More experiments with Amharic and technology

In my last post I found that it was possible to turn a PDF full of images of Amharic text into recognised electronic text using Google Drive, and then get some translation of the results into English using Google Translate.

There were some extremely interesting comments made on the post, which I have been reading.  I have also prepared a PDF of the whole text of the Life of Garima by Yohannes, and run that through the Google Drive process.

Where we started was in trying to read a passage of this text, in which – supposedly – God stopped the sun so that St Garima could copy the bible in one day.  The summary of the work  given by Rossini (instead of a proper translation, drat him), indicates that this was on lines 356-60 of his text, which turns out to be the last line of p.161 and the first three of p.162.  Here they are:

The output from the OCR is good, but you still have to compare the characters carefully.  Errors can often be picked up just by dumping the raw scan output into Google Translate, which shows things like numerals.

Here we have a character that is plainly wrong, and coming out as a numeral “4”.  It looks like an “o” with a hat and two dots under.  The two dots under are legs in another copy of Rossini.

I’m guessing that it’s a “ge” character, from looking at the Wikipedia article, but I can’t be sure. The script isn’t an alphabet, but a syllabary, based on syllables.  Each character is a consonant followed by a  vowel, which makes for a lot more characters.  There’s a table of the characters on the Wikipedia article, consonants down the left, vowels across the top.  I’ve not really looked at this.

The Google translate output is also interesting because of the choice of “detected language” – Tigrayan, rather than Amharic.  If you force it to Amharic, you get a lot less meaning.

One awkward part of using Google Drive to do the OCR is that it doesn’t preserve the line breaks.  That makes comparing the lines more awkward.   So you have to manually do this:

፬ ፡ ወኮነ ፡ በአሐቲ ፡ ዕላት ፡ ወነሥአ ፡ መጽሐፈ ፡ ወቀለመ ፡ ወወጠነ፡
ይጽሐፍ ። ወተንሥአ ፡ ለጸሎት በሰርክ ። ወጸሐፉ ፡ ሎቱ : መላእክት ፡ ወንጌ ለ ፡
በ፬ ፡ ሰዓት ፡ ወትርጓሜሁ ። ወመላእክተ ፡ እግዚአብሔር ፡ ወትረ ፡ ይት ለአክዎ ፡
ወእግዚእነሂ ፡ ክርስቶስ ፡ ያንሶሱ ፡ ምስሌሁ ። ወተሰምዐ ፡ ዜናሁ :
ውስተ ፡ ኵሉ ፡ ሀገር ። ጸሎቱ ፡ ወበረከቱ ፡ የሀሉ ፡ ምስሌነ ።

The Wikipedia article mentioned earlier gave me a list of punctuation marks.  There are two sorts of punctuation visible in here.  The colon mark is actually word division, which means that some words above go over two lines.  I’ve chosen not to split words above.  The double colon mark “::” is the full stop.  Interestingly Google Translate gives different results if you remove the spaces!

Going through the electronic text, removing spaces, I notice that sometimes the word-separator isn’t detected by the OCR.  So I added that in.  Sometimes it put a Roman colon instead, so I replaced that.  Finally I split on sentence:

ወጸሐፉ፡ሎቱ፡መላእክት፡ወንጌ ለ፡በ፬፡ሰዓት፡ወትርጓሜሁ።

And run it again and I get this:

But this still is not good enough to do much with.  If we didn’t have an idea what the text said, this would not tell us.

All this fiddling about would certainly get to into contact with the language, and start you on a journey to learning it.  But it’s not good enough a translation for other purposes, although intriguing.

One suggestion that was made in the comments to the last article was that ChatGPT gave better results.  The output quoted was indeed produced, and was very smooth and seemed to be a series of liturgical prayers.  But… I don’t think that this is actually the content.  These AI tools are really only an improved version of the text prediction tools you get on messaging on a mobile phone.  So it was pumping out garbage.

Anyway I tried it on this passage, and it crashed GPT very effectively!  At the moment I can’t get any reply of any sort, not even to “hello”.

I don’t think that I will do more here.  Clearly the technology is almost, but not quite good enough to be useful.


Listening to Hard Rock helps Egyptologist make Middle-Kingdom Papyrus Discovery

In 2010 a doctoral student at Johns-Hopkins University in Baltimore named Marina Escolano-Poveda was present at a conference of Egyptologists in Mallorca. While there she visited the small and obscure local museum.  There she discovered some papyrus fragments written in demotic.  Over time she studied these and found them to belong to the early Middle Kingdom.  In the end, her repeated efforts to find out what they were bore fruit.  She was working on her doctorate, on a different period, so work on the fragments had to take place at night. She recalls the moment:

I remember perfectly the moment when I had this revelation. It must have been 3 o’clock in the morning. I was listening to the song ‘Salir’ by the group Extremoduro. I then realised that the papyrus I was looking at was by the same author as that of La Dispute![1]

The fragments belonged to a famous text, the Dialogue of a Man with his Ba, preserved as part of a set of 4 papyrus rolls in Berlin.  These came from a notorious 19th century dealer at the court of the khedive Ismail.  His real name is unknown, but he called himself Jean d’Anastasie.  He offered them for sale at Sothebys in March 1837, claiming that they originated at Thebes/Luxor, as no doubt they did.  This was the same dealer who also uncovered the library of Greek magical papyri.  The four rolls were purchased by Lepsius in 1842 who lodged them in the Berlin museum. Today it is P. Berlin 3024.

There was some damage to the start of the roll, as is not uncommon.  So it is likely that some portions split off, and were subsequently sold separately by the enterprising dealer.  Some have since been identified in the Amherst collection. Nothing is known about how they came to Mallorca.

The fragments are P. Mallorca I and II.  I found a photograph on this paywalled site:

Papyrus Mallorca I – fragment of P. Berlin 3024, Dialogue of a man with his Ba.

The find was published, and thankfully this is both in English, and open-access: Marina Escolano-Poveda, “New Fragments of Papyrus Berlin 3024,” in: Zeitschrift für Ägyptische Sprache und Altertumskunde, 144 (2017) pp. 16-54 (accessible here).

There’s stuff out there, people.  If you look, you find it.  It doesn’t matter who you are, so long as you persist.

  1. [1]There are various tabloid accounts online.  This one is from Euronews, “Alicante Egyptologist Solves Enigma Of 4,000-Year-Old Manuscript”.

Is it possible to read editions of Amharic texts? An experiment

In my last post I mentioned how the Life of St Garima in Ethiopian was printed by Rossini, but without a translation.  In fact it has never been translated into any modern language, to my knowledge.  I don’t know any Ethiopian, and I doubt that I ever will.

But we live in an age of wonders, when it comes to unfamiliar languages.

So… is it possible to work with Ethiopian language editions, even if you know no Ethiopian?  What about Google Translate?  Ethiopian is in this heavy unfamiliar script.  Is there OCR for this?  If you can scan Rossini’s edition, can you pop it into Google Translate and get the English?

There are two sorts of Ethiopian out there, I know.  There is Ge`ez, or classical Ethiopian; and there is Amharic, the modern dialect.  Rossini printed his text from a 19th century manuscript.  So it seems likely that this is in Amharic.

A quick Google confirmed; Google Translate knows Amharic!  A bit of googling found me an Amharic news website online, here.  I’m using Chrome, so all I had to do was right-click anywhere and select “Translate to English” and the whole website was rendered into some sort of English.  And… it worked!!  Yay me!  It’s obviously not 100%, but it’s way better than 0%!

So what about OCR?  I was sad to see that Abbyy Finereader apparently doesn’t support Amharic.  That’s a blow.  It was developed originally to handle Cyrillic, so it certainly has the capability.  But it’s not offered.  Drat.

A bit of googling brought me to a dubious-looking website here, claiming to offer a selection of tools which could do Amharic OCR.  The prose felt a bit machine-generated, so I worried that it was bunk, or worse, a malicious site.  But the first option was… Google Drive.

I never knew this, but seems that, if you upload a PDF containing an image of text, and then open it in Drive as a Google Docs document, it OCR’s the content.

Well, I thought, let’s give it a try.  So I extracted the first page of Rossini’s edition, using Adobe Acrobat Pro 9 – no flashy latest-edition stuff going on here!  Here’s a pic:

Then I uploaded it, and opened as a Google document.  And … it just treated the Amharic as an image.  Dang!  But I noticed that it did indeed OCR the Italian at the top of the page!

This is supposed to work.  So I thought maybe I should work over the image a bit.  I imported the one-page PDF into Abbyy Finereader 15, and chopped off the Italian at the top, and the critical apparatus at the bottom.  I then used the image editor in Finereader to “whiten the background”.  This can be flaky, but this time it worked fine, and I got a pure white background.   And I got this:

(I’ve just seen the marginal notes, which I need to chop off as well, so I’ll have to go round the loop again)

I exported the image as a PNG, and I used Acrobat again to create a PDF from the image.  Then I uploaded the new PDF to Google Drive, and opened it as a Google Docs document.  And… it worked!  Sort of…

በስመ : አብ : ወወልድ ‘ ወመንፈስ ፡ ቅዱስ ፡ ፩ ፡ አምላከ ፡ ላዕሌሁ ፡ ተወ ከልኩ፡ ወቦቱ ፡ አመንኩ ፡ እስከ ፡ ላዓለመ ፡ ዓለም ፡ አሜን ።

ድርሳን ፡ ዘደረሰ ፡ ቅዱስ ፡ ዮሐንስ ፡ ኤጲስ ፡ ቆጶስ ፡ ዘአክሱም o ፡ በእንተ ዕበዩ ፡ ወክብሩ ፡ ለቅዱስ ፡ ይስሓቅ = ወይቤ ፤ ስምዑ ‘ ወልብዉ ፡ ኦአኀውየ 5 ፍቁራንየ ፡ ዘእነግረከሙ ። ርኢኩ ፡ ብእሲተ ፡ እንዘ ፡ ይዘብጥዋ ፡ ዕራቃ ወእንዘ ፡ ይሀርፉ ፡ ላዕሌሃ ፡ ወላዕለ ፡ እግዝእትነ ፡ ማርያም ፡ እንዘ ፡ ይብሉ በእንተ ፡ ወልዳ ፡ ክርስቶስ ፤ እምብእሲት ፡ ኪያሁ : ኢተወልደ ፣ ይብሉ ፡ እላ ፡ ኢየአምኑ ፡ በክርስቶስ = ወኮንኩ ፡ እንዘ ፡ እረውጽ ፡ ወአኀዝኩ እስዐም ፡ ታሕተ ፡ እገሪሃ ፡ ለይእቲ ፡ ብእሲት ፡ እንዘ ፡ ትብል ፤ እወ ▪በዝ ፡ አንቀጽ ፡ ወፅአ ፡ ንጉሠ ፡ ሰማያት ፡ ወምድር ። ወሶበ ፡ ትብል፡ ከሙዝ ፡ ወ

That’s… rather astonishing.  No idea what all that is, but it looks sort of right.  Let’s bear in mind that Rossini printed his edition in 1897.  This is not a modern typeface.  So this is rather good.

Next step was to paste it into Google Translate.  It set it to auto-detect the language, and pasted in the first bit.  And… it worked.  In fact it gave a really useful transcription into Roman letters as well, which makes it a LOT easier to manipulate the text.

OK, I’m cheating slightly.  The first time I uploaded, the translation ended at “Spirit”.  But this is a Google Translate bug – it sometimes omits the remainder of a sentence.  If you split the text with a line feed, you often get the rest.  And that’s what I did.  I worked out by experiment where I needed to be, and then I got the above.

I don’t quite believe the translation of the second sentence either.  I suspect I need to play with this a bit to work out what each word is.

I notice all those colons between every word.  It might help if I actually looked up the script online!

But I think you’ll agree that this is quite marvellous – I, who know absolutely nothing about the language, am getting something useful out!



An Ethiopic Legend: Abba Garima copied the Bible in a Single Day

Here’s an interesting one, which I came across today.  There is an early set of gospels in Ethiopia, at the Abuna Garima monastery in Ethiopia’s Tigrai Highlands.  An article in the Independent 6 July 2010 by Jerome Taylor tells us:

The monks have their own legend about how the gospels came into their possession. They believe they were written by Abba Garima, a Byzantine royal who arrived in what was then the kingdom of Axum in 494 and went on to found the monastery. According to the monks, Abba Garima finished his exquisite work in a single day because God stopped the sun from setting while he worked.

This claim is repeated in many places, often based on Wikipedia’s wording, which references the Independent article:

According to tradition, Abba Garima wrote and illustrated the complete Gospels in a single day: God stopped the sun from setting until the Saint completed his work.

“Tradition” is a weasel word.  We have no traditions, not in the modern age, handed down from father to son orally.  What we have are books.  So whenever we see “tradition” mentioned, we need to ask what the literary source is.

Fortunately information is not far to seek.  In Judith McKenzie and Francis Watson, The Garima GospelsEarly Illuminated Gospel Books from Ethiopia, Oxford (2016), (preview here), we find the following statement, p. ix, n.25:

23 There is a tradition that while Abba Garima was copying the Gospels, God stopped the sun so that he might complete the task in a single day (Heldman 1993: 129). This incident is not included in Carlo Conti Rossini’s edition of the text, where it is reported instead that, while he began the task, angels completed the work for him in four hours (Conti Rossini 1897: 161-62). There is, however, an illustrated manuscript at the monastery showing in a single miniature Abba Garima copying the Gospel below both a full and a setting sun, under the inscription “As Abba Garima wrote the Gospel in the land of Atäret” (fig 20 here). Atäret is one of the places said to have been granted to the monastery by Gabra Masqäl (A. Bausi, “Ǝnda Abba Garima,“ EAE 2: 284:. I am grateful to Denis Nosnitsin and Nafisa Valieva for their assistance in the identification of this story.

– M. Heldman, “The Heritage of Late Antiquity,” in: R. Grierson (ed), African Zion: The Sacred Art of Ethiopia, (1993), pp. 117-32, esp. 129-30.
– C. Conti Rossini, “L’omilia di Yohannes, vescovo d’Aksum in onore di Garimâ,” in: Actes du Onzième Congrès des Orientalistes. Paris – 1897. Quatrième section. Hébreu – Phénicien – Aramée – Éthiopien – Assyrien, Paris (1898), pp. 139–177.

The Garima Gospels: Early Illuminated Gospel Books from Ethiopia, p.18, fig.20.

I was only able to access the preview, but this extraordinarily interesting book must be the source for what appears online.  But let’s look deeper into the references.

The Heldman reference is as follows:

According to tradition, the monastery from which this manuscript [=Abba Garima Gospels I] takes its name was founded by Isaac, also known as Abba Garima, one of the Nine Saints who came to Ethiopia from the Roman Empire in the late fifth century. Monks of Abba Garima’s monastery reported to Donald Davies that Abba Garima himself had copied this manuscript, and that he had completed the task in a single day. God stopped the sun’s course one afternoon so that Abba Garima could finish copying his Gospels before night fell.[1]

  1.  Conversation between the author and Donald Davies, 1986.

So the story is circulating in just this form in Ethiopia currently.

The Rossini “section 4” is at Gallica here.  The article turns out to be the publication of a text in Ethiopic, the Gadla Abba Garima, called by Rossini the “Homily of Yohannes, bishop of Axum, in honour of Garima.” He printed it from a manuscript in the French National Library (A = Paris BNF et. 132, 19th c.) and one in Berlin (= B, with an impenetrable shelfmark, 16th c.).  The oldest MS is 15th century.  “Yohannes” is of course John, and a bishop of that name did arrive in Ethiopia from Egypt around 1439.[1]

Unfortunately the homily is given without translation (!).  The best the editor could do was to give a summary of what it says.  Basically Abba Garima was a Byzantine prince named Isaac, who came to Ethiopia and took the name of Garima, and is one of the Nine Saints who appear in Ethiopian literature in the 14-15th century and are unknown before then.[2]

I’ve run Rossini’s summary over into English.  Rossini states:

Here is a brief summary of the narration contained in the homily:

And then we get this:

After years of historic marriage, Masfyânos, king of Rome, and Sefengyà, his pious wife, have a son, Yeshâq, «whose name means “pearl”»; while the patriarch is baptizing him, a great supernatural light foretells his future glories (v. 14-36.) — At the age of twelve the child is sent to school, where he progresses rapidly: still growing up, his parents would like to give him a wife, but a celestial vision distracts them (v. 36-48).— With the death of Masfyânos, reluctant Yeshâq is placed on the throne. After seven years of peaceful reign, he secretly flees to Ethiopia, called there by a letter from the saint Pantalêwon of Somâ`t: the angel Gabriel transports him there in four days, while the messengers of Pantalêwon take ten months to return and four days. There Yeshâq received the monastic habit from Pantalêwon, and he remained with his teacher for a year (v. 48-108) — Then, having heard of his departure, Liqânos of Constantinople, Yem`atâ of Qosyât, Schmâ of Antioch, Gubâ of Cilicia, Afsê of Asia, Malâ` of Romyâ, and `Os of Caesarea also went to Ethiopia, and with great abstinence, and in great holiness, live in the same house with Pantalêwon and Yeshâq (v. 108-122). — While they are like this, a governor of Aksum announces to them how the country is dominated by a huge snake, Arwè, venerated as a God, and to which, in addition to infinite animals, a girl is given daily: assured of the fact thanks by sending of Yeshâq and `Os, who was very frightened at the sight of the monster, the nine saints with great prayers obtain from God the death of the serpent (v. 123-284). — Then Ethiopia is filled with tumults and disorders; until God, seeing the righteousness of the faith of that land, and hearing the prayers of the saints, invoking a king of David’s lineage, places Kâlêb on the throne. (v. 285-288). —  After thirteen years, and many prodigies performed by them, a poor monk, Melkyânos, joins them, whose humility they despise; whereupon, to punish them, God takes away from them a mysterious face, which used to come down to illuminate their meals. Having obtained the pardon of this, the saints divided, and Yeshâq retired to Madarâ (v. 288-309). — There he performs great miracles: he frees a possessed person, heals a woman who has been suffering for thirty years from an uninterrupted flow of blood, etc., etc.

At the same time he is made head of the priests of Madarâ (v. 310-345). — One day, he sows a grain of wheat: in a short time this germinates, grows, produces a very rich harvest, which the saint distributes to the poor (v. 346-355). — Another day, having ceased writing in order to pray, the angels, who always served him, copy for him the gospel and his interpretation of it (v. 356-360). — Heals a girl invaded by an evil spirit (v. 360-442). — Visited by two monks, he feeds them, but puts away his portion, whereupon, a little later, invited, he celebrates the Eucharistic sacrifice, for which, his companions not knowing that he was still fasting, he is accused to Pantalêwon. The latter calls him to an interview, and, having met, he invites him, in order to be able to take him back in secret, to have his companions leave: «not only men, but let the trees of the wood and the stones move away from us!» exclaims Yeshâq: the trees and the stones obey, so, recognizing his innocence, Pantalêwon shouts: «Garamkani, you have amazed me!» and from this Yeshâq takes the name of Garimâ (v. 443-491). — Returning to the convent, one day Garimâ stops the sun to be able to fulfill his prayers (v. 492-496). —  The donkey that used to serve him and bring him the gospel and food having died, he mourns him bitterly (v. 497-507). — After writing under a tree and having spat on a large stone, he makes a healthy spring gush forth (v. 507-511). — Having come across a village that does not observe Sunday rest, he scolds them, is badly beaten, and launches terrible curses against it (v. 512-527)» — King Gabra Masqal, hearing the saint’s wonders, visits him in Modani or Bèta Masqal, receives his blessing, has a church erected there in honour of the saint, and to this and to the convent of Garimâ he donates the land of Tâfâ, `Adwâ, Mesâh(?), Sebe`ito(?), and Maya Lehekuet (v. 528-556). —  One time the saint sows a grape: immediately it germinates, and he draws the juice for the mass. Gathered around him many men, he gives rules and precepts for the community (v. 557-665). —  Having descended into the heart of the mountain, he causes a wonderfully healthy spring to gush from it (v. 566-569). — When his pen falls while he is writing, it becomes a plant (v. 569-571). — Informed of all this, the king gives him the land of Atarêt and seven other cities (v. 571-575). —  The saint, while going with Yem’âtâ, stops a large boulder, which Satan rolls against him to kill him, meets one last time with Pantalêwon, by whom he is comforted for the beatings given to him by the violators of Sunday rest (v. 576-592). — And finally, warned by God of his imminent end, he obtains from Him great promises for those who venerate him, bids farewell to his brothers and disappears on the 17th of the month of Sene (v. 593-640). — One of his disciples then has a vision of future painful events in the locality sanctified by Garimâ, where a wicked people will settle (v. 641-645).[3]

It is a pity that we do not have a translation of pp.161-2, because the summary does NOT give us the story as Judith McKenzie summarises it.  I expect that Dr. McKenzie is correct.  But note how this 15th century text has two episodes that each have part of the story, rather than just one?  This suggests that the narrative we have today was assembled somewhere since this text was written.  Hagiography is often revised over time.

I was unable to discover whether any of the Lives of the Nine Saints have been translated into any western language.  If not, this is rather a shame.  The Life of Garima is 23 pages; surely not beyond the powers of any native English speaker who knows Ethiopian?  Is there anybody out there?

  1. [1]Or so I gather from a rather marvellous article: Stuart Munro-Hay, “Saintly Shadows”, in: Walter Raunig, Steffen Wenig (edd.), Afrikas Horn, Series: Meroitica 22, Otto Harrassowitz Verlag, (2005) 137-168; p. 162 (preview here).
  2. [2]Stuart Munro-Hay, p.140.
  3. [3]

    Ecco un breve compendio della narrazione, contenuta nell’ omilia :

    Dopo anni di storile matrimonio, Masfyânos, re di Roma, e Sefengyà, piissima sua moglie, hanno un figlio , Yeshâq, «il cui nome significa mar­garita»; mentre il patriarca lo battezza, una grande luce soprannanaturale ne preannuncia le future glorie (v. 14-36.) — A dodici anni il fan­ciotto è mandalo a scuola, ove rapidamente progredisce: cresciuto ancora, i suoi genitori vorrebbero dargli moglie, ma una visione celeste ne li dis­toglie (v. 36-48).— Morto Masfyânos, Yeshâq riluttante è posto sul trono. Dopo selle anni di pacifico regno, fugge di nascosto in Etiopia, chiama­tovi da una lettera del santo Pantalêwon di Somâ`t: l’angelo Gabriele ve lo trasporta in quattro giorni, mentre i messi di Pantalêwon impiegano nel ritorno dieci mesi e quattro giorni. Colà Yeshâq receve da Pantalêwon l’abito monacale, e col suo maestro rimane un anno (v. 48-108) — Allora, avuta notìzia della sua andata, passano in Etiopia anche Liqânos di Costantinopoli, Yem`atâ di Qosyât, Schmâ di Antiochia, Gubâ della Cilicia, Afsê dell’ Asia, Malâ` di Romyâ, `Os di Cesarea, e con grandi astinenze, e in grande santità vivono in una sola casa con Pantalêwon e Yeshâq (v. 108-122). — Mentre così stanno, un governatore di Aksum annuncia loro come il paese sia dominato da un immane serpente, Arwè, venerato come un Dio, e in pasto al quale, oltre a infiniti animali, si dà giornalmente una fanciulla: sinceratisi del fatto mercè l’invio di Yeshâq e di `Os, il quale ultimo assai si spaventa alla vista dei mostro, i nove santi con grandi preghiere ottengono da Dio la morie del serpente (v. 123-284). — L’Etiopia allora si empie di tumulti e di disordini; sino a che Dio, vedendo la rettitudine della fede di quella terra, ed esau­dendo le preci dei santi, invocanti un re della stirpe di Davide, pone sul trono Kâlêb. (v. 285-288). — Dopo tredici anni, e compiuti da loro nu­merosi prodigi, un povero monaco, Melkyânos, si unisce ad essi, i quali ne vilipendono l’umiltà; onde, per punirli, Dio lor toglie una face mis­teriosa, che soleva scendere a illuminarne i pasti. Ottenuto di ciò il per­dono, i santi dividonsi, e Yeshâq ritirasi in Madarâ (v. 288-309). — Ivi egli compie grandi miracoli: libera un ossesso, guarisce una donna, da trenta anni sofferente per ininterrotto flusso di sangue, ecc., ecc.

    Intanto, è fatto capo dei sacerdoti di Madarâ (v. 310-345). — Un gior­no, egli semina un acino di grano: in breve ora questo germina, cresce, produce una ricchissima messe, che tutta il santo distribuisce ai poveri (v. 346-355). — Un altro giorno, avendo cessato di scrivere per pregare, gli angeli, che sempre lo servivano, gli copiano l’evangelo e la sua in­terpretazione (v. 356-360). — Sana una fanciulla invasa dallo spirito ma­ligno (v. 360-442). — Visitato da due monaci, egli dà loro da mangiare, ma ripone la sua parte, onde, poco di poi, invitato, celebra il sacrifizio eucaristico, di che, ignorandosi dai suoi compagni com’ egli si conservasse digiuno, è accusato presso Pantalêwon. Questi lo chiama a colloquio, e, incontratolo, lo invita, per poterlo riprendere in segreto, a far allontanare i suoi compagni: «non gli uomini soltanto, ma gli alberi del bosco e le pietre si scostino da noi!» esclama Yeshâq: gli alberi e le pietre obbe­discono, onde, riconosciuta l’innocenza, Pantalêwon grida: «Garamkani, mi hai stupito!» e da ciò Yeshâq trae il nome di Garimâ (v. 443-491). — Tornato al convento, un dì Garimâ ferma il sole per poter compiere le sue preghiere (v. 492-496). — Essendo morto l’asino che soleva ser­virlo e portargli l’evangelo ed il cibo, lo piange amaramente (v. 497- 507). — Stando a scrivere sotto un albero e avendo sputato su un gran sasso, ne fa sgorgare una fonte salutare (v. 507-511). — Imbattutosi in un villaggio che non osserva il riposo domenicale, lo redarguisce, è in malo modo percosso, e contro di esso lancia terribili maledizioni (v. 512- 527)» — Il re Gabra Masqal, intesi i prodigi del santo, lo visita in Modani o Bèta Masqal, ne riceve la benedizione, fa in onore del santo colà erigere una chiesa, e a questa ed al convento di Garimâ dona la terra di Tâfâ, ‘Adwâ, Mesâh(?), Sebe`ito(?), Maya Lehekuet (v. 528-556). — Una volta il santo semina un acino d’uva: subito questo germina, ed egli ne trae il succo per la messa. Adunatisi intorno a lui molti uomini, egli dà regole e precetti per la comunità (v. 557-665). — Sceso nel cuor del monte, fa da esso zampillare una fonte mirabilmente salutare (v. 566- 569). — Cadutagli la penna mentre sta scrivendo, essa diviene una pianta (v. 569-571). — Informato di tutto ciò, il re gli dona la terra d’Atarêt ed altre sette città (v. 571-575). — Il santo, mentre va con Yem’âtâ, ferma un grande macigno, che Satana, per ucciderlo, gli roto­lava contro, incontrasi un’ultima volta con Pantalêwon, dal quale e confortato per le percosse dategli dai violatori del riposo domenicale (v. 576-592). — Ed infine, avvertito da Dio della prossima sua fine, e ottenute da Lui grandi promesse per quelli che lo venereranno, sa­luta i suoi fratelli e scompare al 17 del mese di sanê (v. 593-640). — Un suo discepolo ha, poscia, una visione circa futuri dolorosi eventi della località santificata da Garimâ, ove si stabilirâ un popolo malvagio (v. 641-645).

Theotokos: A review of the scholarship on the paleographical date of P.Ryl. III 470 (“Sub Tuum Praesidium”)

In 1929 papyrologist C.H. Roberts published a papyrus fragment from Egypt.[1]  The text is in Greek, and is a Christian prayer, containing the word “theotokos”.  The fragment is held by the John Rylands Library in Manchester, where it has the shelfmark P.Rylands 407 (online here).  Here is the excellent image from that site:

P.Rylands III 470, via Manchester Digital Collections.

The statements made by Roberts have dominated all subsequent discussion of this papyrus.  Here is a page-grab of it.

This is a catalogue entry, not a journal article, so it is inevitably concise, and the details are provisional.  Roberta Mazza (2019) adds to this description:

The manuscript in question was purchased in Egypt by J. Rendell Harris in 1917 and published by C. Roberts in 1938.3 It is a small fragment that presents many aspects of interest for its material shape and the text it carries. In his edition, Roberts proposed a date in the fourth century with a question mark due to the many doubts the papyrus raised. Roberts’s discussion in the introduction addressed both the handwriting and the material features of the object, as well as the textual content. Written along the fibre of a leaflet, the back of which is blank, was clearly a hymn or prayer to the Mother of God, as indicated by the vocative θεοτόκε at line 4. The shape of the capital letters – tall and compressed, adorned by decorative serifs – was briefly discussed and compared to the chancellery style of that used in a letter of the prefect Subatianus Aquilas, securely dated to 209 CE (SB I 4639).[2]

The letter of Subatianus Aquila dates to 209 AD, and looks like this (this wonderful image from Berlin, online here):

P.Berl.11532. Edict of the prefect, Subatianus Aquila, and signed by him personally.

I would have liked to add images of the other papyri mentioned below, but I have not found these online.

This “chancellery style” mentioned here is referred to by Stegmüller, below.  Likewise the “peculiar Alpha found in 470” mentioned by Roberts is something that Hans Förster was later to pick up on.  It is clear that cautious old Colin Roberts is not at all comfortable with the date insisted on by E. Lobel.

Before we examine how this publication has been received down the years, it is probably best to say something about methodology.

The process of assigning a date to a papyrus fragment, or indeed any manuscript, relies on two independent pieces of data.

  • The first is the paleographical date.  The trained paleographer consults handbooks filled with examples of papyri where the text itself contains a date, or where the papyrus can certainly be assigned a date on some other grounds.  He then compares these with the papyrus under inspection, and assigns it a date based on similarities of letter shape, abbreviations, and so on.  It is inevitably a subjective process, but rewarding if done carefully.
  • The second is the historical date that emerges from the text, and particularly the history of doctrine.  The Christian religion has elaborated its teachings considerably since the looser apostolic age, in response to various pressures and indeed fashions, while proceeding to the very careful language of Trent.  Texts may therefore be placed somewhere along that timeline, based on whether they are similar to other doctrinal expressions characteristic of a period of history.

In both cases, it is necessary to start from what is certainly known, and assess the undated object against it.  The undated object should not be dated before the earliest evidence of something, nor after the latest evidence of it.  It is generally wisest to place it closer to the middle of the bell curve, in the absence of any other evidence, rather than close to an outlier.

A Bell Curve for probability, showing occurrences over time.

Following the 1938 Roberts publication, the next step came in 1939 from Father F. Mercenier O.S.B., who recognised – there are advantages to having priests in papyrology – that the text was an early version of the Marian prayer, “Sub Tuum Praesidium” (“Under thy Compassion”).  But on the date Mercenier, like so many since, merely repeated the judgement of Lobel and Roberts.[3]

1940 featured an over-excited claim by Ortiz de Urbina, S.J., stating, “we also note that ἀειπαρθένος – still rare – appears in Didymus of Alexandria and in the Sub Tuum Praesidium,” although in fact the fragment does not contain it, nor is it part of the earliest text, but a later addition.[4]  Again this simply accepted the date given by Roberts and Lobel.

The first reassessment of the paleography of Lobel appeared in 1952 from Otto Stegmüller.  It reads as follows (via Google translate – German in the footnote).  I have added emphasis.

Our joy at finding the Sub tuum praesidium on an old papyrus is dampened by the difficulty of dating the piece. The date of origin of the papyrus is of great interest to us, especially as evidence of the veneration of the Mother of God before the Ephesinum is very sparse. Are we coming back with our fragment much beyond that point in time? The editor, Mr. Roberts, first acquaints us with the judgment of the paleographer Lobel. “Lobel would be unwilling to place 470 later than the third century. But such individual hands are hard to date, and it is almost incredible that a prayer addressed directly to the Virgin in these terms could he written in the third century”. Roberts concludes that our papyrus cannot have been written before the second half of the 4th century.  P. Mercenier brings in his work the dating of Lobel and that of Roberts, without deciding on one. All further mentions of the papyrus therefore place it in the third or fourth century.

Is this dating justified?  In the first instance, the paleography must be heard. But it must be explained from the outset that it cannot give a reliable judgement. The writing belongs to the so-called chancellery style, which has remained essentially the same over the centuries. The firm starting point for the dating is the letter of the governor Subatianus Aquila of the year 209 AD. Our writing certainly belongs to this style, which was developed and practiced in the Alexandrian chancery. Individual letters, especially some ornaments, suggest that our writing is a little later than the Subatian letter. We know very similar writings from the Book of the Iliad in the Morgan Collection (end of the 3rd or beginning of the 4th century; Schubart, Paleography fig. 95) and from the Stockholm forgers’ book (first half of the 4th century; Schubart fig. 96 ). These writings, originating in Egypt, give us a justification for taking the beginning of the 4th century as the earliest point for our writing. Paleography has to be content with that. To this day, it is unable to decide whether the script was written in the 4th, 5th or 6th century. The Coptic liturgy gives us a certain term ante quem. Texts that exist in the Byzantine and Coptic liturgy must be dated before the turn of the 5th and 6th centuries, since no Greek text was included in the Coptic liturgy after this point in time.[5]

In 1969 there was a paper by G. Giamberardini, “Sub tuum praesidium” e il titolo “theotokos”, in: Marianum: Ephemerides Mariologiae 31 (1969) 324-362, often referenced in the literature.  This criticises severely the reconstruction of Mercenier, but there is no attempt at a paleographical dating.  Instead it merely endorses that of Lobel, and supporting it by a non-paleographical argument from supposed early uses of the word “theotokos”; where himself admits (p.352):

“Having suspected the authority of Origen, the tendency has arisen to consider interpolated and spurious all the documents prior to the Athanasian period, in which the Θεοτόκος appears.”[6]

In 1995 the first of two papers by Austrian papyrologist Hans Förster appeared, which form the first serious study of the papyrus.[7]  I cannot very well give this excellent paper in full here, but it is quite difficult to access, and I thought that it would help if I translated a few sections into English and put them here.  Unlike all previous papers, there is a section (p.186-7) specifically to the paleography.  I’ve run it through Google Translate (the original German is in the footnote):

Digression on the Exact Dating of the Manchester fragment13:

The first thing that catches the eye is the comparatively narrow and tall way the letters are written. The streaks of the ε, for example, are so short that in line 10 Roberts reads a ρ instead of an ε. Despite the uncial handwriting, the μ is not written like a capital Latin “M”, but rather resembles two adjacent “1”s in Latin cursive; the crossbar of the η is in the upper third of the letter.  The υ, like the α, is very peculiar and corresponds to a Latin “V”, while the λ corresponds to an upside down “V”; the usual extension of the right diagonal to the top left does not take place.  The κ consists of two independent parts. There is a noticeable gap between the vertical downstroke and the two diagonals, which are drawn in one sweep.  This form of the κ can be found, for example, in a psalm fragment from the first half of the 8th century, P.Amst. I 2114.  The crossbar of the η written very high up is also found in a liturgical calendar from the year 535/6, the P.Oxy. 135716.  The form of the μ on that papyrus also resembles the form found in P.Ryl. III 470.  The  κ in this papyrus, however, in contrast to P.Ryl. III 470, is written as a continuous letter.  The δ in P.Oxy. 1357 has a diagonal stroke to the top left, while in the fragment P.Ryl. III 470 it is written like a triangle.  The previously mentioned form of the κ, an identical form of the π, the above-mentioned form of the μ and an ε with similarly short strokes can also be found in P.Berol 13269, which is assigned to the 7th/8th Century16.  P.Lond. 1817, dated in the 6th century, still alternates between the uncial and the above-mentioned form of the μ, and similarities to the Manchester fragment exist with regard to the ε and the κ, because the gap between the two parts of the κ in P. Lond. 1817 is not as large as in P.Ryl. III 47017.  In summary it can be said that a dating to the 3rd or 4th century must be regarded as very improbable. Based on the comparative texts cited, P.Ryl. III, 470 belongs between the 6th and 7th centuries. The 5th century is unlikely as the date of origin, but cannot be completely ruled out. The use of brown ink, of which there is no evidence prior to the 4th century, also points in this direction18.

13.  As for paleographic dating, it should be noted that majuscule manuscripts, as Roberts himself writes in his commentary, are extremely difficult to date. At the ends of the strokes of κ, λ, ι, ε, and σ are small dots of ink that give the writing an artificial, or, as Lobel put it, decorative appearance. For a similar phenomenon in magical texts, cf. p. 7 col. 17; pt 10; p. 36 col. 2; P. 36 Col. 7 in: K. Klassendanz, Papyri Graecae Magicae. Die griechischen Zauberpapyri (=The Greek Magical Papyri) II, Stuttgart 2nd ed. 1974. Lobel cites the letters ο, ι and ε as the important comparative letters for his dating.

14.  H. Maehler, G. Cavallo, Greek Bookhands of the Early Byzantine Period. A. D. 300-800, London 1987, 188.

15.  Maehler, Cavallo, Bookhands (see note 14) 68.

16.  See R. Seider, Paläographie der Griechischen Papyri II 2: Literarische Papyri, Stuttgart 1970, 179.

17.  Seider, Paläographie II 2 (see note 16), 175f.

18.  I am indebted to H. Harrauer for the discussion. See also V. Gardthausen, Griechische Paläographie 1: Das Buchwesen im Altertum und im byzantinischen Mittelalter, Leipzig 2nd ed. 1911, 203-217. He writes (p.205) that around the 5th century AD “a brownish, also metallic ink came into use” and quotes Schubart in this context in note 5 on the same page, who speaks of a brownish-red ink , which can be found in the papyri from the 4th century AD. At the same time, the more valuable brown ink was only used for important texts.[8]

The paper is full of relevant material.  Förster notes that there are folds in the papyrus, which mean that originally it was folded into a very narrow strip.  This, he infers, together with the “almost monumental” letters and decorated form, the brown ink, the folds, and the blank reverse, suggests that it was an amulet.  This would explain the “unique script” observed by Roberts.  Note 8 discusses the form of the alpha, which Roberts felt was characteristic of inscriptions:

8.  This form of the α is common in inscriptions. An example of a template for an inscription showing an α with a broken crossbar is P. Vindob. G. 26.013, cf. Sijpesteijn, P.J., Wiener Melange, II. Christliches, a) Ein Trishagios-Hymnus, ZPE 40 (1980) 92-95. However, the form of the μ on the inscriptions is not identical to that shown on the Manchester papyrus. C. M. Kaufmann, Handbuch der altchristlichen Epigrafik, Freiburg i. Br. 1917; this form of the α has, for example, a grave inscription from the 5th century (p. 75), an inscription above the burial chamber door of Abbess Thekla (p. 290f), the writing samples from Antioch from the 5th and 6th centuries on p 413 and the Greek inscriptions on pp. 414ff. The Coptic Shenute epitaph, which is cited on p. 75f, also has an α with a broken crossbar. However, all the inscriptions cited show a form of the μ, which corresponds to a capital Latin “M”, while there are magical papyri in which the α has the form described. See P. Mil. Vogl. 127 from the 2-3rd century: I. Gazzaniga, M. Vandoni, Papiri della Universitä degli Studi di Milano, Milan 1965, pp. 59f. This also speaks against its use as a template for an inscription and more for its use as an amulet.[9]

He ends with this summary:

After new paleographic study of P.Ryl. III 470, the early dating of this fragment – to the 3rd century -,  which has hitherto been generally accepted in theological scholarship, cannot be maintained. The dating to the 6th or even 7th century means that its value as a source of historical evidence that has often been assigned to this fragment falls apart. A Viennese fragment that transmits the same text means that the previously considered possible supplements are almost entirely based on the Byzantine form of the antiphon constricted. The question arises as to whether this antiphon could possibly have found its way into the Eastern liturgy from the Western liturgy. The variety of forms known in the West[47] supports this, while only two Greek forms of this antiphon are known, each of which shows only marginal deviations from one another. The papyrus from Manchester was used as an amulet. A sometimes assumed use as a template for an inscription is ruled out.

47.  See on this Giamberardini, „Sub tuum praesidium“ (note 10), 331-336.[10]

The paper was noted in the review of recent publications, C. Römer, “‘Christliche Texte II”, in: Archiv für Papyrusforschung 44 (1998), 129-39, which said (p.138):

“A new dating to the 6-7th century of P. Ryl. III 470 (not convincing to me) attempted by H. Förster…”

983 Eine Neudatierung auf das 6.-7. Jh. n. Chr. des P. Ryl III 470 (für mich nicht überzeugend) versucht H. Förster, Zur ältesten Überlieferung der marianischen Antiphon “Sub tuum praesidium”, Biblos 44,2, 1995, 183-192”

Ten years later, in 2005, Förster returned to the subject in a further paper, this time taking account of paleographical evidence from Coptic papyri.[11]  Here is an extract, p.106 f.:

On the Paleographical Dating

Lobel uses only a few letters to justify his dating. However, the overall impression of the text seems to contradict its dating. Of course, the “singular alpha”, which is said to be used primarily in inscriptions, is a particular problem in paleographical dating.  In this respect, a look beyond the Greek texts is required: such an alpha can be found in a whole series of Coptic texts, where it is found primarily in Greek of Greek. Reference need only be made to Brit. Mus. Or. 6782, fol.1.  The inscriptions of the depiction of Saints Theodoros and Mercurius in Vat. Copto 66, fol. 210 and 287 also offer an alpha within the Greek passages, as is also found in the Greek text that contributed to the early dating of the origin of the antiphon Sub tuum praesidium.[27] However, this alpha is neither recorded in the Paläographie by Cramer[28] or in Stegemann’s Paläographie.[29] This does not mean, however, that this particular alpha is only found in the Greek passages of Coptic literary texts.[30]  The inscription of the depiction of Mary on Pierpont Morgan Lib. M. 597 also offers the “strange” Alpha in a Coptic context[31]. The same Alpha form can also be found on a mummy tablet  from the Coptic period.[32]

In this context it is noticeable that a whole series of other “noteworthy” aspects in the manuscript of the papyrus from Manchester can also be explained by a closeness to the Coptic manuscript tradition.  We neeed only refer to the point-like thickening of the ends of many strokes.  For Greek literary texts from the period mentioned, this can rightly be described as eccentric, but for Coptic literary texts a large number of documents can be found, especially from the 8th and 9th centuries[33]. Also the Upsilon and the Mu can probably be better explained by their closeness to this writing tradition than by assigning the text to the 2-3, or 4th century. In this context, it must be particularly emphasized that, for example, the alpha of the comparative texts cited by Roberts could not be more different. In the case of the comparative texts cited for the early dating, one can therefore only speak of very selective points of contact, but this does not seem sufficient for an early dating, if at the same time the points of contact with the Coptic scribal tradition of the 8th and 9th centuries are so conspicuous that this must be described as a close relationship.

So if you take the overall impression of the text, a later date does not seem unjustified, the parallels to the Coptic manuscripts leave a date of the 8-9th century as more probable than the dating of the manuscript to the 2-3, or even the 4th century.

As a consequence of these considerations, it must be stated that not only circumstantial evidence, but very clear parallels can be cited against the often over-enthusiastic classification of the Manchester papyrus as the oldest witness of the antiphon Sub tuum praesidium.  In this respect, the Vienna copy of this antiphon[34], which is from the 6/7 century, must considered to be the oldest copy.[12]

27. See J. Leroy, Les Manuscrits Coptes et Coptes-Arabes illustres. BAH 94. Paris 1974, plate 105.
28. See M. Cramer, Koptische Paläographie. Wien 1964. Table I
29. See Stegemann, Koptische Paläographie. 25 Tafeln…, Heidelberg, 1936.
30. For example, P. Vindob. K. 7589. The edition of which is currently being prepared by the author is such an alpha: cf. the corresponding section of the parchment on Plate 4.
31. See Leroy, Manuscrits, plate 35.
32. See H. Froschauer, “Tradition im koptischen Bestattungswesen. Ein christliches Mumientafelchen aus den Beständen Tamerit in der Papyrussammlung der Österreichischen Nationalbibliothek.” Eirene 40 (2004) 91-100; 96: “The […] Alpha is […] characterized by a jagged middle branch with a descender. When the letter is rotated 180°, it takes on the shape of an omega. However, whether this spelling was consciously intended as a ligature of alpha and omega and thus perhaps even connected to the sign of the cross of the Christian symbolism of beginning and end, is rather doubtful and cannot be treated as more than speculation.” But the letter is also part of a Greek word in the passage quoted by Froschauer – in this respect it cannot represent the sign of the cross with “Alpha and Omega”; the cross is also very often to be found at the beginning of the text on a small mummy tablet; the “strange” alpha described above is usually not to be found in the vicinity of a sign of the cross; and, ultimately the rotation of an entire codex or an inscription on a tombstone, required for the recognition of the speculatively possible ligature in the first place, is rather impractical. So this alpha must be taken solely as a decorative variant of this letter, which is typical of the Coptic period, and is therefore of interest as a help in dating an undated text. On the Viennese parchment leaf, too, the alpha is to the left of the column and is somewhat enlarged, so here the decorative character of the letter is obvious.[13]
33.  Cramer. Palaographie, table 13
34.  This is P. Vindob. G. 17.944; see K.Treu and J. Diethart, Griechische literarische Papyri christlichen Inhaltes II. Wien 1993, 56 and plate 16, there listed as fragment 29.

This seems to be the latest paleographical discussion.  There are many other papers which repeat the Lobel dating, always in passing, but none of them seem to be aware of Förster’s arguments.

In 2014 Anne-Marie Luijendijk, Forbidden Oracles? The Gospel of the Lots of Mary, Tübingen: Mohr Siebeck 2014, p.30, n.87 mentions the question thus:

87. P.Ryl. Gr. III 470 (ed. Roberts). … The best edition is Stegmuller, “Sub tuum praesidum”. The date of the papyrus is disputed. Stegmuller (ibid., esp. at 82) places it at the end of the fourth century. Förster proposes a much later date in several publications: „Zur altesten Uberlieferung“, idem, „Fruheste Zeugnisse der Marienfrommigkeit“, and idem, „Die alteste marianische Antiphon“. Romer („Christliche Texte“, 138) deems Förster’s arguments “not convincing” („fur mich nicht uberzeugend“). …

A collection of papers in 2015, Presbeia Theothokou: The Intercessory Role of Mary across Times and Places in Byzantium (4th-9th Century), (online here) contained two relevant papers.

Theodore de Bruyn wrote (p.116, n.18):

18. Förster’s article of 2005 has, to my knowledge, not been considered in any of the more recent studies of the origins of the cult of Mary. Some studies also do not take his article of 1995 into account, e.g., Johnson, “Sub tuum praesidium,” 254–55; Price, “The Theotokos,” 89 n. 4 (corrected in Price, “Theotokos: The Title,” 56 n. 1, but without substantial discussion). A full consideration of Förster’s arguments is beyond the scope of this paper; I hope to discuss them elsewhere. Suffice to say, however, that any argument for a third- or fourth-century date for the antiphon must take as its point of departure the paleographical considerations of Förster (as well as Stegmüller) allowing for a later date for P.Ryl. III 470. Lobel’s brief paleographical remarks in P.Ryl. III 470 intro. cannot remain the principal basis for assigning the papyrus an early date. The issue is now no longer whether a prayer referring to Theotokos can be assigned to the third century; Roberts’ argument on that point has, obviously, been refuted. The issue is whether an antiphon whose earliest witnesses may be assigned to the sixth or seventh centuries or later originated in the third century.[14]

And Arne Effenburger wrote (p.50) that Förster had ‘convincingly explained, the papyrus – in the present material form probably a “protective amulet – can only have been created “between the 6th and the 7th century” due to paleography and the use of brown ink, which is why “a dating to the 3rd or 4th century as very improbable”’.[15]

A 2021 paper in Polish by P. Towarek, “Prayer „Sub Tuum praesidium”: Time of Origin, Place in Liturgy and Reception in Musical Culture. Outline of the Issues,” Vox Patrum80, 239–268 (online here, with English abstract).  The abstract states:

In the discussion on the question of its dating, many researchers pointed, for example, to the 3rd century (Giamberardini, Starowieyski). It turns out, however, that in the light of the latest palaeographic research, this time should be moved to the 6th/7th or even 8th/9th century (Hans Förster, Theodore de Bruyns, Arne Effenberger).

The Trismegistos website for the papyrus here gives it the reference TM 64320 / LDAB 5541 with a date of AD 700-900, and references the Förster 2005 paper, “followed by Mihalyko, p. 353”.  The latter is A. T. Mihálykó, The Christian Liturgical Papyri, Studien zur Antike und Christentum 114, Tübingen (2019), p. 353 no. 267:

A. T. Mihálykó, The Christian Liturgical Papyri (Studien zur Antike und Christentum 114) p. 353 no. 267

The most recent paper known to me is H. Munoz, “Papiro griego Rylands 470: notas a una de las más antiguas oraciones a la Virgen / Greek Papyrus Rylands 470: Notes to One of the Oldest Prayers to the Virgin”, in: Estudios Clásicos 163 (2023), 59-67 (online here with bibliography) which is not paleographical.  It mentions the Förster dating, and adds (n.4) that it is “contested by Römer 1998 and Luijendijk 2014,” which rather misleads the reader.

It seems that it may be a while before knowledge of the question-mark over the date of P.Ryl.III 407 becomes general.  Let us hope that the question will attract more professional paleographers to examine the date again.

  1. [1]C. H. Roberts , Catalogue of the Greek and Latin Papyri in the John Rylands Library Manchester, Volume III, Theological and Literary Texts (Nos. 457-551), Manchester: Manchester University Press, (1938), p.46-47 (p.62 of the PDF).  Downloadable from here.
  2. [2]R. Mazza, Dating Early Christian Papyri: Old and New Methods – Introduction, “Journal for the Study of the New Testament” 42/1 (2019), p. 49-50.  She also calls the Lobel dating “a shaky provisional date”.
  3. [3]P.-F. Mercennier, “L’Antienne mariale grecque la plus ancienne.,” Le Muséon, 52 (1939), pp. 229-233.  Unfortunately I have not been able to access this, so am reliant on the account given by Stegmüller.
  4. [4]I. Ortiz de Urbina S. J., “Lo sviluppo délia Mariologia nella Patrologia Orientale,” Orientalia Christiana Periodica 6 (1940), 54: “Notiamo inoltre che ἀειπαρθένος — ancora raro — fa capolino in Didimo Alessandrino e nel Sub tuum praesidium”.  Again I have not been able to access this, so rely on Stegmüller.
  5. [5]Otto Stegmüller, “Sub tuum praesidium. Bemerkungen zur ältesten Überlieferung.,” Zeitschrift für Katholische Theologie, 74 (1952), pp. 76-82: “Die Datierung. — Unsere Freude darüber, daß wir das Sub tuum prae sidium schon auf einem alten Papyrus finden, wird durch die Schwierigkeit der Datierung des Stückes gedämpft. Die Entstehungszeit des Papyrus ist für uns von größtem Interesse, zumal Zeugnisse einer Muttergottesverehrung vor dem Ephesinum sehr spärlich sind. Kommen wir mit unserem Stück wesentlich über diesem Zeitpunkt zurück? Der Herausgeber, Mr. Roberts, macht uns zuerst mit dem Urteil des Paläo graphen Lobel bekannt. „Lobel would be unwilling to place 470 later than the third Century. But such individual hands are hard to date, and it is almost incredible that a prayer addressed directly to the Virgin in these terms could be written in the third Century”7. Roberts kommt zum Schluß, daß unser Papyrus nicht vor der zweiten Hälfte des 4. Jahrhunderts geschrieben sein kann. P. Mercenier bringt in seiner Arbeit die Datierung von Lobel und die von Roberts, ohne sich für eine zu entscheiden. Alle weiteren Erwähnungen des Papyrus setzen diesen daher ins dritte oder vierte Jahrhundert. Ist diese Datierung berechtigt? Als erste Instanz ist die Paläographie zu hören. Aber sie muß von vornherein erklären, daß sie kein sicheres Urteil abgeben kann. Die Schrift gehört zum sogenannten Kanzleistil, der über Jahrhunderte hinweg im wesentlichen gleich bleibt. Der feste Ausgangspunkt für seine Datierung ist der Brief des Statthalters Subatianus Aquila vom Jahre 209 n. Chr8. Unsere Schrift gehört sicher zu diesem Stil, der in der alexandrinischen Kanzlei ausgebildet und geübt wurde. Einzelne Buchstaben, besonders einige Verzierungen legen nahe, unsere Schrift etwas später als den Subatianusbrief anzusetzen. Ganz ähnliche Schriften kennen wir aus dem Iliasbuch der Sammlung Morgan (Ende des 3. oder Anfang des 4. Jh.; Schubart, Paläographie Abb. 95) und aus dem Stockholmer Fälscher buch (erste Hälfte des 4. Jh.; Schubart Abb. 96). Diese aus Ägypten stammenden’ Schriften geben uns ein Recht, den Anfang des 4. Jahrhunderts als frühesten Terminus für unsere Schrift anzunehmen. Damit muß sich die Paläographie begnügen. Mit ihren Mitteln kann sie bis heute nicht entscheiden, ob die Schrift im 4., 5. oder 6. Jahrhundert geschrieben ist. Einen gewissen Terminus ante quem gibt uns die koptische Liturgie. Texte, die in der byzantinischen und in der koptischen Liturgie vorhanden sind, müssen vor der Wende des 5. und 6. Jahr hundert angesetzt werden, da nach diesem Zeitpunkt kein griechischer Text mehr in die koptische Liturgie übernommen wurde9.”
  6. [6]Messa in sospetto l’autorità di Origene, si è creata la tendenza a ritenere interpolati e spuri tutti i documenti anteriori al periodo atana- siano, nei quali si nomina la Θεοτόκος.
  7. [7]H. Forster, “Zur ältesten überlieferung der marianischen Antiphon Sub tuum praesidium”, in: Biblos:  Osterreichische Zeitschrift fur Buch- und Bibliothekswesen 44 (1995) 183-192.
  8. [8]Exkurs zur genauen Datierung des Fragmentes aus Manchester13:  Zunächst fällt die vergleichsweise schmale und hohe Schreibweise der Buchstaben auf. Die Ausstriche des ε beispielsweise sind so kurz, daß Roberts in Z. 10 anstelle eines ε ein ρ liest. Trotz der unzialen Handschrift ist das μ nicht wie ein großes lateinisches „M“ geschrieben, sondern ähnelt eher zwei nebeneinanderliegenden „1“ in der lateinischen Schreibschrift; der Querbalken des η findet sich im oberen Drittel des Buchstaben. Das υ ist, ebenso wie das α, sehr eigen und entspricht einem lateinischen „V“, während das λ einem umgedrehten „V“ entspricht; die sonst übliche Verlängerung der rechten Diagonale nach links oben findet nicht statt. Das κ besteht aus zwei unabhängigen Teilen. Zwischen dem vertikalen Abstrich und den beiden Diagonalen, die in einem Schwung durchgezogen werden, findet sich ein merklicher Zwischenraum. Diese Form des κ findet sich zum Beispiel in einem Psalmenfragment aus der 1. Hälfte des 8. Jh., P.Amst. I 2114. Der sehr weit oben geschriebene Querbalken des η findet sich auch in einem liturgischen Kalender aus dem Jahr 535/6, dem P.Oxy. 135715. Auch die Form des μ auf diesem Papyrus gleicht der Form, wie sie sich in P.Ryl. III 470 findet. Das κ wird in diesem Papyrus jedoch im Gegensatz zu P.Ryl. III 470 als zusammenhängender Buchstabe geschrieben. Das δ hat im P.Oxy. 1357 einen Ausstrich der Diagonalen nach links oben, während es bei dem Fragment P.Ryl. III 470 wie ein Dreieck geschrieben wird. Die angesprochene Form des κ eine identische Schreibung des π, die beschriebene Form des μ sowie ein ε mit ähnlich kurzen Ausstrichen findet sich auch in P.Berol 13269, der in das 7./8. Jh. eingeordnet wird16. P.Lond. 1817, in das 6. Jh. datiert, wechselt noch zwischen der unzialen und der beschriebenen Form des μ, Ähnlichkeiten zum Fragment aus Manchester bestehen hinsichtlich des ε und des κ, wobei der Spalt zwischen den beiden Teilen des κ in P. Lond. 1817 nicht so groß ist wie bei P.Ryl. III 47017. Zusammenfassend läßt sich sagen, daß eine Datierung in das 3. bzw. 4. Jh. als sehr unwahrscheinlich zu gelten hat. Aufgrund der angeführten Vergleichstexte dürfte P.Ryl. 111,470 zwischen dem 6. und dem 7. Jh. entstanden sein. Das 5. Jh. ist als Entstehungszeitpunkt unwahrscheinlich, kann jedoch nicht völlig ausgeschlossen werden. Die Verwendung von brauner Tinte, für die es keine Belege aus der Zeit vor dem 4. Jh. gibt, weist ebenfalls in diese Richtung18.
  9. [9]8.  Diese Form des α ist bei Inschriften verbreitet. Ein Beispiel für eine Vorlage für eine Inschrift, die ein α mit gebrochener Querhaste aufweist, ist P. Vindob. G. 26.013, vgl. Sijpesteijn, P. J., Wiener Melange, II. Christliches, a) Ein Trishagios-Hymnus, ZPE 40 (1980) 92-95. Allerdings ist auf den Inschriften die Form des μ nicht identisch mit der, die der Papyrus aus Manchester zeigt. Vgl. hierzu C. M. Kaufmann, Handbuch der altchristlichen Epigraphik, Freiburg i. Br. 1917; diese Form des α hat zum Beispiel eine Grabinschrift aus dem 5. Jh. (S. 75), eine Inschrift über der Grabkammertür der Äbtissin Thekla (S. 290f), die Schriftproben aus Antiochien aus dem 5. und 6. Jh. auf S. 413 sowie die griechischen Inschriften auf S. 414ff. Auch die koptische Schenute-Grabschrift, die auf S. 75f angeführt wird, hat ein α mit gebrochener Querhaste. Alle angeführten Inschriften weisen jedoch eine Form des μ auf, die einem großen lateinischen „M“ entspricht, während sich magische Papyri finden, in denen das α die beschriebene Form hat. Vgl. hierzu P. Mil. Vogl. 127 aus dem 2/3. Jh.: I. Gazzaniga, M. Vandoni, Papiri della Universitä degli Studi di Milano, Mailand 1965, S. 59f. Auch das spricht gegen die Verwendung als Vorlage für eine Inschrift und für eine Verwendung als Amulett.
  10. [10]Nach einer erneuten paläographischen Untersuchung des P.Ryl. III 470 ist die bisher in der theologischen Wissenschaft allgemein vertretene Frühdatierung dieses Fragmentes in das 3. Jh. nicht zu halten. Der historische Zeugniswert, der diesem Fragment häufig zugesprochen wurde, fällt mit der Datierung in das 6. oder sogar 7. Jh. Durch ein Wiener Fragment, das den selben Text überliefert, werden gleichzeitig die bisher erwogenen Ergänzungsmöglichkeiten fast vollständig auf die byzantinische Form der Antiphon eingeengt. Es ist zu fragen, ob diese Antiphon möglicherweise aus der westlichen in die östliche Liturgie Eingang gefunden haben könnte. Dafür spräche die Vielfalt der Formen im Westen47, während nur zwei griechische Formen dieser Antiphon bekannt sind, die jeweils nur marginale Abweichungen voneinander zeigen. Der Papyrus aus Manchester wurde als Amulett verwendet. Eine bisweilen angenommene Verwendung als Vorlage für eine Inschrift scheidet aus.
  11. [11]Hans Förster, «Die älteste marianische Antiphon – eine Fehldatierung? Überlegungen zum “ältesten Beleg” des Sub tuum praesidium», in: Journal of Coptic Studies 7 (2005), pp. 99-109.
  12. [12]Lobel verwendet nur einige wenige Buchstaben, um seine Datierung zu rechtfertigen. Allerdings scheint der Gesamteindruck des Textes seiner Datierung eher zu widersprechen. In besonderer Weise ist natürlich das “singuläre Alpha”, das ja angeblich vor allem bei Inschriften Verwendung findet, ein Problem bei der paläographischen Datierung. Insofern ist der Blick über die griechischen Texte hinaus geboten: Ein derartiges Alpha lässt sich für eine ganze Reihe koptischer Texte belegen, dort findet es sich vor allem in griechischen Passagen. Es sei hierfür nur auf Brit. Mus. Or. 6782. fol. 1 verwiesen. Auch die Beischriften der Darstellung der Heiligen Theodoros und Merkurios auf Val. Copto 66. fol. 210 und 287 bieten innerhalb der griechischen Passagen ein Alpha, wie es sich auch in dem griechischen Text findet, der zur Frühdatierung der Entstehung der Antiphon Sub tuum praesidium beigelragen hat27. Allerdings wird dieses Alpha weder in der Paläographie von Cramer28 noch in der Paläographie von Stegemann verzeichnet29. Dies bedeutet jedoch nicht, dass dieses besondere Alpha nur in den griechischen Passagen koptischer literarischer Texte zu finden wäre30. Auch die Beischrift der Darstellung der Maria auf Pierpont Morgan Lib. M. 597 bietet innerhalb eines koptischen Textzusammenhanges das “merkwürdige” Alpha31. Auch auf einem Mumientäfelchen aus koptischer Zeit findel sich die beschriebene Form des Alpha32.

    In diesem Zusammenhang fällt auf. dass auch eine ganze Reihe anderer „merkwürdiger“ Aspekte in der Handschrift des Papyrus aus Manchester durch eine Nahe zur koptischen Handschriftentradition erklärt werden kann. Es sei nur auf die punktartigen Verdickungen der Enden vieler Hasten verwiesen. Für griechische literarische Texte aus dem angesprochenen Zeitraum darf dies wohl mit Recht als exzentrisch bezeichnet werden, für koptische literarische Texte lässt sich jedoch eine Vielzahl von Belegen vor allem aus dem 8. und 9. Jahrhundert finden33. Auch das Ypsilon oder das My kann durch eine Nahe zu dieser Schrift-tradition wohl besser erklärt werden als durch eine Einordnung des Textes in das 2/3. bzw. 4. Jahrhundert. Besonders betont werden muss in diesem Zusammenhang, dass zum Beispiel das Alpha der von Roberts angeführten Vergleichstexte wohl nicht unterschiedlicher sein kann. Man kann also bei den für die Frühdatierung angeführten Vergleichstexten nur von sehr punktuellen Berührungen sprechen, dies scheint jedoch für eine Frühdatierung nicht ausreichend, wenn gleichzeitig die Berührungen mit der koptischen Schriftüberlieferung des 8. und 9. Jahrhunderts so auffällig sind, dass dies als Naheverhältnis bezeichnet werden muss.

    Wenn man also den Gesamteindruck des Textes nimmt, so scheint eine Datierung in spatere Zeit wohl nicht ungerechtfertigt, die Parallelen zu den koptischen Handschriften lassen eine Einordnung in das 8/9. Jahrhundert als wahrscheinlicher gelten als die Datierung der Handschrift in das 2/3. oder selbst das 4. Jahrhundert.

    Als Konsequenz dieser Überlegungen muss damit festgehalten werden, dass nicht nur Indizien, sondern sehr eindeutige Parallelen gegen die oftmals begeisterte Aufnahme des ältesten Zeugnisses der Antiphon Sub tuum praesidium angeführt werden können. Insofern darf ab sofort der Wiener Beleg dieser Antiphon34, der aus dem 6/7. Jahrhundert stammt, als der älteste Beleg gelten.

  13. [13]Vgl. Froschauer. „Tradition“. 96: „Das […] Alpha ist durch […] eine gezackte Mittelhaste mit Unterlänge gekennzeichnet. Diese erhalt bei Drehung des Buchstabens um 180° die Form eines Omega. Ob mit dieser Schreibweise allerdings bewußt an eine Ligatur von Alpha und Omega und damit vielleicht sogar im Zusammenhang mit dem Kreuzzeichen an die christliche Symbolik von Anfang und Ende gedacht wurde, ist eher zu bezweifeln und kann über den Status einer Spekulation nicht hinausführen.“ Da der Buchstabe auch an der von Froschauer zitierten Stelle Teil eines griechischen Wortes ist — insofern kann er nicht das Kreuzzeichen mit „Alpha und Omega“ repräsentieren —, da ferner das Kreuz sehr häufig am Anfang des Textes auf einem Mumientafelchen zu finden ist, da darüber hinaus das beschriebene, „merkwürdige” Alpha meistens nicht in der Nähe eines Kreuzzeichens zu linden ist und da Letztlich die Drehung eines ganzen Kodex oder auch einer Inschrift auf einem Grabstein, die überhaupt erst zum Erkennen der spekulativ für möglich gehaltenen Ligatur führt, eher unpraktisch ist, wird man dieses Alpha einzig für eine dekorative Variante dieses Buchstabens halten müssen, die für die koptische Zeit typisch und damit als Datierungshilfe eines undatierten Textes interessant ist. Auch auf dem Wiener Pergamentblatt steht das Alpha links neben der Kolumne und ist etwas vergrößert, der dekorative Charakter des Buchstabens ist also offensichtlich.

  14. [14]T. de Bruyn, “Appeals to the intercessions of Mary in Greek Liturgical and Paraliturgical texts from Egypt” in: L. Peltoma etc, Presbeia Theothokou: The Intercessory Role of Mary across Times and Places in Byzantium (4th-9th Century), Wien (2015), p.115-129; p.116 and n.18.  Online here (PDF), and at JSTOR here.
  15. [15]A. Effenburger, “Maria als Vermittlerin und Fürbitterin”, in: L. Peltoma etc, Presbeia Theothokou: The Intercessory Role of Mary across Times and Places in Byzantium (4th-9th Century), Wien (2015), 49-108; p.50 f.

From My Diary

I intend to write a post about the often repeated date of a papyrus, R.Ryl. III 407, which uses the word “Theotokos”.  As part of this, I’ve been collecting journal articles in order to find out what the actual arguments are, and what scholarship has been done.  It begins to look very much as if only three people have ever made a serious attempt at a paleographical dating – E. Lobel in the original publication in 1929, Otto Stegmüller in 1952, and above all Hans Förster in 1995 and 2005.

But retrieving the literature is not a simple task, even though most of the articles are decades old.  Indeed I have rather given up attempting to access two articles:

  • F. Mercennier, “L’Antienne mariale grecque la plus ancienne.,” Le Muséon, 52 (1939), pp. 229-233
  • G. Giamberardini, “Sub tuum praesidium” e il titolo “theotokos”, in: Marianum: ephemerides Mariologiae 31 (1969) 324-362.

Mercennier’s paper is today of no importance, I believe.  Hopefully Giamberardini does not contain anything essential.  A kind colleague supplied me with Stegmüller.  But I was obliged to write to Dr. Förster last year in order to obtain his second article.  The first article – I didn’t know it existed when I wrote to him – just cost me $20 for a PDF containing 6 page-scans, courtesy of a research library.  At least they scanned and sent it within 24 hours.  There’s a way to go before scholarship is available to us all.

But reading through what I have, I find that none of the other papers sometimes cited qualify as serious discussion.  This is because they just do not carry out the serious comparison of letter forms and dated/dateable manuscripts.  Indeed I have just read one article in which the author casually asserts a random dating without troubling to offer either reference or argumentation.  This won’t do.  How many dates for papyri rest on very slender scholarship?

One of the articles that I could obtain was not really useful, but the author cited the scholarship on this very papyrus as an example of how the paleography of Greek papyri is mostly rubbish and all the dates need to be completely rethought.  The author will not want for examples, I am sure.  There may be something in this, but as a non-paleographer I cannot know.  But unfortunately I do know that the author is one of a small group of US academics making this demand whose motives are less than scholarly.  Indeed I can think of two people involved – there are probably others – who fairly seethe with partisan hatred.  It’s easy to destroy, but hard to build.

It sometimes seems to me almost as if there are many people in the humanities, especially in the USA, who should not hold teaching posts.  They just do not seem to love scholarship for its own sake.  They do not seem concerned to advance scholarship, nor to increase knowledge, or even to uphold the reputation of the humanities.  Rather they behave like wreckers, or else researchers for some political-religious agenda.  The universities themselves seem to be in the hands of administrators rather than scholars, which leads to strange distortions of all kinds, many of which work to the disadvantage of researchers and indeed students.  The financial side seems horrific, of wealthy clerks and starving scholars.  But I see nothing of the desire for improvement anywhere; only to apportion the spoils of office differently.  The younger generation seem likely to be the victims.  It makes me very glad to be retired!  These are strange times, aren’t they.

But we are also blessed in ways unimaginable.  This morning I went into search of my old Thesaurus Linguae Graecae disks, and fired up Diogenes, and did a word search for “theoto/ko”.  It produced a huge number of results, mostly late, of course.  It produced the catena fragments of Origen on Luke that I was talking about a few days ago.  I wish that I’d thought of it then!

I don’t intend to spend an endless amount of time on the theotokos issue, but I’d like to verify the date of the papyrus. I’m not a paleographer, and I don’t intend to dabble.  But it should be possible to report what the scholarship actually is.


A photo of the Meta Sudans from the Palatine

Here’s yet another picture of the conical stub of the now vanished Roman fountain, the Meta Sudans, in front of the Colosseum. It was posted online here, where it is described as “late 19th century”.

I wish we had more 20th century photographs of the Meta Sudans.  There ought to be legions of them.



The Last Known Senatus Consultum of the Roman Senate?

In Massimiliano Vitiello, Amalasuintha: The Transformation of Queenship in the Post-Roman World, University of Pennsylvania Press, (2017), p.93, we learn of the problem of bribery in Papal elections during the Ostrogothic rule of Italy.  He quotes a letter in 533 AD to Pope John II, preserved in Cassiodorus Variae, Book 9, letter 15, from which we learn that the kings tried to stop the selling of bishoprics.  Dr V. continues:

In addition, in order to avoid a future repeat of the problem, King Athalaric ordered the urban prefect to issue a decree of the Senate, the text of which was to be engraved on a marble table and displayed in front of the atrium of the Church of Saint Peter, with copies posted in public places and assemblies for a thirty-day period. This is the last known Senatus Consultum, although the Senatorial sigla SC (= Senatus consulto) on a series of bronze coins minted in Rome in 536 in the name of King Theodahad indicates that this was not the last deliberation the Senate ever made.

The senate can only have been little more than a town council by this date.  King Athalaric was in fact a minor, only 18 when he died in 534 AD.  The government was in the hands of his mother Amalaswintha.  Athalaric was succeeded by his cousin Theodahad (534-536), whose weakness and ineptitude led to Justinian’s invasion of Italy.  Theodahad in turn was replaced by Vittigis, who himself was captured and shipped off to Constantinople in 540.  By this time Rome was deserted and in ruins.

A selection of letters from the Variae have been translated in the Liverpool TTH series, including letter 15, but not letter 16, which is our source for the inscription.  The letter is addressed to one of the very last Urban Prefects, Salventius.

Here’s the letter from the edition of Mommsen in 1894, online here, p.281 (p.474 of the PDF), who dates it to the end of 533 AD.  I’ve added a couple of full-stops and split up the paragraph.

Salventio Vir Inlustris Praefecto Urbis Athalaricus Rex.

Grata res est cunctis profutura vulgare, ut generale fiat gaudium, quod potuit esse votivum, alioquin laesionis causa noscitur, si beneficia potius occulantur. Dudum siquidem senatus amplissimus, ab splendore suo cupiens maculam foedissimae suspicionis abradere, provida deliberatione constituit, ut in beatissimi papae consecratione nullus se abominabili cupiditate pollueret. Poena etiam constituta, qui talia praesumere temptavisset.

Quod nos laudantes et augentes inventum, ad beatissimum papam direximus constituta, quae his antelata praefulgent, ut ab honestate sanctae ecclesiae profanus ambitus auferatur. Hoc vos** ad notitiam senatus et Romani populi volumus sine aliqua dilatione perducere, quatenus cunctorum figatur cordi, quod cupimus omnium studio custodiri.

Verum ut principale beneficium et praesentibus haereat saeculis et futuris, tam definita nostra quam senatus consulta tabulis marmoreis praecipimus decenter incidi et ante atrium beati Petri apostoli in testimonium publicum collocari; dignus enim locus est, qui et gloriosam mercedem nostram et senatus amplissimi laudabilia decreta contineat, in quam rem illum direximus, quo redeunte noscamus impleta quae iussimus. Incertum enim videtur habere quod praecipit, cui rerum effectus tardius innotescit.

(** “vos” is what the edition says, but surely it should be “nos”?)

Here’s a quick translation.  Corrections welcome:

King Athalaric, to Salventius, Vir Inlustris, Prefect of the City.

A welcome thing is profitable to publish to everyone, so that there is general rejoicing that it has been able to be fulfilled as promised.  Otherwise it is thought of as the cause of harm if the benefits are hidden away. For a while ago the great senate, desiring to remove from its splendour the stain of the most hideous suspicion, decided with a prudent resolution, that in the consecration of the most blessed pope no one should pollute himself with abominable greed. A penalty was also established for anyone who attempted to presume such things.  This is not a wrong, because then merit is truly sought in the one chosen, when money is not loved.

This, praising and extending the plan, we have brought to the attention of the most blessed pope the decrees, which were brought before him, so that the profane sphere might be removed from the honour of the holy church.  We wish to bring this to the notice of the senate and the people of Rome without any delay, so that what we desire to be guarded by the zeal of everyone is fixed in the hearts of everyone.

However, in order that the principal benefit may be effective both now and in the future, we order that both our decision and the decrees of the senate are engraved carefully on marble tablets and placed in front of the atrium of the church of blessed Peter the Apostle as a public testimony; for the place is worthy, which contains both our glorious gift and the laudable decrees of the great senate, on which matter we instructed it, so that when we return we may know that what we ordered has been fulfilled. For no-one can be certain whether what he orders has been done if the evidence of it only becomes known much later.**

(** Lit.: For he seems to consider as uncertain what he commands, to whom the effects of things become known more slowly.)

Dr Vitiello’s statement about posting copies for thirty days is not referenced, unfortunately.

It’s not certain whether any actual senatorial action was involved in producing the coins.[1]  But the bronze coins of Athalaric already feature the “SC” inscription, such as this one, and this is continued under Theodahad, such as this 40 nummi follis (found here; another example with description here as The Final Sestertius):

D N THEO-DAHATVS REX, moustached and mantled bust right, wearing ornate Spangenhelm; mantle decorated with pectoral cross / VICTORIA PRINCIPVM, S C across field, Victory standing right on prow of galley, holding palm across left shoulder and wreath in outstretched right hand.

A selection of other coins by Theodahad may be found here, all produced by a mint in Rome, and most with SC upon them.  His successor, Vittigis, minted in Ravenna and his coins do not use SC.

  1. [1]See H. T. Hindrum, Political Culture of Ostrogothic Italy in a Numismatic Perspective, thesis (2014), p.54.  Online here. (PDF)

Every one of us has two guardian angels, a good one encouraging, and a bad one tempting – patristic sources

Via Twitter I learn of this fascinating statement by Prof. Tom Ward:

I just learned through Peter Lombard (Sent II.XI) that Gregory of Nyssa and Origen thought that each human soul not only has a (good) guardian angel, but a (bad) tempting angel (=demon). I’d always thought this was just a cartoon thing.

I enquired about the source:

… the translator of Lombard (Silano), offers only this: “Gregory of Nyssa, De vita Moysi, following Origen, In Lucas, hom.12.”

Let’s look up the sources.  It’s good to have this here, since material is disappearing from twitter all the time at the moment.

Gary Hartenburg looked at the Gregory of Nyssa reference:

Gregory of Nyssa: there is a trustworthy teaching from the tradition that after the fall, God “appointed an angel with an incorporeal nature to help in the life of each person and…he also appointed the corrupter who, by an evil and maleficent demon, afflicts the life of man” (bk. 2, ¶45).

That’s from The Life of Moses, trans. Malherbe and Ferguson; p.165 n71 has a number of further references.

That footnote reads:

… The guardian angel is already in Hermas, Vis. 5.1-4. See Danielou, The Angels and their Mission according to the Fathers of the Church (Westminster, Md., 1957), pp. 68-82. The “two spirits” is in Hermas Mand. 5.1.1-4, Rabbinic speculation (G.F. Moore, Judaism, 1, pp. 479-493), the Dead Sea Scrolls (1QS 3-4); cf. the angels of God and of Satan in Barnabas 18. Philo, Quaes. Ex. 1.23, states, “Into every soul at its very birth there enter two powers, the salutary and the destructive.” See article on “Demon,” Dict. Spir. 3: 160-168.

However these references do not seem very relevant to our question of two angels, and neither of the Hermas references is relevant.

The Origen, Homilies on Luke, homily 12, (FOC translation by Thomas Scheck) reads:

4. It was indeed a great joy to these shepherds, to whom the care of men and provinces had been entrusted, that Christ had come into the world. The angel who administer ed the affairs of Egypt received a considerable advantage after the Lord came down from heaven, for the Egyptians could become Christians. It profited all who governed the various provinces, for example, the guardian of Macedonia, the guardian of Achaea. and the guardians of the other regions. It is not right to believe that wicked angels govern individual provinces, and good angels do not have the same provinces and regions entrusted to them. I think that what Scripture says about individual provinces should also be believed more generally about all people. Two angels attend each human being.  One is an angel of justice, the other an angel of iniquity.13 If good thoughts are present in our hearts and justice springs up in our souls, the angel of the Lord is undoubtedly speaking to us. But. if evil thoughts turn over in our hearts, the devil’s angel is speaking to us. Just as there are two angels for individuals. so. I believe, there are different angels in individual provinces, some good and some evil.

5. For example, very wicked angels kept watch over Ephesus, on account of the sinners who lived in that city. But, because there were many believers in that city, there was also a good angel for the church of the Ephesians. What we have said about Ephesus should be applied to all the provinces. Before the coming of the Lord and Savior, those angels could bring little benefit to those entrusted to them, and their efforts were not powerful enough to bring about success. What indicates that they were hardly able to help those under their charge? Listen to what we say.14 When the angel of the Egyptians was helping the Egyptians, hardly one proselyte came to believe in God. And this took place when an angel was administering the Egyptians.

In the same text, homily 35 contains this:

3.  I have to touch on some more hidden matters, that we might understand that the adversary is of one sort, while the three other persons—that is, the ruler, the judge, and the debt collector—are of another sort.  We read that the angel of justice and the angel of iniquity argued about Abraham’s salvation and his loss, as each of the camps wished to claim him for itself.  The condition is. of course, that someone should be willing to accept a writing of this kind.7

But. if it displeases anyone, he should go to the book entitled The Shepherd.  There he will find that two angels are present to every man: a wicked angel, who exhorts him to wrongdoing: and a good angel, who urges him to do everything good.8   Elsewhere, too. it is recorded that two angels attend a man. for good and for evil. The Savior, too, mentions the good angels when he says, “Their angels always see the face of my Father, who is in heaven.”9

You should also ask whether the angels of those who are little children in the Church “always see the Father’s face,” while others’ angels do not have the liberty to behold the Father’s face. For, we cannot hope that everyone’s angel always sees “the face of the Father, who is in heaven.”  If I am in the Church, no matter how very little I am, my angel enjoys the liberty and the trust always to see “the face of the Father, who is in heaven.”10

4. But. if I am an outsider, and not a member of that Church “that has neither spot nor wrinkle, nor anything of that sort.” and the facts prove that I am not a member of such a congregation, then my angel does not enjoy the trust of beholding “the face of the Father, who is in heaven.”  For this reason the angels care for good people. They know that, if they guide us well and lead us to salvation, they too will enjoy the trust of seeing the Father’s face. If salvation is secured for men by their care and diligence, they always behold the Father’s face. So too, if someone perishes through their negligence, they realize that the matter is a danger to them.

A good bishop, the best steward of a church knows that, if the sheep of the flock entrusted to him are kept guarded, it is because of his meritorious service and virtue. Realize that the same is true of the angels.  If someone who was entrusted to an angel sins, the angel is disgraced. And the opposite is also true. If someone entrusted to an angel, even the least person in the Church, makes progress, it redounds to the angel’s glory. For, they will see “the face of the Father, who is in heaven,” not sometimes, but “always.”  Other angels never see it. For, according to the merit of those whose angels they are, the angels will contemplate the face of God either always or never, little or much. God has clear and certain knowledge of this matter. So does someone found to be instructed by Christ, rare as that is.

7. See J. T. Milik, “4 Q Visions de ‘Amram et une citation d’Origene,” Revue Biblique 79 (1972) 77-97, who studied an apocryphal writing from Qumran on Amram, the father of Moses and Aaron.  Milik believes that the text was translated into Greek and read by Origen, who alludes to it here. Hence “Abraham” is an incorrect reading for “Amram.”

8. Cf. Homily 12.4-6 and Hermas, Shepherd 36.2-10.

Which is certainly an interesting point of view!  Some of it is clearly coming from apocryphal literature.  But let’s look at the references to the Shepherd of Hermas.  A number of English translations are at the Early Christian Sources Project, here.  I’ve chosen the Synder translation.

Here’s Mandate 6, chapter 2 (= Shepherd, 36) (online here):

The two angels in man (36:1-10)

Mandate 6, Chapter 2

36.1. “Now hear,” he said, “about faith. There are two angels with man, one of righteousness and one of evil.” 2. “How then, sir,” I said, “will I know their powers, since both angels live with me?”

3. “Listen,” he said, “and understand. The angel of righteousness is delicate and modest and meek and quiet. So whenever this one rises up in your heart, he speaks with you at once concerning righteousness, sincerity, reverence, contentment, and every righteous deed, and every glorious virtue. Whenever all these things come up in your heart, know that the angel of righteousness is with you. So these are the works of the angel of righteousness. So trust this one and his works.

4. Now observe the works of the evil angel. First of all, he is ill tempered and bitter and foolish, and his works are evil, destroying the servants of God. So whenever this one comes into your heart, know him by his works” [cf. Matt. 7:16].

5. “Sir,” I said, “I do not understand how I will recognize him.” “Listen,” he said. “Whenever ill temper or bitterness comes over you, know that he is in you; then comes the desire for many affairs and the extravagance of many kinds of foods and intoxicating drinks, much carousing and various and unnecessary indulgences and desires for women, and covetousness and great arrogance and pretention and whatever things resemble and are similar to them. So whenever these things come to your heart, know that the angel of evil is in you.

6. So when you recognize his works, shun him and do not trust him, for his works are evil and harmful to the servants of God. Here you have the powers of both angels; understand them and trust the angel of righteousness. 7. But shun the angel of evil, because his teaching is evil in every case. For if any man is faithful and the thought of that angel comes to his heart, that man or woman will surely commit a sin. 8. But again, even if a man or a woman is very wicked, and there comes to his heart the deeds of the angel of righteousness, of a necessity he will surely do something good.

9. So you see,” he said, “that it is good to follow the angel of righteousness, to bid farewell to the angel of evil. 10. This commandment shows things concerning faith, so that you might believe the works of the angel of righteousness, and by doing them live to God, but believe that the works of the angel of evil are bad, so by not doing them you will live to God.”

Whoever knew that this cartoon trope goes all the way back to Hermas; indeed probably earlier?


Theotokos: Pierius of Alexandria

Our next possible candidate for the earliest use of the term “Theotokos” is Pierius of Alexandria, who died “after 309 AD”.  Our source isn’t great for this, for it is some fragments, which may or may not be by Philip of Side (ca. 380-431+), whose vast history of the early church has otherwise perished.

Back in 2010 – gosh isn’t that a long time ago – Andrew Eastbourne kindly went through the mess of fragments of Philip of Side and translated the lot for us.  It’s online here.  Among them is this:

Fragment 4.7.  Pierius, a presbyter of Alexandria, flourished at this time, and in Pontus, Meletius the bishop—men who were amazing with respect to their learning.  And Pierius too, in his first discourse[111] of those On the Pascha, asserts strongly that Paul had a wife and dedicated her to God for the sake of[112] the Church, renouncing his association with her. And I also read quite a number of his other indispensable works and especially the one Concerning the Mother of God and the one On the beginning of Hosea.  And Theodorus, a certain court-pleader in Alexandria, writing in epic verse, says in his 13th book that Pierius and Isidorus his brother suffered martyrdom and have a very large shrine in Alexandria.  And in his discourse On the Life of the Holy Pamphilus, Pierius himself provided very much help in the divine Scripture.

111. Or “book” (Gk. λόγος); but from Jerome, De viris illustribus 76, this work appears to have been homilies on Hosea orally delivered at Easter; and Photius (Bibl. cod. 119) speaks of 12 λόγοι (of which he particularly mentions the one “on the Pascha and Hosea”) contained in one βιβλίον.

112. Gk. διά (Cf. Sophocles, Lexicon s.v. διά 3).

Since these are “fragments”, where does this come from?  Well, from manuscripts full of miscellaneous extracts of this and that.  In this case, these are found in MS Oxford Bodleian Barocci 142, on fol. 212r-216r; and in MS Oxford Bodleian misc. 61 (= Auct. E.4.18), on fol. 136r-143r (which material is, however not published anywhere).

Now Barocci or Barozzi was a Cretan Greek who collected manuscripts and sold his collection to an Englishman.

Dr Eastbourne noted:

In the translations below, the italicized material is directly from Eusebius, whether verbatim or paraphrased; the normal text represents the additions made by our author to Eusebius’ history.

The Greek text for this material, taken from that Barocci manuscript, was printed by C. de Boor, “Neue Fragmente des Papias, Hegesippus und Pierius in bisher unbekannten Excerpten aus der Kirchengeschichte des Philippus Sidetes,” TU 5.2 (1888), pp. 169-71.  Thankfully a list of the volumes of Texte und Untersuchungen is on German Wikipedia here.  Vol. 5, p.170-171 is:

There’s “theotokou” nice and clear at the start of p.171 line 2.

De Boor’s preface says that the series of historical extracts runs from the birth of Christ to the close of the Church History of Socrates, i.e. in 439 AD.  Scholars have tended to suppose that these are from Philip of Side, since this lost 5th century church history is the obvious source for early material.  Philip is the last writer to know Papias, for instance.  But the manuscript does not name the compiler, and we don’t know who he was or when he wrote.  The Bodleian website states that the manuscript itself is 13-14th century and the material in it was compiled by Nicephorus Callistus as source material for his own Ecclesiastical History.

Barocci 142 is online, so we can inspect the folios ourselves.  Here’s folio 216r:

Sadly I find the book-hand completely impenetrable.  I presume the red headings are for each extract?

So… what do we make of this?  Is this a valid witness?

I’m slightly inclined to feel that it is.  The extracts come from someone with genuine access to early material.  The fact that the extracts terminate in the 5th century suggests that the source work did also.  Whether it is indeed Philip of Side, or some other, now forgotten compiler of the period, is not of importance, really, compared to the date.

But the compilation cannot date prior to the toxic Nestorian disputes, in which the use of – or failure to use – the word “theotokos” suddenly became a matter to kill for.  A reference to a book title – we’re told that Pierius wrote On the Theotokos – is something less than knowing the content of the book.  Ancient book titles are fluid things, more about indicating content, in many cases.  This may not be the original title either.  There’s plenty of room for zealous tampering, if the title did not please the copyist or excerptor, or needed to be “improved”.

But in the end, the negatives are just speculation.  We do have an ancient source that Pierius wrote a book “On the Mother of God”.  Perhaps indeed he did.

What it said, of course, is another matter.