial consideration, culminating in a quote from Bob Rucker (based rather loosely on a contentious article in PLoS Med, in 2005, by John Ioannidis called ‘Why Most Published Research Findings are False’) that “about one-half of the papers published in peer-review journals are basically wrong.” Curious, then, that during the course of his paper, Joe mentions Ray Roger’s “peer-reviewed” paper in Thermochimica Acta, his own “peer-reviewed” article in Chemistry Today, Riani and Atkinson’s “peer-reviewed” paper in Statistics and Computing, Tristan Casabianca’s “peer-reviewed” paper in Archaeometry, Paolo di Lazzaro’s “peer-reviewed” paper in the Journal of Imaging Science and Technology, and Larry Schwalbe’s “peer-reviewed” paper in the Journal of Archaeological Science. By the law of averages, three of these, it seems, are “basically wrong.”

My own special mention, apart from a comment that I admitted I would be hard put to counter every one of Joe’s mountain of assertions and references he thinks support his hypothesis, concerns the Quad Mosaic images:

“One shroud sceptic has pointed out that all four corners of the Shroud are green, but it’s certainly possible that if the C-14 corner had been manipulated, the other three corners could have been repaired simultaneously.”

For those unfamiliar with them, here are the four Quad Mosaic Images positioned as they were taken, in their original colours and, below, slightly enhanced for clarity.

Quad Mosaic images from shroud.com

As can be seen, the skeptic did not point out that all four corners of the Shroud are green, but that all four Quad Mosaics have green corners. If it be true that a green corner indicates a medieval repair, then it is strange that these medieval repairs were so neatly spaced out along the length of the Shroud; but worse, if it be true that a change in colour represents a change in chemical composition, then what are we to make of the two red and three blue bands right across the cloth, including Jesus’s face? In truth, of course, the colours of these images are dominated by artefacts of the lighting conditions under which they were made, so that the actual difference in colour between the different areas of the Shroud makes considerably less contribution to the result.

It is a pity that Joe uses these images as evidence of medieval repair, as their obvious irrelevance detracts from the simple observation, that anybody can make, that the radiocarbon corner is indeed noticeably darker than the rest of the Shroud in ordinary photographs. Unfortunately, this grubbiness does not support Joe’s hypothesis. He quotes an extremely detailed and intricate analysis of the area – of a UV fluorescence photo rather than a visible light photo but the effect is the same – by John M. Morgan III, which actually refutes it. (See: ‘Digital Image Processing Techniques Demonstrating the Anomalous Nature of the Radiocarbon Dating Sample Area of the Shroud of Turin,’ Scientific Research and Essays Vol. 7(29), 2012). Joe’s assessment of this paper, “the author concluded that there is evidence of cotton in the C-14 sampling area,” is simply wrong. Morgan took the idea that there was cotton in the radiocarbon corner, and speculated that the colour variation of the UV photo could relate to it. Had he been told the anomalous material was nylon, his findings would have been exactly the same. His conclusion actually included,

“Benford and Marino (2008) demonstrated that the C-14 samples were taken from an area of the TS that had been expertly restored by medieval artisans by reweaving a cotton patch into the original linen of the TS. According to Brown (2005), the artisans then attempted to dye the newly rewoven area to match the aged appearance of the remainder of the TS. The image segments created using the first component of the PCA (Figure 14) may provide a way to identify unique chemical composition features on the TS. Also, the unsupervised classification image produced via K-MEANS clustering (Figure 13) may provide a way to identify the exact area where the cotton patch and the portion of the TS adjacent to the cotton patch was dyed to match the older linen material of the TS.”

The least confusing, and clearest indicator of the putative “way to identify” either the patch or the dye is Figure 12 in the paper.

Figure 12. Z-score image calculated from the first component of the PCA
with overlay of the C-14 sample area.

The green area represents either the patch or the dye, or both, and to make it even more obvious, here are Figures 13 and 14, referenced in the quote above:

From Figures 12 and 13, which have the radiocarbon sample areas overlaid, it is very clear that any apparent anomalies, patch, dye or both, increase in proportion towards the base of the image. This enlargement of the tested portion only, demonstrates it very clearly indeed.

The pale blue and yellow component of this image increases proportionately from the top (Arizona) portion to the third (Oxford) segment. So clear and irrefutable was this demonstration that it enabled Robert Schneider to say:

“When you compare the laboratory carbon dates with the Z-Score resulting from a Principle Component Analysis (PCA) of the UV-Fluorescence data you get a coefficient of determination of 0.9986 which corresponds to a correlation coefficient of 99.93%, an extraordinarily high value given that what it says is that the average carbon date is almost perfectly predicted by the UV-Fluorescence.”

‘Dating The Shroud Of Turin: Weighing All The Evidence,’ St Louis Conference, 2014

However, the dates of the samples actually get older across the same stretch of cloth, so if it is true that the observed colour difference reflects the proportion of admixed “cotton patch” or “dye,” then the patch or dye make the apparent date of the Shroud older, not younger, than its actual date. This is a powerful refutation of the “medieval repair” hypothesis, not a justification, but Joe quotes it just the same.

In similar vein, Joe is proud to announce that he has three books about invisible reweaving, “Invisible Mending” (1951), “Invisible French-Re-Weaving Simplified” (Two vols, 1954), and “The Frenway System of French Reweaving” (1962). The first two may, for all I know, be serious accounts of the history or techniques of an ancient skill, but the third is a popular manual dispatched, with appropriate tools, to anybody who signed up for William Bogolub’s enterprising mail-order “Guild” to enable people to master a technique which would provide them with endless employment, at home or in a business. His method, the introduction explains “does not require formal education, does not require great strength or unusual physical attractiveness, does not require expensive office space or elaborate equipment, and does not require a special knack for dealing with people.” There is an inevitable inference that his marketing was targeted at precisely this kind of person, and not surprising that the preamble to his instruction book was considerably overhyped.

“Probably the reason this art of reweaving has gone relatively unnoticed is the great secrecy which has heretofore kept all but a few people in the world in ignorance of the techniques involved. These secrets have been closely guarded handed down from generation to generation to a select few. The only exceptions were people who paid huge sums in order to receive knowledge of the art. Every novice reweaver had to spend years as an apprentice.”

This is salesman’s puff, specifically written to encourage his buyers to feel that they have acquired, at little cost, an esoteric secret that will enable them to earn huge sums of money themselves. For Joe to quote it as historical evidence is misguided. In 1955 Bogolub was constrained by a court order not to make the claim that anybody could make “huge sums” from his course.

That’s not to say that there was anything wrong with the technique or the tools provided. “The Fabricon Invisible Reweavers Guild” was popular for about fifty years, and even today sets of tools, members’ badges or the instruction manual are frequently offered for sale on eBay. However, what is abundantly clear is that, from the point of view of a Shroud student, this invisible reweaving method is not, under close scrutiny, invisible. The interwoven threads must, for some distance above, below and to either side of the hole to be patched, run alongside the original thickness, doubling the density of the cloth, and the ends of both original and new threads must be visible on one side or the other. Several companies specialise in this technique, and I commissioned two to reweave holes in some fabric I sent them. The better of the two was the work of Michael Erhlich’s company, Without A Trace, of Chicago, and here it is from both sides:

Each photo is about five centimetres across, and the hole about one centimetre.

This is astonishingly fine work to someone not thoroughly familiar to the task, and I’m not surprised that it takes hours and is very expensive. However, if the Shroud had been repaired like this, it would, like this, have become apparent under microscopic scrutiny. Joe counters the fact that the mend is clearly visible from behind with a quote from Jeanette Hauser’s 1954 book:

“Occasionally you may be asked to reweave a damage invisibly on both sides. […] Invisibility is achieved on both sides by applying the face-weaving method to the back side of the fabric, with a slight variation in the lock-in procedure. After pick backs have been made, the needle is inserted some distance to the right of the first one, and then glided thru the fabric between the top and bottom surfaces. The needle must be kept hidden within the fabric until brought out at the pick- back. On the left side of the damage the procedure is reversed. The needle enters at the point of lock-in, and then glides thru the fabric and comes out some distance to the left. Thus when the protruding ends of the replacement thread are clipped close to the surface the lock-ins will not show from either side.”

Without Hauser’s book to hand, it is not easy to identify the precise meanings of her “pick-backs” and her “lock-ins,” but I think the last sentence is enough for us. The protruding ends of the replacement threads are clipped. However neatly, and however closely, there is no doubt that they will still be visible under a microscope, and no such ends, nor the doubled thickness of the threads where they run alongside each other, are observable on any of the photos of the radiocarbon samples.

So has Joe any evidence at all? Indeed he does, and it must be examined carefully. Thibault Heimburger and Giulio Fanti independently took small length of threads from the radiocarbon corner apart, and found that there were cotton fibres spun in amongst the flax. Heimburger’s analysis is at shroud.com/pdfs/thibaultr7part1.pdf (also part2 and part3), and Fanti’s at shroud.com/pdfs/fantir7appendix.pdf. Both authors admit that their conclusions cannot be precise because of the difficulty of observing and identifying every fibre, but suggest that about “10%-20%” (Heimburger) and “2.1%” (Fanti) of the thread fragment they examined was cotton. If it could be demonstrated that only the radiocarbon corner contained threads blended with cotton, then that would be good evidence for interpolation, but no threads from other parts of the Shroud have been analysed similarly.

I have already addressed (see ‘FTIR and the Shroud’ on this site) Alan Adler’s comment….

“The patterns […] are all distinguishably different from one another, clearly indicating differences in their chemical composition. In particular, the radiocarbon samples are not representative of the non-image samples that comprise the bulk of the cloth.”

…. and I’m afraid I find him over selective. In ‘Further Spectroscopic Investigations of Samples of the Shroud of Turin’ (The Orphaned Manuscript), he publishes 34 FTIR spectra from the Shroud, all generically similar, including 12 from the radiocarbon sample and 4 “non-image.” In ‘Updating Recent Studies on the Shroud of Turin,’ he selects one of each, and notes a clear difference between them. In fact there is much more variation among the spectra from each area than there is between them.

Of the rest, little needs be said. Joe finds a Roman tunic with cotton blended amongst the flax, an interesting comment about the seams of the Bayeux Tapestry to the effect that “the seams are almost undetectable on the upper side,” which is irrelevant to the Shroud, and several comments about the patches sewn on by the Poor Clares in 1534, and the stitching of the Holland cloth, which are irrelevant to the radiocarbon corner.

Towards the end of the paper, Joe assembles his evidence, not in terms of observations, but in terms of the “multiple experts” whose observations and opinions “all point in one direction.” Well, obviously. His army includes people who have never inspected the area closely, and even “experts” who have inspected it and explicitly deny there was any mending! In 1988, Gabrial Vial (Centre International d’Etude des Textiles Anciens) and Franco Testore (professor of Textile Technology at the Polytechnical School of Turin) examined several areas of the Shroud very closely before deciding where to take the sample from, and Giovanni Riggi di Numana actually cut it. In 2002, Mechthild Flury-Lemberg unstitched the backing cloth and replaced it. All these people were specifically looking for unrepresentative areas of the cloth and had the experience and expertise to find them, but didn’t, and said so. A fragment of the Arizona sample was studied in microscopic detail by Rachel Freer-Waters and Timothy Jull, who found no evidence of blended thread. Joe’s response to their opinions is instructive. Of Vial and Testore, Joe says: “Some have claimed that if there had been a repair, they would have detected it.” Well, quite. So they would. Of Giovanni Riggi, Joe quotes some comments Riggi made about the Holland cloth as if it was about the radiocarbon corner. Of Flury-Lemberg, “we have already seen evidence that seemingly contradicts that assertion and more will follow,” followed by the reasonable statement that, “if it is asserted that an invisible repair can’t be done on linen, it is sufficient to disprove by finding one instance in which it was.” Quite so, but no such instance has been found. And finally, of Freer-Waters and Jull, Joe accuses them of deliberately hiding an incriminating piece from Barrie Schwortz so that repairs would not be made public. “Might Arizona have discovered that the sample showed some indication of repairs?” Mad, bad, and dangerous to know.

What about Joe’s supporters? The three people to whom he originally sent a photo of the Zurich sample replied:
David Pearson (French Tailors, Columbus, Ohio): “It is definitely a patch.”
Donna Campbell (Thomas Ferguson & Co. Ltd): “There is something different left versus right.”
Louise Harner (Albany International): “The float is different on either side of the sample.”

Here is the photo Joe includes in his paper, and a better version of it published elsewhere:

Without being able to inspect different areas of the Shroud, it is not possible to say if these observations are unique to this area, or whether, as is in fact the case, different diameters of thread, often quite extreme differences, can be found all over it.

Finally, two experienced experts on medieval tapestries have not actually seen any invisible mending because… its invisible.

Robert Buden, President of Tapestries and Treasures: “Is there such a thing as an invisible repair? Yes, I have seen it, or more appropriately, not seen it, in several types of textiles…”

Thomas Campbell, Associate Curator, Metropolitan Museum of Art: “It is very difficult to identify such repairs. They certainly must exist – the 16th century weavers were magicians, but I can’t think of any documented examples.”

On the face of it, these quotes seem slightly silly: how can these experts be so sure that invisible mending exists if they’ve never seen any? but I don’t think that’s fair. Their expertise was in tapestries, in which, of course, the weave of the fabric is always invisible. Where they become threadbare, or the backing textile frays, they can repaired – and contemporary documentation mentions repair – such that although the fill-in stitching is perfectly obvious from behind, the replacement colour thread can match the old perfectly.

Much is made of Ray Rogers’s (“peer-reviewed,” of course) paper in Thermochimica Acta, specifically his finding of cotton threads on some sticky tape samples of the area, and his discovery of “a plant gum containing alizarin dye present in two forms” coating the threads he examined. Of the cotton, I have spoken above. The coating, however, is interesting, and I’m afraid I dispute Rogers’s observation that “there was absolutely no coating with these characteristics on either the Holland cloth or the main part of the shroud.” The sticky tape slide of the Holland cloth, labelled 1FH, is thick with what appears to be mould growing on a organic matrix, red particle agglomerates, and particle covered fibres, as can be seen on Eugenia Nitowski’s collection of microphotographs.

Representative photos from 1FH, at x325, x325 and x125, by Eugenia Nitowski

Nothing quite like it is seen throughout the rest of her collection. What’s more, visible-light and UV-fluorescence photos of the area show that the white, fluorescent material of the Holland cloth is completely obscured by some kind of covering, which must have been applied after the Holland cloth was stitched on, as its original whiteness is very evident where the Raes sample was cut away.

The Radiocarbon corner in visible and ultra-violet light, both by Vernon Miller.
The area of Holland cloth exposed by the cutting of the Raes sample is very evident.

THE BOTTOM LINE

The evidence that there is interpolation of 16th century threads into the 1st century Shroud boils down to three things.
One: surface contamination, such as can be seen on photographs and was detected by Ray Rogers, can be rejected on the grounds that there isn’t enough, and that what there is appears in inverse proportion to its putative effect on the date.
Two, that the radiocarbon corner contains cotton whereas the rest of the Shroud does not, has never been tested, let alone confirmed.
Three, “sixteenth century weavers were magicians.”