We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 26 September 2008
Farað witodlice and Iærað ealle þeoda and fulligeaþ hig on naman fæder and suna and þæs halgan gastes, and lærað þæt hig healdon ealle þa ðing þe ic eow behead (Matt. XXVIII.19–20).
With these words at the end of the gospel according to Matthew, Jesus sends out his disciples to spread the words and deeds of the Lord to all peoples. With respect to the Anglo-Saxons, this order was impressively executed by the earliest translation of the Vulgate gospels into a vernacular, the West Saxon Gospels (WSG). This text, from the late tenth or early eleventh century, survives in four complete manuscripts (A, B, C, Cp) and two fragments (F, L) from the Old English period and two complete manuscripts from the late twelfth century (R and its copy H).
1 The major editions of the WSG are The Holy Gospels in Anglo-Saxon, Northumbrian and Old Mercian Versions, Synoptically Arranged, with Collations Exhibiting all the Readings of all MSS., ed. Skeat, W., 4 vols. (Cambridge, 1871–1900)Google Scholar and now The Old English Version of the Gospels, ed. Liuzza, R. M., EETS os 304 (Oxford, 1994).Google Scholar
2 For the sigla, see the Appendix, below, pp. 175–8. The sigla for the manuscripts of the WSG correspond to those chosen by Skeat and Liuzza, with the exception of F for the ‘Yale Fragments’ (‘Y’ in Liuzza's edition).
3 For detailed descriptions, see Ker, N. R., Catalogue of Manuscripts containing Anglo-Saxon (Oxford, 1957Google Scholar; reissued with addenda 1991), no. 20, Liuzza, , The Old English Version, pp. xvii–xxGoogle Scholar, and The West-Saxon Gospels. A Study of the Gospel of Saint Matthew with Text of the Four Gospels, ed. Grünberg, M. (Amsterdam, 1967), pp. 19–28Google Scholar. For the rubrics, see also pl. III.
4 For detailed descriptions of the fragments, see Ker, , Catalogue, no. 1Google Scholar, Liuzza, , The Old English Version, pp. xli–xliiGoogle Scholar and in particular Liuzza, R. M., ‘The Yale Fragments of the West Saxon Gospels’, ASE 17 (1988), 67–82.Google Scholar
5 Pericope (from Greek π↦ρικόπτ↦ιν ‘to cut around’) is the technical term for a liturgical reading of set length. It was only introduced by Protestant theologians of the sixteenth century. In the Middle Ages, the Latin terms were capitula euangelii or sequentia euangelii and meto-nymical euangelium (cf. ModE Gospel (of the day) and German Evangelium). See Lenker, U., Die west-sächsische Evangelienversion und die Perikopenordnungen im angelsächsischen England, Texte und Untersuchungen zur Englischen Philologie 20 (Munich, 1997), 85–93.Google Scholar
6 Grünberg, , The West-Saxon Gospels, p. 369.Google Scholar
7 In his Preface to the edition of 1571, which was prepared under the supervision of Archbishop Parker (by John Joscelyn?), John Foxe explicitly links the Old English translation with the aims of the Reformation; see also Liuzza, , The Old English Version, pp. xiii–xv.Google Scholar
8 The rubrics to the WSG have attracted very little scholarly attention; the most important investigations are Tupper, F., ‘Anglo-Saxon Diæg-mæl’, PMLA 10 (1895), 111–241CrossRefGoogle Scholar, the notes to Bright's, J. edition of the Gospel according to John (Euangelium Secundum lohannem. The Gospel of Saint John in West-Saxon (Boston, 1904), pp. 115–82Google Scholar) and Frere, W. H., Studies in Early Roman Liturgy: 2. The Roman Gospel-Lectionary, Alcuin Club Collection 30 (Oxford, 1934), 221.Google Scholar
9 For the relationship of the manuscripts, see Liuzza, , ‘The Yale Fragments’, pp. 75–80Google Scholar, Liuzza, , The Old English Version, pp. xliii–lxxiiiGoogle Scholar, and Lenker, , Perikopenordnungen, pp. 23–7.Google Scholar
10 Liuzza, , The Old English Version, pp. lvi–lvii.Google Scholar
11 The relationship of R and H to the other copies is fairly clear – R is a copy of H, H itself a copy of B.
12 Liuzza, , The Old English Version, p. lviii.Google Scholar
13 The Kentish origin is obvious, for example from the spellings <yo> instead of West Saxon <eo> for /e:o/ (e.g. dyofel, syocnessa (Mark 1.39)) and forms in -an, -ade and -ad (geclœnsad Mark 1.42) instead of West Saxon -on, -ode and -od (see Campbell, A., Old English Grammar (Oxford, 1959), §§ 297 and 757Google Scholar). Anwald without breaking is attested in Mark 1.27. See also Liuzza, , ‘The Yale Fragments’, p. 75.Google Scholar
14 Cf. pl. III.
15 The later addition of liturgical rubrics in lines left blank for this purpose is not unique to A but is also found in a manuscript of Ælfric's Catholic Homilies: ‘At the beginning of each homily the scribes left a space for a large initial capital and also one or two blank lines into which the heading and pericope incipit were later inserted’ (Clemoes, P., ‘Description of the Manuscript’, Ælfric's First Series of Catholic Homilies. British Museum, Royal 7. C. XII, Fols. 4–218, ed. Eliason, N. E. and Clemoes, P., EEMF 13 (Copenhagen, 1966), 22).Google Scholar
16 See below, pp. 147–9.
17 Without any obvious reason the scribe changed his method for the last four rubrics (John XX. 19, XXI.1, XXI.15 and XXI.19), where the Latin follows the Old English.
18 Similarly, two lines are left blank before Matt. XX.29, where the Saturday in the summer Embertide is described in a long phrase: ‘þys sceal on sæternesdæg on þære pentecostenes/wucan to þam ymbrene’.
19 Only the first instances in each gospel are cited. The Old English term for the pericope is god-spell, except for the four cases when passio refers to the Passion (Matt. XXVI.2, Mark XIV.1, Luke XXII.1 and John XVIII.1).
20 Homilies of Ælfric. A Supplementary Collection, ed. Pope, J. C., 2 vols., EETS os 259–60 (London, 1967) 1, 230 and 288.Google Scholar
21 For a detailed analysis of the different characteristics such as breaking, palatal diphthongization etc., see Lenker, , Perikopenordnungen, pp. 216–19.Google Scholar
22 Other late features are dative plural forms in -on instead of -um, e.g. wucon (Mark XI.1, Luke XIX.29 and John 1.15) and dagon (Luke XI.5, John XV.12 and XV.17). Cf. also the spelling <mænies> (Matt. XXIV.42) instead of <mæniges> (Luke XII.35) which points towards a vocalization of /j/.
23 Other instances of syncope of the genitive ending are sœtertidœg in Mark IX.2 and Luke III.1.
24 The form ucan is attested five times (Luke XVI.10, XXI.20, John XVI.5, XX.11 and XXI.1). For a similar phenomenon of loss of initial w before u, see the forms uton for wuton (Campbell, , Old English Grammar, § 471Google Scholar). The change is probably triggered by w being a half-vowel, i.e. an unsyllabic u, which was lost before the vowel /u/.
25 Cf. Theodulfi Capitula in England. Die altenglischen Übersetzungen zusammen mit dem lateinischen Text, ed. Sauer, H., Texte und Untersuchungen zur Englischen Philologie 8 (Munich, 1978), 190–1, 377 and 393.Google Scholar
26 The form ucan is found in Luke XVIII. 12 (‘Ic fæste tuwa on ucan’) in the three south-western manuscripts of the WSG, Cp (Bath), C (Malmesbury?) and B (unknown origin); in A, however, the form appears as wucan. Another instance of ucan is recorded in the poem Seasons of Fasting (c. 1000). Sisam attributes this form to the ‘South Midlands’ and comments on its use in Old English: ‘The sound-change is concealed by the conventional spelling which spread with the Late West Saxon Literary dialect. It appears commonly in the Domesday record of 1086 where the traditional spelling is no longer followed’ (Sisam, K., Studies in the History of Old English Literature (Oxford, 1953), p. 52).Google Scholar
27 For a detailed summary of the attestations of þæge and their critical assessment, see Förster, M., ‘Die spätae. deiktische Pronominalform þæge und ne. they’, Anglia Beiblatt 52 (1941), 274–80Google Scholar and ‘Nochmals: ae. þœge’, ibid. 53 (1942), 86–7, and in particular Hofstetter, W., Winchester und der spätaltenglische Sprachgebrauch. Untersuchungen zur geographischen und zeitlichen Verbreitung alteng lischer Synonyme, Texte und Untersuchungen zur Englischen Philologie 14 (München, 1987), 563–7.Google Scholar
28 See instances 5, 6, 7 and 8 in Förster, ‘Die spätae. deiktische Pronominalform þœge and exam ples 2, 3 and 4 in Hofstetter, , Winchester und der spätaltenglische Sprachgebrauch, p. 563.Google Scholar
29 The origin and provenance of the manuscript in Exeter is apparent from both palaeographical evidence and the inventory of books procured by Bishop Leofric for the church of Exeter. One copy of this inventory was originally part of A and is now found in a quire prefixed to the Exeter Book (Exeter, Cathedral Library, 3501, fols. 0, 1–7). The manuscript itself can be identified as ‘þeos englisce cristes boc’ in this list; in another copy (Oxford, Bodleian Library, Auct. D. 2.16, fols, iv, 1–6) it is called ‘englisc Cristes boc’. See Lapidge, M., ‘Surviving Booklists from Anglo-Saxon England’, Learning and Literature in Anglo-Saxon England. Studies presented to Peter Clemoes, ed. Lapidge, M. and Gneuss, H. (Cambridge, 1985), pp. 33–89, at 64–9.Google Scholar Leofric's donation is also documented in an entry on 1r: ‘Hunc textum euangeliorum dedit leofricus episcopus ecclesiae sancti petre apostoli in exonia ad utilitatem successorum suorum. Đas boc leofric biscop gef sancto petro and eallum his æftergengum into exancestre gode mid to ðenienne.’
30 For detailed descriptions of the manuscripts, see Liuzza, , The Old English Version, pp. xvii–xlii.Google Scholar
31 Abbreviations: ‘Vul’ refers to the text of the Vulgate (Nestle-Aland. Novum Testamentan Latine, ed. , K. and Aland, B. (Stuttgart, 1982))Google Scholar, ‘WSG to the Old English translation (ed. Liuzza) and ‘A’ to the text of the rubrics in the manuscript A. For the text of Latin lectionaries, the incipits cited in the gospel-lists Qe and Sa were chosen (see the Appendix, below, pp. 175–8).
32 The relevant phrases are marked by italics or bold letters.
33 The exemplar of the Latin incipits in A, however, is as yet unknown. Their comparison with both the incipits attested in extant Anglo-Saxon witnesses and the textual varieties documented in the critical edition of Latin gospelbooks by Fischer (which includes the texts of Roman and Milanese lectionaries) showed too many discrepancies. Their closest textual relative is Paris, Bibliothèque de l'Arsenal, 612 (Rheims, c. 850). See Fischer, B., Die lateinischen Evangelien bis zym 10. Jahrhundert, 4 vols. (Freiburg, 1988–1991).Google Scholar
34 Cf. the chapter-headings before Matt. V.13, V.27, VI.1, VI.7, VI.22, VII.6, VII.13, VII.22, VIII.5, VIII.28, IX.23, IX.35, X.7, X.11, X.22, Mark 1.9,1.14,1.16,1.22,1.29, XII.41 and John III.22, V.24, VI.5 (for VI.1) and VI.37. For chapter-headings in general, see Meyvaert, P., ‘Bede's Capitula Lectionum for the Old and New Testament’, RB 105 (1995), 348–80, at 349–52.Google Scholar
35 Cf. the lack of a proper name or noun as the subject of the sentence (‘egredientes, uenerunt’); the adverb ‘protinus’ is not deleted.
36 These chapter-headings were a very useful tool in a time when there was no chapter and verse division of the gospels. In the gospels according to Luke and John, major parts of the beginning chapters are used as liturgical readings so that the original chapter-headings were replaced by the rubrics.
37 For the blank lines, see pl. III (before Matt. XVII.22 ‘Đa hig wunedon …’). There are twenty blank lines in Mark, twenty-one in Luke, eight in Matthew and only two in John. Only four of them are found in the neighbourhood of chapter-headings.
38 See Grünberg, , The West-Saxon Gospels, p. 21Google Scholar and Liuzza, , The Old English Version, p. xx.Google Scholar
39 See the index of biblical lessons in Lenker, , Perikopenordnungen, pp. 529–35.Google Scholar
40 For general surveys of the subject, see Jungmann, J., Missarum Sollemnia. Eine genetische Erklärung der römischen Messe, 5th ed., 2 vols. (Freiburg, 1962) I, 501–83Google Scholar; C. Vogel, Medieval Liturgy. An Introduction to the Sources, trans, and rev. W. Storey and N. Rasmussen; and Martimort, A.-G., Les Lectures liturgiques et leurs livres, Typologie des sources du moyen âge occidental 64 (Turnhout, 1992), 15–20.Google Scholar
41 For the development of the elements which surround the readings of the gospel of the day, i.e. the chants, collects, the homily and the Credo, see Jungmann, Missarum Sollemnia, I; Young, K., The Drama of the Medieval Church, 2 vols. (London, 1933; corr. repr. 1962) I, 21–9Google Scholar; and Harper, J., The Forms and Orders of Western Liturgy from the Tenth to the Eighteenth Century. A Historical Introduction and Guide for Students and Musicians (Oxford, 1991), pp. 109–26.Google Scholar
42 Evidence for the first reading in the Anglo-Saxon eucharistic service is provided by the Old English gloss pistelrœdere for Latin subdiaconus in the Regularis Concordia (see Die Regularis Concordia und ihre altenglische Interlinearversion, ed. Kornexl, L., Texte und Untersuchungen zur Englischen Philologie 17 (Munich, 1993), lines 798, 801, 1011 etc., cf. p. ccxxxix).Google Scholar For the documents for the first reading, see Frere, W. H., Studies in Early Roman Liturgy: 3. The Roman Epistle Lectionary, Alcuin Club Collection 32 (Oxford, 1935)Google Scholar and for the lack of witnesses from Anglo-Saxon England, see Gneuss, H., ‘Liturgical Books in Anglo-Saxon England and their Old English Terminology’, Learning and Literature in Anglo-Saxon England, ed. Lapidge, and Gneuss, , pp. 91–141, at 110.Google Scholar
43 See Jungmann, , Missarum Sollemnia I, 566–7.Google Scholar For Anglo-Saxon England, the reading of the gospel by the deacon is attested in a number of sources, e.g. the Regularis Concordia: ‘Diaconus uero, antequam ad euuangelium legendum accedat’ glossed by ‘se diacon ær þam to godspelle to rædenne toga’ (Regularis Concordia, ed. Kornexl, , line 803Google Scholar) and Ælfric's pastoral letters:‘Diaconus is gecweden <þegn> se þenað þæm mæssepreoste oþþe þam bisceope æt þære mæssan and godspel ræt’ (Die Hirtenbriefe Ælfrics in altenglischer und lateinischer Fassung, ed. Fehr, B., Bibliothek der angelsächsichen Prosa 9 (Hamburg, 1914; repr. with a Supplement to the Introduction by P. Clemoes, Darmstadt, 1966), 108Google Scholar). Cf. also the sign for the deacon in the Monasterialia Indicia ‘Đonne þu diacon abban wille þonne stric þu ealgelice mid þinum scyte fingre and wyrc cristes mæl on þin heafod foran on þæs halgan godspelles getacnunge’ (Monasteriales [sic] Indicia. The Anglo-Saxon Monastic Sign Language, ed. Banham, D. (Pinner, 1991), p. 48 (no. 124)).Google Scholar
44 The present study is restricted to the reading of the gospel of the day as it appears in the context of the eucharistic liturgy. The epistle readings and the lessons included in the daily Office and the occasional offices demand special treatment as they developed independently; for the books for the daily Office, see Gneuss, , ‘Liturgical Books’, pp. 110–35Google Scholar; for their readings, see Martimort, , Les Lectures liturgiques, pp. 71–103.Google Scholar
45 See the inventory of manuscripts in the Appendix, which is based on Gneuss, H., ‘A Preliminary List of Manuscripts Written or Owned in England up to 1100’, ASE 9 (1981), 1–60Google Scholar and in particular Gneuss, , ‘Liturgical Books’, pp. 91–141.Google Scholar See now also Pfaff, R. W., ‘Massbooks: Sacramentaries and Missals’, The Liturgical Books of Anglo-Saxon England, OEN Subsidia 23 (1995), 7–34.Google Scholar
46 Unfortunately, sacramentaries and missals are neither distinguished in Latin, nor in OE or ModE: the terms missale, mœssebo and missal can denote both books. Cf., for example, the codex known as the ‘Missal of Robert of Jumièges’ (Rouen, BM, 274Google Scholar (Y.6)) which is actually a sacramentary. Cf. Gneuss, , ‘Liturgical Books’, pp. 99–100Google Scholar and Pfaff, , ‘Massbooks’, pp. 7–8.Google Scholar
47 See Gneuss, , ‘Liturgical Books’, pp. 104–6Google Scholar and Hartzell, K. D., ‘Graduals’ in The Liturgical Books of Anglo-Saxon England, ed. Pfaff, , pp. 35–8.Google Scholar
48 See Gneuss, , ‘Liturgical Books’, p. 108Google Scholar and, in addition, eight notes in Cambridge, University Library, Kk. 1. 24 (Nc) and four in the Stonyhurst Gospel(Nk).
49 The sigla are chosen according to a system in which the first letter of each siglum shows what kind of source the document is.
50 See Gneuss, , ‘Liturgical Books’, p. 109Google Scholar and, in addition, the capitularies in Cambridge, Fitzwilliam Museum, 45–1980 (Pc), Hanover, Kestner-Museum, WM XXIa. 36 (Sx), London, BL, Loan 11 (Sc), New York, Public Library, 115 (Qc) and the fragment of a capitulary in Oxford, Bodleian Library, Bodley 381, fols, i–ii (Px). Another extremely important old witness, the epistle- and gospel-list in Würzburg, Universitätsbibliothek, M. p. th. f. 62 (Oa), is included, although its Anglo-Saxon origin is not certain (see below, n. 79). Two lists in Anglo-Saxon man uscripts mentioned by Gneuss (Besançon, BM, 14; no. D. 21) and Klauser (Antwerp, Plantin-Moretus Museum, lat. 194) are post-Anglo-Saxon and were added to these gospelbooks in the twelfth or thirteenth century (for the latter, see Klauser, T., Das römische Capitulare Evangeliorum. Texte und Untersuchungen zu seiner ältesten Geschichte. 1. Typen, Liturgiegeschichtliche Quellen und Forschungen 28 (Münster, 1935), p. xxxviii (no. 11)).Google Scholar
51 For the transmission of the Old Testament in general and for liturgical notes in Old Testament manuscripts, see Marsden, R., The Text of the Old Testament in Anglo-Saxon England, CSASE 15 (Cambridge, 1995), 40–1.Google Scholar
52 For the general development, see Frere, , The Roman Gospel-Lectionary, pp. iii–iv and 59–61Google Scholar; Klauser, , Capitulare Evangeliorum, pp. x–xxiiGoogle Scholar; Martimort, , Les Lectures liturgiques, pp. 15–58Google Scholar; Vogel, , Medieval Liturgy, pp. 293–304 and 315–20Google Scholar, and Lenker, , Perikopenordnungen, pp. 94–132.Google Scholar
53 The notes to the capitula-tables in the ‘Lindisfarne Gospels’ are ed. Skeat, , The Holy Gospels, pp. 16–22Google Scholar (Matt.), 2–5 (Mark), 3–11 (Luke) and 3–8 (John). They are described and analysed by Brown, T. J., ‘The Latin Text’, Euangeliorum quattuor codex Lindisfarnensis Musei Britannici codex Cottonianus Nero D.IV permissione Musei Britannici totius codicis similitude expressa. Book I. Part II: The Latin Text, ed. Kendrick, T. D., Brown, T. J. et al. (Olten, 1960), pp. 35–6Google Scholar and his edition of The Stonyhurst Gospel of Saint John, Roxburghe Club (Oxford, 1969), pp. 25–7.Google Scholar
54 These ‘quasi-capitularies’ are merely tables of names of church festivals. They are not arranged in the liturgical order and do not indicate which pericopes they refer to. They are thus useless as a means of finding the gospel of the day. Only comparison with the marginal notes to ‘Burchard Gospels’ allows their allocation to certain liturgical texts. For more detailed descriptions, see Brown, , ‘The Latin Text’, pp. 35–8Google Scholar, Chapman, J., Notes on the Early History of the Vulgate Gospels (Oxford, 1908), pp. 52–63Google Scholar and Lenker, , Perikopenordnungen, pp. 136–40 and 387–96.Google Scholar
55 For details, see Klauser, , Capitulare Evangeliorum, pp. xiii–xiv and xxx–xxxviGoogle Scholar, Martimort, , Les Lectures liturgiques, pp. 22–6Google Scholar, Vogel, , Medieval Liturgy, pp. 315–16Google Scholar and Lenker, , Perikopenordnungen, pp. 102–6 and 387–412.Google Scholar
56 An exception to this rule are the notes in the ‘Burchard Gospels’, which are found at the upper margin of the page. The beginning of the pericope in the text is indicated by a cross.
57 In some manuscripts this is due to later bookbinders who, when rebinding the books and trimming the margins, cut off some of the marginal notes; cf. the seventh-century notes in Oxford, Bodleian Library, Auct. D. 2.14 (Ng/Nh) and those in Cambridge, University Library, Kk. 1. 24 (Nc) or London, BL, Add. 40000 (Na).
58 For details, see Frere, , The Roman Gospel-Lectionary, pp. 59–214Google Scholar, Klauser, , Capitulare Evangeliorum, pp. xiv–xviii and xxxvii–lxxGoogle Scholar, Vogel, , Medieval Liturgy, pp. 316–18Google Scholar and Lenker, , Perikopenordnungen, pp. 107–14 and 413–56.Google Scholar
59 Cf. pl. IV.
60 The beginning of a list with the first mass of Christmas or the vigil mass at None of Christmas Eve is one of the characteristics which distinguishes the Roman types 2 and 3 (see below, pp. 160–3). In all strictly Anglo-Saxon witnesses investigated here, one of the Christmas masses is chosen as the beginning of the liturgical year, although the shift to the first Sunday in Advent is commonly dated between the eighth and the tenth century (see e.g. Vogel, , Medieval Liturgy, p. 312Google Scholar and Ælfric's statement in his homily on the Circumcision: ‘Sume ure þeningbec ongin-nað on aduentum domini’ (Ælfric's Catholic Homilies. The First Series. Text, ed. Clemoes, P., EETS ss 17 (London), 1997)), 228)).Google Scholar
61 In the old Roman plan (‘Gregorian sections’) certain saints’ days are chosen as fixed points for counting of the Sundays after Pentecost, namely the feasts of Peter and Paul (29 June), Laurence (10 August), Cyprian (14 September) or Michael (29 September). Accordingly, the Sundays are called ‘dominica post natale S(s) Apostolorum (Petri et Pauli) Laurentii, Cypriani, Angeli’). For details, see Lenker, , Perikopenordnungen, pp. 72–3, 164–6 and 506.Google Scholar
62 Thus Tb (Oslo, Riksarkivet, Lat. fragm. 201 + Universitetsbiblioteket, Lat. frag. 9) contains only parts of the Passion according to Matthew without reference to the liturgical day.
63 For details, see Frere, , The Roman Gospel-Lectionary, pp. 214–20Google Scholar, Gneuss, , ‘Liturgical Books’, p. 107Google Scholar and Lenker, , Perikopenordnungen, pp. 115–17 and 457–76.Google Scholar
64 London, BL, Stowe 944 (Va), Warsaw, Biblioteka Narodowa, i. 3311 (Vc) and London, College of Arms, Arundel 22, fols. 84 and 85 (Vf).
65 See the descriptions in the manuscript catalogues and hence the classification in Gneuss, , ‘Liturgical Books’, p. 107.Google Scholar
66 The fragments Malibu, John Paul Getty Museum 9 (Ud), in spite of the liturgical formulas in ilio tempore, do not belong to this group, as their text can be shown to follow the pure Vulgate tradition. See Lenker, , Perikopenordnungen, pp. 463–4Google Scholar and for a manuscript similar in textual character, London, BL, Royal I.A. XVIII (Ua). E. C. Teviotdale has pointed out to me that more Vulgate manuscripts of this kind have survived.
67 Private devotion seems most likely in the case of Oxford, Bodleian Library, Lat. liturg. f. 5 (S. C. 29744), the ‘St Margaret's Gospels’, which belonged to Margaret of Scotland, wife of Malcolm III (1057–93).
68 For details, see Gneuss, , ‘Liturgical Books’, pp. 99–102Google Scholar, Pfaff, , ‘Massbooks’, pp. 7–34Google Scholar and Lenker, , Perikopenordnungen, pp. 118 and 477–92.Google Scholar
69 For the suggestion that Bishop Leofric himself was the executing scribe of the marginal notes, see Drage, E. M., ‘Bishop Leofric and the Exeter Cathedral Chapter 1050–1072: a Reassessment of the Manuscript Evidence’ (unpubl. DPhil dissertation, Univ. of Oxford, 1978), pp. 139–41.Google Scholar
70 See Wormald, F., ‘Fragments of a Tenth-Century Sacramentary from the Bindings of the Winton Domesday’, Winchester and the Early Middle Ages. An Edition and Discussion of the Winton Domesday, ed. Biddle, M. (Oxford, 1976), pp. 541–9Google Scholar and Lenker, , Perikopenordnungen, pp. 487–8.Google Scholar
71 This inventory was compiled from Klauser, , Capitulare Evangeliorum, pp. xx, xxxv, xxxvi, xxxvii, lxxi, lxxxi, xc and cxiv.Google Scholar The Anglo-Saxon witnesses are included in this list. Altogether, Klauser lists and briefly describes about 1300 medieval manuscripts from the seventh century to the fifteenth (pp. xxx–cxx).
72 The items in parentheses are witnesses which are not capitularies in the strict sense and thus appear under the rubric ‘notes’ in my diagram of Anglo-Saxon sources, such as the ‘quasi-capitularies’ in London, BL, Cotton Nero D. iv (Mx) and London, BL, Royal 1. B. VII (My).
73 Klauser, , Capitulare Evangeliorum, pp. xiii–xiv.Google Scholar
74 The rubrics in A and F might be added to this column, as they share more similarities with notes than any other method. See below, p. 170.
75 Cf. the ratio of gospelbooks in two volumes of the ‘Survey of Manuscripts Illuminated in the British Isles’ (Temple, E., Anglo-Saxon Manuscripts 900–1066 (London, 1976)Google Scholar and Kauffmann, M., Romanesque Manuscripts 1066–1190 (London, 1972))Google Scholar: ‘Twenty-nine out of the 106 catalogue entries [in Temple], over a quarter, are gospelbooks or Gospel lectionaries, whereas in the post-Conquest volume in the same series the 106 entries include only seven gospelbooks’ (Heslop, T. A., ‘The Production of de luxe Manuscripts and the Patronage of King Cnut and Queen Emma’, ASE 19 (1990), 151–95, at 152).Google Scholar
76 Moreover, lectionaries recording non-Roman traditions are quite numerous from the sixth, seventh and eighth centuries; for an inventory, see Vogel, , Medieval Liturgy, pp. 320–3Google Scholar and Martimort, , Les Lectures liturgiques, pp. 37–9.Google Scholar Strikingly, most of them are palimpsests.
77 See above, pp. 149–50.
78 See below, p. 160.
79 From Anglo-Saxon England, the only trace of such a document is the epistle-list in Oa (Würzburg, Universitätsbibliothek, M. p. th. f. 62). The Anglo-Saxon origin of this manuscript is, however, highly disputed, so that Gneuss, ‘Preliminary List’, does not include it. See Lowe, , CLA IX, no. 1417Google Scholar and Bischoff's summary: ‘Es besteht also durchaus die Möglichkeit einer Abschrift durch einen Boten oder Begleiter Burghards in Rom neben der Herkunft aus England’ (Bischoff, B. and Hofmann, J., Libri Sancti Kyliani. Die Würzburger Schreibschule im VIII. und IX. Jahrhundert (Würzburg, 1952), p. 96 (no. 10)).Google Scholar
80 For an inventory of Anglo-Saxon bible manuscripts, see Marsden, , The Text of the Old Testament, pp. 40–1.Google Scholar
81 The context in the will of Ælfwold, bishop of Crediton (997–1016), is particularly instructive: ‘… and in to Crydian tune þreo þeningbec: mæsseboc, and bletsungboc and pistelboc’. Masseboc here refers to the sacramentary (see Lapidge, , ‘Surviving Booklists’, p. 56Google Scholar and Gneuss, , ‘Liturgical Books’, pp. 99–101).Google Scholar
82 Lapidge, , ‘Surviving Booklists’, pp. 33–89.Google Scholar That the booklists provide helpful evidence can be shown by the fact that the numbers and ratios tally with the extant codices in the case of sumptuous books such as psalters, gospelbooks and bibles: while psalters are listed in eleven, and one or more gospelbook(s) are recorded in nine of them, we find only one instance of a bible.
83 Monasteriales [sic] Indicia, ed. Banham, , p. 24 (no. 10).Google Scholar
84 Banham's translation ‘(because one reads) the word of God …’ (ibid.) is misleading. It would translate a genitive construction (Codes spet) but not the compound godspel found here.
85 For a first suggestion in this direction, see Gneuss, , ‘Liturgical Books’, p. 110.Google Scholar
86 For the similar phrases in the sign for the deacon who reads the gospel of the day, see above, n.43.
87 Hirtenbriefe, ed. Fehr, , 13Google Scholar, § 52 (MS O): ‘He sceal habban eac þa wæpna to þam gastlicum weorce, ær-þan-þe he beo gehadod, þæt synd þa halgan bec:saltere and pistolboc, godspellboc and mæsseboc, sangboc and handboc, gerim and pastoralem, penitentialem and rædingboc.’ For minor variations in other copies and the difficulties with the identification of the manuscripts, see ibid. pp. lxxxvi-lxxxvii and 126–7, § 157 and Gneuss, ‘Liturgical Books’, p. 121.
88 Hirtenbriefe, ed. Fehr, , p. 51, § 137Google Scholar: ‘Presbyter debet habere etiam spiritalia arma, id sunt diuinos libros, scilicet missalem, lectionarium, quod quidam uocant epistolarium…’
89 Most of the extant lectionaries have therefore only survived as palimpsests (see above, n. 76) or as fragments used as binding-strips or flyleaves in later manuscripts. See Lenker, , Perikopenordnungen, pp. 457–8.Google Scholar
90 Hartzell similarly argues for the transmission of missals: ‘That few missals survive must not convince us they were not written. The missal is, par excellence, the missile of the expanding church, the missionary church, the crusading church, the church in motion…’ (Hartzell, K. D., ‘An Eleventh-Century English Missal Fragment in the British Library’, ASE 18 (1989), 45–97, at 46).Google Scholar
91 See Klauser, , Capitulare Evangeliorum, pp. x–xiiiGoogle Scholar, Martimort, , Les lectures liturgiques, pp. 15–20Google Scholar, Vogel, , Medieval Liturgy, pp. 293–304Google Scholar and, for the early development, Dijk, S. van, ‘The Bible in Liturgical Use’, The Cambridge History of the Bible. 2: The West from the Fathers to the Reformation, ed. Lampe, G. (Cambridge, 1969), pp. 220–51, esp. 225Google Scholar, and Willis, G., St Augustine's Lectionary, Alcuin Club Collections 44 (London, 1962).Google Scholar
92 For the development of the liturgical year, see Vogel, , Medieval Liturgy, pp. 304–14Google Scholar, Harper, , The Forms and Orders of Western Liturgy, pp. 45–57Google Scholar and Lenker, , Perikopenordnungen, pp. 60–80.Google Scholar
93 The term ‘Temporale’ refers to the Sundays and ferial days as they run throughout the liturgical year without any regard for their date or interruption by saints’ days. The Sanctorale entries contain information about the lessons for days of individual saints or for the Commune Sanctorum.
94 Examples are taken from the inventories for the items for the days of the Temporale (#) and the Sanctorale (‡) in Lenker, , Perikopenordnungen, pp. 298–351 and 352–83.Google Scholar The affiliation of a reading to different Roman traditions is recorded by the symbols ° (type 2) and ▪ (type 3). See below, pp. 161–3.
95 The end of the lesson cannot be determined here as the closing words of the pericope are recorded neither in the capitula-tables (Mx, My) nor in the marginal notes to the ‘Burchard Gospels’ (Mv). See above, pp. 152–3 and n. 95.
96 For the late introduction of the Thursdays in Lent, see below, pp. 167–8.
97 The second reading is usually introduced by Item aliud. This method is chosen by the common ancestor of Sa, Sb, Sc, Sd and Se for liturgical days whose readings show variation in the tenth and eleventh centuries. See Lenker, , Perikopenordnungen, pp. 445–6.Google Scholar
98 The number of these Sundays changes according to the date of Easter, which can fall between 22 March and 25 April. Hence their number varies between one to six after Epiphany and twenty-three to twenty-eight after Pentecost. See the tables in Vogel, , Medieval Liturgy, pp. 404–10Google Scholar and Lenker, , Perikopenordnungen, p. 505.Google Scholar
99 Chavasse, A., ‘Les plus anciens types du lectionnaire et de l'antiphonaire romains de la messe. Rapports et date’, RB 62 (1952), 1–91.Google Scholar The classification is based on Hesbert, R., ‘Les èvangiles des dimanches après la pentecôte’, Le Codex 10673 de la Bibliothèque Vaticane. Fonds Latins. Graduel bénéventaine (XI' siècle), Paléographie musicale 14 (Tournai, 1931), 129–44.Google Scholar
100 For the complicated history of different types of sacramentaries and their designations, see Gamber, K., Sakramentartypen. Versuch einer Gruppierung der Handschriften und Fragmente bis zur Jahrtausendwende, Texte und Untersuchungen 49–50 (Beuron, 1958)Google Scholar, and Pfaff, , ‘Massbooks’, pp. 8–9.Google Scholar
101 The ‘Comes’ is ed. Wilmat, A., ‘Le Comes de Murbach’, RB 30 (1913), 25–69Google Scholar; for a description, see also Vogel, , Medieval Liturgy, p. 347.Google Scholar
102 Frere, The Roman Gospel-Lectionary and Klauser, Capitulare Evangeliorum. Investigation into the origin and organization of the liturgical readings has been pursued from the mid-nineteenth century; see the now dated studies by Ranke, E., Das kirchliche Pericopensystem aus den ältesten Urkunden der Römischen Liturgie dargelegt und erläutert (Berlin, 1847)Google Scholar and Beissel, S., Entstehung der Perikopen des Römischen Meβbuches, Ergänzungshefte zu den Stimmen aus Maria Laach 96 (Freiburg, 1907; repr. Rome, 1967).Google Scholar
103 See Chavasse, , ‘Les plus anciens types’, p. 6Google Scholar: ‘Ces trois variétés [Klauser's II, Λ, Σ] se distinguent l'une de l'autre par leur sanctoral plus ou moins riche, mais leur temporal demeure identique, à peu de choses près, et c'est pourquoi nous disons qu'elles constituent, en réalité, trois variétés d'un même type fondamental.’
104 Frere's analyses are comparable to Klauser's in their concentration on changes in the Sanctorale, but his closer investigations anticipate Chavasse's work. For the ‘Alternative Ferias’, see below, p. 166.
105 See above, n. 79.
106 The eight notes in this gospelbook cannot be attributed with certainty. Their only distinctive feature is the provision of two lessons for the second Sunday after Epiphany (#21)) which also occurs in capitularies of the mixed types (e.g. Pg).
107 See the pericopes for Benedict Biscop (‡1) in Bede's homiliary (Xa), for Cuthbert (‡24, ‡109) in the ‘New Minster Missal’ (Wa) and for Swithun (‡67) in the ‘Red Book of Darley’ (Wh).
108 The only real exception are the seventeen marginal notes (s. viii) in Durham, Cathedral Library, A. II. 16 (Me) which follow an Old Gallican tradition. The notes are ptd Turner, C. H., The Oldest Manuscripts of the Vulgate Gospels. Deciphered and Edited with an Introduction and Appendix (Oxford, 1931), p. 217.Google Scholar
109 Naples is suggested by the lessons for the feasts of St Januarius (‡118, ‡119) and a feast for the main church of Naples (‘dedicatio basilicae Stephani’; ‡301). Lapidge also presumes a Naples origin for ‘In dedicatione sanctae Mariae’ (‡301) and ‘Et in dedicatione fontis’ (‡302); see Bischoff, B. and Lapidge, M., Biblical Commentaries from the Canterbury School of Theodore and Hadrian, CSASE 10 (Cambridge, 1994), 157.Google Scholar
110 For its readings, see Morin, G., ‘Les notes liturgiques de l'Èvangèliaire de Burchard’, RB 10 (1893), 113–26Google Scholar, Brown, , ‘The Latin Text’, 34–43Google Scholar and Chapman, , Notes on the Early History, pp. 52–63.Google Scholar For further details, see Lenker, , Perikopenordnungen, pp. 133–46 and 387–99.Google Scholar
111 For details, see Chapman, , Notes on the Early History, pp. 159–61Google Scholar and Bischoff, and Lapidge, , Biblical Commentaries. 166–7.Google Scholar
112 The actual use of this system has sometimes been doubted because of the difficult and unsystematic recording of the lessons in the capitula-tables and ‘quasi-capitularies’ or, more generally, because of the attested Northumbrian loyalty to Rome, which would not allow the use of a Neapolitan system. These arguments are, however, somewhat anachronistic, as the sources only testify to liturgical improvisation, by their search for the best method of marking pericopes and by the incompleteness of the lessons. For details, see Lenker, , Perikopenordnungen, pp. 141–6.Google Scholar
113 Even its basis should probably rather be described with Chavasse as ‘la vieille organisation romaine’ (Chavasse, , ‘Les plus anciens types’, p. 74, n. 1).Google Scholar
114 The only cases with more or less full agreement are two groups of capitularies (Qa, Qb, Qc and Sa, Sb, Sc, Sd, Se), each of which derives from a common ancestor.
115 For this new series, six readings for the Sundays were added to the original twenty readings of type 2 (Klauser's Σ and Frere's ‘Standard’), to adjust the tradition to the actual number of Sundays after Pentecost. See Chavasse, , ‘Les plus anciens types’, pp. 11–16Google Scholar and Lenker, , Perikopenordnungen, pp. 163–6.Google Scholar
116 The ten readings provided for the Sundays after Epiphany in type 2 are always too many, as the largest possible number of Sundays is six. This means that the need to modify the system was not as urgent as in the case of the too few Sundays after Pentecost – the spare readings were just not used. Thus the capitularies of the mixed types (Ph, Qa, Qb, Qc) give ten Sundays, the group Sa, Sb, Sc, Sd and Se six, and Qe, in accordance with the ‘Comes of Murbach’, five. For details, see Lenker, , Perikopenordnungen, pp. 167–8.Google Scholar
117 Frere, , The Roman Gospel-Lectionary, pp. 119–23.Google Scholar Since Chavasse only describes modifications in the Sundays after Pentecost (and partly Epiphany), the description of the other features is based on the analyses of Frere and the readings in the ‘Comes of Murbach’ (ed. Wilmart).
118 For other days with distinctive readings, see Lenker, , Perikopenordnungen, pp. 168–72.Google Scholar
119 Cf. the readings for the second Sunday in Lent, above p. 160–1.
120 Frere, , The Roman Gospel-Lectionary, p. 65.Google Scholar
121 The selected passages in the older series are Matt. XV.21–8, Luke XVI.19–31, Luke IV.38–44, Luke VII.11–16 and Luke VII.36–50, in the new series John VIII.31–47, V.30–47, VI.27–35, V.17–29 and VII.40–53; cf. the full entry of the fifth Thursday, above p. 161. Maundy Thursday follows a different tradition.
122 Pg shares some features with Frere's type ‘Vitus-4’ (Klauser's Δ), Ph with Frere's type ‘Vitus-15’. An overall similarity is therefore notable; for details, see Lenker, , Perikopenordnungen, pp. 420–8.Google Scholar
123 For details, see ibid. pp. 430–7.
124 These manuscripts only record the augmented series for the Sundays after Pentecost (type 3) or readings for other indistinct days.
125 See below, p. 169.
126 See above, p. 156.
127 From the tenth century, such notes are found, for example, in Oxford, Bodleian Library, Auct. D.2.14 (Ni; s. vi–vii) and Cambridge, University Library, Kk. 1. 24 (Nc; s. viii).
128 On this scribe, see Dumville, D., English Caroline Script and Monastic History: Studies in Benedictinism, A.D. 950–1030, Studies in Anglo-Saxon History 6 (Woodbridge, 1993), 111–40Google Scholar (in particular 120–2) and Pfaff, R., ‘Eadui Basan: Scriptorum Princeps?’, England in the Eleventh Century. Proceedings of the 1990 Harlaxton Symposium, ed. Hicks, C., Harlaxton Medieval Stud. 2 (Stamford, 1992), 267–83.Google Scholar
129 For details, see Lenker, , Perikopenordnungen, pp. 193–5 and 442–56. Se (Cambridge, St John's College 73) is a later copy (after 1081?) of Sd.Google Scholar
130 Heslop, , ‘The Production of de luxe Manuscripts’.Google Scholar
131 For a detailed analysis of the readings and an explanation of the unequal numbers of Temporale and Sanctorale items, see Lenker, , Perikopenordnungen, pp. 263–70.Google Scholar
132 The so-called ‘Synoptic Gospels’ (Matthew, Mark and Luke) are very similar in contents and structure. Matthew and Luke knew and used the Gospel according to Mark and only added additional material from other sources into its general framework. John is different in both structure and contents, so that there are only a few parallel passages.
133 As A does not mark the end of the readings, it can here only be surmised from the parallel passages.
134 There is no full agreement in this case: the parallel passages are Mark X.46–52 and Luke XVIII.35–43 (‘Jesus heals a blind beggar’). A lacks Luke XVIII.31–4 (‘Jesus speaks a third rime about his death’).
135 In Mark there are only twenty-three rubrics compared to seventy-three in Matthew, forty-six in Luke and fifty-seven in John. Seven of those in Mark are parallel passages, so that only sixteen of the rubrics in A are used in the pure Roman tradition.
136 For a group of capitularies which were copied in the German-French area (s. x/xi), see Frere, , The Roman Gospel-Lectionary, pp. 137–9.Google Scholar
137 For a description of the capitulary, see Frere, , The Roman Gospel-Lectionary, pp. 200–1Google Scholar, Klauser, , Capitulare Evangeliorum, p. lxGoogle Scholar (no. 292) and Lenker, , Perikopenordnungen, pp. 277–9.Google Scholar
138 In lectionaries and a number of gospelbooks, liturgical signs are recorded in the Passion pericopes which are read on Palm Sunday (Matt.), Tuesday (Mark), Wednesday (Luke) and Friday (John) of Holy Week. These signs – c (‘celeriter’) for the commentator, s (‘sursum’) for the Jews and t (later †; ‘tenere, trahere’) for Jesus – tell the deacon in which voice and rhythm the part has to be proclaimed. See Young, , The Drama of the Medieval Church, pp. 101 and 550Google Scholar and in particular, Stäblein, B., ‘Passion’, Die Musik in Geschichte und Gegenwart 10 (1962), 886–97.Google Scholar
139 Homilies of Ælfric, ed. Pope, I, 259–85.Google Scholar
140 Ælfric's Catholic Homilies, ed. Clemoes, , p. 525.Google Scholar
141 Ælfric's Catholic Homilies. The Second Series. Text, ed. Godden, M., EETS ss 5 (London, 1979), 56.Google Scholar
142 ‘For the contemporary congregations these translations must have had the advantage of novelty to add to their normal interest, for the corresponding lessons were read in Latin. Their presence alters the proportions and the emphasis of the homilies, not only making it desirable to shorten the exegesis but giving some encouragement to simplified interpretations of which the chief function is to emphasize the direct meaning of the gospel itself’ (Homilies of Ælfric, ed. Pope, I, 150).Google Scholar
143 Drage, , ‘Bishop Leofric’, p. 282.Google Scholar
144 The Old English Version of the Enlarged Rule of Chrodegang together with the Latin Original. An Old English Version of the Capitula of Theodulf together with the Latin Original. An Interlinear Old English Rendering of the Epitome of Benedict of Aniane, ed. Napier, A. S., EETS os 150 (London, 1916), 50 (ch. xlii).Google Scholar
145 The dates and places of origin given in square brackets refer to later added (marginal) notes. ° refers to type 2, ▪ to type 3 of the Roman traditions. See above, pp. 160–3.
