Augustine, City of God

Augustine, City of God*

'Explicui tuos libros; neque enim tam languidi aut inertes erant, ut me aliud quam se curare paterentur: iniecerunt manum, ereptumque aliis solicitudinum causis suis vinculis illigarunt ..., ut ego anceps sim quid in illis magis mirer, sacerdotii perfectionem, philosophiae dogmata, historiae plenam notitiam, an facundiae iucunditatem. ... Et usus es validissimo exemplo recentis calamitatis, quo licet firmissime causam muniveris, tamen si utrumvis licuisset, id tibi nolueram suffragari. Sed quando orta inde fuerat convincendorum stultitiae querela, necesse fuit inde argumenta veritatis accersere.' (ep. 154.2, Macedonius ad A. [413/4])

These are the words of the first reviewer of the first installment of ciu.[[1]] Vicar of Africa in the delicate year 413-414, after the uprising of Heraclian and the purge that followed, Macedonius was the ideal reader for the 'magnum opus et arduum,' poised between empire and church: he first wrote to A. (ep. 152) asking the bishop's views on the delicate matter of episcopal intervention in the affairs of state. A. wrote back at length (ep. 153), passing along a copy of the first three books of ciu.

It is good to heed the first reviewer's praises (for the happy combination of erudition and argument, doctrine and eloquence), and his regret (that the work dilates on the late catastrophe). His views give a useful perspective from which to consider the events of 410, and their echoes in Africa. That will lead naturally to a consideration of the contents and sources of the work as we have it; but first it will be useful to review the external facts as we have them for the composition and transmission of the work.

I. Dates of composition

The composition of ciu. occupied A. for at least a decade, perhaps fifteen years; cf. retr. 2.43.1. There are numerous indications of his progress on the work, from which a tentative schedule may be constructed. The following outline summarizes the available indications. Evidence for the dates is briefly indicated, but for fuller discussion one should consult the note by G. Bardy in the BA edition of the retr. (12.587-588).

Books 1-3: No later than 1 September 413; probably no earlier than late 412; book 1 may have circulated independently before the other two.[[2]]

Books 4-5: In circulation by spring 415 (ep. 169).

Books 6-10: Available by 417 (Orosius, hist., praef.; ep. 184a.).[[3]]

Books 11-13: By 417/18 (ep. 184a; cf. ciu. 12.10 and trin. 13.9.12).

From this point, opinions diverge. Some think that all 22 books were finished by c. 422; others insist on postponing 18 to at least 425 (citing ciu. 18.54.72: but others take the arithmetic suggested there literally and insist on postponing that book until 429); but it seems clear that the work was complete and out of A.'s hands by the time he wrote retr. 2.43.1 (426/7): 'Hoc ... grande opus tandem viginti duobus libri est terminatum.'

II. Manuscripts and editions

The first important document of the manuscript tradition of ciu. is the letter to Firmus (now ep. 1A in CSEL 88). The purport of this letter for the manuscript tradition may be summarized thus: (a) the 22 books were first prepared in separate quaterniones; (b) together the quaterniones would be too bulky as a single codex; (c) hence A. recommends that two codices be made, divided to incorporate the first ten books in one manuscript, the last twelve in another. If further division is required, five codices are recommended, to contain books 1-5, 6-10, 11-14, 15-18, and 19-22 respectively. It is also clear that the individual books had come into circulation separately.

The last line of the letter to Firmus may indicate further important information: 'Quantum autem collegerit viginti duorum librorum conscriptio missus breuiculus indicabit.' Following H.I. Marrou,[[4]] it has become the practice to treat the list of chapter headings that occurs in many manuscripts (and which Eugippius seems to have known in the sixth century when he made his excerpts from this work) as though it were the breviculus here mentioned and to print the list at the head of the work. Though this is an undoubted advance over the earlier practice of inserting the individual headings in the text of the work itself, the headings may be post-Augustinian, and may even have been written by Eugippius himself. In that case, whatever the 'breviculus' may have been, we do not have it.[[5]]

Pending completion of the HUWA catalogue, we know of 394 medieval manuscripts of all or part (or excerpts) of ciu., more even than of the Confessions. The most recent analysis shows six manuscripts (and seven fragments) earlier than the ninth century; thirty-one (plus twelve fragments and three collections of excerpts) survive from the ninth century. More than half the ninth century manuscripts have not been used in any critical edition. The most venerable manuscripts (on which our editions rely most heavily) are: Lyon, Bibliotheque Municipale, 607 (L: North Italy, 6th century, containing books 1-5); Veronensis XXVIII(26) (V: early fifth century [!], North Africa [!], containing books 11-16), and Paris Bibliotheque Nationale lat. 12214 (C: Italy, sixth century, containing books 1-9: to be supplemented with book 10 from the same codex, now held at Leningrad, Publichnaja Biblioteka Q.1.4).[[6]] It will be seen that these MSS offer impressively ancient testimony for the first sixteen books. Nevertheless, it would be dangerous to assert that considerable improvement in the text, or at least confidence in its foundations, could not be derived from a fuller examination of the manuscript tradition than has yet been undertaken. Whether it is useful to think of two different recensions of the work owing to A.'s own hand (see CCL 47.VII), which arises from the differences between the readings of C and L and is further fueled from mention in the letter to Firmus of the copy there transmitted as having been 'relectos,' could well be resolved if we knew the MSS tradition better.[[7]]

The earliest printed edition of ciu. appeared at Subiaco in 1467; there are other incunabula, of which the most important is that of Johannes Amberbach (Basel 1489). In the following century, the edition with ample commentary by the Spanish humanist Ludovicus Vives takes pride of place, but mention must also be made of the edition of A.'s works prepared by the theologi Lovanienses (Antwerp 1576).

As with other works of A., the contribution of the Maurists