A recent interview with Giulio Fanti (‘Shroud Wars: Panel Review (Part 3B) – Scientific Dating Evidence,’ at youtube.com/watch?v=2YHkaqLawTw&t=2796s) reiterates his opinion that the chances that Byzantine coins showing the face of Christ did not derive from the Shroud are billions to one against. In the transcript accompanying the podcast, he says:
“The new probabilistic model produced the following result: the Byzantine engraver had only 1 in 10 billion billion chance of getting that face on the Justinian II coin without having seen the image of Jesus on the Holy Shroud. However, the range of resulting values varies from 5 probabilities in a billion billion billion to 2 probabilities in 10 million billion […] This confirms that the Byzantine engravers, who molded the face of Christ on the gold solidus of Justinian II, certainly saw the Holy Shroud or a good copy of it.”
Without having the ‘new probabilistic model’ in front of me just at present, I’ll have a look at the previous one, which was published in ‘Numismatic Dating of the Turin Shroud through the Analysis of Byzantine Coins,’ which can be found at academia.edu. In Section 6 (P[r]obabilistic Analysis Applied to One Coin), Fanti selects a typical example of one particular style of minting, finds twelve features on it which correspond to similar features on the Shroud, and by assigning probabilities to each one and plugging them into ‘Bayes formula,’ finds that the probability of the engraver not having modelled his coin on the Shroud is “seven chances in one billion of billions” against.
This is the type of coin Fanti chooses:
And this is his specific example, enlarged on the left, and merged with the Shroud on the right:
The lower two images, truncated here, show the Shroud successively clearly.
Already a significant bias has emerged. Some features supposedly of the coin can only be seen on the merged image, and vice versa, and some have been seriously distorted. A line through the moustache (8) is completely missing on the coin, and a W-shaped beard on the coin has been reduced to an inverted U for comparison with the Shroud. It is arbitrarily assumed (in contradiction to Diana Fullbright’s paper ‘Forelocks in Early Christian Tradition’ at shroud.com/pdfs/n59part4x.pdf) that the epsilon blood-mark was represented as two or three little wisps of hair. The engraver, of course, could not have copied from the negative, with all the illusory depth and detail we have come to recognise, but must have used, if he used it all, the faint sepia image on the cloth
Fanti arbitrarily assigns a probability to each of twelve features on the coin; the probability that the craftsman could have included it in his portrait without reference, either by direct observation or from a copy, to the Shroud. Although he admits his probability is of necessity arbitrary, he has some confidence that he cannot be far wrong.
So let’s have a look at them, shall we?
Here we have Fanti’s coin, coloured coded to help the discussion below, a rather better version of the same style of coin, and the Shroud, slightly enhanced in contrast for clarity.
1. HAIR (wavy blue lines)
Fanti describes this as: “Wavy shoulder-length hair, longer on the left side, similar to “payot”, the side curls [of] Orthodox Jews. The right side of the hair is less bushy, maybe because partially torn, like the beard.” It is Fanti’s opinion that that there is only a 1% chance that the hair on these coins was not based on the hair on the Shroud. In my opinion almost the opposite is true. I don’t think anybody with any familiarity with the Shroud at all could possibly have decided that wavy tresses of hair, curving in a quasi-circle around a wide face with prominent ears, was an appropriate way to copy it. Some commenters see a kind of bulb of hair on the right side of the image, which Fanti also sees on the right side of the coin. This would suggest that the craftsman saw the Shroud as a direct image of Christ, like a painted portrait, and not as an imprint, where left and right would be reversed.
2. ARCHED EYEBROW (red arch above eyes)
“High-arched left eyebrow due to a hematoma,” and assigned an improbability of 98%. I think this is confused. For a start the mark on the coin looks more like an imperfection such as a scratch or dent derived from usage, and for a second the elevated eyebrow, if such it is, appears to be on the right side of the coin, while any suggestion that one eyebrow is higher than the other appears on the left side of the Shroud image. This is sometimes explained by the craftsman realising that the Shroud image is an imprint, and reversing its features for his portrait. This is the opposite of what we observed with the hair, and is a classic example of the “have-your-cake-and-eat-it” argument, which is often invoked in considerations of whether the Shroud derived from the Pantocrator image, whereby the craftsman can either simply have copied the Shroud, or reversed left and right, knowing that the Shroud image was an imprint. This occurs arbitrarily with different features, and gives an enthusiast two opportunities to find a comparison between a Byzantine image and the Shroud, selecting whichever fits best.
3. “REVERSED 3” (red lines at top of brow)
“At least one lock of hair at the centre of the forehead referring to the reversed “3”-shaped forehead wound. It must be remembered that the Trullan Council forbade to reproduce the signs of the Passion, so the bloodstains were [deliberately] confused with locks of hair.” Diana Fullbright demonstrates that numerous images not of Christ also include these locks, and numerous images of Christ don’t. Also there is nothing in the Canons of the Trullan Council to suggest that signs of the Passion were forbidden. To Fanti, there is only a 1% chance that the craftsman did not derive his forelock from the Shroud. To me, there is a much greater improbability that he did.
4. EYES (yellow squares)
The eyes are not given any particular description in Fanti’s paper other than “big,” but are given a 90% probability that they derive from the Shroud. They are little buttons in square sockets. The Shroud image is nothing like them at all.
5 – 6. RIGHT EYE’S CONTUSION and SWOLLEN CHEEKBONES (no emphasis on the image above)
The cheek swellings are classic examples of the “have-your-cake-and-eat-it” argument, and can be interchanged left/right according to which attribute most resembles the Shroud. Fanti mentions “protruding cheekbones, the right more than the left one, because of the swelling caused by the blows.” However, many of these Byzantine coins do have asymmetrical cheeks of one kind or another, and so does the Shroud image. This is the first of Fanti’s features that stimulates any curiosity, although I don’t think it is evidence that the coins must have derived from the Shroud. A huge number of portraits, Byzantine mosaics and everything else, have asymmetrical cheeks, even if they are full face in aspect. And it is perfectly likely that a medieval craftsman could give his portrait bruised cheeks without reference to a particular previous image. Nevertheless, Fanti gives them improbabilities of 5% and 2% respectively.
7. NOSE (no emphasis on the image above)
“Broken nasal cartilage, therefore nose is slightly flattened and turned to the right side,” Fanti says. In his own reference coin, the nose has been worn flat, not modelled that way, and in the central image, it is neither flat not turned. Fanti thinks the probability that the coin craftsman did not see the Shroud is 2%.
8. MOUSTACHE (purple arch)
The moustache is thin, and sharply downturned. It looks nothing like the Shroud moustache at all. Fanti thinks the probability that the coin craftsman did not see the Shroud is 2%.
9 – 11. SPARSE BEARD ON THE RIGHT SIDE, BEARD’S SHAPE, & BEARD’S GAP BELOW THE LOWER LIP (green)
Chances that the craftsman had not seen the Shroud, 1%, 2% and 20% respectively. The real beard is certainly bipartite, as are many similar Byzantine beards, not just Christ’s, but there is no Shroud justification for the coins’ very prominent cheekpiece on the man’s right, or the slightly less prominent cheekpiece on his left. This is as likely to be artistic freedom as most of the other features, and cannot be said to derive from the Shroud image at all.
12. SAME FOLD UNDER THE NECK AS THE EDGE OF THE GARMENT
I’ve not bothered with an image for this. I have previously demonstrated that the fold visible in the Enrie photograph is not present in the Pia photograph, and was an artefact of the rolling up of the Shroud between the 1898 and 1931 ostentations. Although Fanti thinks the chances of the hem of Jesus’s tunic not having been copied from the Shroud are 10% (which is generous for him), in fact there is no possibility that it could have been at all.
Fanti then wanders into Bayesian probability, but really he might just as well have multiplied all his improbabilities together, which comes to a chance of about 3 billion to one against the craftsman having operated independently. This is ludicrous. Let us make our own assessment, based on a craftsman with a blank slate, and a picture of the Shroud next to him, and instructions to copy it.
- Hair. It is most unlikely that this was copied from the Shroud. 10 to 1 against.
- Arched eyebrow. This is mostly unnoticeable. 2 to 1 against.
- Reversed “3.” Quite impossible. 100 to 1 against.
- Eyes. Very poor copies if they are copies. 2 to 1 against.
- Right eye contusion. Just possible. Evens.
- Left eye swelling. Not noticeable. 2 to 1 against.
- Nose. Looks like any nose. Evens.
- Moustache. Nothing like the Shroud. 5 to 1 against.
- Beard weaker on the left. Possible. Evens.
- Beard forked. Quite likely. 2 to 1 for.
- Gap below lower lip. Quite likely. 2 to 1 for.
I’ve not bothered with 12, as adding infinity or zero to the list of multiplicands is not helpful. The probability of Fanti’s chosen coins having been copied from the Shroud is about a thousand to one against.
Bizarrely, there are several Byzantine designs that look a lot more like the Shroud than the ones Fanti chose for his comparison, but although some features of some of them are common to the Shroud, to claim that any of them must have been copied from it is absurd. In fact, it is obvious that the image on all the coins derives from Pantocrator iconography, dating back to the 4th century or so, and probably specifically from one particular Pantocrator image. I shall explore this in another article.
————————-
Another common statistical error, which Fanti really shouldn’t be making, is that of relying on “points of congruence” between two images, and imagining that a certain number guarantees identity. These were originally championed by Alan Whanger (‘Polarized Image Overlay Technique: a New Image Comparison Method and its Applications’, Applied Optics, March 1985) and rarely re-assessed, either in numerical or qualitative terms. What exactly a “point of congruence” is has never been defined. Some authenticists claim that sixteen or so points of congruence is all that is needed for two fingerprints to be considered identical, and others that thirty or so are enough to identify a suspect caught on camera. The number quoted varies, and the truth, of course, is that it is all nonsense. What follows may not be what Alan Whanger or his followers exactly meant, but in the absence of their explaining, or even knowing, what they meant themselves, it is as good an interpretation as any.
Here is a random scribble (blue), a freehand copy (green), and another random scribble (red). Below are two overlays, of the copy and the second scribble, on the original. Points of congruence are marked in grey.
The green copy, it turns out, has fourteen points of congruence with its original, and the random scribble has nineteen. Following “points of congruence” logic, one must infer that the red scribble is more likely to have been derived from the blue one than the green copy.