¹ Lindelöf, U., Der Lambeth-Psalter: eine altenglische Interlinearversion des Psalters in der Hs 427 der erzbischöflichen Lambeth Palace Library: I, Text und Glossar, II, Beschreibung und Geschichte der Handschrift; Verhältnis der Glosse zu anderen Psalterversionen; Bemerkungen über die Sprache des Denkmals, 2 vols., Acta Societatis Scientiarum Fennicae 35.i and 43.iii (Helsinki, 1909–1914)Google Scholar. Hereafter cited as ‘Lindeöf’ with volume and page number.

² For descriptions of the manuscript, see Lindelöf II, 1–2, and 14–15; James, M.R. and Jenkins, C., A Descriptive Catalogue of the Manuscripts in the Library of Lambeth Palace (Cambridge, 1932), pp. 588–90Google Scholar; and Ker, N.R., Catalogue of Manuscripts Containing Anglo-Saxon (Oxford, 1957), pp. 432–3Google Scholar (no. 280).

³ Wanley gives no foliation references in his description of the manuscript (Librorum veterum septentrionalium catalogus, in Hickes, G., Linguarum vett. septentrionalium thesaurus, 2 vols. (Oxford, 1705) II, 268–9Google Scholar), presumably because the present foliation had not yet been entered. A century later the present foliation appears in Todd's, H. J.A Catalogue of the Archiepiscopal Manuscripts in the Library at Lambeth Palace (London, 1812), p. 54Google Scholar, who mentions (p. x) that ‘many Codices also have now received the advantage of foliation’.

⁴ This gathering has a fifteenth-or sixteenth-century supply leaf (fol. 78), incorrectly identified by Ker, , Catalogue, p. 343Google Scholar, as ‘12⁶’ instead of ‘10⁶’.

⁵ Incorrectly given as ‘25’ by James, , A Descriptive Catalogue, p. 588.Google Scholar

⁶ Not in Lindelöf, but ed. Förster, M., ‘Die altenglischen Beigaben des Lambeth-Psalters’, ASNSL 132 (1914), 328–35Google Scholar, at 329–31.

⁷ Ed. Förster, ibid. pp. 328–9.

⁸ The Lambeth fragment is ed. Logeman, H., ‘Anglica Minora’, Anglia 11 (1889), 103Google Scholar; Wülker, R., Bibliothek der angelsächsischen Poesie II (Leipzig, 1894), 211–12Google Scholar; and Dobbie, E. van K., Anglo-Saxon Minor Poems, ASPR 6 (New York, 1942), 94–6Google Scholar, who supplies only its variant readings as part of an edition of the full text from London, British Library, Cotton Julius A. ii. Dobbie (p. lxxxvi) thinks that the Lambeth Psalter once contained a complete text (seventy-nine lines) of the prayer, but that ‘a folio or more, containing the missing matter, has been lost after fol. 183’. He notes that the text immediately following, the first canticle, has no rubric and speculates that it may have been a casualty of the same loss. However, since fol. 183 marks the end, and fol. 184 the beginning, of integral gatherings of eight leaves, Dobbie's theory is doubtful.

⁹ Discussed by Korhammer, M., Die monastichen Cantica im Mittelalter und ihre altenglischen Interlinearversionen: Studien und Textausgabe, Texte und Untersuchungen zur Englischen Philologie 6 (Munich, 1976), 239–40Google Scholar, who suggests that it originated in a monastery of southern England. The litany is ed. Lapidge, M., Anglo-Saxon Litanies of the Saints, HBS 106 (London, 1991)Google Scholar, no. XXVII.

¹⁰ Compare the Salisbury Psalter, where at the bottom of 151 v the original litany after the canticles was erased and replaced by a twelfth-century litany; see , C. and Sisam, K., The Salisbury Psalter, EETS os 242 (London, 1959)Google Scholar, 2; this litany is ed. Lapidge, , Anglo-Saxon Litanies of the Saints, no. XLIII.Google Scholar

¹¹ Ker, , Catalogue, no. 281 (p. 343)Google Scholar. These leaves were added to the Lambeth Psalter probably at the same time as (or after) no. 5, as indicated by the presence of their opening words among the fifteenth-century scribblings on 209v.

¹² Dr T. Webber (pers. comm.) would date the hand of 3r to the first quarter of the twelfth century and that of 4r to the late-eleventh/early-twelfth century. Note also the foliation ‘I.10’ (seventeenth-century hand?) at the bottom of 3r, which does not harmonize with either the ‘A. 2’ numbering of the gathering immediately preceding or the ‘B’ numbering of that following.

¹³ Marking, respectively, the first part of a tripartite division of the 150 psalms; the beginning of a series of psalms used for daily Vespers in the Roman and Benedictine Office; and the end of the psalms proper.

¹⁴ As indicated by the individual rubric and by the generous spacing which he allotted to it.

¹⁵ The acephalous Letter, which now begins fol. 1, would require eleven more lines (twelve if a title were included) to be complete. This suggests that at least one folio, with space for text of seven to eight lines on its verso and eighteen lines on its recto, is now missing.

¹⁶ As suggested by the tidy arrangement whereby the psalms begin on the recto of the first folio of the first full quire, with a large decorated initial ‘B’ (of BEATUS, ps. I.1). Such an appropriate position would be hard to arrive at if the scribe had been encumbered by the preceding texts. Furthermore, Jerome's Letter has none of the readings which might be expected from its association in the Lambeth Psalter with a Carolingian recension of the Gallicanum (see below, pp. 147–8); e.g. in its readings ‘uirgulam uiderit’ (Carolingian texts of the Letter invert the order and supply an illustration of uirgulam: ÷); ‘prespexerit’ (Carolingian, ‘perspexerit’), ‘et uobis’ (Carolingian, ‘uobis’), ‘de turbulente’ (Carolingian, ‘e turbulente’). These differences suggest that the Lambeth scribe supplied the Letter from a source other than the exemplar for the psalms, after he had written the latter. For a similar practice with gospels, compare York, Minster Library, Add. 1 (s. x/xi), where the first gathering, containing prefatory (gospel) matter, was added later; and see Mc Gurk, P., Latin Gospel Books from A.D. 400 to A.D. 800, Publications de Scriptorium 5 (Paris, Brussels, Antwerp and Amsterdam, 1961), 8.Google Scholar

¹⁷ The Lambeth Psalter shares roughly the same ratio of length of written space to number of lines with three psalters containing an interlinear Old English gloss. Thus, Lambeth Psalter, 166 (mm.):16 (10.38); Oxford, Bodleian Library, Junius 27, 185:20 (9.25); London, British Library, Royal 2. B. V, 202:19 (10.63); and Stowe 2, 227:20 (11.35).

¹⁸ Which has eighteen lines, rather than sixteen, per page in the same written space, arguably because here there was no need to allow for an Old English gloss.

¹⁹ Lindelöf II, 16–18. For example, relying on Migne's edition of the Gallicanum, where sed tu is the reading for ps. VI.4, Lindelöf labelled the Lambeth Psalter's corresponding reading et tu as a Romanum contamination; the latter, in fact, is also the Gallicanum reading. Likewise, he labelled Lambeth's original reading uirtutem tuam (ps. LXVII.29) as Romanum and the correction uirtuti tuae as Gallicanum, whereas the opposite is true.

²⁰ See , C. and Sisam, K., The Salisbury Psalter, pp. 48–9.Google Scholar

²¹ ‘…ubicumque virgulam [ = obelus] viderit praecedentem, ab ea usque ad duo puncta quae inpressimus sciat in Septuaginta translatoribus plus haberi; ubi autem stellae similitudinem perspexerit, de hebraeis voluminibus additum noverit, aeque usque ad duo puncta …’ (Biblia Sacra iuxta Latinam Vulgatam Versionem ad Codicum Fidem: Liber Psalmorum (Rome, 1953) X, 3–4Google Scholar:‘…wherever he sees the preceding virgule (scil. obelus), let him understand that from there up to the two dots which we have marked represents what the Septuagint has in excess; but where he notices the symbol of the asterisk, let him recognize an addition from the Hebrew books, which likewise extends to the two dots’.

²² See Fischer, B. et al. , Der Stuttgarter Bilder-Psalter (Stuttgart, 1968), p. 265.Google Scholar

²³ Identified by collating the Lambeth Psalter against the Benedictine critical edition of the Gallicanum (Biblia Sacra, as cited above, n. 21). This edition has been criticized for its excessive rigour in omitting doubtful signs; see Fischer, , Der Stuttgarter Bilder-Psalter, pp. 265Google Scholar and 286, n. 31.

²⁴ Many other obeli in the Lambeth Psalter, distinguished from those of the original scribe by their thin hand and frequently by their location above the line and on the margin, are of uncertain date, but probably from the Anglo-Saxon period, since Gallicanum psalters of the twelfth century and later, in keeping with a new, more scientific approach to scriptural studies, tend to demonstrate a conservative attitude towards the use of the critical signs.

²⁵ See below, p. 153.

²⁶ Probably added by the rubricator at the same time as he added the biblical tituli. The latter was done after the completion of the main text, as evident from numerous instances where the tituli had to be continued onto the margin, e.g. pss. V, VIII, XV, because the main scribe did not allow enough space for them.

²⁷ See Schneider, H., Die Altlateinischen Biblischen Cantica, Texte und Arbeiten 29–30 (Beuron, 1938), esp. 46–50 and 75–88.Google Scholar

²⁸ See Mearns, J., The Canticles of the Western Church (Cambridge, 1916), pp. 62–7Google Scholar. Throughout the present paper the numbering of Canticles in the Lambeth Psalter follows that of Mearns.

²⁹ No doubt under the influence of the Benedictine Reform, as evident from textual agreements between the English canticles and those found in later Carolingian psalters. For example, in the Quicumque, they share the readings tertia die resurrexit (v. 38) instead of surrexit of the textus receptus; sedet (v. 39) instead of sedit; Dei patris omnipotentis (v. 39) instead of patris; see Turner, C. H., ‘A Critical Text of the Quicumque Vult’, JTS 11 (1900), 401–11, at 404Google Scholar. In the Te Deum, they agree on uerum el (v. 12) instead of uerum, suscepturus (v. 16) instead of suscepisti; see Julian, J., A Dictionary of Hymnology, 2nd ed. (London 1907), pp. 1120–1.Google Scholar

³⁰ On the relative merits of the two readings, see Julian, , Hymnology, p. 1123b.Google Scholar

³¹ Abbo Floriacensis: Quaestiones Grammaticales, ed. Guerreau-Jalabert, A. (Paris, 1982), p. 263 (§42) and n. 189.Google Scholar

³² On this psalter, see Korhammer, P.M., ‘The Origin of the “Bosworth Psalter”’, ASE 2 (1973), 173–87.Google Scholar

³³ Oxford, Bodleian Library, Douce 59 and Rheims, Bibliothèque Municipale, 4, both ninth-century from France.

³⁴ See C., and Sisam, K., The Salisbury Psalter, p. 5, n. 3.Google Scholar

³⁵ Omitted from the present discussion are the numerous construe glosses (dots and signs) in the Lambeth Psalter, on which see Robinson, F. C., ‘Syntactical Glosses in Latin Manuscripts of Anglo-Saxon Provenance’, Speculum 48 (1973), 443–75, esp. 454–7CrossRefGoogle Scholar; and Korhammer, M., ‘Mittelalterliche Konstrukionshilfen und Altenglische Wortstellung’, Scriptorium 34 (1980), 18–58, esp. 38–9.CrossRefGoogle Scholar

³⁶ Except for twenty-three examples given in Lindelöf II, 2–3.

³⁷ Pace Lindelöf I, 115, n. 3,‘ecclesiae’, written directly over SION, is in the same hand as the Old English gloss.

³⁸ Note that Latin and Old English gloss even correspond in number (singular) and case (genitive).

³⁹ Wieland, G. R., The Latin Glosses on Arator and Prudentius in Cambridge University Library MS Gg. 5. 35 (Toronto, 1983), p. 48.Google Scholar

⁴⁰ Sometimes the Old English equivalent, eala, is also supplied.

⁴¹ Old English did not have a separate inflection for the vocative; see Wieland, , The Latin Glosses, pp. 51–3.Google Scholar

⁴² Cf. Magni Aurelii Cassiodori Expositio Psalmorum I-CL, ed. Adriaen, M., CCSL 97–8 (Turnhout, 1958), 362, lines 348–9Google Scholar, ‘utique in hoc mundo’.

⁴³ Glossaria Latina, ed. Lindsay, W.M. et al. , 5 vols. (Paris, 1926–1931) III, 15.Google Scholar

⁴⁴ ibid. III, 62.

⁴⁵ Corpus Glossariorum Latinorum, ed. Goetz, G., 7 vols. (Leipzig, 1888–1923) IV, 178, line 28, and 570, line 30.Google Scholar

⁴⁶ The possibility that he found all of his material already combined in a single source is fairly remote. Certainly, of the major Latin alphabetical glossaries which have survived from Anglo-Saxon England no one glossary by itself can account for all of the Lambeth glosses. On these glossaries, see Lapidge, M., ‘The School of Theodore and Hadrian at Canterbury’, ASE 15 (1986), 45–72Google Scholar, and Pheifer, J.D., ‘Early Anglo-Saxon Glossaries and the School of Canterbury’, ASE 16 (1987), 17–44.Google Scholar

⁴⁷ A similar type of gloss in London, British Library, Harley 110 is noted by Page, R.I., ‘The Study of Latin Texts in late Anglo-Saxon England [2]: The Evidence of English Glosses’, Latin and the Vernacular Languages in Early Medieval Britain, ed. Brooks, N. (Leicester, 1982), pp. 141–65, at 151Google Scholar.

⁴⁸ It was certainly being copied at this time, as evidenced by London, British Library, Harley 603 and Arundel 155, and Paris, Bibliothèque Nationale, lat. 8824. See further C., and Sisam, K., The Salisbury Psalter, p. 49, n. 1.Google Scholar

⁴⁹ Cf. respectively, Sancti Aurelii Augustini Enarrationes in Psalmos I-CL, ed. Dekkers, E. and Fraipont, J., CCSL 38–40 (Turnhout, 1956), 66, lines 7–8, and 98, line 2Google Scholar; and Cassiodorus, , Expositio Psalmorum, ed. Adriaen, , p. 188, line 1Google Scholar.

⁵⁰ Ed. PL 26, cols. 863–1382.

⁵¹ On Theodore's exegesis, see Devreesse, R., Essai sur Theodore de Mopsueste, Studi e Testi 141 (Vatican City, 1948), 55–78.Google Scholar

⁵² Both ed. De Coninck, L., Theodori Mopsuesteni Expositionis in Psalmos, Iuliano Aeclanensi Interprets in Latinum Versae quae supersunt, CCSL 88A (Turnhout, 1977)Google Scholar. From this edition are taken the citations of Julian's work and its epitome, which follow.

⁵³ It is not possible to determine from these few examples whether the Lambeth Psalter glossator had access to Julian's work or its epitome, or a combination of both, such as is attested in certain Hiberno-Latin manuscripts.

⁵⁴ See Glossa in Psalmos: The Hiberno-Latin Gloss on the Psalms of Codex Palatinus Latinus 68 (Psalms 39:11–151:7), ed. McNamara, M., Studi e Testi 310 (Vatican City, 1986), 48–51.Google Scholar

⁵⁵ First established by Ramsay, R. L., ‘Theodore of Mopsuestia in England and Ireland,’ Zeitschrift für celtische Philologie 8 (1912), 452–97CrossRefGoogle Scholar, at 480–5. See further O'Neill, P. P., ‘The Old English Introductions to the Prose Psalms of the Paris Psalter: Sources, Structure and Composition,’ SP 78 (1981), 20–38, esp. 22–3Google Scholar.

⁵⁶ The former may have passed to the Continent at an early date (see Glossa in Psalmos, ed. McNamara, , p. 18Google Scholar) and in any case tends to borrow interpretations rather than individual phrases such as those found in the Lambeth Psalter. The latter has Theodoran interpretations in an Old English paraphrase.

⁵⁷ Dempsey, G. T., ‘Aldhelm of Malmesbury and the Paris Psalter: a Note on the Survival of Antiochene Exegesis’, JTS 38 (1987), 368–86.CrossRefGoogle Scholar

⁵⁸ This is the name used for tituli not originating in the biblical text in Les “Tituli Psalmorum” des manuscrits latins, ed. Salmon, P., Collectanea Biblica Latina 12 (Vatican City, 1959).Google Scholar

⁵⁹ To judge by ps. LXXIII. 1, the Latin text of which begins on the upper margin of 91 r with the Old English gloss above, and the Christian titulus in turn entered above the latter. An Old English addition (uel utaneddest pu) to the main gloss (utadræfdest pu) of the first line, is squeezed in before the Christian titulus on the far left margin, far away from its referent. Normally, such additions to a first-line gloss are entered in the upper margin immediately above the word being elaborated (e.g. para on 92v, ps. LXXIII.22) but not in this instance because the space was already occupied by the titulus.

⁶⁰ The use of different colours recalls a similar practice for marking the initials of liturgical rubrics and psalms; see Ker, , Catalogue, pp. xxxvii–xl.Google Scholar

⁶¹ They were edited by Lawlor, J., ‘The Cathach of St. Columba’, Proc. of the R. Irish Acad. 33C (1916), 241–437, appendix iv (pp. 413–36)Google Scholar, but neither accurately nor fully. For example, Lawlor incorrectly distinguished the green and violet tituli as different series while lumping together two different sets as a single series (‘the black’).

⁶² Edited by Ramsay, , ‘Theodore of Mopsuestia in England and Ireland’, pp. 493–5Google Scholar. Ramsay did not mention the indecipherable tituli in green at pss. XCV and CXLIII, or the insertion of the word ‘propheta’ in blue before the green titulus to ps. LXXIII. In his identification of the source for the titulus to ps. LVII, read ‘Arg. (b)’ not ‘Arg. (a)’.

⁶³ Of uncertain date and origins, perhaps seventh-century Ireland. See Fischer, B., ‘Bedae de Titulis Psalmorum Liber’, Festschrift Bernhard Bischoff zu seinem 65. Geburtstag, ed. Autenrieth, J. and Brünholzl, F. (Stuttgart, 1971), pp. 90–110, esp. 93–6 and 107Google Scholar. The Argumenta are edited in PL 93, 483–1102, and again by Bright, J. W. and Ramsay, R. L., The West-Saxon Psalms, Being the Prose Portion, or the ‘First Fifty,’ of the So-Called Paris Psalter (Boston, 1907).Google Scholar

⁶⁴ Thus sixty-two are borrowed from the allegorical, seven from the historical, and two from the moral interpretations of the Argumenta.

⁶⁵ The West-Saxon Psalms, ed. Bright, and Ramsay, .Google Scholar

⁶⁶ For example, for ps. IV, the Paris Psalter has ‘Vox in cruce quando positus fuit’ the Lambeth Psalter ‘Aliter, Deus [iusti]cie exaudiuit [in cru]ce positum [filium] suum’. (Words in square brackets are no longer legible, but conjecturable.)

⁶⁷ For example, for pss. XXVI, XXXVII and LVIII, Paris uses the historical interpretation, Lambeth the allegorical; conversely, at pss. XLV and XLIX.

⁶⁸ For example, ps. XV, ‘Vox ad patrem; Ezechias orauit Dominum in egritudine’, which combines the allegorical and historical.

⁶⁹ Even when using the historical interpretation of the Argumenta, Set I often qualifies it. For example, in ps. XLIII, ‘Profeta sanctorum pressuras s[uppli]cationes[que comme]morat’, the vague ‘sanctorum’, which could apply to any Christian, replaces ‘Machabaeorum’ of the source, presumably because the historical application to anyone other than David was deemed objectionable. Likewise, in ps. XLV, ‘[Ex per]sona canitur sanctorum [proliber]atione sua gratias [agen]tium’, ‘sanctorum’ replaces the specific historical reference to the Two Tribes (‘duarum tribuum’) in the original. See further Ramsay, , ‘Theodore of Mopsuestia’, p. 496.Google Scholar

⁷⁰ Including its historical interpretations for pss. VII and CL.

⁷¹ See Les, “Tituli psalmorum”, ed. Salmon, , pp. 49 and 51.Google Scholar

⁷² Likewise ps. LXXIII, ‘Aliter uox aecclesiae de Iudeis’, where the main tradition has ‘Vox Christi ad patrem’.

⁷³ In two different scripts, though possibly by the same hand. The first is distinguished by heavily written, consistently sized capitals, the second by slenderly shaped, mixed capitals.

⁷⁴ For example, ps. CIV, ‘Vox apostolorum de Iudeis’, where the main series has ‘Vox Christi ad apostolos de Iudeis’ ps. CXXX, ‘Vox sanctae Marie’, where the main series has ‘Vox ecclesiae rogantis’.

⁷⁵ Compare for ps. CXLV, Lambeth's ‘Hic anima ipsam ortatur et ipsas sibi respondit’ with Brev., ‘Ipsa se hortatur, et ipsa responded (PL 26, 1323D).

⁷⁶ See above, p. 153.

⁷⁷ Most in a thick hand, the remaining four in a small neat hand similar to that of the marginal Latin glosses.

⁷⁸ For example, ps. XXI, ‘Hic humanitas loquitur Christi’ ( = Brev. 931C).

⁷⁹ In more than twenty instances, e.g. pss. VII, LIV and CXXIII.

⁸⁰ Twelve psalms lack a Christian titulus, e.g. pss. LXVIII–LXIX and CX.

⁸¹ Likewise, ps. CXLIX, ‘Vox Christi ad fideles’, was made into the full text of the Argumentum (allegorical) and Columban Series by the addition in green, ‘de futuro et resurr[ectione]’.

⁸² He also emended Set I at pss. LXXXIV and CL.

⁸³ Obviously, they did not perceive their work in these terms. They may not even have been aware that the allegorical Argumenta and Series I were separate works, since the two have so much in common (the latter is the source of the former; see Les, “Tituli psalmorum”, ed. Salmon, , pp. 47–8).Google Scholar

⁸⁴ Listed by Gneuss, H., ‘Liturgical books in Anglo-Saxon England and their Old English Terminology’, in Learning and Literature in Anglo-Saxon England: Studies presented to Peter Clemoes on the Occasion of his Sixty-Fifth Birthday, ed. Lapidge, M. and Gneuss, H. (Cambridge, 1985), pp. 91–141, at 114–16.Google Scholar

⁸⁵ On the Winchester origins of these three psalters, see, most recently, Hofstetter, W., Winchester und der spätaltengliscbe Sprachgebrauch, Texte und Untersuchungen zur Englischen Philologie 14 (Munich, 1987), 69–73Google Scholar; and ‘Winchester and the Standardization of Old English Vocabulary’, ASE 17 (1988), 139–61, at 157–8Google Scholar. The close textual relationship between the Christian tituli of these three Psalters was recognized by Ramsay, R. L., ‘Theodore of Mopsuestia’, pp. 496–7Google Scholar, though he was unaware of the separate existence of a Series I or of its Carolingian recension.

⁸⁶ See Fischer, B., Der Stuttgarter Bilder-Psalter, pp. 261–2.Google Scholar

⁸⁷ Especially in the use of the rather high ‘e’, as in the readings AD TE and EX EO listed below. I am grateful to Dr T. Webber for advice on the script.

^88. As did Lindelöf II, 17–18.

⁸⁹ Compare the readings of the psalters with sigla V, D and Ω in the Benedictine critical edition.

⁹⁰ PL 93, cols. 485–1098, at 888C. The true authorship of this commentary was plausibly established by Morin, G., ‘Le pseudo-Bède sur les Psaumes et l'opus super Psalterium de Maître Manegold de Lautenbach’, RB 28 (1911), 331–40Google Scholar; see also Fischer, , ‘Bedae de Titulis’, p. 90Google Scholar. For Anselm's dependence on Manegold, see Smalley, B., The Study of the Bible in the Middle Ages (Notre Dame, IN, 1964), p. 49.Google Scholar

⁹¹ PL 116, 193–696, at 452C, where the commentary is erroneously attributed to Haimo of Halberstadt. For its true authorship, see Wilmart, A., ‘Un Commentaire des Psaumes restitué à Anselme de Laon’, Recherches de théologie ancienne et médiévale 8 (1936), 325–44Google Scholar; Smalley, , The Study of the Bible, p. 67, n. 2Google Scholar; and Flint, V. I. J., ‘The “School of Laon”: A Reconsideration’, Recherches de théologie ancienne et médiévale 43 (1976), 89–110, at 92, n. 11Google Scholar. The reading OMNES still had some standing in England in the thirteenth century, as witnessed by the testimony of the English biblical scholar William de la Mare (d. 1290) who felt the need to explain why he rejected it: ‘neque ante id quod dicitur REGES premittunt hebr. et grec. quod dicitur OMNES’ (cited in the apparatus to ps. LXXV.13 of the Benedictine Gallicanum).

⁹² Probably not, since Anselm's floruit (c. 1100) is somewhat later. In any case the psalter text used by Anselm was probably already in circulation when he wrote his commentary.

⁹³ See Glunz, H. H., History of the Vulgate in England from Alcuin to Roger Bacon (Cambridge, 1933), pp. 86–7.Google Scholar

⁹⁴ See above, p. 150. Another example in the Lambeth Psalter is the marginal gloss, ‘o inique’, which proposes an alternative grammatical reading of the lemma INIQUE (ps. XLIX.52) as a vocative noun rather than an adverb. Cf. Anselm, ‘EXISTIMASTI INIQUE, vel vocativus vel adverbium’ (PL 116, col. 367D).

⁹⁵ Enarrationes in Psalmos, p. 1052, lines 40–1.

^96. See above, p. 158 and n. 85.

⁹⁷ The Stowe Psalter, perhaps reflecting its origins in a different Winchester house (see Kimmens, A. C., The Stoat Psalter, Toronto Old English Ser. 3 (Toronto, 1979), xixGoogle Scholar), does not have OMNES. The Tiberius Psalter (in its corrected state) is closest to the Lambeth Psalter, sharing with it all but one of the thirteen listed emendations. The one reading which they do not have in common, the addition EX EO (ps. LXXIV.9), is also absent from the other three Winchester psalters.

^98. History of the Vulgate, pp. 133–48.Google Scholar

^99. ibid. p. 141.

¹⁰⁰ It is hardly suprising that the psalter should come in for such close textual scrutiny at Winchester, since after the gospels it was the most widely used book of the Bible, especially in a monastic setting as part of the Divine Office.

¹⁰¹ The one exception, the decorated initial ‘B’ of BEATUS (ps. I), which combines animal heads, plaited interlace, and acanthus, is a type normally found in non-liturgical manuscripts; see Wormald, F., ‘Some Decorations from Manuscripts of the Winchester School’, Archaeologia 91 (1945), 107–35, at 122CrossRefGoogle Scholar.

¹⁰² See Der altenglische Regius-Psalter, ed. Roeder, F., Studien zur englischen Philologie 18 (Halle, 1904), xi–xii.Google Scholar

¹⁰³ In both the Roman and Benedictine Office. The first page of ps. CIX, a protected inner leaf, shows the kind of wear compatible with daily use. Unfortunately, it is not possible to determine whether the wear occurred during the Anglo-Saxon period or later. See also the marginal note accompanying ps. CXVIII, discussed below, p. 165.

¹⁰⁴ See Wieland, G.R., ‘The Glossed Manuscript: Classbook or Library Book?’, ASE 14 (1985), 153–73.Google Scholar

^105. See above, p. 153.

^106. See Wieland, , The Latin Glosses, pp. 191–2.Google Scholar

¹⁰⁷ See Robinson, , ‘Syntactical Glosses’, p. 465Google Scholar; and Korhammer, , ‘Konstrukionshilfen’, pp. 39, 48 and 50Google Scholar.

¹⁰⁸ Since he must have known Latin syntax very well in order to do his work.

¹⁰⁹ If so, probably for students at a fairly advanced stage of study. The possibility that the Lambeth Psalter might have been used as reading material for monks - a suggestion made for certain glossed Anglo-Saxon manuscripts of well-known Christian poetic texts, by Lapidge, M., ‘The Study of Latin Texts in late Anglo-Saxon England: the Evidence of Latin Glosses’, Latin and the Vernacular Languages in early Medieval Britain, ed. Brooks, , pp. 99–140Google Scholar – is very unlikely, since it does not have the clean appearance of these ‘library books’ and in any case the psalter was a text which every monk would be expected to know by heart.

¹¹⁰ See Gneuss, H., Hymner und Hymnen im englischen Mittelalter (Tübingen, 1968), pp. 186–7CrossRefGoogle Scholar, and ‘The Origin of Standard Old English and Æthelwold's School at Winchester’, ASE 1 (1972), 63–83Google Scholar at 77; and Korhammer, , Die Monastichen Cantica, pp. 235–45.Google Scholar

¹¹¹ See Hofstetter, , Winchester Sprachgebrauch, p. 86Google Scholar: ‘weist PsI [Lambeth Psalter] ganz auffallende Übereinstimmungen mit den in Winchester entstandenen Psalterversionen PsF, PsG und PsJ auf’.

¹¹² In particular Korhammer, , Die Monastichen Cantica, pp. 236–7Google Scholar, notes relatively frequent occurrences (as percentages) in the Lambeth Psalter of weoruld and syndon instead of West Saxon woruld and synt. However, patterns of distribution may tell more than absolute percentages. Thus woruld is the normal form in the first part of the Lambeth Psalter (pss. I–XLV); while syndon does not occur in pss. LIII–LXXVIII.3 or in the Canticles (both constituting scribally discrete sections), and infrequently (5 x out of 37 x ) in the marginal Old English glosses. This evidence suggests that the occurrence of non-West Saxon forms depends either on the scribe of an individual section or on his source and does not necessarily reflect the usage of the scriptorium where the Lambeth Psalter was written. For a similar conclusion based on textual and lexical evidence, see next note.

¹¹³ See C., and Sisam, K., The Salisbury Psalter, p. 72Google Scholar, who describe the Lambeth Psalter gloss as the product of ‘a process of collection and accretion’. Likewise, Hofstetter, , Winchester Sprachgebrauch, p. 87Google Scholar, would attribute the non-West Saxon features of vocabulary to dependence on non-West Saxon sources.

¹¹⁴ At the bottom of 149r in a hand similar to that which entered some of the Set IV tituli.

¹¹⁵ ‘Prima hora dominica, dicenda quattuor capitula psalmi centesimi octaui decimi’ (‘At Prime on Sunday the first four verses of ps. CXVIII are to be recited’: Regula S. Benedicti, ch. 18). The Roman Office begins the same Hour with ps. L11I.

¹¹⁶ See C., and Sisam, K., The Salisbury Psalter, §6Google Scholar. In the Lambeth Psalter no division, either at v. 36 or 40, was marked by the main scribe, presumably because the psalter was not meant for liturgical use. But someone subsequently added a cross before v. 40 and a marginal note (indecipherable to me), both in dry-point.

¹¹⁷ Namely, the scribes of (i) the tituli with Carolingian readings; (ii) the ‘Winchester Psalter’ readings; and (iii) the Old English gloss.

¹¹⁸ Thus, the exemplar of the Old English gloss was probably not that which supplied the ‘Winchester’ readings, since the former does not reflect the innovations of the latter; and the Set I and II Christian tituli, with their different readings for the same psalms, almost certainly came from different sources.

¹¹⁹ One potential objection remains: if the Lambeth Psalter was at Winchester, it seems odd that later Old English glossators (of the Vitellius, Tiberius and Arundel Psalters) who are believed to have worked there, did not use such an excellent aid. The most obvious answer is the evidence that these glossators did, in fact, use the best of the Lambeth Psalter, but in the form of its underlying Winchester source; see C., and Sisam, K., The Salisbury Psalter, pp. 67–71Google Scholar, Berghaus, F.-G., Die Verwandtschaftsverhältnisse der altenglischen Interlinearversionen des Psalters und der Cantica (Göttingen, 1979), pp. 76–90Google Scholar, and Hofstetter, , Winchester Sprachge-brauch, pp. 69–74Google Scholar. Another possible explanation is that the Lambeth Psalter had already left its original (Winchester) home by the mid-eleventh century; see Wildhagen, K., ‘Das Psalterium Gallicanum in England und seine altenglischen Glossierungen’, Englische Studien 54 (1920), 35–45, at 40–2Google Scholar, though his evidence is dubious and inaccurate. For example, he claims that the manuscript was at Lanthony as early as the eleventh century, and that the inclusion of St Mildred in item 6 (which he did not recognize as a fifteenth-century addition to the Lambeth Psalter) indicated that the manuscript was previously owned by a woman. I am grateful to the American Council of Learned Societies for a grant-in-aid (Summer, 1983); to the Librarian of Lambeth Palace Library for permission to publish from the contents of the Lambeth Psalter; and to Dr M. Lapidge for corrections and editorial suggestions.