The OT Canon: Augustine and Cassiodorus

REVUE BÉNÉDICTINE48

THE STRUCTURE OF THE COLLECTIONSTHAT MAKE UP THE SCRIPTURES:

THE INFLUENCE OF AUGUSTINE ON CASSIODORUS

The sequence in which biblical books were read, and the internal connections that were understood to exist between them, was, until recently, a matter of importance in Christian exegesis. Moreover, in earlier times it was also a practical matter: sequences, expressed as lists, were a significant factor in the formation of the variety of western views on the canon,1 and more, mundanely, they acted as guides to how the various books were bound together into codices in the period before ‘the pandect’2 – a single codex containing all the canonical books – become possible or common. In studies of this question of the interrelationship between lists / canons / gatherings two figures always have prominence: Augustine (354-430), whose most influential list is to be found in the De doctrina christiana 2,8,13,3 and Cassiodorus (485/90-c.580) whose list can be found in the Institutiones 1,1-9. However, the relationship of Cassiodorus to Augustine on this question is not at all clear. Cassiodorus explicitly presents us with a list which he claims is that of Augustine (Institu-tiones 1,13,1), but this list is not that of the De doctrina christiana4; while in very next sentence Cassiodorus declares that … Augustinus secundum praefatos novem codices … secundo libro de Doctrina Chris-tiana Scripturas divinas LXXI librorum calculo comprehendit (1,13,2). While it is unclear whence Cassiodorus obtained the list he attrib-utes to Augustine (1,13,1); this paper will demonstrate that his latter

1. See T. O’Loughlin, ‘Inventing the Apocrypha: The Role of Early Latin Canon Lists,’ Irish Theological Quarterly 74 (2009), pp. 53-74.

2. An object that we hear of in Cassiodorus, Institutiones 1,5,2; I am using the edition of R.A.B. Mynors, Cassiodori Senatoris Institutiones (Oxford 1937).

3. I am using the edition of J. Martin, CCSL (1962) 32, pp. 39-40.4. See Table 2 below.

DOI : 10.1484/J.RB.1.2462

T. O’LOUGHLIN 49

statement (1,13,2) is factually correct: that Cassiodorus’s proposed division of the biblical books into nine codices follows De doctrina christiana 2, and that his rationale for the division of the Scriptures is based upon Augustine’s understanding in De doctrina christiana.5

Apart from being witnesses to the common Latin Old Testament canon, Augustine’s list in the De doctrina and Cassiodorus’s direc-tions for the nine codices of the ideal library, and almost certainly that of his own codex grandior,6 do not seem to have much in com-mon. Augustine’s concern is that of the canon as such and his writ-ing is staccato, brief, and little more than a list. Hovering in the background of this list is the dark shadow of Jerome, whose views on the large Septuagint canon were already published, and his attempt to exclude certain books or, at least, to reduce them to lesser status. The list in the De doctrina can be seen as a reply to such attempts to promote the notion that the canon is coterminous with the Hebrew Bible by not only rejecting a linguistic in favour of an ecclesial criterion,7 but by setting out by name all the various books and thus making plain that while the list could be broken up into categories such as ‘history’ or ‘prophecy’ they all stood on a level as books of ‘the scriptures.’ Cassiodorus, on the other hand, would seem to be acting as the paradigmatic ‘schoolmaster’ when he sets out his list.8 He seems little troubled by the question of the canon, but he is con-

5. Commenting upon 1,13,2, Richard Marsden (‘The Codex Amiatinus and a sister pandect’ in The Text of the Old Testament in Anglo-Saxon England, Cambridge 1995, p. 133) stated that ‘this seems to confirm that the nine volumes [of Cassiodorus’s codex grandior] followed the Augustinian order of books. It cannot refer to Cassiodorus’s general discussion of the books of scripture in I.i-ix, for there he neither follows nor recommends the Augustinian order.’ This paper will show that the exact opposite is the case: (1) Cassiodorus does not follow the order that he labels ‘that of Augustine’ in Institutiones 1,13,1 (see table 2); and (2) that in 1,1-9 Cassiodorus does follow an order derived from Augustine (De doctrina christiana 2), and, thereby, recommends it (see table 1).

6. If we assume that the picture of Ezra – Cassiodorus in front of an armarium containing nine codices now found in the Codex Amiatinus, fol. Vr, is derived from an earlier image in the codex grandior (see Marsden, ‘The Codex Amiatinus and a sister codex,’ where the case for such an assumption is well made), then that nine volume bible probably reflects the actual bindings for the biblical books in Vivarium. We shall return to this point in the conclusion of this paper (and see table 3). On the drawing, see R.L.S. Bruce-Mitford, The Art of the Codex Amiatinus (Jarrow 1967).

7. De doctrina christiana 8,12 and 9,14 (these form the ‘brackets’ around the list); this relationship of Augustine’s list to Jerome’s various comments on the canon is pursued in ‘Inventing the Apocrypha,’ pp. 60-2.

8. On the notion of Cassiodorus as a ‘schoolmaster,’ see J.J. O’Donnell, Cassiodorus (Berkeley 1979); and T. O’Loughlin, Teachers and Code-Breakers: The Latin Genesis Tradition, 430-800 (Turnhout 1998), pp. 46-9.


cerned with the library of the Christian, indeed the monastic,9 school: that it should have a curriculum of biblical learning, should possess a complete set of biblical commentaries, that the books should be properly arranged for the readers, and there is even the touching note that he has added some blank quires (… puros etiam quater-niones …) to these special teaching codices so that if more material comes to hand, it can be inserted later.10 Now while it is clear that Cassiodorus is familiar with the exegetic writings of Augustine, what has not been pointed out is that when he declares that he is arrang-ing his commentary materials in nine codices, each relating to one of the nine sections into which the biblical books are divided, that Cassiodorus was taking over this structure from Augustine’s De doc-trina christiana.

While Augustine is primarily concerned with providing a list of those books that are to be included within the range of Scripture (totus autem canon scripturarum … his libris continetur), he does so in a particular order, and the manner in which he narrates that order-ing reveals that he thought of the list, and indeed may have recalled it from memory, as series of groups of books. The first items on the list are, as we would expect, the books of the law:

Quinque Moyseos, id est Genesi, Exodo, Leuitico, Numeris, Deuterono-mia; et uno libro Iesu Naue, uno Iudicum, uno libello, qui appellatur Ruth, qui magis ad Regnorum principium uidetur pertinere;

Augustine’s note on the nature of Ruth calls for some comment. His statement that the Book of Ruth was actually more related to the time of the kings of Israel (than to that of the Judges), dem-onstrates that here he wants to locate it with the preceding seven books or he would not have made this point. This linking of Ruth with the preceding books is, however, a departure from the view of the links between these biblical books that he expressed elsewhere. Augustine commented twice on the first seven books, the hepta-teuch, as a collection,11 and on those occasions there is no hint that he saw Ruth as belonging with them. Moreover, the almost contem-porary list (from 405) found in Innocent I’s letter to Exsuperius of Toulouse indicates that Innocent too thought of Ruth linked with

9. See Institutiones 1,8,14.10. Institutiones 1,2,12.11. In his Locutionum in Heptateuchum (J. Fraipont ed., CCSL 33 (1958), pp.

379-465) and his Quaestionum in Heptateuchum (ibid., pp. 1-377).

T. O’LOUGHLIN 51

Kings for the reference to this book reads: Regnorum Libri IIII simul et Ruth.12 So why does Augustine, acknowledging the problem, place Ruth here: the answer probably lies in the fact that in many codices (e.g. Codex Vaticanus) this is the location of Ruth, and Augustine follows the sequence he encountered in actual practice.13 But Ruth apart, the seven books form a conceptual unity in that they relate to the time before there was a king in Israel, and as such they form the earliest historical stratum within the canon, but as a matter of practice those seven books are encountered along with Ruth.

That the list from Genesis to Ruth is a grouping is confirmed by Augustine’s use of deinde to introduce the next list of books:

deinde quattuor Regnorum, et duobus Paralipomenon, non consequenti-bus, sed quasi a latere adiunctis simulque pergentibus. Haec est histo-ria, quae sibimet annexa tempora continet, atque ordinem rerum; sunt aliae tamquam ex diuerso ordine, quae neque huic ordini neque inter se connectuntur, sicut est Iob, et Tobias, et Esther, et Iudith, et Machabae-orum libri duo, et Esdrae duo, qui magis subsequi uidentur ordinatam illam historiam usque ad Regnorum uel Paralipomenon terminatam;

While Augustine sees all these books as relating the history of Israel after the foundation of the monarchy, and as such forming a group within the canon, it is also clear that he thinks of ‘history’ as comprising two sub-groups. Samuel, Kings and Chronicles form the first sub-group because they relate their story in a connected and orderly narrative. The other books, although histories, are lack-ing those qualities, with the exception of the ‘two books of Esdras’14 which do contain an orderly account up to the time of the conclusion of books ‘of Kings and Chronicles.’

12. This decretal (‘Consulenti tibi’) was edited by H. Wurm, ‘Decretales selectae ex antiquissimis Romanorum Pontificum epistulis decretalibus’ in Apollinaris 12 (1939) 40-93, and the reference to Ruth is found on p. 75.

13. The fact that Cassiodorus places Ruth in his first codex can be seen as corollary confirmation that he is following De doctrina christiana in his arrangement of the biblical books in groups; furthermore, in the explicitly ‘Augustinian’ list in Institutiones 1,13,1 Ruth is placed after Judges and before Kings (see Table 2 below).

14. Augustine’s reference to ‘the two books of Esdras’ needs, in the light of later developments, some comment. Augustine thought of the two texts Ezra and Nehemiah (as found in the Vulgate) as a single book; and the other book of Esdras is the text we refer to as ‘I Esdras’ (Augustine considered I Esdras’s unique element, a riddle about the world’s strongest forces: wine, kings, women, or truth, to be a part of the canon of scripture, and saw it as a prophecy of Jesus as ‘the truth’ in De ciuitate Dei 18,36). On this topic, see P.-M. Bogaert, ‘Les Livres d’Esdras et leur numérotation dans l’histoire du Canon de la Bible Latine,’ Revue Bénédictine 110 (2000) 5-26 at 12.


Again, a new larger grouping of books is introduced by deinde and called ‘prophets’:

deinde prophetae, in quibus Dauid unus liber psalmorum; et Salomonis tres: Prouerbiorum, Cantica canticorum, et Ecclesiastes. Nam illi duo libri, unus qui Sapientia, et alius qui Ecclesiasticus inscribitur, de quadam similitudine Salomonis esse dicuntur: nam Iesus Sirach eos conscripsisse constantissime perhibetur. Qui tamen quoniam in auctori-tatem recipi meruerunt, inter propheticos numerandi sunt. Reliqui sunt eorum libri, qui proprie prophetae appellantur, duodecim prophetarum libri singuli, qui connexi sibi met, quoniam numquam seiuncti sunt, pro uno habentur: quorum prophetarum nomina sunt haec: Osee, Ioel, Amos, Abdias, Ionas, Michaeus, Naum, Abbabuc, Sophronias, Aggaeus, Zacharias, Malachias. Deinde quattuor prophetae sunt maiorum uolu-minum: Esaias, Hieremias, Danihel, Ezechihel.

Within this very large grouping we can easily distinguish what Augustine thinks of as three real sub-groups. The most obvious sub-group is ‘the major prophets’ which are seen to come after the others and, by contrast with the preceding list are treated as sepa-rate writers and works. Then we can isolate another grouping: the minor prophets, who, though there are twelve names in the list, are never thought of separately but seen as connected with one another, and treated as one. These two sub-groups can be distinguished from the rest of the group in that they, together, can be considered as ‘properly called prophets.’ The remainder, form the first sub-group: Psalms, Proverbs, Song of Songs, Ecclesiastes,15 Wisdom and Sirach.16 This sub-group is conceived formally as made up of the writings of David (Psalms) and Solomon (the other five books), despite the fact that Augustine is explicit that he considers both two ‘books of wisdom’ to be the work of Jesus ben Sira. The fact that he is explicit in his statement of the non-Solomonic authorship of these books, but equally insistent that they be counted among the prophets in this location indicated that this listing we find in the De doctrina christiana is not simply a list of books’ titles, but a formal organizing system that he has inherited.

Having pointed out that it is in those forty-four books that the authority of the Old Testament is to be found, Augustine crosses the major division in his canon and turns to the New Testament:

15. I.e. Qoheleth.16. I.e. Ecclesiasticus.

T. O’LOUGHLIN 53

Noui autem quattuor librorum euangelio: secundum Matthaeum, secun-dum Marcum, secundum Lucam, secundum Iohannem; quattuordecim epistolis Pauli apostoli, ad Romanos, ad Corinthios duabus, ad Galatas, ad Ephesios, ad Philippenses, ad Thessalonicenses duabus, ad Colos-senses, ad Timotheum duabus, ad Titum, ad Philemonem, ad Hebraeos; Petri duabus; tribus Iohannis, una Iudae, et una Iacobi; Actus Apos-tolorum libro uno, et Apocalypsis Iohannis libro uno.

In contrast to the Old Testament listing, Augustine does not give us clear guidance as to how he imagined the books of the New Tes-tament to fall into groups, probably because, in contrast to the Old Testament (where there is more possibility of confusion and where there was already some dispute as to the status of some of the works he included), he expected his audience to be sufficiently familiar with the material that they would break it into gospels, let-ters, and other books. The most striking feature to modern eyes is the location of Acts in the list, almost at the end, and from this we can assume that since these two books (Acts and Apocalypse) do not belong to either of the genera mentioned (gospels and letters) that they form a final group on their own. With regard to the gos-pels, it is worthy of note that Augustine does not simply call them by their authors’ names (as he does with the major prophets) but repeats ‘secundum’ and thus makes a subtle point that though there are ‘books of the gospel’ (librorum euangelio) there is only one gos-pel in four accounts.17 Likewise he asserts, without further comment, that Paul wrote fourteen letters and simply lists Hebrews as the last of them.18 We can presume that the three New Testament groups are ‘gospels,’ ‘letters,’ and ‘what is left.’ Lastly, again in contrast to the Old Testament list, Augustine does not give us a final tally of books, presumably it was enough to number of Pauline letters; and, in any case, there was no disagreement or danger of ambiguity over this New Testament list.

Cassiodorus’s preface makes clear that his world is that of the edu-cational provider: he wants Christian schools, teachers, and instruc-tion – and the prerequisite for this is books, and, in particular, intro-

17. See T. O’Loughlin, ‘St Augustine’s view of the place of the Holy Spirit in the formation of the gospels,’ in D. V. Twomey and J. E. Rutherford eds, The Holy Spirit in the Fathers of the Church (Dublin 2010), pp. 86-95.

18. See T. O’Loughlin, ‘Jerome’s De uiris illustribus and Latin Perceptions of the New Testament’s Canon’ in J. E. Rutherford and D. Woods eds, The Mystery of Christ in the Fathers of the Church: Essays in Honour of D. Vincent Twomey SVD (Dublin 2012), pp. 55-65.


ductory books,19 which come to the student in a properly arranged, i.e. corrected, form.20 So the whole work could be described as aim-ing at a distillation of the writings of the great teachers into a sys-tem of readings,21 and one of the books that he cites by name in the preface is the De doctrina christiana – so we do not have to look far to demonstrate the plausibility of a direct influence on the arrange-ment of biblical studies in a series of well-structured codices.22

The first of these codices concerned what Cassiodorus labels the ‘Octateuch’23 which contains Genesis through to Ruth. Most of Cas-siodorus’s interest is focussed on Genesis, with the other seven books being mentioned in just one paragraph of the chapter,24 and his main concern is to set out a list of the commentaries to be studied.25 The commentaries, we are led to believe, have all been arranged in a single codex by him and this chapter is a guide to that codex. Whether or not the codex also contained the biblical texts is not made explicit, but it is clear that the order of instructional codices follows the arrangement that is natural to the biblical books, and, by implication, the way they were gathered into books.26 Therefore, by observing the order of the arrangements of the commentaries, we see the underlying biblical order. What we must consider is this: nine codices with the biblical books in this order, and then commentary codices with a parallel order.

19. Institutiones, 1, praefatio 1: … ut ad uicem magistri introductorios uobis libros … .

20. Institutiones, 1, praefatio 2: … auctoritatem diuinam in codicibus emendatis … .21. See T. O’Loughlin, Teachers and Code-Breakers: The Latin Genesis

Tradition, 430-800 (Turnhout 1999), pp. 245-71.22. Institutiones, 1, praefatio 7; and there is the reference to it in 1,13,2

(see above). Cassiodorus also mentions the importance of the De doctrina on another occasion: in his list of the ‘introductores’ in 1,10,1 (see T. O’Loughlin, ‘Early Medieval Introductions to the Holy Book: Adjuncts or Hermeneutic?’ in R.N. Swanson ed., Studies in Church History 38: The Church and the Book (Woodbridge 2004), pp. 22-31).

23. See P.-M. Bogaert, ‘Eptaticus: le nom des premiers livres de la Bible dans l’ancienne tradition chrétienne grecque et latine’ in J.-C. Fredouille, M.-O. Goulet-Cazé, P. Hoffmann, and P. Petitmengin eds, Titres et articulations du texte dans les oeuvres antiques: Actes du colloque international de Chantilly, 13-15 décembre 1994 (Paris 1997), pp. 313-37.

24. 1,1,9.25. See T. O’Loughlin, Teachers and Code-Breakers, pp. 321-4.26. That the codices do contain the biblical books is confirmed by the Ezra

picture in Codex Amiatinus; to which we will come in due course.

T. O’LOUGHLIN 55

The second codex is entitled ‘on kings’ (de regum) and comprises works on the four books of Kings,27 and the two books of Chroni-cles.28

Cassiodorus – who wants to present everything in an ordered way – then makes a somewhat surprising move: his next chapter is enti-tled ‘On prophets’ (de prophetis), but this third chapter deals with the fifth codex.29 It is concerned with Isaiah, Jeremiah, Ezekiel, and Daniel (note that Augustine placed them in the order: Daniel fol-lowed by Ezekiel) for each of whom specific commentaries are men-tioned.30 This fifth codex is also concerned with the ‘remainder’: the minor prophets (residuos … prophetas … quos sermo uulgus propter breuitatem librorum suorum minores appellat). Cassiodorus then lists them but in a slightly different order to Augustine (the places of Obadiah and Joel are interchanged, and Micah comes second from last). Cassiodorus was listing not only the biblical books, but the commentaries upon these by Jerome, and a possible explanation of the unusual order would be this: we know that Jerome wrote these commentaries at various times,31 and these had already been assem-bled within Cassiodorus’s library in a different manner to that found in canon lists, and this was perpetuated in the Institutiones – what-

27. 1,2,5.28. 1,2,13.29. The sequence of chapters / codices can best be seen in tabular form:

Institutiones’ chapter Codex number

1 12 23 54 35 46 67 78 89 9

30. 1,3,2 (Is); 1,3,3 (Jer); 1,3,4 (Ez and Dan).31. See De uiris illustribus, 135 (A. Ceresa-Gastaldo ed., Girolamo: Gli

Uomini illustri / De Viris Illustribus, Florence 1988, pp. 232-4) where Jerome states that he has composed commentaries on Micah, Nahum, Habakkuk, Zephaniah, and Haggai [note that this is the standard biblical order, although it shows that Jerome began in the middle of the set of twelve] and that he is proceeding with the work (and see M. Adriaen, ‘Praefatio’ in M. Adriaen ed., CCSL (1969) 76, p. v).


ever, the explanation, this order given by Cassiodorus is idiosyn-cratic.

Then in the fourth chapter, Cassiodorus turns to the third codex which is devoted solely to the psalms (de psalterio); and in the fifth chapter to the fourth codex which is devoted to the works of Solo-mon: Proverbs, Ecclesiastes, the Song of Songs, Wisdom, and ‘Eccle-siasticus’.32 He notes that Jerome thought that Philo was the author of Wisdom and Sirach, but for Cassiodorus that was an exceptional opinion, and with a rhetorical flourish we are given the impression that ‘most,’ including the fathers Augustine and Ambrose, have accepted Solomonic authorship of all five books. The orderly teacher wanted to present an orderly reality.

Now Cassiodorus can turn to the hagiographa (de agiographis), and in the sixth codex are found the works on Job, Tobit, Esther, Judith, Maccabees (and we are given the basis to infer that there are two books of Maccabees), and the ‘two books of Esdras.’33 We are given no hint as to a problem with this last designation, nor any clue as to what actual texts were intended by the phrase: in Hesdrae uero libris duobus … . The reason for the unusual order in which Cassi-odorus narrates objects he explicitly labels with a sequence of car-dinal numbers is not hard to find: the sequence already existed in the De doctrina christiana, but his own need to arrange Augustine’s sub-groups of ‘prophets’ into distinct codices entailed narrating them out of sequence so that the more important, from his perspective on the overall values of books to a Christian student, preceded the less important. Cassiodorus was therefore willing to adapt his nar-ration to his ultimate subject, the scriptures, but was not willing to arrange the order of materials, the codices, in an order which broke with that which he inherited from Augustine.

The final three codices are less problematic for Cassiodorus: the seventh deals with the gospels,34 the eighth contains ‘the letters of the apostles,’35 and, finally, the ninth codex is devoted to the Acts of the Apostles along with the Apocalypse of John.36 The one problem is that Cassiodorus explicitly states that Paul wrote thirteen letters

32. 1,5,1 (Prov); 1,5,3 (Qo); 1,5,4 (Song); 1,5,5 (Wis); and 1,5,6 (Sir).33. 1,6,1-3 (Job); 1,6,4 (Tob, Est, Jth, Macc); and 1,6,6 (Esdras).34. 1,7.35. 1,8.36. 1,9.

T. O’LOUGHLIN 57

(in epistolis tredecim sancti Pauli),37 – ignoring the number, fourteen, given by Augustine – and it would appear that he omits Philippians. The situation is, however, more complex than a simple omission. We might approach the question by eliminating the letters that are not problematic: Romans, Corinthians (x 2), Galatians, Ephesians, Thes-salonians (x 2), Colossians, Philemon, Timothy (x 2), and Titus.38 Cassiodorus also mentioned James,39 the first letter of Peter (so we can infer there is at least a second),40 and he refers to both the first and second letters of John (but there is no hint that there is a third),41 but, by contrast, there is no mention of the letter of Jude. Does this mean we should assume that he does not consider those not mentioned canonical? Since the canonical status of these works was not questioned within the churches, such a conclusion would be ill-founded: we must simply assume that he passed over them in silence, probably because their lack of bulk (15 and 25 verses respectively) does not push them to prominence in the imagination. Now to Paul, have mentioned that he wrote ‘thirteen letters,’ and seemingly hav-ing concluded his comments on Paul (the listing comes much later in the chapter), Cassiodorus simply mentions the fact that he has found a good commentary on Hebrews by John Chrysostom,42 and so it would appear that he had some information about the ques-tion of the Pauline authorship of Hebrews (such as he could have obtained from Jerome’s De uiris illustribus43). So faced with two pieces of information: (1) that Paul wrote thirteen letters; and (2) the doubt about the author of Hebrews, Cassiodorus has sought to retain all his authorities intact and by obfuscation avoided having to make a decision between conflicting authorities. So Paul wrote thirteen letters, but only twelve letters are explicitly attributed to him; Hebrews is mentioned slightly apart from Paul, and not attrib-uted to him; Philippians in not mentioned in the list, but a com-monly used phrase from it (Phil 3:20) is quoted in the Institutiones

37. 1,8,1; and he repeats this assertion that Paul wrote thirteen letters in 1,8,2.

38. These are mentioned by name, and, where applicable, by quantity in 1,8,12-14.

39. 1,8,4.40. 1,8,4.41. 1,8,4 and 1,8,7.42. 1,8,3.43. See T. O’Loughlin, ‘Jerome’s De uiris illustribus and Latin Perceptions

of the New Testament’s Canon.’


2,3,5 – so the possibility is there for the audience to add it to the other explicitly mentioned, making thirteen, as if its omission was simply an oversight!44 Moreover, lest there be any doubt, the whole twenty-seven books of the standard New Testament canon are listed in the ‘Augustinian’ list in Institutiones 1,13,1. The whole message is greater than the sum of the details: and that message which Cassi-odorus most wishes to deliver to his students is that the tradition is a comprehensive whole which speaks with a single, consistent voice.

Cassiodorus sought at the outset of the Institutiones a curriculum for Latin biblical studies; and, as this paper has shown, the structure and sequence of that programme came from Augustine’s De doctrina christiana. However, he also took that schema for books to heart in Vivarium and the bibles that were produced there. The evidence for this is somewhat indirect, but nonetheless convincing. Let us return to the image mentioned above of Ezra - Cassiodorus in front of his armarium which is found in the Codex Amiatinus. On the spines of these codices are written their contents,45 and each corresponds to the nine codices outlined in Institutiones 1,1-9; and they are arranged in the same sequence as the biblical books are mentioned in the De doctrina christiana (see table 3).46 The seventh-century Northumbrian image thereby offers us an insight into a detail of Augustine’s influ-ence in sixth-century southern Italy.47

University of Nottingham Thomas O’Loughlin

44. The solution that Cassiodorus, or any theologian in the sixth century, simply did not know or recognise Philippians seems too far-fetched to be considered, except momentarily.

45. See Marsden, ‘The Codex Amiatinus and a sister codex,’ p. 134.46. Marsden, ibid., says that they ‘follow in outline the Augustinian order’

[presumably meaning the order found in the ‘Augustine list’ in Institutiones 1,13,1; but as can be seen in table 3: they follow the order found in the De doctrina christiana exactly.

47. I wish to record my gratitude to the editor and his readers for many helpful comments on an earlier draft of this paper.

T. O’LOUGHLIN 59

TABLE 1:

THE STRUCTURAL GROUPINGS OF BIBLICAL BOOKS SET OUT IN DE DOCTRINA CHRISTI ANA AND THAT PROPOSED BY

CASSIODORUS IN INSTITUTIONES 1,1-9.

Augustine Cassiodorus

Group 1:pre-kings history

The First Codex:The Octateuch

Genesis, Exodus, Numbers, Leviticus, Deuteronomy, Joshua, Judges, and Ruth

Group 2 / 1:orderly history

The Second Codex:On Kings

Kings (x 4), Chronicles (x 2)

Group 2 / 2:other history

The Sixth Codex:The Hagiographa

Job,Tobit,

Esther,Judith,

Maccabees (x 2),‘Esdras’ (x 2)

Group 3 / 1:prophets

The Third Codex:The Psalter

Psalms,

Proverbs,Song,

Ecclesiastes,Wisdom,

Sirach

Psalms

The Fourth Codex:Solomon’s Works

Proverbs,Song,


Sirach


Group 3 / 2:minor prophets

The Fifth Codex:Prophets (implicitly in two groups:

major and minor)48

Hosea,Joel,

Amos,Obadiah,

Jonah,Micha,

Nahum,Habakkuk,Zephaniah,

Haggai,Zachariah,

Malachi

Isaiah,Jeremiah,Ezekiel,Daniel,

‘and the 12 minors’Hosea,

Obadiah,Amos,Joel,

Jonah,Nahum,

Habakkuk,Zephaniah,

Haggai,Zachariah,

Micha,Malachi

Group 3 / 3:major prophets

Isaiah,Jeremiah,

Daniel,Ezekiel

Group 4:gospels

The Seventh Codex:The Gospels

Matthew, Mark, Luke, John

Group 5:epistles

The Eighth Codex:The Epistles of the Apostles

Romans,Corinthians (x 2),

Galatians,Ephesians,

Philippians,Thessalonians (x 2),

Colossians,Timothy (x 2),

Titus,Philemon,Hebrews

Peter (x 2),John (x 3),

Jude,James





Titus,Philemon,Hebrews

Peter (x 2),John (x 3) [Jn 3?],

Jude [?],James

48. Since the second sub-group is labelled by Cassiodorus, and thought of as a collection, we can infer that the first four were thought of as ‘major’ and so forming a sub-group.

T. O’LOUGHLIN 61

Group 6:[no designation can be inferred]

The Ninth Codex:[no designation given]

Acts of the Apostles,the Apocalypse of John

TABLE 2:

A COMPARISON OF THE GROUPINGS OF BIBLICAL BOOKS IN DE DOCTRINA CHRISTIANA 2 AND THAT ATTRIBUTED TO

AUGUSTINE IN INSTITUTIONES 1,13,1.

Augustine inDe doctrina christiana

Augustine inInstitutiones 1,13,1

Group 1:pre-kings history

History in 22 books

Genesis,Exodus,

Numbers,Leviticus,

Deuteronomy,Joshua,Judges,

Ruth

Moses (x 5)

Joshua,Judges,

Ruth

Kings (x 4),Chronicles (x 2)

Job,Tobit,

Esther,Judith,

‘Esdras’ (x 2),Maccabees (x 2)

Group 2 / 1:orderly history

Kings (x 4),Chronicles (x 2)

Group 2 / 2:other history

Job,Tobit,

Esther,Judith,

Maccabees (x 2),‘Esdras’ (x 2)

Group 3 / 1:prophets

Prophets in 22 books


Psalms,

Proverbs,Song,


Sirach

Psalms,Solomon (x 3),Sirach (x 2),

Major Prophets (x 4)Isaiah,

Jeremiah,Ezekiel,Daniel

Group 3 / 2:minor prophets Minor Prophets (x 12)

Hosea,Joel,

Amos,Obadiah,

Jonah,Micha,

Nahum,Habakkuk,Zephaniah,

Haggai,Zachariah,

Malachi

Hosea,Joel,

Amos,Obadiah,

Jonah,Micha,

Nahum,Habakkuk,Zephaniah,Zachariah,

Haggai,Malachi

Group 3 / 3:major prophets

Isaiah,Jeremiah,

Daniel,Ezekiel

Group 4:gospels

The New Testament

Matthew,Mark,Luke,John

Matthew,Mark,Luke,John

Group 5:epistles





Titus,




Colossians, Timothy (x 2),

Titus,

T. O’LOUGHLIN 63

Philemon,Hebrews,


Jude,James

Philemon, Hebrews,


Jude,James

Group 6:[no designation can be inferred]



22 + 22 + 27 = 71 (Institutiones 1,13,2)

TABLE 3 :

THE RELATIONSHIP OF THE IMAGES OF THE NINE CODICES IN THE PICTURE OF EZRA - CASSIODORUS IN CODEX

AMIATINUS49 TO THE NINE CODICES DESCRIBED IN THE INSTITUTIONES 1,1-9.

In this comparison it is assumed that the codices lie in the armarium in this order:50

Codex 1 Codex 2

Codex 3 Codex 4

Codex 5 Codex 6

Codex 7 Codex 8

Codex 9

Codex Amiatinus Cassiodorus

Codex 1OCT . LIB . LEG

The First Codex:The Octateuch

49. The texts on the spines in the drawing follow Marsden, ‘The Codex Amiatinus and a sister pandect,’ p. 134 (Marsden follows, while correcting a misprint, R.L.S. Bruce-Mitford, The Art of the Codex Amiatinus, p. 9).

50. We can be certain that this is the correct order in which to consider the arrangement of the codices from the sequence on the two lowest shelves where we have gospels (on left of shelf), epistles (on right of shelf), and the final codex on the left beneath the gospels.


This identification is certain: the codex contain the Octateuch and this includes the books of the law.

Codex 2REG . PAR . L . VI

The Second Codex:On Kings

Again, the identity is certain as both the image and the list contain six books made up of Kings (REG), Chronicles (PAR).

Codex 3HIST . LIB . VIII

The Sixth Codex:The Hagiographa

While the names given to the group do not coincide, the number of books in codex image and list are identical.

Codex 4PSAL . LIB . I

The Third Codex:The Psalter

The identity is again certain.

Codex 5SAL . LIB . V

The Fourth Codex:Solomon’s Works


Codex 6PROP . L . XVI

The Fifth Codex:Prophets

The identity is again certain: the ‘prophets’ codex contained 16 books in total.

Codex 7EVANG . L . IIII

The Seventh Codex:The Gospels


Codex 8EPIST . AP. XXI

The Eighth Codex:The Epistles of the Apostles

The identity is again certain: the ‘epistles’ codex contained 21 books in total.

Codex 9ACT . AP . APOC . IS

The Ninth Codex:[no designation given]

The image names the two books that comprise this codex in the list.

The OT Canon: Augustine and Cassiodorus

Documents