Thursday, July 30, 2009

Master's Thesis: Chapter 2.5 An analysis of Stephen Carlson's handwriting analysis

Are there really "signs of a forgery" to be found in the handwriting of the Theodore-letter? The first question we need to ask concerns the applicability of the principles of QDE (Questioned document examination) for assessing the handwriting of an 18th-century MS - originally, QDE was created to identify 20th-century handwritings, produced while using 20th-century equipment.1 The Theodore-letter, at least in its appearance, looks to be copied during the 18th-century, with (presumably) 18th-century equipment. Are these methods of spotting a forgery useful for all kinds of handwritings?2 Furthermore, how should the fact, that Morton Smith, if he really did forge the MS, did not have to imitate any one specific handwriting by any one specific person, but did need to write in a general way as an 18th-century copyist would have written, be taken into account? In the famous QDE-cases - in which the matter is ultimately settled by the problems found in the handwriting - the situation seems usually to be about the imitating of a specific handwriting of a specific individual.3

If we accept the applicability of QDE to the assessment of an 18th-century MS4, we have another question to settle: the sources, Morton Smith's B/W photographs he took in 1958 with his pocket camera, that Carlson uses for his analysis. Professional forensic document examiners prefer the original documents, because photographs (or worse yet, photocopies) can bring their own extras e.g. to the line quality of the handwriting. A professional in the field of QDE, Hannah McFarland, warns his clients of the dangers of not using original documents: "Photocopies do not reproduce the finer features of handwriting such as pressure, ductus, and line crossing."5 Since photocopying and Smith's photographing in 1958 use similar techniques - both are based on the information carried by light photons, though this information may never be captured in its completeness - the warning chimed by McFarland about the dangers of using photocopies when assessing the handwriting, has to be taken seriously into consideration with photographs as well. Only the originals are good enough for QDE analysis.

With these preliminary remarks out of the way - keeping in mind that every further conclusions drawn from the photographs applying the principles of QDE are not resting on a solid foundation - we must turn our attention to the comparative material Carlson provides in his book from the other 18th-century MSS from Mar Saba. For some reason, every one of these seem to have lots of the same "signs of a forgery" present as what Carlson claims to find in the Theodore-letter. To take some specific examples, Carlson observes the sign of the cross, in the beginning of the Theodore-letter, to have been drawn very slowly, with two strokes in such a way that three out of four endings of the line (according to Carlson, all four endings of the line) have blunt ends.6 However, from the comparative material provided by Carlson, I see similar blunt endings and beginnings of the line in e.g. Sabas 518 MS, folio 2, recto, line 5 in the letter tau of the word τοῦ, and in the letter lambda of the word δούλου, and in Sabas 523 MS, folio 4, recto, line 11 in the letter pi of the word πνικάς, and in the letter lambda of the word πόλε.7 When Carlson sees the lifting of the pen in the first line of the Theodore-letter, in the transition from the letter epsilon to the letter kappa of the word ἐκ, I see a similar blot of ink in e.g. Sabas 452 MS, folio 1, recto, line 11 in the word ἐκ.8 The fixing of the letters afterwards is noted by Carlson e.g. in the first line of the Theodore-letter, in the stigma ligature9, but I do not see any essential difference when comparing this ligature to the Sabas 523 MS, folio 4, recto, line 11 letter pi of the word πόλε.10 From these observations we must ask, if all these "signs of a forgery" are in reality conformances with the law, naturally present in MSS for reasons relating to their conditions of copying, e.g. in the leaving of the pen point to the MS or lifting of the pen in the middle of writing a letter or a combination of letters, while turning one's gaze to look at the original text?

The only empirical evidence readily available for assessing this question comes from a research conducted in blogosphere in 2005 and 200611, where a group of scholars, pastors, and students all proficient in ancient Greek, attempted to copy by hand (without accents or punctuation) the Second Epistle to Timothy. The results of this experiment may grant some necessary perspective for the judging of the handwriting of the Theodore-letter: some letters were harder to copy than others, and for different individuals different letters proved to be the more laborious; gradual descent into numbness due to working for a long period ended up with a much increased tendency towards making errors, and with a need to correct letters afterwards, which in turn transformed to ink blots and problems (shakiness) in the quality of the drawn line; regularly, the first letters written just after the pen had been dipped into the ink bottle came out as containing too much ink, which in turn produced more ink blots over the lines of the letters; the numbness due to working for a long period manifested itself also with diversity in the use of nomina sacra when the work had began to be mechanic without an effort to comprehend the meaning of the Greek text; in addition, the placement of the copied text (the attempt to get it into organized lines) had an effect to the way words were abbreviated and ligatures exploited, and the beginning of the text contained the bulk of the textual variants.12 And, if the copyist notices an error he has just drawn, the only option is to continue forwards - the ink must be given time to dry completely before any attempted corrections, otherwise the text becomes virtually unreadable with all the extra ink making a total mess of the passage.13

Another welcome point of view is provided by Roger Viklund in his 2009 online-article "Reclaiming Clement's Letter to Theodoros: An Examination of Carlson’s Handwriting Analysis", where Viklund analyzes the handwriting of the letter from the colour photographs that Father Kallistos Dourvas took in the late 70s (as opposed to the B/W photographs Carlson uses for his analysis). After scrutinizing the quality of the line of the handwriting, Viklund concludes that - following the criteria spelled out by Carlson for the "signs of a forgery" - these signs are to be found everywhere, and the argument proposed by Carlson that the text gets better towards the end of the letter14, does not hold water. The observations concerning "the forger's tremor" are also difficult to substantiate, for Carlson does not provide any criteria for defining just when a correct case of "the forger's tremor" is spotted. In addition, the observation made by Carlson, that there are incidents where the point of the pen has been lifted from the paper in the middle of a letter drawing, cannot be confirmed by Viklund: for every case reported by Carlson, it is not, according to Viklund, undoubtedly clear that the lifting of the pen has actually happened, and for some of the cases Viklund cannot see anything even remotely resembling the possible lifting of the pen. In the end, the argument proposed by Carlson, that the forger drew the letters slowly at first (this would manifest itself in the forger's tremor) but had the capability for some precise correction of the letters afterwards, cannot be the case here: if the forger had the ability to make precise corrections, why would there be a problem in the quality of the line in the first place? Viklund comes to the conclusion that the handwriting of the Theodore-letter derives from the 18th-century, from the hand of an elderly person.15

The "signs of a forgery" present in other 18th-century Mar Saba MSS16, the (suggestive) results of the blogosphere experiment17, and the observations made by Viklund about the regularity of the "signs of a forgery" and the subjectivity of assessing "the forger's tremor", put all together, suggest that the methods commonly used as part of QDE are not directly applicable for assessing the authenticity of MSS.

We need to draw two more question marks over Stephen Carlson's comparison between the handwriting of the Theodore-letter and the other 18th-century MSS from Mar Saba. The observations made by Carlson, like the varying use of the nomina sacra, are probably irrelevant, as Scott Brown has argued, for uniformity in the handwritings practically disappears in the East after the 15th-century.18 In addition, the monastery of Mar Saba was an exception - this point is relevant only if the text of the Theodore-letter was copied into Voss' book in Mar Saba and not somewhere else - in its choosing of applicants, for only experienced monks were admitted, monks that had already developed their writing habits (and, consequently, their handwriting) somewhere else. The divergence in the handwritings of the monks in Mar Saba must have been exceptionally great.19

Also, the remark made by Carlson, that the Theodore-letter is the only MS written by a very narrow pen point, may not be an anomalous sign of a forgery, either. Had the writer of the Theodore-letter used a pen point as wide as the writer of the aforementioned Sabas 518 MS, the text could hardly have been made to fit in the back pages of Voss' book, since its dimensions were only 198x148x23 mm - a small-sized text may be produced only with a small-sized pen point.20 On the other hand, a small-sized text may more easily have problems in the quality of the drawn line (ink blots etc.), interpreted as "signs of a forgery", when the observer is especially looking for these "signs".21 The bottom line, at this point, is simply that Carlson's analysis is not very persuasive in itself when he compares the Theodore-letter to the other 18th-century MSS from Mar Saba. Next we need to turn our attention to the claimed similarities between the handwriting of the Theodore-letter and the handwriting of Morton Smith, the purported forger of the letter.

Unfortunately, in Stephen Carlson's comparison of the handwriting of the Theodore-letter and the handwriting of Morton Smith, he cuts too many corners short.22 First of all, Carlson's comparison material from Smith is too limited in its size, because "as a working minimum, four or five pages of carefully selected continuous, natural writing" is required for a reliable analysis between two different handwritings.23 Carlson, however, instead of four or five pages, confines himself to two pages that feature Smith's handwriting in Greek in marginal notes.24 Greater care would have been preferable, since "under no circumstances... can identity be established by one, two, or even several 'unusual' characteristics".25 Instead, a positive identification of the author of two handwritings can be made only when the both handwritings in question have "a combination of a sufficient number of points of agreement without any fundamental dissimilarities".26 When two writers are using the same alphabet, there must necessarily be some similarities in their forming of the same letters - otherwise, the correct identification of the letters would become a chore. This is perfectly understood in QDE, too, and the standards for identifying two different handwritings as deriving from the same author are strict: there has to be numerous small, but extremely individualistic characteristics present in both handwritings - a general similarity found in the handwritings is not enough for establishing their derivation from a single author.27 As Scott Brown has argued, the keywords here seem to be "highly individual habits".28 How clearly are these to be found in Carlson's comparison of the letters lambda, tau, and theta between the Theodore-letter and Morton Smith's own handwriting?

In his 1951 work, "Tannaitic Parallels to the Gospels", Morton Smith wrote every Greek letter with his own hand. This large sample of Smith's handwriting, augmented with a letter by Smith from 1955, and with marginal notes - Brown performs a thorough comparison of the handwriting of the Theodore-letter and the 1951 handwriting of Smith. Regarding the letter lambda, Smith seems to prefer writing it in two strokes (from the sample Brown can distinguish only eight instances where the letter lambda looks as if written in one stroke only), whereas the majority of the lambdas in the Theodore-letter look to have been written in one stroke. In Smith's two-stroke lambdas, when the first stroke has produced the hat and the right leg, the left leg of the letter is situated in relation to the middle point of the first stroke as follows: beneath the middle point 55%, approximately at the middle point 35%, above the middle point 10%. In the Theodore-letter, the point of contact is beneath the middle point in 96% of the cases.29 Consequently, the similarities between the lambdas in the handwritings cannot be due to a lapsing back into the forger's own handwriting, as Carlson suggested, because the forger, following the principles of QDE, would do this "progressively at various points", but not all the time.30

In the lambdas written with a single stroke there is even less similarity present. Smith's one-stroke lambda leans to the left, has a height comparable to other "high" letters, has not much retracing after the drawing of the right leg before commencing the left, and the hat and the left leg do not twist to point at each other. The one-stroke lambda in the Theodore-letter, on the other hand, leans to the right, has a height comparable to a "short" letter, has a long retracing after the drawing of the right leg before commencing the left, and the hat and the left leg are curled to point at each other.31 In the same vein, there is not much regarding the letters tau and theta for establishing Morton Smith as the author of the Theodore-letter. A one-stroke tau has been extremely common for centuries, and there are no "highly individual habits" in its forming for positive identification.32 In the thetas Smith wrote, there seems to be such diverging forms in existence that Carlson should not have had any trouble for finding specimens that came close to the thetas found in the Theodore-letter - here Carlson's method is named as "highly dubious" by Brown.33

Following the principles of QDE to the letter, there does not seem to be any room for firm conclusions to be drawn from the variants of theta Smith used. The handwriting of Morton Smith cannot be linked to the handwriting of the Theodore-letter on the grounds presented by Carlson, for the suggested similarities between these two are in too general a level. Brown summarizes the situation effectively:

"If these letterforms are easy to reproduce or appear consistently in MS 65 [the Theodore-letter], then the similarity to Smith's writing is probably coincidental. If they are normal for eighteenth century, then there is no basis for treating them as suspicious. If they are commonplace in twentieth-century Greek handwriting, then there is no basis for ascribing them specifically to Smith. And if Smith normally wrote these letterforms in a different way, then Carlson's examples of Smith's Greek handwriting are unrepresentative, and there is no way to explain why forms that are atypical for Smith's writing would appear with any regularity in MS 65."34

To summarize: the handwriting of the Theodore-letter cannot be linked to the handwriting of Morton Smith, and Morton Smith cannot be identified as the writer of the Theodore-letter by comparing their respective handwritings, because 1) the comparison material utilized by Carlson is too small in size, 2) there are no "highly individual habits" present in both handwritings, and 3) the handwriting in the Theodore-letter does not "lapse back" into the handwriting of Smith "progressively at various points"; instead, the Theodore-letter maintains its palaeographical identity with consistency.35

Scott Brown's argumentation related above is persuasive in its other aspects but in his choice of the sample material, which does not qualify for a rigorous analysis using the methods of QDE. A professional in the field, Hannah McFarland, directs the application of QDE thus: "A person's handwriting or handprinting habits can change over time. Consequently, the exemplars and questioned writing should be contemporary with each other. Ideally, both bodies of writing should be written with a year or two of each other."36 As we have seen above, the main body of the comparison material used by Brown was from Morton Smith's 1951 work "Tannaitic Parallels to the Gospels", and from his letter from 1955. The date of the additional marginal notes is not mentioned by Brown; presumably the date cannot be established for these notes.37 Smith did not return to the monastery of Mar Saba after 1958, and Guy G. Stroumsa et al. found the letter from the monastery in 1976. At the latest, Smith would have had to forge the letter in 1958, while he was studying the MSS in Mar Saba. How much earlier could he have done the writing of the letter to Voss' book? Not much, I would argue, but, in any case, the temporal vicinity to the year 1958 of the samples of Smith's handwriting would greatly strengthen the reliability of the conclusions drawn from their comparison.

Though commonly not known, Morton Smith decided to publish the Theodore-letter before the end of the year 1958 along with his English translation in order to secure the rights to the text's "proper" publication, which did not happen until 1973. Smith's 1958 publication of the Theodore-letter was a rather ascetic self-publish, 13 sheets of A4-sized notebook paper, containing a short foreword, the Greek text of the Theodore-letter, and Smith's English translation.38 Notwithstanding the looks of the self-publication, it was still a part of Smith's extensive bibliography as "Manuscript Material from the Monastery of Mar Saba, discovered, transcribed and translated by Morton Smith. New York: Privately published, pp. I + 10".39 The publication year - and consequently the year of its writing - is 1958, and over four sheets of A4-sized pages of Smith's "continuous, natural writing" in Greek fulfills nicely the criteria set down by Ordway Hilton.40 What does the comparison between Smith's 1958 handwriting and the Theodore-letter reveal? Does it reaffirm the conclusions made by Brown that were based on Smith's handwriting from 1951 and 1955?

On the whole, the answer to this question is positive: from the very first page onwards it is clear that Smith still prefers to draw his lambdas in two strokes. The consistency we find from the Theodore-letter, that 96% of the lambdas situate the intersection of the two feet below the middle point of the letter41, is not present in Smith's handwriting, where the point of the intersection varies preferring the below only marginally. Regarding his taus, Smith writes them varyingly with one and two strokes - the Theodore-letter, on the other hand, prefers the one-stroke variant. The greatest change has happened in the way Smith writes the letter theta - whereas in 1951 he drew it with one stroke and a loop (ϑ), the 1958 version is drawn varyingly with one and two strokes (preferring one), but without the loop: instead, the brim of the letter forms an unbroken circle (θ). Naturally, this variant of theta has no similarity with the variant used in the Theodore-letter, and, consequently, the linkage observed by Carlson - that the drawing of the theta begins with a short horizontal line from left, from about the middle point in relation to the height of the letter42 - has vanished. With the comparison of these three letters Smith cannot, following the principles of QDE, be linked to the author of the Theodore-letter. In fact, the samples discussed above offer material for a very distinct conclusion: that Morton Smith has not written the Theodore-letter.

Significant differences observed between the handwriting of the Theodore-letter and the handwriting of Morton Smith may even contain such "fundamental dissimilarities" that clearly establish the writer of the Theodore-letter not to have been Morton Smith.43 One the early pioneers of QDE, Ordway Hilton, published an article in 1954 titled "Handwriting Identification vs. Eye Witness Identification", that offers an excellent case study of the fact how repeated differences establish the authors of two different handwritings to have been separate individuals. The repeated differences are deemed to be especially weighty if they occur in inconspicious details.44 Following Hilton, these differences are 1) the drawing of the letters in such a manner that the pen is lifted (or is not lifted) in different places, 2) the inclusion (or lack) of small extra "hooks" or loops, 3) the moving (or not moving) of the pen along the line of the letter that has already been drawn, 4) the crossing of the line (or lack of crossing), 5) the variation in the sizes of the letters compared to other letters, and 6) the variation in the sizes of the parts of the letters compared to other parts of the letters.45

The most important differences, according to Hilton, are the variations in the use of the loops, and the variation in the moving of the pen along the line of the letter that has already been drawn.46 The trust put into QDE is first-rate in other aspects, too, as Hilton gives an example of a criminal case where a testimony of 13 (!) eyewitnesses was disqualified on the basis of a handwriting analysis of this kind. Or, as he himself said it: "The eye witnesses may be correct, but scientific examination of the writing is certainly a more accurate method for establishing the facts."47 What kinds of conclusions may be drawn from applying the principles of Hilton's handwriting analysis to the handwriting of Smith from 1958 and to the handwriting of the Theodore-letter?48

Numerous repeated differences are to be found. When Smith draws the circumflex in the form of a semi-circle, sometimes with a pointed top (â), the Theodore-letter has a wavy line (ã) instead. The letter sigma, as the last letter of the word, has its bottom pointing left in Smith's handwriting, whereas the Theodore-letter variant turns to point clearly straight down. Spiritus lenis in Smith has its end points pointing to the left, when in the Theodore-letter they point down. As we have already seen above, the letter theta is drawn by Smith in 1958 in such a way that the brim of the letter forms an unbroken circle, whereas the Theodore-letter features the traditional cursive variant. The bottom of the letter iota, on the other hand, tends to lean to the right in Smith, but the iota in the Theodore-letter points sometimes straight down, and at other times turns either to left or right with a hook. As for the letter gamma, Hilton would be delighted in its consistent variation including both the differing loop and differing way of moving the pen along the line of the letter that has already been drawn: for Smith the furrow of the letter - formed from the line of the letter beginning from the topmost right high point and the topmost left high point of the letter - is not very deep, when the line of the letter is already coming up after turning around with the help of a loop, whereas in the Theodore-letter the furrow of the letter is very deep in a striking manner, and the loop is altogether missing - instead, from the low point of the letter we go back up with riding quite liberally along the line of the letter that has already been drawn, but in Smith's gamma this phenomenon does not happen at all.49 The last differences are in the letter kappa, where Smith begins his drawing from high without a hook, but the Theodore-letter from low with a hook, and the letter mu, where Smith has a straight left leg, but the Theodore-letter has a hook to the right.

Applying the principles of QDE as Hilton used them, the conclusion seems evident that Morton Smith has not written the text of the Theodore-letter.50 On the other hand, there are no undisputed "signs of a forgery" present in the handwriting, either. And, as I have stated above, it should always be remembered that the application of QDE for assessing the authenticity of MSS may contain some serious flaws - everything I have written in the past 11 pages may potentially be absolutely irrelevant for the question of the MSS authenticity. If all of the above put together is not yet enough for making the reader convinced of the problematic nature of Carlson's hoax hypothesis regarding the assessment of the handwriting of the Theodore-letter and the handwriting of Smith, one of his arguments for the hoax is still remaining. The Madiotes-clue, presented in Chapter 2.4, and the MS Smith put into his catalogue as number 22 (afterwards: MS22), turns against Carlson's case in the end, when one of the defenders of the authenticity of the Theodore-letter, Allan J. Pantuck, managed to obtain the original negative of the photograph where Carlson found the Madiotes-clue.51 The good thing with the original negative was the fact that it had not been cropped in any way - quite a different case than with the photograph Carlson had access to.

The previously published photographs of the MS2252 were cropped in such a way that of the folio 1, recto only the left part was visible, as the cropping was vertical in its orientation.53 The unfortunate line of deduction by Carlson must have gone the following way: the uppermost handwriting of the folio 1, recto of the MS22 looks to be from the 18th-century, and according to Carlson this is the same handwriting as the one in the Theodore-letter, with all the "signs of a forgery" present in both.54 On the other hand, Morton Smith had began his literary description of the handwritings on this page from the autograph of M. Madiotes, whose handwriting he estimated to be from the 20th-century.55 Carlson supposed this estimation to have been a deliberate clue left by Smith, as the first handwriting in the picture was clearly not from the 20th-century. The autograph itself, Μ. Μαδιότης, Carlson could not, however, see, because of the way the photograph was cropped.56

As these assumptions of Carlson turn out to be unfounded, the Madiotes-clue crumbles as well, for in the original negative Pantuck found it is clearly seen that the autograph of M. Madiotes has not been written with the same handwriting as the first handwriting on the page, the one Carlson interpreted to have been linked to M. Madiotes - there are, in effect, five different handwritings present on the page. Smith seems to have been right: Μ. Μαδιότης is, according to the assessment of Pantuck and Brown, from the 20th-century. Respectively, Carlson seems to have been right, too: the first handwriting on the page looks to be from the 18th-century.57 It is just that the Μ. Μαδιότης autograph and the first handwriting on the page have nothing to do with each other: additionally, the autograph's orientation is upside down compared to the first handwriting on the page.58 The root of the confusion is easy to locate: it is Morton Smith, though he certainly did not mean it.

Why does Smith mention in his catalogue published in Νέα Σιών only three names, when the original negative shows clearly that there are five different handwritings on the page? The reason must be, quite simply, that two of these handwritings do not have a personal name attached to them - they are inscriptions - like the first handwriting on the page that begins with the words το παρōν βιβλειον (with a wrong accent) translating roughly as "The present book" - but without any personal names attached Smith did not refer to them in his catalogue.59 For Carlson's defence it must be remarked that the description of Smith in his catalogue really gives the impression that there are only three different handwritings on the page, and that the personal name Μ. Μαδιότης is linked to the first of these.

As we have seen, however, this first handwriting has no personal name attached: Μ. Μαδιότης is positioned upside down beneath it, but does not contain any other writing.60 The suggestion Carlson makes - that the name is a pseudonym (as he did not find it from an online-directory)61 - is not very persuasive, either, in light of the original negative, as the spelling variant "Μ. Μαδιότης" found in Νέα Σιών looks to be a printer's error: for the offprints Smith received he had made a correction, and, spelled correctly, the name should have been Μ. Μαδεότας.62 According to Pantuck and Brown the correct spelling may very well be Μ. Μοδέστος, a name that has been very commonly used for centuries.63 After these observations, the Madiotes-clue by Carlson is far less persuasive than before the close scrutiny of the photograph in question. Brown, however, did not stop here: with the same application of the methods found in QDE, as used by Ordway Hilton, and as used by myself above to argue, based on the "repeated differences" found in the handwritings of the Theodore-letter and the handwriting of Morton Smith, that the Theodore-letter has not been written by Smith, Brown studies the first handwriting on the MS22 and the handwriting of the Theodore-letter, arguing that they have not been written by a single author.

For the very first thing, Scott Brown rejects the claims of Stephen Carlson, that a single author was responsible for the handwritings in both the Theodore-letter and the MS22, on the basis of inadequate amount of comparison material: the cropped photograph of the MS22, which was used by Carlson, contains only 16 individual symbols, including nine different letters and a single ligature, and the size of the sample is far too small for making a positive identification between the author of the Theodore-letter and the author of the first hand of the MS22, as the minimum sample size should be "four or five pages of carefully selected continuous, natural writing".64 Contrary to Carlson's use of the methods of QDE, Brown cites Hilton some more: "Repeated small differences establish clearly that two specimens are the work of two individuals despite a great number of general similarities."65 Even with the extra letters revealed in the original negative of the picture, unearthed by Allan Pantuck, the size of the sample remains extremely narrow, but the observations made by Brown could still be considered suggestive.

It is a long list of differences between the handwritings. Compared to the handwriting of the Theodore-letter, the first hand of the MS22 1) does not put an accent on top of the letter omicron in the definite article το, whereas the Theodore-letter manages to put its accents right every time, 2) does not bind the letter tau to the letter omicron following it, the letter alpha to the letter rho following it, or the letter rho to the letter omicron following it, even though the Theodore-letter binds together all these letter combinations, 3) binds the letter pi to the letter alpha following it with a very thin line without a loop, whereas the Theodore-letter binds these two with a strong line, and with a loop, 4) draws the circumflex with a straight line, while the Theodore-letter uses the wave form, 5) draws the letters beta and gamma in a very dissimilar manner compared to the Theodore-letter, 6) tends to tilt the letter lambda to the left without turning the high point and the end point of the left leg towards each other, whereas the Theodore-letter tends to tilt the letter lambda to the right and turn the high point and the end point of the left leg towards each other, 7) binds the combination πο together with a line drawn on top of the cursive pi, while the Theodore-letter does not (and, furthermore, the Theodore-letter manages to draw the cursive pi in a dissimilar manner, resembling a lot of the letter epsilon), 8) uses a ligature for the epsilon-iota combination in the word βιβλειον, which makes the spelling of the word different than one used in the Theodore-letter (βιβλιον).66

In addition to these differences, Brown remarks of the general disparate nature of the two handwritings: the letters in the MS22 are drawn individually without connecting them to the letters preceding and following, whereas the Theodore-letter connects letters on a regular basis. If the handwriting in the MS22 is supposed to be a deliberate clue of the hoaxer's true identity, why was it not drawn undisputedly with the same handwriting as found in the Theodore-letter?67 Furthermore, as we have seen, the principles of QDE dictate that where there are no "highly individual habits" present, there we cannot conclude that the two texts are written by the same author - in the handwritings of the Theodore-letter and the MS22 these are not found. All things considered, the conclusion from the first hand of the MS22 is the following: as far as it is possible to conclude anything reliably from the small sample size we have, the handwriting of the MS22 is not 1) the handwriting of M. Madiotes, 2) the handwriting of the Theodore-letter, nor 3) the handwriting of Morton Smith.68

Two more observations should be made. First, the first hand of the MS22 uses a narrow pen point, suggested to be an anomaly among the MSS in Mar Saba by Carlson.69 Because the first hand of the MS22 is undoubtedly from the 18th-century70 - as Carlson himself concluded - but is not penned by the author of the Theodore-letter, it is an independent witness of the use of the narrow pen point in Mar Saba during the 18th-century, and not an anomaly, as suggested by Carlson. Second, the claims of Carlson concerning the "signs of a forgery" present in the first hand of the MS22 suddenly reappear to us in an uncanny light: when Carlson claims to have found "signs of a forgery" from a text that does not seem to have any connection with Madiotes, the Theodore-letter, or Smith, what should we conclude of Carlson's general competence for spotting and interpreting these "signs" correctly? Whatever it is he sees in the handwriting of the MS22, it looks to be a natural feature of the quality of the drawn line as they are found in MSS, and not an indication of sinister motives from the part of the copyist.71

1This question is raised by e.g. Foster 2005, 67-68.
2Carlson does not question the suitability of QDE at all, although a professional forensic document examiner, Emily J. Will, directs the search for the "signs of a forgery" from foreign language documents thus: "the examiner must first learn about the characteristics of the written language and how that writing is taught"; [Further note: the emphasis on the way a given (foreign) writing is taught sounds interesting. How was the writing of Greek taught for the orthodox monks in the 18th-century? What kind of significance does this detail play?]
3Carlson, too, gives an incident concerning the forging of an autograph, as an analogous case; Carlson 2005, 26-27.
4After the publication of "The Gospel Hoax" Carlson hired a professional examiner, Julie C. Edison, whose testimony affirms every observation of the handwriting by Carlson as valid;
5; Brown, in addition, remarks of the problems relating to varying resolutions of the photographs; Brown 2006c, 145 n. 4.
6Carlson 2005, 27-28.
7Carlson 2005, Figure 2B and Figure 2C.
8Carlson 2005, Figure 2A. Note, that - in my estimation - only the narrower pen point used in the Theodore-letter makes the blot more easily discernible.
9Carlson 2005, 29.
10Carlson 2005, Figure 2C. More examples are offered in Shandruk 2008a. A much clearer picture of the similarity between the two ἐκ words is offered in Paananen 2008a, 93. It should be noted, however, that Viklund 2009 observes from the much better (high-resolution) photographs taken by Father Kallistos Dourvas that letters epsilon and kappa, in the first line of the Theodore-letter forming the word ἐκ, are in reality written separately (not tied together), with the latter letter (kappa) just happening to hit the "tail" of the first letter (epsilon). The ink blot that has formed in this crossing of the lines, has therefore not come into existence from a lifting of the pen (in the process of forging the letter), as proposed by Carlson; Carlson 2005, 28; Viklund 2009.
11The beginning of this endeavour was Rick Brannan's blog post on December 23th, 2005;
12Carlson argues that the text of the letter gets better towards the end; Carlson 2005, 30-31. The blogosphere experiment seems to support this observation - however, after a close scrutiny of the text of the Theodore-letter, Viklund concludes that this phenomenon is not present in the text; instead, the text of the letter is of uniform quality throughout; cf. Viklund 2009. Additionally, Viklund asks, should Carlson's observation be the other way around, if he holds the text of the letter to be a slowly drawn imitation of an 18th-century hand, for the exhaustion due to the slow process of imitation would more naturally be manifested in a tendency to put more errors in the text towards the end, and not less.
14Carlson 2005, 30-31.
15Viklund 2009.
16In addition, Foster observes that some of the "signs of a forgery" are present in the Akhmïm fragment that contains the Gospel of Peter; Foster 2005, 67-68.
17I have given space to the blogosphere experiment, even though I am at present unsure about its applicability to the case at hand, as the only known empirical (though still non-controlled properly) experiment where the difficulties relating to the copying of the MSS are observed. I freely admit that the conclusions drawn from this experiment are (at least somewhat) unreliable. The results contained lots of the "signs of a forgery", in the same manner as Carlson observed these, and general illegibility, and it is virtually certain that an inexperienced copyist produces irregular handwriting. What happens to the quality of the handwriting and to the numbness and all the effects it encourages when the copyist gets more experienced, we lack knowledge of. The matter could be studied in a more professional manner, although, at the same time, we must acknowledge Rick Brannan for his exemplary way of putting into practice the best ideals of Web 2.0 ideology.
18Brown 2006a, 299. Brown cites Barbour 1981, xxii, xxiii as authoritative.
19Brown 2006a, 299; Peristeris 2001, 171. Also, Morton Smith observes that the handwriting of the Theodore-letter contains Western European influences. The reason for this inclination is, no doubt, the general practice of the Eastern monasteries during the 17th-century and the 18th-century, the collecting of patristic texts printed in the West; Smith 1973, 2-3.
20Smith 1960a, 251.
21According to his own words, Carlson went off to find the "signs of a forgery" from the handwriting, as "it is hard to find evidence of forgery if one is not looking for it”; Carlson 2005, xviii, 24.
22In my analysis of Carlson's comparison between the handwriting of the Theodore-letter and the handwriting of Morton Smith I rely heavily on the arguments developed by Scott G. Brown; Brown 2006a; Brown 2006c.
23Hilton 1993, 300; I borrow this quote from Brown, who uses it as a basis for his critique in Brown 2006a, 300 and Brown 2006c, 144-149.
24Carlson 2005, 46.
25Hilton 1993, 153-154; I borrow this quote from Brown, who uses it as a basis for his critique in Brown 2006a, 300 and Brown 2006c, 144-149.
26Hilton 1993, 153-154; cursive mine; I borrow this quote from Brown, who uses it as a basis for his critique in Brown 2006a, 300 and Brown 2006c, 144-149.
27Hilton 1993, 200. Here Hilton talks about modern autographs; the applicability of the methods developed for modern autographs is, at present, uncertain when dealing with manuscripts.
28Brown 2006a, 301.
29Brown 2006a, 302-303. If the first chapter of the "Tannaitic Parallels to the Gospels" is removed from the sample, the figures are 40% beneath, 60% rest - in this instance Smith's handwriting reaffirms the conclusions drawn from the experiment made in the blogosphere, that the writer produces more textual variants in the beginning of his text, and less towards the ending.
30Brown 2006a, 301-303; the citation is from Conway 1959, 27, whom Brown cites as authoritative.
31Brown 2006a, 303-304.
32Brown 2006a, 304. Additionally, Smith drew the letter tau varyingly with one and two strokes.
33Brown 2006a, 305.
34Brown 2006a, 302.
35Brown remarks that the idea of "lapsing back" into the forger's own handwriting is related to forging autographs in situations, where the forger cannot copy the autograph from a model, but has to resort to memory (e.g. forging a check under the nose of a bank clerk). If Smith forged the Theodore-letter, would he have done it from a model he had prepared beforehand? Would the "lapsing back" into his own handwriting take place in this scenario, at all?; Brown 2006a, 300-301.
36Bold original;
37Brown 2006a, 302.
38Contrary to what some scholars have speculated, the self-publication did not contain photographs of any sort; this is, however, presumed e.g. by Jeffery 2007a, 263 n. 65.
39Kraabel 1975.
40Hilton 1993, 300. Important in QDE, along with the temporal vicinity of the samples that are to be compared, is the sufficient size of the sample material, as Brown remarks; Brown 2006c, 145; Brown cites e.g. Huber 1999, 249. At this point, my comparison of Smith's 1958 handwriting and the handwriting of the Theodore-letter is meant to supplement the comparison by Brown, related above. An even more reliable conclusion, following to the principles of QDE, would require the use of numerous samples of Smith's handwriting, all written in 1958 +/- two years.
41Brown 2006a, 303.
42Carlson 2005, 47.
43A similar analysis is made by Brown, where it is made clear that the writer of the Theodore-letter is not the same person as the writer of the unidentified handwriting in the manuscript Smith catalogued as number 22; for details of this analysis cf. Brown 2006a; Brown 2006c; Pantuck 2008.
44Hilton 1954, 209.
45Hilton 1954, 209-212.
46Hilton 1954, 211.
47Hilton 1954, 212.
48I wish to point out here that every observation and conclusion made from the handwriting of Smith from 1958 is my own, even though I have never received formal training for applying the principles of QDE to handwritings. If any one detail seems to be a bit out of place in this discussion, it is the fact that none of the individuals taking stances regarding the handwritings - that would include Carlson, Brown, Viklund, and myself - have received formal training in this field, or have access to the original material, as the Theodore-letter seems to have completely vanished after 1990. Only Carlson has confirmed his analysis with the help of a professional in QDE, but the exact details of this confirmation are lacking;
49Appendix 1 contains examples of the first six gamma letters from the Theodore-letter, and of their transcription made by Smith in 1958.
50One possible objection to the comparison of the handwritings would draw attention to the fact that the handwriting of Smith is not cursive, as the handwriting in the Theodore-letter. It may be the case, that an even better argument would be built on the basis of multiple pages of Smith's handwriting that he would have written in the 18th-century cursive hand, on request. Unfortunately, to make any sort of comparison between the handwritings, we have to, as both Carlson and Brown have done, use whatever material we have, and the increase in the reliability of the conclusions should not be overestimated, either. To be on the safe side, my own comparison has focused on those letters and symbols that are very similar in both the Theodore-letter and the handwriting of Morton Smith in 1958. The perceived differences between these letters and symbols that look very similar at first glance are, following the principles of QDE laid out by Hilton, all the more telling when they happen in small and inconspicious details.
51This photograph has been printed in Pantuck 2008.
52Smith 1960b, 177; Smith 1985, 37.
53For some reason Carlson concentrated his efforts to the picture published in Smith 1960b, although the picture that was used as an illustration in "The Secret Gospel" (1973) was cropped in a different way, revealing a few more letters of the handwritings on the page.
54Carlson 2005, 42-43.
55Smith 1960a, 119.
56The same reconstruction of Carlson's probable reasoning, as he was looking at the cropped photograph, is also done by Pantuck 2008, 112.
57Pantuck 2008, 118.
58antuck 2008, 117.
59So concludes also Pantuck 2008, 118.
60Pantuck 2008, 121.
61Carlson 2005, 43.
62It is also possible that Smith had changed his mind regarding the spelling of the name written in unclear cursive hand; Pantuck 2008, 112-115.
63Pantuck 2008, 122-123.
64Hilton 1993, 300; the quote is used as a basis for the critique in Brown 2006a, 300 and Brown 2006c, 144-149.
65Hilton 1993, 161; the quote is offered as cited in Brown 2006c, 144-149.
66Brown 2006a, 297-298; Pantuck 2008, 120-121. Brown 2006c offers an even more detailed analysis of the differences in the handwritings.
67Brown 2006c, 148.
68The comparison between the handwriting of the first hand of the MS22 and the handwriting of Morton Smith is possible with the same reservation as applies to any comparison utilizing an inadequate amount of comparison material: as was the case in the comparison between the handwriting of the Theodore-letter and the handwriting of Smith, there are no "highly individual habits" to be found, but "repeated differences" abound.
69Carlson 2005, 32-33, 46.
70Pantuck 2008, 118.
71These two conclusions are made by Pantuck 2008, 124, as well.

No comments:

Post a Comment