Codicotextual Analysis of the Gospel Manuscript (Ath.62) from Oshki on the example of Matthew’s Gospel

Old Georgian translations of the Gospel from the viewpoint of editing show rather mixed picture. The Gospel manuscript (Ath.62) from Oshki has attracted our attention by the fact that the trace of correction is observed on it. It was interesting to establish what redaction  the primary and corrected text belong to and what relation they have to other redactions of the Four Gospels known for today.

The Gospel manuscript from Oshki is kept in the collection of the manuscripts at Iveron Monastery on Athos. Its photocopy and microfilm with number Ath.62 are kept at the National Centre of Manuscripts. The manuscript dates to the 11th century. For the first time it was described by Robert Blake. The description offered by Blake almost a century ago is very important from the viewpoint that the scholar had seen this manuscript directly, the opportunity of which we were deprived of. We can judge only with the help of the photocopy.

According to Blake's description the manuscript contains the text of Giorgi Mtatsmindeli's edition of the Gospels; it is written on the 11th-century grayish white, thick and already yellow parchment. The manuscript is not dated. It is supposed that the place of its creation is Athos. For writing there was used brownish ink of bad quality. On the first four pages of the manuscript in the columns the canons of Eusebius and the letter  to Eusebius Carpianus are written. The arches are fulfilled with the colors of poor quality (brick-red, orange, blue, violet). The text is written with Nuskhuri script of average size inclined to the right, in two columns, the number of the lines is 19, paper size is 204x160, column size is 132x98, the space between the columns is 11 mm, the manuscript comprises 315 pages, covered with violet leather cover, the last page is torn off, the ornaments are absent. R.Blake supposes that the manuscript preserves the text corrected by Giorgi Mtatsnindeli according to the Greek original in order to create his own version of the Gospel. In his opinion, this could be the text of the Gospel edited by St.Euthymius [Blake, 1932:146].

According to the photo film lr of the manuscript is empty. As an introduction to the text of the Gospels is written in the arches Eustathius law (1v-5v), the letter to Eusebius Carpianus (6r) and short description of each chapter of Matthew's Gospel (6v-8v). Besides this, there is later inscription on 6v in Mkhedruli script about a bookbinder, Makary Kartveli the monk ("The Gospel was bind by Georgian monk Makary"). At the beginning of the text the words: "First chapter, the Gospel of Mathew. Book of Origin", is done in Asomtavruli script and then it continues in Nuskhuri script.

The Oshki MS of the Four Gospels contains the Gospel of all four Evangelists:

1.      9r-91v - The Gospel of Matthew. At the end of the chapter there is an inscription: "The Gospel of Matthew was written in Jerusalem on the eights year from the Resurrection according to Hebrew.  On 41 year from the birth of Christ ".

2.      92r-146v  - The Gospel of Mark. The postscript: "The Gospel of Mark was written on the twelfth year from the Resurrection. It is read during Lent."

3.      147r-238v  - The Gospel of Luke. The postscript: "The Gospel of Luke was written on the fifteenth year from the Resurrection.

4.      239r-296r - The Gospel of John. The postscript: "Saint John, the evangelist and theologian, have mercy on me, Diagon Ivane"].

The back page is half torn. Probably, the manuscript had a colophon but it was damaged and has not come down to us. Each Gospel has introduction, short description of the chapters.

It is hard to judge about the manuscript according to the photo film. In spite of this, as a result of closer examination of the manuscript it has been revealed that two people were working at the manuscript, a scribe and an editor. The scribe writes the main, initial text, his Nuskhuri letters are thin and small, supposedly, done in light brown ink, and the editor corrects the main text, erases or adds in it the readings of other edition. His letters supposedly written in black ink are comparatively large, angular and inclined to the left (we have marked the text of the lower layer of the Oshki MS with conventional sign o#, the secondary, corrected by an editor  - with the mark o*).

At the beginning of the paragraphs and chapters large, Mtavruli oil letters are used. Of punctuation marks we often come across colons at the end of the sentences and three dots - at the end of the paragraphs. For indication of the end of the chapters special separation sign is used.

The editor of the manuscript in the main text introduces the changes of the following type:


1.  He diligently scrapes and erases the word, in some cases even the whole sentence and writes new variant instead of it (Fig.1.).

2.  Reduces body text: erases the letters, words, phrases, sometimes the whole paragraph and leaves the place empty (Fig.2).

                                                                                 Fig.1                                                                             Fig.2


3.      In text part he adds the word, phrase or letter which he writes on the top of the line (Fig.3).




4.     If it is possible, he changes the words by shifting the letters or displacement. For example: in text part the word gesma was written and the editor changed it with the word gasmies[Matt.5:21]; there was a word daimarilos and the editor wrote the word sheimarilos instead of it [Matt.5:13].

The question has, naturally, been raised as to: which edition the initial text contains, according to what it was corrected by another scribe or editor. Why did it become necessary? Which edition did he take as an original and whether he took into consideration Giorgi Mtatsmindeli's texts recognized as Vulgate.

Unfortunately we do not have an academic edition of Georgian translation of the Books of Gospel, in which the reading of all manuscripts would accumulate. The preparation of such kind of edition has been started by Professor Zurab Sarjveladze. Under his leadership there was created a group composed of E. Giunashvili, M.Machkhaneli, D.Tvaltvadze and S.Sarjveladze. They prepare academic edition of Georgian translation of Matthew and Mark Gospels. It is through the use of the material prepared by this group that we tried to restore the primary text of the Oshki manuscript.

We compared the initial text of the Oshki Gospel (o#)  with  Urbnisi (F), Palestine (G), Jruchi (D), Parkhli (E), Alaverdi Gospels. Of them the first two were issued by I.Imnaishvili  [Georgian ...1979], and Jruchi and Parkhli Gospels were issued by A.Shanidze [Georgian ...1945], the changes introduced by an editor and different readings (o*) were compared to Vani (H), Ejmiadzin (I) and Gelati (K) Gospels. The latter was issued by I.Imnaishvili [Georgian... 1979].

The collation of the initial text (o#) of the Oshki Gospels MS with the text of Alaverdi (a) Gospel, Opizuri (DE) and the manuscripts containing the so-called Euthymius editions (FG) revealed an interesting picture: it completely coincides with the readings of Alaverdi (a) Gospels; o# is not identical to Opise (DE) edition. Also it does not coincide with the readings of Urbnisi and Palestine Gospels recognized as Euthymius Athonite (FG) redaction in spite of the fact that it often follows them. Each of that reading of the initial text of the Oshki Gospel similar to which there has not been found either in Jruch-Parkhli (DE) or Urbnis-Palestine (FG) Gospels coincided with the readings of Alaverdi (a) Gospels.

Below we give examples to evidence that the readings of o# and (a) oppose to DEFGHIK readings. E.g.,  in Alaverdi (a) and Oshki o# Gospels we have such reading: vitar-igi acro ars bche da sachirvel gzai, romeli miiyvanebs cxorebasa, da mciredni arian, mavalni mass shina [Mth.7:14].  And in o*DEFGHIK instead of the phrase mavalni mass shina we have - "romelni hpoeben mas". We give two of many similar examples:


o#a: vitarca ikhilai esu eri mravali garemo misa, ubrdzana carsvlai nier kerdzo; [Matt.8:18].

o*DEFGHIK: caisvlai

o#a: da ukuetu eshmaki eshmaksa ganaskhams, tavsa tvissa ganevlta; ver mtkicears meupebai misi? [Matt.12-26].

o*HIK: vitar damtkicnes.

DE: vitar mtkice ars.

o#a: xolo ese natesavi verarait shesadzlebel ars gansvlad, garna locvita da markhvita [Matt.17-21].

o*HIK: ararait ganvals.

DE - here this article is absent at all.

o#a: xolo raita ara davabrkolnet igini, carved zghuad da shtaagde samcheduri, da, romeli pirvelad aghmohdes tevzi, moighe da agughe piri misi da hpoo mis shoris satiri, moighe igi da miec mat shentvis da chemtvis [Matt. 17-27].

o*HIK: statiri.

E: sascori.

D: sasmeli.


o#a: rametu movida zde kacisai modziebat carcymedulisa [Matt. 18:11].

o*HIK: cxorebad.

DE - this article is absent.

o#a:  hrqua mas: romelni mcnebani? Xolo iesu hrqua mas, raita ara kac-hkla, ara imrusho, ara cili scamo[Matt.19:18].

o*DEFGHIK: vitarmed.

o#a: xolo tquan zeda ara egre iqos, aramed, romelsa undes tquan shoris didqopai, iyos tqvenda msaxur [Matt.20:26].

o*FGHIK: iyavn.

DE- this article is absent.

o#a: vin arta amatganman yo nebai mamisa tvisisai? Hrques mas: pirvelman man. Hrqua mat iesu: amen getqv tquen, rametu mezuereni da medzavni cinagidzgodian tquen sasupevelsa catasa [Matt.21:31].

o*DEFGHIK: ghmrtisasa.

The above examples prove that there are many different readings between the initial text (o#) of the Oshkian manuscript of the Gospel and other editions (DEFGHIK). As we have already mentioned the editor in some cases erases the letters, words, phrase, sometimes even the whole paragraph, and leaves the place empty. In spite of the fact that he diligently scratches out the text, in the majority of cases we still managed to restore the erased text and, as it turned out, the reading of the initial text of the Oshki Gospels (o#) are again  identical to Alaverdi (a) Gospel. We give examples when we had different reading in the initial text similar to which is not attested neither in Opiz nor in the so-called Euthymius. The editor of the Oshki Gospel has taken them off from the main text in order to keep it on one level with Giorgi Mtatsmindeli. For instance, in o# we had: da iyvne tkuen modzulebul yoveltagan carmartta saxelisa chemisatvis, xolo romelman daitminos igi sruliad, tsxovndes [Matt. 10:22]. The editor erases the word carmartta. As it turned out none of other editions (o*DEFGHIK) have it and it is only found in Alaverdi (a) edition. Similar cases are frequently attested. For example:

o#a:  rametu romelsa akundes, mieces da miematos, da romelsa ara akundes, da romelsa ara akundes, da romelgha-igi hkonies, vitarmed akuis, mo-ve-egos mas [Matt. 13:22].

o*DEFGHIK: - hkopnies vitarmed.

o#a:  da hrqua monata tvista: ese ars iovane matlismcemeli, romeli me movkal, igi aghdgomil ars mkudretit, da mistvis ikmnebian dzalni ese mis tana [Matt. 14:2].

o*DEFGHIK: - romeli me movkal.

o#a: da sheipqresigi da ganiqvanes gareshe venaxisa mis da mokles da ganagdes igi [Matt. 21:39].

o*HIK: -da ganagdes igi.

o#a: amistvis getyvi tquen: migeghos tquengan sasupeveli ghmrtisai da mieces natesavssa ucxosa, romelni yofden nayopsa missa [Matt. 21:43].

o*DEFGHIK: ucxosa.

As is seen the editor, on the one hand, takes the words from the text, on the other, adds them, sometimes even the whole phrase in order to level the text with the edition of Giorgi Mtatsmindeli. In this case too the readings of the initial text (o#) of the Oshki Gospel and Alaverdi (a) coincide with each other. For example, one paragraph of the Oshki (o#) and Alaverdi (a) Gospel is read in this way: vitar igi shevida saxlsa ghmrtisasa da purni igi shesaciravtani shechamna, romeltai ara jer-iyo chamad, garna mghdelta xolo [Matt.12:4]. The editor added the word chamadis later: misa arca mistanata mat (o*FGDEHIK) and we get such variant of this paragraph: vitar igi shevida saxlsa ghmrtisasa da purni igi shesaciravtani shechamna, romeltai ara jer-iyo  chamad missa arca mistanata mat, garna mghdelta xolo Matt.12:4].  We have misa arca mistanata mat in Urbnisi (F), Palestine (G), Jruchi (D), Parkhli (E), Vani (H), Ejmiadzin (I) and Gelati (K) Gospels too. Similar examples:

o#a: mravalta mrquan me mas dghesa shina: upalo, ara saxelita shenita vtsinatsarmetqvelebita da saxelita shenita eshmakni ganvassxenit da saxelita shenita dzalni mravalni vqmnenit? [Matt.7:22].

o*FDHIKDE: upalo] + upalo.

o#a: xolo tquenni tavisa tmanica ganracxil arian [Matt.10:30].

o*FGHIKDE: tmanica] + yovelni.

o#a: xolo iovanes rai esmnes sapyrobilesa shina sakmeni kristesni, miavlinna mocapeni [Matt.11:2].

o*FGHIK: mocapen] + tvisni

DE: misni.

o#aDE: xolo me getyvi tkuen: koveli sityvai uqmi, romelsa ityodian kacni, micen sityuai mistvis dghessa mas sajelisasa  [Matt.12:36].

o*FGHIK: tkuen] + rametu.

o#aDE:da ubrdzana ersa mas dasxdomai  tivasa zeda. Da moigho xuti igi puri da ori tevzi, aghixxilna zecad da akurtxa da misca puri igi mocapeta tvista, da mocapeta mat misces ersa mas [Matt.14:19].

o*FGHIK: akurtxa]+ da gantexa.

o#a: miugo simon-petre da htqua mas: shen xar kriste dze ghrtisa [Matt.16:13].

o*FGDEHIK: ghrmtisa + tsxovelisai.

This and other similar examples evidence that the initial text (o#) of the Oshki Gospel and the text of the Alaverdi (a) Gospels follow each other. The ground for this supposition is once more expressed by the fact that each reading of the initial text of the Oshki Gospel similar to which has not been found neither in Jruch-Parkhli (DE) nor Urbnis-Palestine (FG) Gospels, coincided with the readings of Alaverdi (a) Gospels.

In the majority of other texts the initial text (o#) of the Oshki Gospel manuscript follows the readings of the Opiz and so-called Euthymeus edition and the secondary follows that of Giorgi. Here are the relevant examples:





Sada ars, axalshobili meupe huriata? [Matt.2:2].  



romeli igi ishva

Rametu shengan gamivides mtavari [Matt.2:6].  




Da aghaghes sapaseta matta [Matt.2:11].  




Acadeac, rametu esred jer-ars aghsrulebad yoveli  simartle [Matt.3:15].  



shuienis chuenda

Carved chemgan   martlukun, eshmako [Matt.3:15].  




Da romelni sxdes bnelsa da achrdilsa sikudilisata [Matt.4:16].  




Da romelman hrquas dzmaa tvissa raka, romel ars sadzagel [Matt.5:22].  

romel ars sadzagel

romel ars sadzagel


Ver gamoxvide mier, vidre ara misce dangisa kotori [Matt.:26].

dangisa kotori

dangisa kotori

Ukuanaiskneli kodrati

Hrqua saxlisa upalman exoismodzghuarsa tvissa [Matt.20:8].

saxlisa upalman

saxlisa upalman

Upalman savenapisaman

Dzal-gica shesunad asumeli, romeli chemda shesumad ars [Matt.20:8].

chemda shesumad ars

chemda shesumad ars

Mme megulebis shesmad

Da xvalisagan mo-rai-vidoda    [Matt.21:18].      




Xolo mezuereta da codvilta hrcmena misi  [Matt. 21:32].         




Xolo dghisa mistvis da jamisa aravin icis, arca angelozta cisata, arca dzeman [Matt.24:36].

arca dzeman

arca dzeman


Romel-igi ganmzadebul ars eshmakisaTvis da msaxurta mista [Matt.25:41].




Mashin movida iesu mat tana dabasa [Matt.26:36].




It is known that Giorgi Mtatsmindeli (1009-1065) with the purpose of getting Georgian translation closer to the Greek original three times returned to the text of Gospel. This fact has already been attested with arguments and therefore we do not stop on it here.

According to Darejan Tvaltvadze's research, in the Alaverdi Gospels rewritten in 1054 we maintain the intermediate stage of Giorgi Athonite's work  at the final edition of Georgian translation when the text of Georgian translations had been compared with the Greek one by him only twice and the process of editing has not yet been finished. Hence, neither Vulgate text has final form because the third and final comparison as a result of which we have got the text recognized by Georgian church as vulgate has not yet been realized [Tvaltvadze, 2008: 12-20]. The same information is given by the scribe of the Alaverdi Gospels: "I have compared with Greek and Georgian Gospel twice. Neither numbers nor laws lack. I did not write the bibliography because the contents fulfill the same function. There's no need to double work. Besides this, the margins of the book would be overloaded. Please, excuse me" [Georgian...1986:212].

As was mentioned above the readings of the initial texts of the Oshki MS coincide with the readings of the Alaverdi Gospels and both contradict to Giorgi's edition. Hence comes logical conclusion that the body text (o#) of the Gospel of Oshki manuscript as well as the Alaverdi Gospels, represents the version obtained through double comparison with the Greek original. It should be mentioned that the Alaverdi Gospel (A-484) is not supplied with Giorgi Mtatsmindeli's known colophon. Neither Oshki MS(Ath.62) of the Gospel is supplied with it. This colophon was supplied by Giorgi Athonite to the final text after the comparison for the third time.

The corrections existed in the Oshki MS of the Gospel made us think whether it is done directly by Giorgi Mtatsmindeli. Moreover that it was rewritten on the Athos and dates by 10th century. However, close scrutinize on the manuscript evidence that this must not have been this way. The Oshki manuscript of the Gospel doe not represent the so-called working version of the Athonite father. The scribe or editor knows well what exactly is to be corrected, what passage is to be replaced. He performs this job very carefully and accurately without intrusion into text. If this manuscript had been issued directly from Giorgi Mtatsmindeli we would have quite different situation. We would have been able to trace the process editing by St.Father.

As a result of textual study of the Oshki manuscript (Ath.62) of the Gospel we can conclude:

The initial text (o#) of the Gospel MS from Oshki and the text of the Alaverdi (a) Gospels follow each other. To our mind, they represent the version obtained by means of two-fold comparison with the Greek original of the Gospel by Giorgi Mtatsmindeli. The process of edition has not been finished yet. The comparison of the initial text (o#) of the Oshki Gospel MS with the Urbnisi (F), Palestine (G), Jruchi (D), Parkhli (E) Gospels evidences that it often follows it but is not identical to any of them. Each of these readings of the initial text (o#) of the Oshki Gospel MS which analog has not been found neither in Jruch-Parkhli (DE) nor Urbnis-Palestine (FG) Gospels, coincided with Alaverdi (a) readings of the Gospel.

The comparison of the second text of the Oshki Gospel MS with the texts of the Vani (H), Ejmiadzin (I) and Gelati (K) Gospels has proved that editor's corrections occurred already according to the final third comparison with the Greek original. Hence, the second text is identical to Giorgi's edition.


Tvaltvadze D.
The text of which edition does the Alaverdi Gospel contain? “Religion”, #1, 12-20.
Description …
Description of Georgian manucripts collection of the former church museum (A). Vol. II1 under E.Metreveli’s edition. Tbilisi.
Two Final…
Two final editions of Georgian Gospels. Text edited and supplied with research Ivane Imnaishvili. Tbilisi.
Two Old Editions…
Two Old Editions of Georgian Gospels according to three Shatberd Manuscrips (the years 897, 936 and 973). Issued by A.Shanidze. Tbilisi.
Robert P. Blake
Catalogue des Manuscrits Georgiens de la Bibliothèque de la laure D’ Iviron au mont antos, I, Paris, Librairie auguste picard, 82, Rue Bonaparte.