I. The state of our text xxxix
TEXTUAL INTRODUCTION'
I. The state of our text B e d e ' s History is one of the very few works written in Latin before the Carolingian renaissance which have come down to us in copies virtually contemporary with their authors; the scribes of our two oldest manuscripts might well have been among the disciples who gathered round the master's deathbed. It is there fore not surprising if its state of preservation is almost impeccable, so that Bede's reader has little more need to trouble himself with the details of textual criticism than the reader of Gibbon or Macaulay. If we invite him to consider the state of our text and its circulation in the Middle Ages, it is that these things have an interest of their own. First, then, for the text.
We know from books like the Lindisfarne Gospels, the Stonyhurst St. John, or the Oxford copy of Bede on Proverbs, M S.
Bodley 819, how high was the tradition of accuracy in the North umbrian scriptoria; and the History is no exception. In the whole work, as it appears in the consensus of our oldest and best copies, there are perhaps thirty-two places (in nearly 300 printed pages) where some defect of sense or syntax suggests that correction is required. But even this small quantum of error is not what it seems. In twenty-six of these places, Bede is transcribing from an earlier source: there are four, for example, in i. 17 and 20 where he is drawing on the Vita sancti Germani, four in i. 27 the Libellus Responsionum of St. Gregory, fourteen in the various papal letters which he has incorporated. That 80 per cent, of these mistakes should occur in quoted documents can hardly be acci dental. Perhaps these defects were there already in Bede's sources; perhaps, when he had a written source, he or his amanuensis transcribed it very accurately, including even its errors, and the 1
This section of the Introduction is based on Plummer's work, and on the Hand-list of Bede Manuscripts (Ithaca, New York, 1943) of M. L. W.
Laistner and H. H. King. It owes much to Professor Bernhard BischofT, Dr. R. W. Hunt, Mr. Neil Kcr, Mr. Paul Meyvaert, and Dr. Michael WallaceHadrill ; and let the kindness of the custodians of many libraries be gratefully recorded.
T E X T U A L IN T R O D U C T I O N
result was faithfully transmitted by the transcribers of the finished work. Three examples will make this almost certain:
In i. 7 (p. 30) there is a clause that makes no sense {quod . . . diuidebatur): he is following the Passio sancti Albani, and the same corruption appears in an independent ninth-century copy of the Passio now in Paris, so it was no doubt already in Bede's own copy.
In ii. 18 (pp. 196-8) the preposition per twice appears governing the ablative instead of the accusative. This is the error of someone to whom late Latin with its changes of syntax and form is still a living language, not of an Englishman who has been drilled in the grammar in his monastic school, and we are reading a letter of Pope Honorius I.
The mistake is probably due to the clerk who made the copy of the letter which was supplied to Bede; and an exact parallel can be found elsewhere in unofficial copies of two letters of Gregory the Great.1
In iii. 19 (p. 272) quartus for quartum, is unexplained, until we find that it is quartus in the Vita sancti Fursei: the source is not here in error, but it has been copied literatim and not adapted to the syntax of Bede's sentence.
There remain six examples of error in all our oldest manu scripts—and so presumably in their common original—where
Bede does not seem to be copying an earlier document. One is the omission of ut in the heading of ii. 5 (p. 118), which seems just a slip. The others may be unprovoked scribal mistakes; but all could be explained if the author had changed his mind and not removed all traces of the change.
In iii. 11 (p. 246), if Bede had written ossa instead of reliquiae, eadem would have been correct: in iii. 14 (p. 256) quam for quod suggests that secretario may have replaced ecclesia or porticu: in iii. 22 (p. 282) qui follows uilla regia as though it had been (say) loco: in iv. 22 (20) (p. 402) promittens se nihil ei mali futurum pro eo looks like the fusion of two alternative phrases se nihil ei mali facturum and nihil mali futurum pro eo: in V. 13 (p. 500) perhaps Bede himself never decided whether the demons were to hold cultra or uomeres or something else.
Of course this is mere speculation on a very narrow base; but these early scribes are seen to be so faithful, that one well may wonder whether in the rare defects of the text they copied, they have preserved for us a glimpse of the author's draft.
The early manuscripts tell us another thing, as Plummer saw (1, pp. xciv-v). He established that they fall into two main classes,
1 See D. Norberg, In Registrum Gregorii studia critica (Uppsala, 1937), 85 n.
T E X T U A L I N T R O D U C T IO N
xli called by him the 'C-type' and the 'M-type', by us c and m, which he distinguishes as follows:
(a) The prayer Praeterea omnes . . . inueniam (below, p. 6) is found after the preface in m, at the end of the whole work in c.
(b) A miracle of St. Oswald (iv. 14) is recounted in m and not in c.
(c) The words in iv. 18 cuius supra meminimus, referring to a previous mention of Benedict Biscop which is not in fact to be found in the text as we have it, are not in c.
(d) At the end of iv. 30, where m has quaedam quae, c has unum quod, as though only one miracle of St. Cuthbert were to follow
(though c, like m, gives two, both iv. 31 and 32), and c in its list of chapters has the headings of iv. 31 and 32 in the reverse order.
(e) In the chronological summary (v. 24), c ends the annal for 731 with archiepiscopus, and then inserts annals for 733 and 734 which are not in m.
(/) In the list of Bede's works at the end of v. 24, c makes no mention of his Capitula on the Prophets (an omission which may of course be accidental).
What is the relation of these two types of text to one another?
The addition in m of iv. 14, which is clearly authentic, and would never have been removed by a reviser, stamps m as the later form; and the appearance in c of annalistic entries for 733 and 734, while m stops at 731, does not (as Plummer thought) tell very strongly the other way, for nothing lends itself so easily to later addition, whether authentic or alien, as annals. But the question is of no great importance, for any hope that we might be able to trace two stages in the writing of the History is soon disappointed.
In this edition, the readings of individual manuscripts are not recorded;1 the m-text (as being the later) is printed, with all c's variations in the notes (which can be given with certainty only in books iv and v, as will appear); and the reader will see that this amounts to very little beyond small changes in the order of words.
Points (c) and (d) in our list give us again a momentary glimpse of Bede at work, changing his mind and forgetting to conceal the traces, and that seems to be all. The two types of text are important for a quite different reason: c held a monopoly in this country,2 while Continental copies with very few exceptions all belong to m.
1 A text with critical apparatus is planned for the series Corpus Christianorum.
2 The compilers of the Northern version of the Anglo-Saxon Chronicle are said (above, p. xviii, n. i) to have used an m-text in the tenth century; this is no doubt possible, but I know no evidence for it.
xlii
T E X T U A L IN T R O D U C T I O N
ii. Our Authorities for the Text1 Our c-text is established from the following copies:
K = Kassel, Landesbibliothek 40 M S. theol. 2 (C LA viii, no. 1140), written in small Northumbrian hands of the late eighth century, with early corrections; no ornament, and the rubrics and chapter-numbers never filled in; the limp wrapper bears a ninth-century title and the fifteenth-century pressmark of the famous abbey of Fulda in Hesse, founded by St. Boniface in 744.
An unpretentious, easily portable book, with a most accurate text, except for the occasional omission of words; it is very greatly to be regretted that only books iv and v survive, the others having been lost—to judge by the rubbing and retracing of the present first leaf—long before the end of the Middle Ages.
C = London, British Museum Cotton Tiberius C. I I (C LA 11, no. 191), of the second half of the eighth century, well written with handsome initials, not in Northumbria but in Southern England.
It was in the south that Old English glosses (Ker, no. 198) were added in the ninth century.
Plummer (1, p. xciii) says 'it is certainly a Durham book', and gives two reasons; (a) the odd mistake meditaturum for meditandum at the beginning of the preface, which is also found in the twelfth-century historian Symeon of Durham: (b) near the end of the preface, nostro is added in an early hand, so as to read patre nostro et antistite Cuthberto.
The first argument can have little force as long as we know so little about Symeon's sources, and because meditaturum is found elsewhere—in a twelfth-century copy from Anchin (below, p. lxi) and as an alternative in a book written in Oxford or Salisbury (p. xlviii). The wording patre nostro is found also in our manuscript O: and unless the date of the addition in C rules this out, one would suppose it to come from their common parent, a Northern book, and to throw no light on their own place of writing or their later home.* 0 = Oxford, Bodleian Library Hatton 43 (4106), which is
1 In what follows, CLA = E. A. Lowe, Codices Latini Antiquiores (Oxford, 1934-69); Ker = N. R. Ker, Catalogue of M SS. containing Anglo-Saxon
(Oxford, 1957).
1 S. M. Kuhn, in Speculum, xxiii (1948), 613-19, and Review of English Studies, N .s. vrii (1957), 355-70, thought of Lichfield as the place of origin, with Worcester as a later home; but his criteria were impugned by K. Sisam in RES, N.s. vir (1956), 1-10, 1 13 -3 1, and viii, 370-4. David H. Wright, in Anglia, L x x x ii (1964), 116, has re-emphasized the relation between C and other illuminated books connected with St. Augustine's, Canterbury.
T E X T U A L IN T R O D U C T I O N
xliii Plummer's Ox (i, p. cxiii), an accurate and handsome book of the early eleventh century, carefully corrected by several later hands, especially in the twelfth. Sir Thomas Hardy's suggestion, followed by Plummer, that it comes from Glastonbury, seems to be without foundation.
C and O are clearly independent, so that O although much later is valuable as a check on the accuracy of C ; but their texts are so close that they must have had a common eighth-century ancestor, which is here called c2. This represents the form of the History current throughout the Middle Ages in these islands; and it is not quite so free from faults as K. In books iv and v, where we have both C2 and K , we can reconstruct the text of c with confidence, for all three copies are very close to their common source. Else where, when c2 differs from m, we cannot always be certain whether we have the true text of c or a variant peculiar to c2.
Fortunately, for one long chapter which is lost from K, we have another c witness, which is Continental and presumably indepen dent of c2t on which it can therefore act as a check. This is i. 27, the Libellus Responsionum, copied by itself, in the form peculiar to the History and with most, but not all, of c2fs characteristic readings (and so perhaps from some manuscript like K), in Zürich,
Zentralbibliothek Rh. 95, of the tenth century, from Rheinau.1
Even where c can be reconstructed, we have given the variants of c2 in full, though often worth little, for they are characteristic of the text current in England in the eighth century.2
Our m-text is based on the following:
M = Cambridge, University Library Kk. 5. 16 (C LA 11, no.
139), known as the Moore M S. because, before King George I gave it to Cambridge in 1715, it had belonged to John Moore, bishop of Ely. It was written in Northumbria in or soon after the year 737, as can be inferred from the retrospective dating at tached to certain historical events. The unusually full use made of the writing-space, the number of small mistakes, the absence of 1 These facts were generously provided by Mr. P. Meyvaert; he is not re sponsible for any misinterpretation of them. See also his article 'Les " Re sponsiones" de S. Grégoire le Grand á S. Augustin de Cantorbéry* in the Revue d'Histoire Ecclésiastique, L iv (1959), 879-92.
2 The Old English version, published by T. Miller for the Early English Text Society, vols. 95, 96, no, h i (1890-8), follows a good c (rather than c2) text; cf. Plummer, 1, p. cxxix, and Professor D. Whitelock, in Proceedings of the British Academy, x l v iii (1962), 57" 9°-
xliv
T E X T U A L IN T R O D U C T I O N
ornament, and other symptoms suggest that the scribe was obliged to be economical of time and material. (Perhaps he was on a visit to Wearmouth or Jarrow, or had the loan of a copy from there in his own monastery for a limited time.) Full details are given by P. Hunter Blair in the complete reproduction published in the Copenhagen Early English M S S in Facsimile, vol. ix (1959).1
At the end are some chronological notes added not later than 737,2 the Old English version of Cædmon's hymn (Ker, no. 25), and some other excerpts, best dealt with when we come to speak of M 's numerous progeny (below, p. lxii).
L = Leningrad, Public Library Lat. Q. v. I. 18 (CLA xi, no.
16 21; Ker, no. 122), a very handsome volume with fine ornament, copied by four scribes, no doubt at Wearmouth or Jarrow, not later than the year 747.3 A complete facsimile was published by O. Arngart in the same Copenhagen series, vol. 11 (1952), and the manuscript has been discussed in Scriptorium, xii (1958), 182207, the hand by E. A. Lowe and the ornament by M. Schapiro.4
The text is very close to that of M, but more accurate. There seems no reason why both should not have been taken from the author's copy—not to say 'autograph', because Bede's own copy may well have been in the hand of an amanuensis.5
U = Wolfenbüttel, Herzog-August Bibliothek Weissenburg 34 (CLA ix, no. 1385) seems, with the two which follow, to represent the text of another, presumably Northumbrian, manuscript now lost, which passed to the Continent and was copied there in the early Carolingian period. This one is a good solid book, of the late eighth century, with illuminated initials in the insular taste, accurately written in an early Caroline minuscule 'attributed to north-eastern France' and 'related to that of the Ada-group'.6 By 1 Add the review b y David H. Wright in Anglia, l x x x ii (1964), 110-17.
2 See P. Hunter Blair in The Early Cultures of N .W . Europe (H. M.
Chadwick Memorial Studies), ed. C. Fox and B. Dickins (Cambridge, 1950),
PP24557-
3 As with M, this is an inference from the retrospective dating used in the margins, the interpretation of which is discussed by David H. Wright in the Revue Bénédictine, l x x i (1961), 266-73, and M. Bévenot, S.J., in Scriptorium,
XVI (1962), 365-9.
4 Mr. P. Meyvaert has shown that the head of a saint in the initial to Book ii on f. 26v of L, identified in an early hand as St. Augustine, is really Gregory the Great: Bede and Gregory the Great (Jarrow Lecture, 1964), pp. 3-4 and pi. 1.
5 MS. Ricketts 177, now in the Lilley Library at Bloomington (Indiana), is half a leaf from an m-text (v. 13) in an eighth-century insular hand.
6 D. H. Wright in Anglia, l x x x ii (1964), 116.
T E X T U A L I N T R O D U C T I O N
xlv the end of the Middle Ages, its home was at Weissenburg in Alsace; for an early copy of it now at St. Gall see below, p. lxv.
E = Würzburg, Universitätsbibliothek M . p. th. f. 118 , written in the second third of the ninth century in a Caroline minuscule of the type used in Würzburg under Bishop Hunbert (832-42), and perhaps copied from the Historia Anglorum recorded in a list of Würzburg cathedral books drawn up about the year 800.1 Later it belonged to the Cistercian abbey of Ebrach, between Würzburg and Bamberg. This is very close in text to the Weissenburg M S .; the third stands a little apart, and now and then seems a more faithful witness than the other two. It is now
N = Namur, Public Library, Fonds de la ville i i , a ninthcentury copy from St. Hubert in the Ardennes, the work of several scribes who differed in skill and sometimes were much puzzled by insular features in the script of their exemplar. Plum mer (1, pp. lxxxvii f.) treats it with an asperity which is most unlike him, and is not warranted; he was not in a position to assess its place in the history of the text, speaks as though the least skilful of the scribes (who was clearly unequal to his task) had written the whole book, and judges it by the severest standards. It has its value, as an independent check on EU in the reconstruction of an early copy of the History, which must have crossed over to Germany in the early days, but unlike K has left numerous descendants, as we shall see.
In practice, our evidence for the m-text, based as it is on two Northumbrian copies written almost in the author's lifetime, has little need of corroboration even from so early a source as this.
Nor need we so much, from the purely textual point of view, regret the loss of another early migrant to the Continent, whose existence we can dimly discern through its twelfth-century descendants
(listed below, pp. lxviii-ix).
We have, then, a reliable text of the History, and can form a working hypothesis of its early circulation. Let us not think that we know the whole story; there are byways still unexplored.
New York, Pierpont Morgan Library M 826 (CLA xi, no. 1662) is a single leaf (formerly Phillipps MS. 36275), containing part of iii.
29-30 in an English hand of the late eighth century. The writing is rather rough, the text inaccurate: its relation to our complete copies
1 See B. Bischof! and J. Hofmann, Libri sancti Kyliani (Würzburg, 1952), pp. 30, 123, 144-