The Patria — the historical works describing the monuments of Constantinople — are ascribed to George Codinus in some of the manuscripts. Averil Cameron states[1] that:
Preger demonstrated in 1895 (op. cit. n.8) that these works belong to the tenth century and are not (as previously supposed) by George Codinus.
The reference given in Cameron tends to confuse the reader as the only works of Preger referenced there are the 1901-7 Scriptores Originum Constantinopolitarum and the 1898 Anonymi Byzantini edition of the Parastaseis. Fortunately a bit of digging reveals the answer. In fact[2] this is a reference to an 1895 study of the textual tradition entitled Beiträge zur Textgeschichte der πάτρια Κωνσταντινουπόλεως, (München 1895). This may be found online at the Universitätsbibliothek Eichstätt-Ingolstadt and indeed elsewhere.
Preger deals with the matter incisively in the very first paragraph of his work (p.4):
It was the opinion of Lambeck, in his De Codini vita et scriptis, repeated also by Krumbacher in his history of literature, a scholar named George Kodinos in the 15th century published three works: 1. Peri twn o)ffikiwn tou= palation Kpolews [De officiis] 2. the Patria Kpo/lews and 3. a low-quality chronicle. Yet Lambeck himself, in his edition of the Patria made use of a manuscript which can be dated at the latest to the 14th century (Paris gr. 854; labelled H by me). In fact the unknown-to-him manuscript Munich 218 is 11th or 12th century. Either Kodinos lived earlier or he is not the author of the Patria.
He continues:
The reason why Lambeck placed Kodinos in the 15th century was because the chronicle written by him finishes with the fall of Constantinople to the Turks. This wretched concoction follows the Patria in many manuscripts, but is anonymous. The fact that the works are transmitted together tells us nothing about the authorship of either. Probably the chronicle has nothing to do with Kodinos. Nor can we get clues about his life from the Officia aulae Cpolitanae. For even in this work, the name of the author is based only on conjecture.
The Officia was first edited in 1588 by Franciscus Junius from a manuscript of Julius Pacius. The title read, “Tou= sofwta/tou kouropala/tou peri\ tw=n o)ffikiali/wn” etc. So it stands also in ms. Munich 247; in Munich 156 the first two words “tou= sofwta/tou” are missing.
Both manuscripts were written by Andreas Damarios. The vague official title “curopalates” only occurs in manuscripts from Damarios — Lambeck claims that he didn’t find it in any manuscript — so the likelihood that is authentic is very weak.
When Junius published the De officia for the second time in 1596, he printed the title as “Georgi/ou tou= Kwdinou= peri\ tw=n o)ffikiali/wn” etc. because, as he indicates on p.7, in one manuscript the title read “Parekbolai e)k th=j bi/blioi tou= xronikou= peri\ tw=n patrw=n th=j Kwnstantinoupo.lewj . . . . . . sunteqei=sai para\ Fewgri/ou tou= Kwdinou=”.
There is a footnote on “in one manuscript”:
On the title he states: “… recens lacunas non exiguas ope Mss. Palatinae bibliothecae, Augustanae et Seileranae supplevit.” There is no manuscript of the Patria among the Augustanae manuscripts. There are two mss (70 and 301) among the Palatine mss., which include the Patria with the above title and the Officia. Where the Seiler manuscript might be found I do not know.
He continues:
So Junius committed a very serious error: he mixed up the titles of two sequential works, the Patria and the Officia, in the same manuscript. Although Gretser, in his edition of the Officia notes this mistake, he still attributes the work to Georgios Kodinos, probably on no other ground than that he believed that the works in the manuscript were written by the same man. He cites no manuscript evidence, and I have not found the name of Kodinos attached to the Officia in any manuscript.
At this point Preger adds a footnote that he has found a manuscript in the Vatican, Barberini gr. 164, where under p.13, is a note associating the two. But this manuscript too is a Damarios production: and he states that no confidence can be placed in any title given by Damarios. To this effect he quotes Muratori (in Latin): “In one word: so dishonest was Andreas Damarios the Epirote, that we should believe nothing from him nor in his book titles.”
The Officia are to us just as anonymous as the chronicle; they seem to have been written during or after the reign of John Cantacuzene (1341-55).
The name of Georgios Kodinos is only found before the Patria here also only in one family of manuscripts which shows other deviations from the original. It seems likely that Kodinos is merely the name of the scholar to whom we owe the recension of the text of the Patria found in the family of manuscripts that I have labelled B. I will say more about this later. At the moment I will say only that the entry on the fall of the porphyry column in 1106 AD (Bekker p. 15,13 ff) is an interpolation from B. This is also the latest date mentioned in the work. All the manuscripts of B are from the 16th century or at most from the end of the 15th century. It is therefore possible that Kodinos lived in the 15th century, although we cannot prove it. Our only certainty is that his lifetime must be placed between 1106 and ca. 1450.
No author is given for the Patria in any manuscript other than the B mss. In the remainder, there is no date later than the reign of Basil II. Bulgaroktonos (976 -1025) (p. 128,1 ff. Bekker).
Unfortunately at this point the author drops into Greek, saying that the text is named on p.114,11. From this reference he infers that the passage must be written between 989-1006 AD. He continues:
In addition on p.145,6 it is stated that since the foundation of Hagia Sophia is 458 years. This church was founded in 537 AD (see Clinton, Fasti Romani II 143 n.) which places this notice in the year 995 AD. It is therefore certain that the author of the Patria Kpolews did not live in the 15th century but at the turn of the first millennium.
The argument as to the date is not quite conclusive; but the mysterious George Codinus is clearly no longer the author.