Mitchell’s mistake

from Remey, ‘Observations’ 1908

I’ve been looking again at an old claim that Abdu’l-Baha’s Will and Testament was not written by Abdu’l-Baha, that it was ‘fraudulent.’ This claim is the foundation for two small Bahai splinter groups that reject the institution of the Guardianship (established by Abdu’l-Baha in his Will and Testament), and it has also been propagated in Germany in anti-Bahai polemics published by the Lutheran ‘Central Office for Questions of Ideology’ (EZW). In looking through the documents, I’ve noticed something that doesn’t seem to have been commented on in the past.

Let me begin by saying that there is no doubt that the Will and Testament is genuine. At the time it was opened and read, there were some hundreds if not thousands of Persian Bahais, including Abdu’l-Baha’s family and secretaries, and also Persian opponents of the Bahai Faith such as Muhammad Ali, or opponents of Shoghi Effendi such as Mirza Sohrab, all of whom were familiar with Abdu’l-Baha’s handwriting. Not one of them has suggested that the Will and Testament is not written in Abdu’l-Baha’s hand – not even those Persian opponents of Abdu’l-Baha and Shoghi Effendi who stood most to gain by casting doubt on its authenticity. Mountford Mills, who saw the Will, said that “It is written entirely in the Master’s own hand [and] It is signed by him.” and that the first two parts bore the seal of Abdu’l-Baha. Abdu’l-Baha’s Persian secretary, who would certainly be in a position to know, confirms this. [note 1]

The claim that the Will and Testament was fraudulent was based entirely on the opinion of Dr. C. Ainsworth Mitchell, an English chemist and document analyst, whose publications include Documents and their scientific examination, with especial reference to the chemistry involved in cases of suspected forgery … and Forensic chemistry in the criminal courts (1938).

What interests me is not whether the Will is genuine, but how a scientist of Dr. Mitchell’s stature could have ventured to give an opinion in a field in which he must have known that he had no special competence. As we will see, he was not even aware that Persian is written from right to left, so he was certainly not familiar with the Persian alphabet. Handwriting analysis, and even the simple reading of a handwritten text, requires familiarity with the way that letters are formed by the hand. I am not a forensic handwriting expert, but I have completed a course in Middle-Eastern codicology, which means learning to read and transcribe the handwriting styles used for Persian and Arabic in the Middle East over the past millennium, and being able to name the most common styles and to date and place texts by style. In attempting to read such handwritten texts for this course, I often found my fingers twitching as if holding a pen. By imagining how the pen moves in sequence, as the scribe writes, one can see what parts of the letter are significant and which are accidents. What looks like an up-stroke might simply be the pen dragging on the paper a little as it is moved to where it must be to begin the next letter. To know that, you have to know where the letter ends (thus, know the alphabet) and where the next letter is begun (thus, to know penmanship). Does the pen begin at A and move clockwise to B, or does it begin at B and move anticlockwise to A? Does the stroke begin at the top and move down, or begin at the bottom, move up and then track back? Only by actually writing in that alphabet, in the way scribes are taught for that style, can one learn to read the style with confidence. So for Mitchell to offer an opinion on a text he could not read, in a script he knew nothing about, is something like someone who knows no English deciding that Dickens could not have written Hard Times, on stylistic grounds.

The problem is that Mitchell appears to have been intelligent, and he had experience in document analysis using both chemical and handwriting analysis, so he knew he was not competent to make an analysis based on a photograph of a text in a language and alphabet he did not know. Why did he venture an opinion then? My suggestion is that he saw some very evident differences, between Abdu’l-Baha’s signatures on the envelope of the Will and Testament and what he was told were authentic signatures from Abdu’l-Baha – differences so large that he was confident that, given the original documents, he or another expert actually familiar with Middle-Eastern handwriting would be able to demonstrate they were not from the same hand. Perhaps he really did see glaring differences: but that could be because what he was given as ‘authentic signatures’ were not authentic, or because Abdu’l-Baha used two quite different signatures.

The background

The sequence of events and references to the sources are presented by Ulrich Gollmer in Chapter 11 of Making the Crooked Straight, in Adib Taherzadeh’s The Covenant of Baha’u’llah (page 299..) and again in his book The Child of the Covenant (page 347..) . The latter two are in Ocean and the relevant sections can be found by searching on ‘Ruth White.’

Briefly, the Will and Testament of Abdu’l-Baha appears to have been written in three stages beginning in 1903 (see Gollmer 685 note 33); as Mountford Mills points out, the fact that the 3rd section was signed without using Abdu’l-Baha’s seal probably means it was written after 1912, when his seal was stolen). When Abdu’l-Baha died in Haifa at the end of November 1921, a Will was found in his safe. When opened, it was found to be addressed to Shoghi Effendi, who was at the time studying in Oxford. It was therefore resealed until Shoghi Effendi’s return to Haifa, in the last days of December. On 3 January 1922 the Will was opened and a Persian believer was commissioned to make copies; on 7 January it was read in public and copies were sent to Iran: in the same month copies in Persian were sent to interested parties, and a preliminary English translation was read in New York in February 1922.

Abdu’l-Baha’s brother, Mirza Muhammad-`Ali, claimed that he, not Shoghi Effendi, was the rightful heir to Baha’u’llah, but he won little support for his claim and – what is important for present purposes – he did not claim that the Will and Testament was not authentic, but rather that Abdu’l-Baha had no legal right to pass over his brother in favour of his grandson, Shoghi Effendi. The British authorities in Palestine investigated the matter and decided in favour of Shoghi Effendi. Over the years since then, various cases regarding the ownership of properties in Israel have been fought, and won, by the Bahais on the basis of the provisions of the Will and Testament. In none of these was the authenticity of the Will disputed. Ahmad Sohrab, who had been Abdu’l-Baha’s secretary and turned against Shoghi Effendi, explicitly vouched for the authenticity of the Will and Testament, as did many others who were familiar with Abdu’l-Baha’s handwriting and style.

The allegation that the Will and Testament was forged originated from an American Bahai, Ruth White, beginning in 1926. She asked the National Spiritual Assembly of the Bahais of the United States for a photocopy of the Will, and was given it, along with photocopies of the four signatures of Abdu’l-Baha on the envelope containing the Will and Testament.  I have posted a PDF file of her photocopies on another page. [2] She also obtained photocopies of two messages Abdu’l-Baha had written in 1912, one in the Bible of the Unity Church in Montclair, (USA) and the other in the Guest Bible of the City Temple, London, and what she claimed were authentic signatures on two letters to Mrs. Stannard and one written to Mrs. Devine. However she does not indicate why she thought these were authentic signatures (she could not read Persian herself, and Abdu’l-Baha’s general practice was not to sign his letters, but to initial them with the letters ع ع `ayn `ayn). She was apparently unaware that the words that Abdu’l-Baha wrote in the Bible of the City Temple end with his signature. We can use this as our own model of an authentic signature by Abdu’l-Baha, with the reservation that it was probably written with a fine-tipped fountain pen whereas the signatures on the Will and Testament would presumably have been written with the broad tipped reed pen used in the Will itself. So she has 3 signatures which she thinks are authentic, and another which really is authentic, but neither she nor her handwriting expert recognise that the last of these is the signature of Abdu’l-Baha. That already tells us that the 3 signatures which she and Mitchell supposed to be authentic did not resemble the one in the City Temple Bible.

Apart from the signatures, she had photographs of the text of the Will and Testament, and the words Abdu’l-Baha wrote in the two bibles. The reason for using the messages written in the two Bibles as samples of Abdu’l-Baha’s handwriting is that the many letters (‘tablets’) which Abdu’l-Baha sent to the Bahais and others in the West were almost never in his own handwriting (he dictated letters to a secretary), and the addressee would receive a translation, prepared either in Palestine or in America by one of the Persian Bahais living there. The texts written in the bibles, and half a page from the Will and Testament, are reproduced in her book ‘Abdu’l-Baha’s Questioned Will and Testament, and I have taken copies from there. She does not reproduce the 3 signatures which she thought were authentic: if she had, our work would be simpler. I will deal with the signatures first, since Mitchell begins with a close analysis of these, and then look at his less detailed remarks about the texts.

The signatures

Ruth White deposited copies of Mitchell’s report with the Library of Congress in the US, and one of the Bahai splinter groups who base their faith on it has put a copy online. His report begins by studying the signatures on the envelope and comparing them with what Mitchell thinks are authentic signatures. He says there are ‘striking differences.’ On page 2 of his report, in the left margin at the bottom, he shows as an example two handwritten L’s which he has written from left to right. He comments “In the authentic signatures the width of these characters, compared to their height, is much greater than in the signatures on the envelope [in which the Will and Testament was found].”

This raises two questions:
– Did he not know that Persian is written from right to left? He has written his two L’s with the pen moving from left to right. Apparently he knew nothing at all of the Persian alphabet, since he never names a letter, and never mentions points and other diacritics (which are crucial in Persian).

Signature in the City Temple bible

– He says he found these letters in the ‘authentic signatures.’ But Abdu’l-Baha’s name does not contain any two vertical letters joined side-by-side: let alone two English L’s, and there is nothing like two looping L’s in the authentic signature of Abdu’l-Baha in the Bible of the City Temple (right). And this signature in the City Temple bible is not unique: if you click on the very first image of Abdu’l-Baha at the top of this page, you should be able to see that the signature on it is similar (it is followed by the `ayn `ayn initials). For a good quality scan, click here. On the face of it, the ‘authentic signatures’ Mitchell was given must have been something else.

Abdu\’l-Baha\’s name

Nevertheless, I thought there was something oddly familiar about Mitchell’s lovely looping ‘L’s. They tug at a memory – it is something to do with Abdu’l-Baha … something I saw in a book … and eventually I found it. Look at the image on the left, which is again the name of Abdu’l-Baha. There you can see those looping L’s again. I will not claim that the two L’s that Mitchell saw must have come from the same hand as these, but there is a definite similarity.

I found my ‘looping-L’ version of the name of Abdu’l-Baha on a photograph of Abdu’l-Baha which has been reproduced in the front of each volume of Badayi’l-athar, a Persian account of Abdu’l-Baha’s journeys to the West, by Mirza Mahmud Zarqani. (The first volume has been translated, as Mahmud’s Diary.) If we zoom out on that image (below left), we can see that the writing says “Abdu’l-Baha, from Isfandiyar Bahram.” I know nothing more about the history of the photograph than what is clear from the inscription: someone called Isfandiyar is sending it to someone, and he or someone else has written the name of Abdu’l-Baha across the photograph. Abdu’l-Baha’s coachman was called Isfandiyar, but it is a relatively common name. There were two brothers called Isfandiyar and Bahram, from Yazd, who farmed one of the Bahai properties at Adasiyyah, but I see no “and” between the two names here. I also found another photograph, with the name of Abdu’l-Baha written in a similar script (below right).

Portrait of Abdu\’l-Baha in Badayi\’u\’l-athar

The second photograph comes from A History of Persia by Percy Molesworth Sykes, and again shows what look like two looping L’s. It may be that this is the photograph of which Lady Blomfield writes: “He signed a photograph, writing His name on the white part of his turban.”

The two L’s that Isfandiyar writes are the “u’l” part in the middle of Abdu’l-Baha’s name, and they should properly be written as two separate letters: a vertical stroke beginning at the top, and then to the left of that a vertical stroke beginning at the top and continuing on to the left, to join the ‘bump’ in the line which is the next letter. The two letters are both vertical strokes, and can only be told apart by the fact that the first, the aleph, is never joined to the following letter, while the second, the lam, always has a tail and joins the following letter. We can see this in Abdu’l-Baha’s signature in the City Temple Bible (below left): he has begun by writing the aleph (numbered 1 in the detail below), and has then lifted his pen off the paper and moved up and right to start the second stroke (the L) at ‘2’. It is nothing like the looping L’s that Mitchell sketches, which he saw in the signatures he had been told were authentic — and very likely were authentic (in which case, Abdu’l-Baha had two habitual signatures).

al-Baha in the signature in the City Temple bible

On the next page of Mitchell’s report he continues his analysis of the signatures, and again he has written some letters, resembling the English letters ‘wc’, to show what he sees in the specimen signatures. What he has drawn is recognisable as the beginning of the signature, the letters `Abd. The fact that he begins his analysis in the middle of the name, with u’l, and then moves on to `Abd, confirms that he was not aware that Persian is written from right to left. In this case his sketch does not resemble either Abdu’l-Baha’s signature nor the handwriting on Isfaniyar’s photograph of Abdu’l-Baha, or any other Persian handwriting I can think of. He shows a stroke curving towards the vertical on the right (which for him is the end of the word, but is in fact its beginning). Here’s Mitchell’s sketch based on the specimen signatures he was given:

and below (left) the corresponding section from Isfandiyar Bahram’s photograph of Abdu’l-Baha, and (right) from Abdu’l-Baha’s authentic signature, in the City Temple bible.

`abd in Isfandiyar Bahram’s handwriting

Two explanations are possible here: the more obvious is that what Mitchell had been told were authentic signatures of Abdu’l-Baha were not authentic. Perhaps a secretary had written something on the bottom of these tablets, saying that the text was a tablet from Abdu’l-Baha to Mrs Stannard or Mrs Devine. Such colophons would normally include the date of writing. Ruth White, who could not read what was written, made an honest mistake, and told Mitchell he could use these as specimen signatures. He saw that there were very large differences between these and the signatures on the envelope containing the Will and Testament, differences so great that he was willing to give an opinion, even knowing he had no special competence to judge.

At this point he should have had second thoughts. Someone trying to forge a Will would surely make the signature on it resemble the deceased’s signature somewhat? Seeing that the four signatures on the Will are like one another, but nothing like the specimen signatures he had been given, he should have suspected that it was the specimens that were not authentic.

However, there is another possibility: Abdu’l-Baha might have had two or more quite different signatures. Although this will seem implausible, especially to people who are used to using their signature to identify themselves when they sign cheques and other documents, there are several reasons why I think it is the more likely explanation.

First, there is the photograph of Abdu’l-Baha that bears the words “Abdu’l-Baha, from Isfandiyar Bahram.” The normal practice for a Bahai writing these words would be to write “His Holiness (Hazrat) Abdu’l-Baha…” and probably not to write over the chest of the figure. But Abdu’l-Baha would not write ‘Hazrat’ in front of his own name.

The Gouran tablet

Second, the signature with the ‘looped L’s” appears at the bottom of a number of tablets, so often that I think it is implausible to suppose that these are all colophons written by the same secretary who has adopted this peculiarity. One of these tablets, right, is written to Mr. Gouran, the father of Munireh Gouran who lived in our Bahai community in Leiden at the end of her life, and was buried here on March 3rd, 2010. A photocopy of the tablet was included in the memorial booklet distributed at her funeral. It is dated April 18, 1920.

Tablet to Graham Pale

The next example is a tablet to Graham Pale in Edinburgh, and is undated. At bottom left is the name of Abdu’l-Baha written with the looping ‘L’s, and on the side the name abdul Baha abbas, with only Baha capitalised, in a copperplate script. This is undated.

The next example is one of the postcards on which the Tablets of the Divine Plan were sent to the United States. Because of the First World War, letters were being opened by censors: sending postcards would have reduced the risk of delay or damage. This comes from the US Bahai archives and was provided by Brent Poirier in the comments to an earlier version of this article (a good example of the advantage of interractive blog publication over journal articles). The whole image is here, I have zoomed in on the signatures, or names of Abdu’l-Baha, in Persian and latin script, the first with the looped ‘L’s, the latter in copperplate with only Baha capitalised. The address “Mrs Helen Goodal, c/o Mr. Jos H. Hannen, 1252 = 8th st N.W., Washington, D.C., U.S. America is written in the same purple ink, in a less elaborate copperplate. When Brent showed us this, I and others commented that since the English address is written in the same ink, and it is hardly likely that Abdu’l-Baha would be writing out an address in English, the Persian signature with the looped ‘L’s must be written by a westerner, or a western-educated person. But that was before I had seen the two examples shown above.

The texts

Mitchell then turns to the text of the Will itself, and concludes (correctly) that pages 9 and 10 at least were written by the person who had signed the envelope. But he also says that the three parts of the Will and Testament were written by three different persons. In fact it was written by Abdu’l-Baha at three different times, presumably with different pens (reed pens are ‘disposables’), and possibly on different paper, with different ink. The first part must have been written around 1904, the last part might have been as late as 1920 or 1921, and handwriting changes with age, especially while one is very young and in old age. So it is hardly surprising that Mitchell detected differences within the Will. Given that he did not know that he was looking at three documents written some years apart, that he had already concluded that the signatures were not genuine, and he could see that pages 9 and 10 were written by the person whom he thinks is the forger of the signatures, it is understandable that he concludes “the writing does not agree with the hypothesis that it was all written by one person.”

Finally, he compares the writing in the Will and Testament, especially on lines 10 to 12 on page 5 of the Will, with the writing in the two bibles, and says he has “failed to detect in any part of the will the characteristics of the writing of Abdu’l-Baha.” Here I think he is simply wrong. He is apparently unaware that there are distinct ‘styles’ of handwriting that have been formalised and are widely used in Persian. In European alphabets, for example, we have copperplate writing and carolingian script and many others. Someone who uses copperplate to write invitations to a dinner party would not use the same script for a shopping list, and a handwriting expert worth his consultancy fee would not try to compare apples with oranges.

Nasta’liq shekasteh

Diagram of ideal letter proportions in nasta\’liq

In the same way, Persian has nasta`liq shekasteh, a flowing, decorative script for formal documents with exaggerated flourishes; nasta`liq, a more moderate version of this; naskh, designed for ease of reading, and also a common handwriting style for everyday personal use, developed for speed of writing. As you can see from the illustration of nasta`liq, in each style, each letter has specific ideal proportions. That means that there is no point in comparing the proportions of letters that are written in different styles: they will differ even in the same writer, because the ‘font’ for each style is different.

Writing of Abdu’l-Baha in the Unity Church bible

Abdu’l-Baha’s handwriting in the Unity Church bible (left) is in a moderately decorative nasta`liq shekasteh.

The Will and Testament (below) is written in much the same handwriting, but with more concessions for speed of writing, which is what one would expect for a long text.

Section from page 5 of Abdu\’l-Baha\’s Will and Testament

For example, in the sample from the Will and Testament one can see, top left, a short horizontal line. This represents two dots: they are usually written this way in people’s everyday handwriting, but they are not written this way in the Unity Church text, which follows the rules for formal decorative writing more closely. Allowing for the fact that the Unity Church text has been written with more care, in a ‘higher’ style, there is nothing here to indicate that the Will and Testament could not have been written by the same person. Most of the various flourishes that appear most striking are simply characteristics of the style, and not particular to Abdu’l-Baha. Notice however where I have inserted an X in both examples, that where the tail of a letter is bent backwards (to the right), Abdu’l-Baha tends to make it angular, whereas in nasta’liq and its variants, and even in everyday handwriting, it would usually be a smooth flourish.

The words Abdu’l-Baha wrote in the City Temple bible in London, in contrast, are more like everyday handwriting, influenced by the flourishes of nasta`liq shekasteh.

Abdu\’l-Baha\’s writing in the City Temple bible

At top left, for example, what looks like an arrowhead pointing upward represents three dots together, and there are several examples of a horizontal stroke that replaces two dots. This is typical of everyday handwriting, rather than calligraphy. It tells us what level of formal handwriting style the writer is aiming at, rather than who is writing. Notice again the angular reversed ‘tail’, marked with an X. Apart from being written in a less formal style, we can see that these words were not written with a flat nib, so that the strokes are all of the same width, unlike those written with a reed nib or the pen used in the Unity Church Bible. The writing is rather shakey: perhaps Abdu’l-Baha was not comfortable with the pen, or was standing in an uncomfortable position. Or perhaps there was no table, and someone is holding the bible up for him as he writes.

We can see that the two texts written in the bibles, both of which Mitchell knew to be the authentic handwriting of Abdu’l-Baha, differ more from each other than either differs from the Will and Testament, which is something between the two in style. How could Mitchell conclude that the Will and Testament was by a different hand, and fail to conclude that, by the same standards, the words in the two bibles would also be written by very different hands?

Summary

My hypothesis is that Mitchell saw the signatures and thought he was facing an open-and-shut case: the ‘authentic’ signatures and those on the Will and Testament were quite different. In this he was correct, but he didn’t know that Abdu’l-Baha had two different signatures, or that the specimen signatures were not in fact written by Abdu’l-Baha. Then he saw that the Will and Testament was not all of one piece. Nobody had told him that the Will was in fact written at different times, and he could not read Persian to see that for himself (Abdu’l-Baha says this explicitly in the text of the Will). This made him confident that there was some sort of fraud afoot, even though he must also have known he was not competent to judge a text written in Persian script. His confidence regarding the signatures and the differences between the three parts of the Will then led him to a serious professional lapse: giving an expert opinion without the required expertise.

All those competent to judge, even the opponents of Shoghi Effendi and Abdu’l-Baha, say that the Will and Testament is written in Abdu’l-Baha’s own hand. It is therefore absurd to claim to follow Abdu’l-Baha, while rejecting the institution of the Guardianship which he bequeathed to us in the Will and Testament, commanding us:

O my loving friends! After the passing away of this wronged one, it is incumbent upon … the loved ones of the Abha Beauty to turn unto Shoghi Effendi … as he is the sign of God, the chosen branch, the Guardian of the Cause of God … He is the Interpreter of the Word of God (The Will and Testament, 10)

~~ Sen McGlinn
http://tinyurl.com/Mitchellmistake

Amended 5 Feb 2010: added the mention of the brothers Isfandiyar and Bahram.