¹ I use the following abbreviations: Ehwald = Ehwald, R., Aldhelmi Opera, MGH, Auct. Antiq. 15 (Berlin, 1913–1919)Google Scholar; Goossens = Goossens, L., The Old English Glosses of MS. Brussels, Royal Library 1650 (Aldhelm's De Laudibus Virginitatis) (Brussels, 1974)Google Scholar; Gwara = Gwara, S., ‘Literary Culture in Late Anglo-Saxon England and the Old English and Latin Glosses to Aldhelm's Prosa de virginitate’, 2 vols, (unpubl. PhD dissertation, Univ. of Toronto, 1993)Google Scholar; Isid. = Isidori Hispalensis etymologiarum sive originum libri XX, ed. Lindsay, W. M. (Oxford, 1911)Google Scholar; Ker = Ker, N. R., Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957)Google Scholar; Lapidge and Herren = Lapidge, M. and Herren, M., Aldhelm: the Prose Works (Cambridge, 1979)Google Scholar; Lapidge and Rosier = Lapidge, M. and Rosier, J., Aldhelm: the Poetic Works (Cambridge, 1985)Google Scholar; Napier = Napier, A., Old English Glosses, Chiefly Unpublished (Oxford, 1900).Google Scholar All citations of glosses derive from my edition.

² Pdv was arguably composed in the 670s. The details of Aldhelm's biography derive mainly from William of Malmesbury's De gestis pontificum anglorum (ed. Hamilton, N. E. S. A., RS 52 (London, 1870)) in which bk V (pp. 330–443)Google Scholar comprises the Vita Aldhelmi. For a full overview of Aldhelm's career – his birth, education, travel, ecclesiastical honours and death – see Lapidge, and Herren, , pp. 5–10Google Scholar as well as Lapidge, and Rosier, , pp. 5–9.Google Scholar

³ See, for example, the Index locorum (II. Loci classici et ecclesiastici) to Ehwald, , pp. 544–6Google Scholar; Lapidge's Appendix III (‘Check-List of Sources of Aldhelm's Exemplary Virgins’, Lapidge, and Herren, , pp. 176–8)Google Scholar; Brown, T. J., An Historical Introduction to the Use of Classical Latin Authors in the British Isles from the Fifth to the Eleventh Century’, Sett Spol 22 (1975), 237–93Google Scholar, and, most recently, Orchard, A., The Poetic Art of Aldhelm (Cambridge, 1994)CrossRefGoogle Scholar and idem, ‘After Aldhelm: the Teaching and Transmission of the Anglo-Latin Hexameter’, Jnl of Med. Latin 2 (1992), 96–133.Google Scholar

⁴ Cf. Gwara, S., ‘The Continuance of Aldhelm Studies in Post-Conquest England and Glosses to the Prosa de virginitate in Hereford, Cath. Lib. MS P.I.17’, Scriptorium 48 (1994), 18–38Google Scholar. On the hermeneutic idiom, cf. Lapidge, M., The Hermeneutic Style in Tenth-Century Anglo-Latin Literature’, ASE 4 (1975), 67–111.Google Scholar

⁵ Cf. Gwara, S., ‘Manuscripts of Aldhelm's Prosa de virginitate and the Rise of Hermeneutic Literacy in Tenth-Century England’, SM, 3rd ser. 35 (1994), 101–59.Google Scholar

⁶ Mone, F., Quellen und Forschungen zur Geschichte der teutschen Literatur und Sprache I (Aachen, 1830) pp. 329–442.Google Scholar

⁷ Hereafter I refer to manuscripts by sigla according to the following list. My sigla differ from Ehwald's, (p. 225) in some respects.Google Scholar A denotes the Yale fragment to which Ehwald had given two sigla: frgt. Oxon.=Bodl. Lib. MS th. d. 24, fols. 1–2; P=Cambr. Add. 3330. In my apparatus P denotes the Hereford Pdv copy, rather than Harley 3013, to which Ehwald assigned no siglum. The London fragment has been incorrectly cited in a number of sources as Add. 50483K, but re-catalogued as of 1997.

R2 London, BL, Royal 7. D. XXIV, 82r-168r: text s. x¹, gll. s. x²;?Glastonbury,?Canterbury

R4 London, BL, Royal 5. E. XI: text s. x/xi, gll. s. xiⁱⁿ-xi^med; Christ Church, Canterbury

S Salisbury, Cathedral Library, 38: text, gll. s. x^ex; Christ Church, Canterbury

C1 Cambridge, Corpus Christi College 326: text s. x², gll. s. x²-xiⁱⁿ; Christ Church, Canterbury

C2 Oxford, Bodleian Library, Bodley 97: text, gll. s. xiⁱⁿ; Christ Church, Canterbury

A Cambridge, University Library, Add. 3330+

New Haven, Yale University, Beinecke Library 401, 401A+

London, BL, Add. 71687+

London, BL, Add. 50483J+

Oxford, Bodleian, Arch. A. fol. 131

Oxford, Bodleian, th. d. 24, fols. 1–2+

Coll. Martin Schøyen MS 187 (olim Aachen, Dr Peter Ludwig's Library; Malibu, J. Paul Getty Museum Ludwig XI 5)+

Philadelphia, Free Library, John Frederick Lewis Coll. ET 121: text s. ixⁱⁿ, gll. s. x²;?Glastonbury,?Canterbury

P Hereford, Cathedral Library, P. I. 17: text, gll. s. xii/xiii;?Abingdon, Cirencester

The dates, origins and provenances of these manuscripts are discussed in Gwara I, pp. xxxvii-clvi.

⁸ Fell, C. E., Edward King and Martyr (Leeds, 1971), pp. xix–xx and 1–16Google Scholar; cf. Rollason, D., ‘The Cults of Murdered Royal Saints in Anglo-Saxon England’, ASE 11 (1983), 1–22, at 2Google Scholar. The sheets are ruled dry-point, ruled nineteen lines to the page. The collation is complex, and Quire III preserves an unusual cancellation: I²(blank)+II⁸(3, 5 canc.)+III¹⁰(4, 5 canc.)+IV⁸-XIII⁸+ XIV⁸(8 canc.) [end of pdv]+XV⁸ (1, 3, 5, 7 canc.)+XVI² (a bifolium, blank).

⁹ Curved elements in b, c, d, e, o, p and q are rotund. Unlike some coeval Canterbury scripts, however, Caroline a has no exaggerated top-stroke, and its back slants at 45°. Letter e has a negligible compartment; g boasts a comparable compartment with a grand, sweeping cauda, resembling that in some scripts of St Augustine's. Although the Caroline element of the Digby script dominates letter-forms, abbreviations and ligatures, the scribe still tends to incorporate æ in Latin, predominantly in final position. On the generic conventions of Style II Anglo-Caroline, cf. Dumville, D., English Caroline Script and Monastic History (Woodbridge, 1993).Google Scholar

¹⁰ According to Napier, John Clyffe or Clyve can be identified in Dugdale's Monasticon Anglicanum and Oxford University Register for 1510 (Napier, p. xiii). Ehwald mistranscribed the name as ‘Clysse’ (Ehwald, , p. 219).Google Scholar

¹¹ Watson, A., ‘Thomas Allen of Oxford and his Manuscripts’, Medieval Scribes, Manuscripts and Libraries: Essays presented to N. R. Ker, ed. Parkes, M. and Watson, A. (London, 1978), pp. 279–314.Google Scholar

¹² Ker, , no. 320.Google Scholar

¹³ I refer to Ker's ‘Groups (i), (ii) and (iii)’ as the products of ‘scribes (i), (ii) and (iii)’ or written in ‘hands (i), (ii) and (iii)’. The terms are meant to be equivalent.

¹⁴ This hand is not identified in Napier's edition of the Old English glosses.

¹⁵ Cf. my discussion in ‘Manuscripts’. Glosses from the first seven layers of Royal 7. D. XXIV (‘Early Strata’) form part of the first layer in Digby 146 (Ker's hand (i) and Napier's ‘second Latin hand’). The Digby text was in fact copied from an apograph of Royal 7. D. XXIV.

¹⁶ Ker, , p. 382.Google Scholar In fact, this hand continues on later folios, e.g. 16r5 (herebecn).

¹⁷ Napier, , p. xiii.Google Scholar

¹⁸ Ker, , p. 382.Google Scholar Napier draws attention (p. xiii) to ‘some few English glosses [which] have also been written … here and there by various different hands’. Thus wrangwise (30v16) and ealswage (70v8) are twelfth-century additions.

¹⁹ For example, the glossing ends after the catalogue of male virgins and part-way into the list of female virgins. A monk might have deemed this part of the book less valuable or pertinent to his intended audience.

²⁰ Respectively, Aldhelm's De Laudibus Virginitatis with Latin and Old English Glosses: Manuscript 1650 of the Royal Library in Brussels (Bruges, 1941)Google Scholar; Ker, , no. 8Google Scholar; ‘De oudengelse Aldhelmglossen in HS. 1650 van de koninklijke Bibliotheek te Brussel’, Handelingen IX der Zuidnederlandse Maatschappij voor Taal-en Letterkunde en Geschiednis 1955, 37–50Google Scholar; Goossens, , pp. 5–8, 28–32, 37–52.Google Scholar

²¹ The volume once belonged to a single codex of 242 folios comprising three items, but now broken up into four fragments. In addition to the Brussels portion there is: (a) Antwerp, Plantin-Moretus Museum MS 47 (Salle iii.68) + (b) London, British Library, Add. 32246 [Ker 2]; (c) Antwerp, Plantin-Moretus Museum MS 190 (Salle iii.55) [Ker 3]. The flyleaves of these manuscripts contain additions of s. xi¹, which Ker used to localize the composite codex to Abingdon. Ker relied most heavily on forty-nine elegiac couplets by an unknown Herbert (of Fleury?) to Wulfgar, identified as abbot of Abingdon 989–1016 (ptd. Dümmler, E., ‘Lateinische Gedichte des neunten bis elften Jahrhunderts’, Neues Archiv 10 (1884), 351–3Google Scholar). Ker insisted that ‘[s]cholia to the Excerptiones, two supply-leaves … the verses [to Wulfgar] and [two] glossaries … are in a pointed, slightly forward-sloping hand which occurs also in [Brussels] 1650’ (Ker, , p. 3Google Scholar). In other words, the earliest stratum of glosses in Brussels 1650 was written in the same hand as the verses to Wulfgar. Speculating that the elegiac verses were addressed to Wulfgar of Abingdon and further theorizing that the eleventh-century provenance of this manuscript was Abingdon, Ker could reach only one conclusion concerning the Brussels glosses: unless the scribe travelled from elsewhere, all of them must have been added at Abingdon. Unfortunately, Ker's reasoning not only contradicts the palaeographical record (cf. Dumville, , English Caroline, pp. 102, 136 and 154Google Scholar) but also makes no sense in the textual history of Pdv manuscripts which I present here.

²² Goossens discusses the palaeography in his volume (pp. 45–52)Google Scholar, and I have not recorded all of his observations here. Hands A, B, C, CD and R appear in chronological order, except for A and B. Scribes A and B never gloss the same words, so their relative chronology cannot be gauged by this formal method. Ker describes these hands in his Catalogue entry.

²³ Goossens, , p. 51.Google Scholar Goossens assumes that the scribes copied an exemplar.

²⁴ Cf. Derolez, R., ‘Zu den Brüsseler Aldhelmglossen’, Anglia 74 (1957), 153–80.Google Scholar

²⁵ Goossens, , p. 48.Google Scholar

²⁶ Ker, , p. 7Google Scholar; Goossens, , p. 50.Google Scholar

²⁷ Cf. Warner, G. and Gilson, J., Catalogue of Western Manuscripts in the Old Royal and King's Collections (London, 1921), p. 136.Google Scholar

²⁸ These features are, of course, characteristic of the script developed by the Canterbury scribe Eadwig Basan and practised at Christ Church from the second quarter of the eleventh century onwards. Eadwig's script appears in the Hanover Gospel Book (Hanover, Kestner-Museum, W. M. XXIa, 36) and in London, BL, Cotton Vespasian A. i, pt ii. By c. 1040 this Canterbury model was imitated in major scriptoria across southern England and formed the basis for the Exeter script of Leofric's scriptorium. Even so, there are appreciable differences between the Exeter script and Eadwig's experimentations: Canterbury books often have a Canonical Capital display script, compressed lower member on a, and the tendency, apparent early at Tours, to link the upper fillips of letters across the bounding line.

²⁹ Bishop, T. A. M., ‘Notes on Cambridge Manuscripts, Part III: MSS. Connected with Exeter’, Trans. of the Cambridge Bibliographical Soc. 2 (1954–1958), 192–9, at 198–9.Google Scholar

³⁰ Drage, E., ‘Bishop Leofric and the Exeter Cathedral Chapter 1050–1072: a Reassessment of the Manuscript Evidence’ (unpubl. D.Phil, dissertation, Oxford Univ., 1978), p. 372.Google Scholar

³¹ The spacing may be coincidental, however, insofar as many Exeter manuscripts have widely separated lines, a characteristic of Eadwig Basan's innovative script; cf. Cambridge, Corpus Christi College 191 (Exeter, s. x3/4).Google Scholar

³² Drage, , ‘Bishop Leofric’, p. 372.Google Scholar

³³ For the Brussels manuscript, Louis Goossens has discerned a number of hands which he terms ‘A’, ‘B’, ‘C’, ‘CD’ and ‘R’. I adopt Goossens's inventory, which is represented in my edition by superscript lower-case letters. Thus, B^c denotes scribe C in the Brussels manuscript. The reference numbers in each gloss citation refer to the number of the lemma in my edition: 13934, the final lemma glossed in any annotated copy of Pdv, indicates the sequential number of that lemma in the list of lemmata. Reference numbers followed by a lower-case letter are simply additions to the corpus, either discovered after the edition was complete or simply corrected.

³⁴ The term ‘group’ went undefined in Napier's book. It seems to mean any manuscripts having a majority of their Old English glosses in common with one another. The evidence of shared error figures only in Napier's ‘Salisbury group’.

³⁵ Napier, , p. xxiii.Google Scholar

³⁶ Ibid. p. xxiv.

³⁷ Ibid. pp. xxiv-xxv.

³⁸ Ibid. p.xxiv.

³⁹ Ibid. p. xxv.

⁴⁰ Ibid.

⁴¹ Glosses 4369 and 8975. Napier found most of these errors by collating Old English glosses in Royal 6. B. VII.

⁴² Altered from stœpa B^a.

⁴³ Some of Napier's putative errors, in particular the grammatical ones, do not illuminate the transmission. In 4637 the lemma SPIRITVS must be genitive singular (cf. Ehwald 258.4), although both Old English glosses are nominative or accusative plural. Similarly, the lemma CVNICVLI (7806) must be nominative (Ehwald 279.25), but it has been glossed as if it were genitive. These semantic errors might have originated with the earliest Pdv glossator, and they are not incontestably spurious. Gloss 7959 is likewise an ambiguous indicator of dissemination, for unparallel higde might have arisen in a manuscript reading NITEBATVR. Higde only appears as an error because of Digby's plural verb. More valuable are 11787 and 11827, although baþana for baþa and 11827 stœþena for staþa could arguably represent morphological changes from strong to weak nouns. Such errors are not necessarily deviations, if morphological.

⁴⁴ Napier, , p. xxiv.Google Scholar

⁴⁵ Ibid.

⁴⁶ eadpearflicum Ο.

⁴⁷ w altered from f B^c.

⁴⁸ Napier, , p. xxiv.Google Scholar

⁴⁹ Because he relied on Bouterwek's edition of the Brussels glosses (Bouterwek, C., ‘Angelsächsische Glossen (1): die ags. Glossen in dem Brüsseler Codex von Aldhelms Schrift De Virginitate’, Zeitschrift für deutsches Alterthum 9 (1853), 401–530)Google Scholar, Napier's examples corresponding to 3014 and 4065 are no longer valid. Gloss 1169 has been shifted from the section dealing with better readings in B against those in Ο; it is erroneously included there.

⁵⁰ Similarly 2499, 2969, 3847, 7627.

⁵¹ See below, p. 151.Google Scholar

⁵² ‘Napier's “Digby group” may suggest that [Digby 146] is the central member of the group, which is not the case. “Abingdon group” is a better term because both [Brussels 1650] and [Digby 146] are Abingdon MSS., whereas [Royal 6. B. VII] and [Hereford P. I.17] are closely connected with them’ (p. 22).Google Scholar

⁵³ Ibid. pp. 23–5.

⁵⁴ ‘Latin and Old English Aldhelm Glosses: a Direct Link in the “Abingdon Group”’, Anglo-Saxon Glossography, ed. Derolez, R. (Brussels, 1992), pp. 141–9.Google Scholar

⁵⁵ Ibid. p. 27. Goossens had earlier qualified his own prodigious efforts: ‘In nearly all cases it will be impossible to establish direct relationships among the Old English glosses: it is good to remember in this respect that for each [manuscript] which has come down to us several others must have been lost in the course of time’ (p. 21).Google Scholar

⁵⁶ Cf. ‘De oudengelse Aldhelmglossen’.

⁵⁷ Only B has this variant.

⁵⁸ Ker, , pp. 382–3.Google Scholar

⁵⁹ Goossens, , p. 26.Google Scholar

⁶⁰ Ibid.

⁶¹ Ibid.

⁶² Goossens, (p. 48) claimed that these do not pertain to the text. In fact, the lemmas to these legitimate glosses appear throughout.Google Scholar

⁶³ Goossens, , p. 23.Google Scholar

⁶⁴ Goossens nowhere cites Drage's work, and the dates of the gloss-hands in the manuscripts (cf. Goossens, , p. 17Google Scholar) rest on outdated descriptions. Moreover, his argument calls for justification on grounds of provenance. If glosses from Brussels 1650 were copied into Royal 6. B. VII, either Brussels 1650 must have been at Exeter or the Royal scribe must have travelled to the foundation housing the Brussels volume, either Canterbury or Abingdon. Neither hypothesis can be verified from known evidence.

⁶⁵ Goossens, , pp. 23–4Google Scholar; idem, ‘Aldhelm Glosses’, p. 142.Google Scholar

⁶⁶ Goossens, , ‘Aldhelm Glosses’, p. 142.Google Scholar

⁶⁷ Ibid. p. 143.

⁶⁸ Ibid. p. 144.

⁶⁹ Ibid. p. 23.

⁷⁰ Cf. Goossens, , p. 24Google Scholar: ‘The agreement between [Royal 6. B. VII] and [Brussels 1650] is much greater here than for the Old English glosses, especially as regards the number of gll. that were taken over. On folio 5R e.g., the glossator of Royal 6. B. VII took over only one OE gl. against 70 Latin gll., on folio 6R no less than 87 Latin gll., but not a single OE gl.’; cf. Goossens, , ‘Aldhelm Glosses’, p. 142.Google Scholar

⁷¹ Goossens, , ‘Aldhelm Glosses’, p. 147.Google Scholar

⁷² Ibid. p. 148.

⁷³ Not to mention: 3927, 3984, 4005, 4333, 4700, 4715, 4763, 4792, 4952, 5224, 5351, 5409, 5554, 5590, 5616, 5734, 5957, 5991, 5993, 6032, 6073, 6099, 6507, 6594, 6668, 6863, 6926, 6990, 7403, 7411, 7525, 7577, 7683, 7738, 7908, 8070, 8402, 8458, 8495, 8578, 8663, 8824, 9067, 9277, 9368, 9807, 9841, 9939, 9987, 10251, 10289, 10480, 10681, 10811, 10852, 10921, 11044, 11816, 12084, 13564, 13575 and 13808.

⁷⁴ Corrected to sermonum B

⁷⁵ For torto uel curuo.

⁷⁶ Altered to curuo B^c

⁷⁷ tropheti altered to trophea Ο.

⁷⁸ And possibly 6439 SVSPICIONVM] iudiciarum R5: iudiciorum B^c: indiciarum Ο.

⁷⁹ Goossens finds categorical alterations in these glosses, but there is no reason to consider these here.

⁸⁰ Altered to nowendes by scribe CD.

⁸¹ topreomum B^a.

⁸² Goossens, , pp. 24–5.Google Scholar

⁸³ Goossens, , ‘Aldhelm Glosses’, p. 149:Google Scholar Obviously, if my assumptions are correct, CD must here be identical with the ‘glossator of [R5]’.

⁸⁴ The hypothesis also explains what Goossens claimed was a ‘striking difference in the way [the Royal glossator] treats the glosses by C as compared with those by A and B’ (Goossens, , ‘Aldhelm Glosses’, p. 146Google Scholar). Glosses in Brussels HANDS A and B appeared ‘changed [in Royal 6. B. VII] more often than not’ (ibid. p. 147) than glosses in HANDC. The corpus of glosses in HANDS A and B circulated in many other manuscripts, and probably in many divergent spellings. Such glosses might have gone unaltered by the scribe who copied from the Digby apograph.

⁸⁵ I, Gwara, pp. clvii–cxciii.Google Scholar

⁸⁶ Ehwald, 238.14: ‘…[he] is made the leader by the effort of his labour’.Google Scholar

⁸⁷ Ehwald, 252.9–10: ‘…[that] they spurned the rich delights of regal feasts in their innocent youth’.Google Scholar

⁸⁸ For sigla, see above, p. 140, n. 7.Google Scholar

⁸⁹ Altered from pearricum B^a.

⁹⁰ I altered.

⁹¹ Cf. Bodden, M. C., ‘Evidence for Knowledge of Greek in Anglo-Saxon England’, ASE 17 (1990), 217–46.Google Scholar

⁹² Conner, P., Anglo-Saxon Exeter: a Tenth-Century Cultural History (Woodbridge, 1993), p. 30.Google Scholar

⁹³ Ibid. pp. 3–8, items 12, 15, 17, 28, 31–3, 46 and 48–9.

⁹⁴ Gwara, , ‘Manuscripts’, p. 157.Google Scholar

⁹⁵ Brooks, N., The Early History of the Church of Canterbury (Leicester, 1984), pp. 55–6.Google Scholar

⁹⁶ By William of Malmesbury's testimony; cf. Hamilton, , Gesta pontificum, pp. 423–5.Google Scholar

⁹⁷ The preceding work derives from my Ph.D. dissertation (cited above), which incorporates an edition of all Old English and Latin glosses to the Prosa de virginitate, fully lemmatized, from the fourteen glossed manuscripts of English provenance. I owe a considerable debt to Professors A. G. Rigg and Michael Herren, who corrected innumerable mistakes and challenged me to refine points I might otherwise have left vague. I have benefited, too, from Professor David Dumville's expert advice on palaeography and dating. I am also grateful to the Dictionary of Old English, University of Toronto, for access to their microfilm manuscript archive. The Centre for Medieval Studies, University of Toronto, and the Associates of the University awarded me research grants to consult most of the Pdv manuscripts in situ.