¹ King Alfred's West-Saxon Version of Gregory's Pastoral Care, ed. Sweet, Henry, EETS o.s. 45 and 50 (London, 1871–1872; repr. 1958), 7Google Scholar, lines 6–8. I refer to this work henceforth as CP. See also The Pastoral Care, edited from British Library MS. Cotton Otho B. ii, by Ingvar Carlson, completed by Lars-G. Hallander together with Mattias Löfvenberg and Rynel, Alarik, Stockholm Stud. in Eng. 34 and 48 (Stockholm, 1975 and 1978).Google Scholar

² King Alfred's Old English Version of Boethius De Consolatione Philosophiae, ed. Sedgefield, Walter John (Oxford, 1899);Google ScholarKing Alfred's Version of St Augustine's Soliloquies, ed. Carnicelli, Thomas A. (Cambridge, Mass., 1969);Google ScholarThe Old English Version of Bede's Ecclesiastical History of the English People, ed. Miller, Thomas, EETS o.s. 95, 96, 110 and 111 (London, 1890–1898; repr. 1959 and 1963);Google ScholarThe Old English Orosius, ed. Bately, Janet, EETS s.s. 6 (London, 1980).Google Scholar I refer to these works henceforth as Bo, Solil, Bede and Or respectively.

³ See Whitelock, Dorothy, ‘The Old English Bede’, PBA 48 (1962), 57–90;Google ScholarLiggins, Elizabeth M., ‘The Authorship of the Old English Orosius’, Anglia 88 (1970), 289–322;Google Scholar and Janet M. Bately, ‘King Alfred and the Old English Translation of Orosius’, Ibid. pp.433–60. Sherman Kuhn's attempts to rehabilitate Bede (‘Synonyms in the Old English Bede’, JEGP 46 (1947), 168–76Google Scholar, and ‘The Authorship of the Old English Bede Revisited’, NM 73 (1972), 172–80)Google Scholar are unconvincing, resting as they do on an attempt to explain away Mercian features, not on an assessment of the work as a whole, and presupposing direct copying without intermediate dictation. A combination of linguistic and literary/historical evidence on the other hand indicates convincingly that CP, Bo and Solil are substantially the work of one person and that this person was King Alfred; see further, below, pp. 77–95. For selective lexical differences between Bede and the genuine works of Alfred, see further, below, pp. 86–9.

⁴ Liber Psalmorum. The West-Saxon Psalms, ed. Bright, James Wilson and Ramsay, Robert Lee (Boston and London, 1907).Google Scholar I cite this work as Ps(P) by psalm and verse; my quotations and figures are based on a partial recollation of the manuscript.

⁵ For a facsimile, see The Paris Psalter, ed. Colgrave, Bertram et al. , EEMF 8 (Copenhagen, 1958).Google Scholar

⁶ The Old English introductions are preserved also in the margins of London, British Library, Cotton Vitellius E. xviii (cited below as MS V).

⁷ I make no attempt to reassess the nature of the Latin text of the psalter used by the translator of Ps(P) or his other sources; Patrick O'Neill, of the University of North Carolina, has been studying these questions.

⁸ Libri Psalmorum Versio Antiqua Latina; cum Paraphrasi Anglo-Saxonica, ed. Thorpe, Benjamin (Oxford, 1835).Google Scholar

⁹ Wülker, Richard Paul, Grundriss zur Geschichte der angelsächsischen Litteratur (Leipzig, 1885)Google Scholar, §§500–1. William of Malmesbury also claimed Alfred's authorship for Bede and Or; see De Gestis Regum Anglorum, ed. Stubbs, W., Rolls Ser. (1887–1889), 1, 132.Google Scholar

¹⁰ Wichmann, J., ‘König Aelfreds angelsächsische Übertragung der Psalmen I-LI Excl.’, Anglia 11 (1889), 39–96Google Scholar, and Bromwich, J. I'A, ‘Who was the Translator of the Prose Portion of the Paris Psalter?’, The Early Cultures of North-West Europe, ed. SirFox, Cyril and Bruce, Dickins (Cambridge, 1950), pp. 290–303.Google Scholar A. S. Cook was cautious: ‘We have seen that Alfred must certainly have translated 1.1, pretty certainly 11.6, not improbably 23.4, and at least possibly the whole of the prose portion of the Paris Psalter. Yet against the last supposition must be set the notable discrepancies of language revealed by the parallel passages adduced above. It will require a more comprehensive and detailed examination to decide whether Alfred is really to be credited with the translation of all the prose Psalms extant’ (Biblical Quotations in Old English Prose Writers (London, 1898), p. xl).Google Scholar See, further, Ibid. pp. xxxvi–xl, and below, n. 60.

¹¹ E.g. Schabram, Hans, Superbia, Teil I (Munich, 1965)Google Scholar, and Seebold, Elmar, ‘Die AE. Entsprechungen von lat. sapiens und prudens’, Anglia 92 (1974), 291–333;Google Scholar see also Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 454–6.Google Scholar

¹² Celia, and Sisam, Kenneth, The Paris Psalter, ed. Colgrave, et al. , p. 16.Google Scholar

¹³ ‘The Prose of Alfred's Reign’, Continuations and Beginnings, ed. Stanley, E. G. (London, 1966), pp. 67–103, at 94–5.Google Scholar

¹⁴ See Bately, Janet M., The Literary Prose of King Alfred's Reign: Translation or Transformation. An Inaugural Lecture (London, 1980), pp. 6–10.Google Scholar

¹⁵ The need for comprehensibility must have been an important factor in determining the translator's methods of rendering the psalms. Close and accurate translation of this difficult devotional text would simply have transferred existing problems from Latin into English: where the thought is hard to grasp in Latin, it would have remained hard in English. What is more, for the author of Ps(P) part of the meaning of the psalms was their relevance to mankind, to the church, to the individual Christian using them: the wise man needs to appreciate this relevance to use the psalter effectively. So the translator not only made clear by expansion or alteration what otherwise might have been obscure but also spelled out the application of each psalm, whether universal or particular. As in the case of Bo and, to a lesser degree, in that of CP, the ‘translation’ is more explicit than the original. See, further, Bately, Ibid.passim. For Alfred's love of the psalter, see Asser, Life of King Alfred, trans. English Historical Documents 1, ed. Whitelock, Dorothy, 2nd ed. (London, 1979), 292–3.Google Scholar For the importance of the psalter in continental education, see, e.g., Lutz, Cora E., Schoolmasters of the Tenth Century (Hamden, Conn., 1977), p. 84.Google Scholar

¹⁶ I am indebted to Professor Peter Clemoes for his constructive criticism of this paper and for his help in the final shaping of it.

¹⁷ Unfortunately a number of lexical items cited as most typical of Ælfric – e.g. gefredan, hreppan and cweartern – not only are themselves absent from Ps(P) but also represent concepts not found there. However, Ps(P) differs from Ælfric's works in using (ge)gearwian (2 ×), not gearcian; pret. for (1 ×), not ferde; and fremde (2 ×) as well as elpeodig (8 ×), but not alfremed. It never represents the concept ‘despise’ by forhogian, though it does use forseon (8 ×); moreover it has earfop (54 ×), not earfopnes as Ælfric, and unlike Ælfricnever uses snottor(nes). For Ælfric's vocabulary, see Homilies of Ælfric. A Supplementary Collection, ed. Pope, John C., EETS 259–60 (London, 1967–1968), 99–103Google Scholar, and for words and mannerisms typical of Wulfstan, Karl Jost, Wulfstanstudien (Bern, 1950), pp. 110–271.Google Scholar It may be noted that, unlike Wulfstan, Ps(P) has (ge)arian as well as beorgan; gescieldan, not werian; and bœlend as well as drybten.

¹⁸ For a useful list of ‘late’ and ‘Winchester school’ words, see Gneuss, Helmut, ‘The Origin of Standard ssOld English and Æthelwold's school at Winchester’, ASE 1 (1972), 63–83, at 76–8.Google Scholar Words absent from Ps(P) include mese (Ps(P) beod 1 ×), bepœcan (Ps(P) beswican 2 ×), besargian (Ps(P) seofian 23 ×, heofan 1 ×, etc.), gedeorf (Ps(P) geswinc 4 ×), scrudnian (Ps(P) smeagan 11 × etc.), pœslic (Ps(P) cyn 2 × and gemetlic 1 ×) and gelaþung (Ps(P) gesamnung 4 × in the sense of ecclesia), while present in Ps(P), but not in ‘Winchester school’ texts, are altare and alter (3 ×, ‘Winchester school’ weofod). See further, The West-Saxon Gospels, ed. Grünberg, M. (Amsterdam, 1967), p. 334;Google ScholarGretsch, Mechthild, Die Regula Sancti Benedicti in England (Munich, 1973), p. 361;Google Scholar and TheodulfiCapitula in England, ed. Sauer, Hans (Munich, 1978), p. 267.Google Scholar Ps(P) also differs from late West Saxon in its preference for tid (11 ×) over tima (1 ×), even in a general context; however, see Ibid. p. 262. It also uses a number of words which late scribes tend to remove as (possibly) obsolete or obsolescent, e.g. andwlita, cneores, gebat, sar and symle; see The Salisbury Psalter, ed. Sisam, Celia and Sisam, Kenneth, EETS 242 (1959), 35.Google Scholar

¹⁹ An Old English Martyrology, ed. Herzfeld, George, EETS 116 (London, 1900);Google ScholarBischof Wœrfertbs von Worcester Übersetzungder Dialoge Gregorsdes Grossen, ed. Hecht, Hans, Bibliothek der angelsächsischen Prosa 5 (Leipzig and Hamburg, 1900 and 1907; repr. Darmstadt, 1965).Google Scholar I refer to these works henceforth as Mart and GD respectively.

²⁰ See, however, below, pp. 81–2 and n. 77. For attempts to identify Anglian dialect forms in these and other texts, see, in particular, Jordan, Richard, Eigentümlichkeiten des anglischenWortschatzes (Heidelberg, 1928);Google ScholarScherer, Günther, Zur Geographic und Chronologie des angelsächsischen Wortschatzes (Berlin, 1928);Google ScholarCampbell, Jackson J., ‘The Dialect Vocabulary of the Old English Bede’, JEGP 50 (1951), 349–72;Google Scholar and Sisam, Celia, ‘AnEarlyFragmentof the Old English Martyrology’ RES n.s. 4 (1953), 209–20Google Scholar, at 216, n. 5. For valuable summaries of earlier findings, with bibliography see, further, Theodulfi Capitula, ed. Sauer, to which, for the sake of economy, I refer frequentlybelow.

²¹ In this paper I have concentrated my attention on the Mercian glosses on the psalms in BL Cotton Vespasian A. i (The Vespasian Psalter, ed. Kuhn, Sherman M. (Ann Arbor, 1965))Google Scholar, which I cite as Ps(A), since these appear to be roughly contemporary with the reign of King Alfred; I have excluded the glosses on the hymns. Where relevant I have given variants from the Junius Psalter (Ps(B)), because of the manuscript's apparent connections with the Lauderdale manuscript of Or, the Tanner manuscript of Bede, and the Parker manuscript of the Anglo-Saxon Chronicle; see Parkes, M. B., ‘The Palaeography of the Parker Manuscript of the Chronicle, Laws and Sedulius, and Historiography at Winchester in the Late Ninth and Tenth Centuries’, ASE 5 (1976), 149–71Google Scholar, and (for the text of Ps(B) Der altenglische Junius-Psalter, ed. Brenner, Eduard (Heidelberg, 1908).Google Scholar For some of the other psalter glosses and their relationships, see Berghaus, Frank-Günter, Die Verwantschaftsverhältnisse der altenglischen Interlinearversionen des Psalters und der Cantica (Göttingen, 1979);Google Scholar see also Gneuss, Helmut, Lehnbildungen und Lehnbedeutungen im Altenglischen (Berlin, 1955).Google Scholar Since I understand that this is the subject of a forthcoming article by Richard Clement, I have not attempted to consider the possible influence of psalter glosses on Ps(P).

²² I refer to this text henceforth as WS. For convenience I have treated the four gospel translations as a unit, although I am not convinced that they are the work of a single translator. For figures for WS I have drawn on the somewhat inaccurate and misleading glossary by Harris, Mattie Anstice, repr. Word Indices to Old English Non-Poetic Texts, with a Preface by Fred C.Robinson (Hamden, Conn., 1974);Google Scholar for the text, see ref. above, n. 18.

²³ For lexical differences between Or and Ps(P), see below, pp. 88–93, and Bately, ‘KingAlfred and the Old English Translation of Orosius’, pp.454–6.

²⁴ For some of the special characteristics of the sections of the Chronicle either compiled in orrelating to King Alfred's reign, see Bately, Janet, ‘The Compilation of the Anglo-Saxon Chronicle 60 B.C. to A.D.890: Vocabularyas Evidence’, PBA 64 (1978), 93–129.Google Scholar

²⁵ Figures for GD are basedon the readings of MSS O and C, not on the revised version in MS H, for which, see Yerkes, David, The Two Versions of Waerferth's Translation of Gregory's Dialogues: an Old English Thesaurus (Toronto, 1979).Google Scholar Where the surviving manuscripts of Bede, GD and Mart have variant readings the form cited is normally that of the base manuscript used by the editors of the work concerned; only selective variants are noted here. Normalized spellings of the head-word are given throughout, mainly following the practice of Hall, John R. Clark, A Concise Anglo-Saxon Dictionary, 4th ed. with suppl. by Merritt, Herbert D. (Cambridge, 1960).Google Scholar For total occurrences in Ps(P) of Old English words cited in this section, see below, pp. 87–8. Unless otherwise stated, the Latin text of the psalter I have used is that of the Paris Psalter, collated with the Latin of the Vespasian Psalter as representative of the Roman text. For Gallican and Hebrew texts of the vulgate psalms, see Biblia Sacra iuxta Vulgatam Versionem, ed. Weber, Robert I (Stuttgart, 1969).Google Scholar

²⁶ I include here the variant untrymnes, which is the preferred form in MS T, with MSS B, Ca and O normally using untrumnes. Miller reports (Bede, pt 11, p. 106) one instance of mettrumnes from the now mainly destroyed MS C, cited in Whelock's edition of the text.

²⁷ On this occasion only I have included instances in Ps(P) without equivalent in the Latin psalter, mettrumnes being confined to the introductions (along with one of the instances of untrumnes). Despite this limited distribution there is nothing in the lexical evidence taken as a whole to suggest separate authorship for these introductions.

²⁸ Cf.Mart, p.180, line14, whereMS B has untrumnes and MS C mettrumnes, and GD, p. 83, line 31, where MSS C and O have mettrumnes and MS H untrumnes. In the Blickling homilies the distribution is untrumnes (7 ×)in homs. xi, xiii, xvii and xviii and mettrumnes (1 ×) in hom. v; see The Blickling Homilies of the Tenth Century, ed. Morris, R., EETS 58, 63 and 73 (London, 1874, 1876 and 1880; repr. 1967)Google Scholar, and Scragg, D. G., ‘The Corpus of Vernacular Homilies and Prose Saints’ Lives before Ælfric’, ASE 8 (1979), 223–77.Google Scholar I refer to the Blickling homilies henceforth as BlHom and follow the numbering of Morris's edition.

²⁹ For Ps(A) unriht Ps(B) has unrihtwisnes.

³⁰ Ps(A) also has unrihtwisnes for iniustitia (4 ×), beside unriht (1 ×), making a total of ninety-nine instances of the Old English noun. In Ps(P) unrihtwisnes occurs a total of fifteen times, rendering not only iniquitas but also impietas, insipientia and iniustitia (each 1 ×). Unriht is usedthirty-onetimes, rendering also iniustitia (3 ×) and nequitia (2 ×). For impietas Ps(A) has arleasnes.

³¹ Twice, both in hom.x.

³² These figures include instances of handgewinn and wiþwinnan where these correspond to labor and laborare in the Latin source. In addition Bede has five instances of gewinn where the Latin uses the corresponding verb.

³³ Geswinc and swincan each occur four times in all in Ps(P). Labor is also rendered by broc and earfoþ (each 1 ×). In Ps(A) geswinc occurs thirteen times and isreservedfor afflictio and tribulatio along with geswencednes (24 ×), besidePs(P) earfoþ (11 ×), geswinc and (nyd) þearf (each2 ×) and unrotnes (1 ×).

³⁴ Eigentümlichkeiten p. 43. See, further, Grinda, Klaus R., ‘Arbeit’ und ‘Mübe’. Untersuchungen zur Bedeutungsgeschichte altenglischer Wörter (Munich, 1975)Google Scholar, and Theodulfi Capitula, ed. Sauer, pp. 260–1.Google Scholar For GD MS H's replacement of winnan and gewinn by swincan and geswinc, see Yerkes, Thesaurus.

³⁵ Ps(B) has offrung (2 ×) and bærning (1 ×), for Ps(A) onsœgdnes.

³⁶ For offrung as apossible West Saxon word, see Rauh, Hildegard, Der Wortschatz deraltenglischen Übersetzungen des Matthaeus-Evangeliums (Berlin, 1936), p. 28;Google ScholarMenner, Robert J., ‘Anglian and Saxon Elements in Wulfstan's Vocabulary’, MLN 63 (1948), 1–9, at 7;Google Scholar and Theodulfi Capitula, ed. Sauer, p. 257.Google Scholar For oblatio Ps(P) has oflata (1 ×), Ps(A) oflate and onsœgdnes (each 1 ×) and Bede lac (2 ×) and onsœgdnes (1 ×). WS has offrung for victima and lac for munus. The only instance of onsœgdnes in BlHom occurs in hom. vi.

³⁷ Ps(B) has geswinc (11 ×) for geswencednes; see, further, Grinda, ‘Arbeit’ und ‘Mübe’, esp. chs. iii, iv and v.

³⁸ See Rauh, , Wortschatz, p. 40Google Scholar, and Menner, Robert J., ‘The Anglian Vocabularyof the Blickling Homilies’, Philoĺogica: the Malone Anniversary Studies, ed. Kirby, Thomas A. and Woolf, Henry Bosley (Baltimore, 1949), pp. 56–64, at 58;Google Scholar also Jordan, , Eigentümlichkeiten, p. 8Google Scholar, n. The single instance of scua in BlHomoccursin hom. vi.

³⁹ With MS variant ymbhringan and a single instance of gegyrewan. See also circumstare, Bede ymbsellan, and Menner, Robert J., ‘The Vocabularyof the Old English Poems on Judgment Day’, PMLA 62 (1947), 583–97, at 593.CrossRefGoogle Scholar Of particular interest in Bedeis the use of ymbhringan by MSS O and Ca in the ‘duplicated’ section, beside MS B ymbsellan 7 ymbhabban and MS T ymbhabban; see, further, Campbell, Jackson J., ‘The Old English Bede:Book III, chapters 16–20’, MLN 67 (1952), 381–6, at 385.Google Scholar

⁴⁰ See also WS's use of betynan and scrydan (each2 ×) and gecnyttan (1 ×) for circumdare.

⁴¹ Wortschatz, p. 16, and see Menner, , ‘The Anglian Vocabulary of the Blickling Homilies ’, p. 58.Google ScholarYmbsellan is found in BlHom 1 and xix. Ps(P) has ymbgittan (2 ×) and behringan (1 ×) without equivalent in the Latin.

⁴² See also p. 108, line 4 geþwœrede.

⁴³ Cf. gegaderian (1 ×)in the VespasianHymns.Bedeand Ps(A) also use (ge)samnian for Latin colligere and convenire: in BlHom this verb is found in homs. iv, vi, vii, viii, xiii, xv and xvii (Latin equivalents unknown).

⁴⁴ A second example of deficere is rendered benumcn 7 bescired in Bede.

⁴⁵ Ps(P) also has gedwœscan and forweorþan (each1 ×) for deficere and once uses the pair geteorian 7 geendian (xxx.11). In all it employs aspringan five times and geteorian fourtimes.Ps(B) has ateorian (3 ×) for Ps(A) aspringan. In WS's figures I include ungeteorod (1 ×)for non deficiens.See also BlHom, where aspringan is foundin hom, vii and geteorian in homs.x and xix.

⁴⁶ Hopian is used in collocation with anbidian and gebidan (each1 ×); see, further, below, n. 106. Ps(A) also uses abidan for Latin sustinere.

⁴⁷ MS Ca abidan (2 ×), MS B gebidan (2 ×)and MS B onbidan (1 ×). Ps(B) similarly has onbidan three times for Ps(A) bidan and twice for Ps(A) abidan. Cf. BlHom, with onbidan (6 ×) in homs. vi and xix, gebidan (2 ×)in homs.xv and xviii and bidan (2 ×) in homs.i and ix.

⁴⁸ Including one instance of fœgnian where the Latin psalter has exaltare, suggesting an intermediate misreading exultare.

⁴⁹ In place of one instance of gefeon in Ps(A) Ps(B) has wynsumian, a word which it and Ps(A) normally use for jubilare (Ps(A) 8 ×). In Ps(P) the single instance of jubilare is rendered by fœgnian 7 myrgan. See also gaudere: Bede gefeon (22 ×, twicepairedwith blissian) alongside efngefeon (2 ×)and blissian 7 gefeon (1 ×)for congaudere; and Ps(A) gefeon (6 ×); besidePs(P) blissian (1 ×); and see GD, p. 69, line 13, where gaudebat isrendered by gefœgnode in MS C but gefeah inOand geblissode inH. In Ps(A) (ge)blissian is reserved for laetari and laetificari, while in Bede blissian is the normal rendering for laetari (9 ×, twice paired with gefeon), though it is not confined to it. For the frequent substitution of fœgnian for what I am assuming, rightly or wrongly, to be ‘original’ gefeon in manuscripts of Bede, see Jordan, , Eigentümlichkeiten, p. 89.Google Scholar For further details of the usage in Ps(P), see below, p. 87 and no.107. In BlHom gefeon is found in homs. i, vi, vii, xiii, xiv, xv, xvii, xviii and xix.

⁵⁰ See Jordan, , Eigentümlichkeiten, pp. 89–90;Google ScholarGretsch, , Regula, p. 336;Google Scholar and Theodulfi Capitula, ed. Sauer, pp. 228–9.Google Scholar

⁵¹ In Ps(A) wunian is found only for demorari, manere and commorari, while bu(i)an does not occur. Ps(B) has occasional eardian, frequent oneardian, for Ps(A) ineardian, Latin inbabitare. In BlHom eardian and wunian are the norm, with pp. gebuen (1 ×)in hom. xi.

⁵² Retribuere is not used in Bede's Historia Ecclesiastica; however, GD has geeadleanian for remunerare.

⁵³ xvi.12 þunribtan wisan. Unribtwis is also used in Ps(P) for iniustus, iniquus, peccator etc. The two instances of arleas occur in the same psalm, xxv. For the usage of the West Saxon Gospel of St Matthew, see ed. Grünberg, pp. 331–2 and 333.

⁵⁴ Soþfœst is not found in Ps(P), which uses ribtwis and ribt for iustus and rectus. See Jordan, , Eigentümlichkeiten, p. 43Google Scholar, and Theodulfi Capitula, ed. Sauer, pp. 258–9.Google Scholar Ps(B) eighttimes replaces ribtwisnes by efnes.

⁵⁵ Ps(B) has gefeob gefeob (1 ×)for Ps(A) wel þe wel þe.

⁵⁶ See Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, p. 127Google Scholar and n. 2. In surviving manuscripts of the ‘first’version of GD, on the other hand, gan is more frequent than gangan.

⁵⁷ See Bately, Ibid. esp. pp.100 and 109–10. An example of deliberate revision is provided by MS H of GD, for which, see above, n. 25.

⁵⁸ ‘The Origin of Standard Old English’, pp. 66–8.

⁵⁹ Attention is not infrequently drawn to the ‘remarkable agreement’ in the rendering of Latin words in Bede and GD; see, e.g., Potter, Simeon, On the Relation of the Old English Bede to Werfertb's Gregory and to Alfred's Translations (Prague, 1931), pp. 17–18.Google Scholar However, in a number of cases the translating word is the obvious one and the same equations are found in texts which have no pretensions to being of late-ninth-century Mercian origin.At the same time preferences are frequently different.

⁶⁰ Although J. D. Bruce had ‘little difficulty in exposing the weakness of Wichmann's evidence’ (The Paris Psalter, ed. Colgrave etal., p. 16)Google Scholar, some of his attacks are completely unfounded; see Bruce, J. Douglas, ‘The Anglo-Saxon Version of the Book of Psalms commonly known as the Paris Psalter’, PMLA 9 (1894), 43–164CrossRefGoogle Scholar, and Bromwich, , ‘Who was the Translator?’, pp. 292–3.Google Scholar Unlike Wichmann, Cook made a deliberate attempt to look for significant differences between Ps(P) and the works of Alfred as well as for similarities; see above, n. 10. However, his corpus is far too small to provide results of any great value and not only do his similarities include forms found in works no longer ascribed to Alfred, namely Or and Bede, but also his differences can be explained away on stylistic grounds. For example, the presence of the expression wolberendum setle in both Ps(P) and CP could, but need not, be evidence of common authorship, while the fact that CP translates Lat ‘Avertisti faciem tuam a me, et factus sum conturbatus’ by ‘Dryhten, ðu ahwyrfdes ðinne ondwlitan from me, ða wearð ic gedrefed’ (ed. Sweet, p. 465, lines 19–20) whereas Ps(P) has ‘þ a awendest þ u þ inne andwlitan fram me, þ a wearð ic sona gedrefed’ (xxix.7) is equally non-conclusive. Indeed, underlying some of the differences are actual similarities of method. Thus CP's addition of Dryhten is not only contextually explicable but paralleled in a number of other verses in Ps(P) as well as CP (see, e.g., CP, p. 413, line 17, Ps(P) xxxviii.14 and Ps(P) xxii.4, where the word alliterates with deaðes and ondrœde); the expression ‘ðu ahwyrf(des) ðinne ondwlitan from me’ is actually found elsewhere in Ps(P) (XII.I); and the addition of sona in Ps(P) conforms with the translator's general practice of attempting to write a balanced and poetic prose, about which, see, further, below.

⁶¹ For Alfred's helpers, who were presumably active on numerous occasions, see the prefatory letter to CP and Asser's Life, chs. lxxvii, lxxix and lxxxviii, trans. EHD 1, 294–8; for differences between CPand Bo, see, further, Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 457–8Google Scholar, and for differences between the various works of Ælfric, see Godden, M. R., ‘Ælfric's Changing Vocabulary’, ES 61 (1980), 206–23.Google Scholar

⁶² For gecoronian, see Kirschner, Josef, Die Bezeichnungen für Kranz und Krone im Altengliscben (Munich. 1975).Google Scholar

⁶³ Cf. CP, Bo, Solil aliesan and CP aliesend. The noun aliesnes is composed of verb stem + nes suffix, a formation usually assumed to be characteristic of Mercian, it being often said that the only verb form to which the suffix could be added in West Saxon was the past participle; see, e.g., Campbell, , ‘Dialect Vocabulary of the Old English Bede’, pp. 367–8.Google Scholar According to Campbell, Ibid. no. 40, ‘The Cura Pastoralis, alone among WS documents, shares this Anglian peculiarity of adding -niss directly to the stem of the verb rather than to the participle. Hence the occurrences in the Past do not prove that the word towesniss was known in the Saxon area.’ For a somewhat less extreme attitude, see Jordan, , Eigentümlichkeiten, pp. 101–2Google Scholar, and The Life of St Chad, ed. Vleeskruyer, R. (Amsterdam, 1953), p. 29;Google Scholar see also Sauer's, summary, Theodulfi Capitula, p. 241Google Scholar, according to which the construction is Anglian or primarily Anglian but occurs ‘noch gelegentlich’ in early West Saxon. A study of all nouns formed from verb + -nes in Ps(P) yields the following:

(a) past participle + -nes: four words, one of which is found also in CP, three are found in Bo and one in Solil, with one apparent hapax legomenon, ymbsetennes (1 ×);

(b) verb stem + -nes: fifteen words, eight of which are found also in CP, two in Bo and two in Solil, with only five not appearing anywhere in the accepted works of Alfred, namely aliesnes (with MS V having the West Saxon aliesednes for two of Ps(P)'s five instances of this word), gehieldnes (apparently a hapax legomenon), trymenes, wyrgnes and generennes (if this represents an original generenes not generednes).

If Ps(P), then, is a West Saxon translation, attributable to King Alfred, the inference to be drawn from Campbell's claim is that, as in the case of CP, Ps(P)'s relative fondness for the construction verbal stem + -nes is to be attributed to the involvement of one or more of Alfred's Mercian helpers. However, neither frequency of occurrence nor distribution patterns will allow us to treat this construction as necessarily untypical of a king whose West Saxon dialect did not have the status of a standard language and whose wife (from whom he might have acquired some of his speech habits) was herself a Mercian. Of the thirteen Ps(P) words not in Bo the majority represent concepts not found there, whilethe remainderoccurin Ps(P) alongsidethe ‘Boethian’ usage:thus acennes (1 ×) beside acennednes (1 ×), herenes (1 ×) beside hering (1 ×) and trymenes (1 ×) beside trymnes (1 ×), untrumnes (1 ×), untrumnes (2 ×) etc.; Bo acennednes (1 ×), hering (3 ×) and untrymnes (1 ×), with CP using acennes, herenes, hering, untrymnes and (un)trumnes, acennes occurring also in Or and in the West Saxon genealogy prefixedto the Anglo-Saxon Chronicle, MS A. The conclusion must be that at least some formations of the type verb stem + -nes were current in early West Saxon as well as in Anglian and that their appearance in works associated with King Alfred, though possibly due to the dialect preferences of his helpers, could represent his own usage.

⁶⁴ Bo and Solil, however, have the noun ofergietolnes.

⁶⁵ I would include here Ps(P) a/on worulde woruld (12 ×)for in saeculum saeculi, since the only similar expressionin the acceptedworksof Alfred, a to worulde, is in the prayer at the end of MS B of Bo, which in my opinion does not belong to the Boethius and does not represent Alfred's work, cf., e.g., this prayer's exceptional use of the preposition toforon.

⁶⁶ Bately, , Literary Prose, p. 14.Google Scholar The translator showed a remarkable restraint in the use of correlation and a remarkable sensitivity to the cadences and unexpressed relationships of the Latin; see, further, above, n. 60, and below, section(1), passim, and cf., e.g., xviii.8 with 111.5.

⁶⁷ Cf., e.g., the use of bœr, not locc or feax, in xxxix.14 ‘Mine fynd wæran gemanigfealdode þæt heora wæs ma þonne hæra on minum heafde’ with Ps(A) rendering Latin capillos by loccas, andxxxvii.8 gesœged 7 gehnœged, for (in)curvatus [Gallican adflictus, ] with Ps(A) gebeged and CP, p. 67, line 18 gebiged (a word found two verses earlier in Ps(P) for Gallican and Hebrew (in)curvatus, Vespasian Psalter turbatus). Assonance and alliteration are indeed very common, while compounds sometimes replace two or more Latin words: e.g., sœfiscas for pisces maris, sœwegas for semitas maris, eorþcyningas for reges terrae and handgeweorc for opera manum; see also gebeorbstow for refugium, and rothwil without equivalent in the Latin psalter. The desire for a tight and balanced structure may explain the absence of certain of Alfred'ssyntactical mannerisms, e.g. swa…swa… swœþer (though see the use of oþertwega oþþe…oþþe and awþer oþþe…oþþe, cited below, p. 90).

⁶⁸ The summary is basedon an exhaustive study of the lexis of Ps(P) inrelation to the usage of the Latin psalter versionsonthe one hand and that of the accepted works of Alfred (and, selectively, other texts of the late ninth century) on the other; it is based also on a detailed study of concepts for which Ps(P) uses a range of Old English words and on a detailed examination of the adjectives and adverbs in Ps(P) and of those in the accepted works of Alfred.I have also investigated every word which occurs frequently in the latter texts but neverin Ps(P), but otherwise I have not attempted an exhaustive study of the lexis of the accepted works of Alfred, either individually or as a whole.

⁶⁹ A number of words in this category are verbs which occur with and without the ge- prefix in one orother corpus. Reference to the textual notes to Carlson's edition of CP will show how much variation of usage could have arisen in the transmission of these forms; see also variations between the two manuscripts of Bo cited in Sedgfield's notes and see Horgan, Dorothy, ‘Patterns of Variation and Interchangeability in some Old English Prefixes’, NM 81 (1980), 127–30.Google Scholar

⁷⁰ Like the instances from the list of contents the verse example is derived from the corresponding passage of prose and so does not strictly represent a separate usage. It should be noted that Sedgefield does not include forms from either the list of contents or the final prayer in his glossary. My figures are based partly on information in Sedgefield's glossary, partly on my own collection of references. I make no attempt to assess the extent to which the verse metres of Bo reflect Alfred's normal usage or indeed the extent of Alfred's responsibility for them.

⁷¹ For a different explanation of rynewœn, see Tinkler, John D., Vocabulary and Syntax of the Old English Version in the Paris Psalter, Janua Linguarum, series practica 67 (The Hague, 1971), 61.CrossRefGoogle Scholar

⁷² Cp goes on to make a distinction between ‘calling’ and ‘shouting’, using clipian and brieman respectively (p. 429, line 1). For the expected noun to translate vulgate clamor WS substitutes the expression man brymde.

⁷³ See also the gloss ‘Conticiscent, silebant: Sicittan’ Anglo-Saxon and Old English Vocabularies, ed. Wright, Thomas and Wülcker, Richard Paul (London, 1884) 1Google Scholar, col. 211, line 43. Conticescere is rendered by geswigian in the Vespasian Hymns.

⁷⁴ If it had been selected, biecnan would, of course, have alliterated with blissian, but at the sametime it would have directed the emphasis outside rather than inside the series of subordinate clauses in which the concept occurs. Wincettað, on the otherhand, alliterates with winnað, which belongs to that series. For Gallican supergaudeant the Vespasian Psalter reads insultent.

⁷⁵ For corona words, see, further, Kirschner, Kranz und Krone, passim. Cynegold is found also in The Phoenix; beafodgold and heafodbeag both occur in the Vercelli homilies.

⁷⁶ See Jordan, , Eigentümlichkeiten, p. 50;Google ScholarScherer, , Geographie und Chronologie, p. 13Google Scholar, and The Old English Orosius, ed. Bately, p. lxxii.Google Scholar Elsewhere Or has the expression gesellan wiþ feo. In late West Saxon tests the most frequently used words for ‘sell’ are (be)ciepan and its varients and (ge)sellan in conjunction with a phrase indicating that the ‘giving’ or ‘handing over’ is in return for money, e.g. to ceape, wiþ weorþe or wiþ feo; see further, Gretsch, , Regula, p. 322.Google Scholar

⁷⁷ Indeed all arguments for provenance based on the concentration of a given word at a given period in texts of the same apparent dialect origin must be treated with a certain amount of suspicion, particularly in view of the absence of published concordances of so much of the small corpus of Old English texts that has come down to us; see, e.g., Sisam, Kenneth, Studies in the History of Old English Literature (Oxford, 1953), pp. 119–39Google Scholar, and the qualifications made to the claims of earlier word-geographers, Theodulfi Capitula, ed. Sauer, pp. 212–76.Google Scholar Given the small corpus and the very restricted number of early West Saxon authors whose work we have, both personal preference and the limited opportunities afforded by the subject matter have to be taken into consideration. In this connection it should be noted that Ps(B) retains bebycgan in both the places Ps(A) has it.

⁷⁸ BlHom, p. 63, line 7, beside pp. 69, line 8, 69, line 13, 75, line 22, and 79, line 23. The two homilies involved, nos.vandvi, contain other allegedly Anglian features; see Menner, ‘The Anglian Vocabulary of the Blickling Homilies ’, pp. 56–64. See Rushworth 1 glosses bebygið vel sellað and sylle vel bebycge (The Holy Gospels in Anglo-Saxon, Northumbrian and Old Mercian Versions, ed. Skeat, W. W. (Cambridge, 1871–1877)Google Scholar, text of Oxford, Bodleian Library, Auct. D. 2. 19, also known as the Macregol Gospels). Of course scribal interference can never be ruled out, as variant readings in Bede, GD, Mart and psalter glosses show: e.g., Bede, p. 130, line 33; GD, p. 64, line 7; Mart, p. 94, line 16; and Berghaus, , Verwandtschaftsverbältnisse, p. 96.Google Scholar

⁷⁹ Ps(P)'s use of bebycgan may also have been influenced by contextual considerations: xliii.14 ‘Þu us be boh test 7 bewrixledest 7 nan folcmidus ne gehwyrfdest ’for Roman and Gallican,xliii. 13 ‘Vendidisti populum tuum sine pretio et non fuit multitudo in commutationibus nostris [var. eorum]’; exploiting the identical prefixes of bebohtest and bewrixledest and the trisyllabic structure of bebohtest and gebwyrfdest. Sellan in the sense ‘give’ occurs freely in both Ps(P) and the accepted works of Alfred. Other words in Ps(P) with an allegedly Anglian flavour are strœl (1 ×, beside fla and flan (4 ×)) and herenes (1 ×, beside hering etc.); see also giornes (1 ×) and cf. Jordan, , Eigentümlichkeiten, p. 75;Google Scholar Menner, ‘The Anglian Vocabulary of the Blicking Homilies ’, p. 58; and GD, ed. Hecht 11, 147, Anm. 7. However, the form giornes is found in Solil and herenes in CP and so, even if we have to suppose the influence of Alfred's Mercian helpers, the presence of similar forms in Ps(P) does not rule out authorship by Alfred for that work. Moreover, though surviving Anglian texts may well have preferred strœl to flan, the current West Saxon word, the Middle English evidence indicates exactly the reverse. In any case in Ps(P) strœl may have been deliberately selected at the expense of the ‘normal' form for the sake of alliteration: xvii. 14 He sende his stræ[las] 7 hi tostencte.’

⁸⁰ Note also how the b of gebeorhstow connects it with behringed and bestanden, linking the opposed ideas of protection and aggression. The Gallican text of the vulgate psalms reads not pressure but tribulatione.

⁸¹ A fifth instance of frofor is found in the prayer at the end of MS B of Bo. Cf. Ps(P) bismrung (1 ×) instead of normal bismer (7 ×) and leahtor (1 ×) instead of leahtrung (3 ×).

⁸² Similarly healsung is used twice for deprecatio instead of normal gebed, because on each occasion the latter is being used in the same verse for Latin oratio: vi.7 ‘7 God gehyrde mine healsunge 7 Drihten onfeng min gebed’, corresponding to the Roman, Gallican and Hebrew psalters, VI.10, ‘exaudivit [var. audivit] Dominus deprecationem meam. Dominus orationem meam suscepit [var. adsumpsit and suscipiet]’ and xxxviii.14 ‘Drihten, gehyr min gebed 7 mine healsunga’, corresponding to the Gallican psalter, xxxviii. 13 ‘exaudi orationem meam Domine et deprecationem meam’; see, further, below, p. 93.

⁸³ I include here instances of hopian with the separable prefix to, taken by Carnicelli and Carlson as the verb to hopian.

⁸⁴ For tobopa beside byht, see below, p. 89.

⁸⁵ Once in the group gehyhte 7 wende. The distribution of forms rendering Latin sperare is gehyhtan Bede (6 ×) and WS (3 ×), (ge)wenan Bede (5 ×) and WS (2 ×), wilnian Bede (1 ×) and WS (1 ×) and bopian WS (1 ×). Cf. GD, p. 27, line 23, where the gebibte of MSS C and O appears as bopode in H.

⁸⁶ See, further, above, p. 75. The pp. adj. gefœgen is not included here.

⁸⁷ For Bede's usage, see above, p. 75. BIHom has gefeon (18 ×), (ge)blissian (11 ×) and wynsumian (6 ×), Latin equivalents unknown.

⁸⁸ Ps(B) on bœcling (1 ×) for Ps(A) on bœc. On hœcling is found also in Bede and BlHom.

⁸⁹ The converse is frequently also true. However, I find misleading Grinda's claim (‘Arbeit’ und ‘Mübe’, pp. 235–6): ‘ In Pl besteht eine relativ feste Bindung von earfoð an tribulatio, während in Alf CP usw. eine ganze Reihe weiterer lat. Wörter durch earfoð übersetzt wird’ and, as a result, if William of Malmesbury's statement that Ps(P) was Alfred's last work is to be believed, ‘so bliebe der schematische Zug an der Wortverwendung in Pl zu erklären’. The accepted works of Alfred certainly use earfoþ to render a range of Latin terms (adversa, adversitas, dolor, flagella, labor, pressura and tribulatio), while the majority of instances of earfoþ in Ps(P) with equivalents in the Latin psalter do indeed render Lat tribulatio. However, Ps(P) also uses earfoþ for labor and pressura (see above, n. 80) and the only other Latin term in Grinda's list which occurs in the first fifty psalms is dolor, which is translated in Ps(P) by sar, the normal rendering also in Alfred's accepted works. So the relatively restricted use of earfoþ in Ps(P) must be attributed to the Latin source, not to the translator.

⁹⁰ Ps(P) also uses aliesan, abreddan and generian in verses where the Roman psalter has eripere, which it normally renders by gefriþian. Ps(A) has aliesan for redimere, generian for eripere and eruere and gefreogan for liberare. Rauh, (Wortscbatz, p. 35)Google Scholar takes forbrytan to be late West Saxon; however, given the relative rareness of this word, its presence in Ps(P) does not make the latter a late text; see, further, Gretsch, , Regula, pp. 337–8.Google Scholar

⁹¹ Cf. Campbell, A., Old English Grammar (Oxford, 1959)Google Scholar, §474. Onliesan and aliesan occur not infrequently as variant readings in the manuscripts of GD.

⁹² Cf. WS, where confringere is rendered by forhrecan, tohrysan and forhrytan (each 1 ×), conterere by tohtrysan (1 ×) and conturbare by gedrefan (3 ×).

⁹³ See Two of the Saxon Chronicles Parallel, ed. Earle, John and Plummer, Charles (Oxford, 1892–1899; repr. 1952) 1Google Scholar, cvi–cvii, and Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, pp. 116–23;Google Scholar for my arguments for an ‘890 Chronicle’, see Ibid.passim. Unless otherwise stated, figures are based on readings from MS A, the Parker manuscript.

⁹⁴ They are not altogether accurate. For instance, a number of words said not to occur in Alfred's works are in fact there (e.g., ofsittan, gesweotolian and geuntrumian in CP; fægernes, gehiernes, genealæcan, swiftnes and geteorian in Bo; and welwilnes in Solil). The list of words in works of Alfred's time could be extended similarly.

⁹⁵ Cf. Bruce's criticism of Wichmann, ‘The Anglo-Saxon Version of the Book of Psalms’, p. 153, n. 1: ‘I will refrain from discussing the laborious comparison of the vocabulary of our text with that of King Alfred's various translations which Wichmann has made … It seems to me impossible to draw any conclusions from such comparisons … At any rate … Wichmann, himself, has not made the slightest attempt to point out in what respect the results of his comparisons tell in favour of Alfredian authorship. One may say, also, of the points of phraseology and diction … that, if they prove anything, it is simply that the text was composed in the same period as those with which it is compared.’

⁹⁶ Pp. 71–7. In this section I give figures for total occurrences of the Old English words, not, as before, linking them to specific Latin words.

⁹⁷ Mettrumnes etc.: Ps(P) 6 ×, CP 24 ×, Bo 2 × in prose and 1 × in verse and Solil 2 ×; untrumnes: Ps(P) 3 ×, CP 9 ×, Bo 1 × in prose and Solil 2 ×; untrymþ: Ps(P) 1 ×.

⁹⁸ Unriht: Ps(P) 31 ×, CP 16 × and Bo 5 × in prose and 1 × in verse; unrihtwisnes: Ps(P) 15 ×, CP 14 × and Bo 5 × in prose. Cf. woh in Ps(P), CP and Bo, a form found also in Or. Eleven of CP's fourteen instances of unrihtwisnes render Lat iniquitas.

⁹⁹ Geswinc: Ps(P) 4 ×, CP 37 ×, Bo 8 × in prose and 3 × in verse and Solil 3 ×; swincan: Ps(P) 4 ×, CP 15 ×, Bo 12 × in prose and 3 × in verse and Solil 2 ×. See also below, n. 101. For gewinn and winnan in the sense ‘fight’, ‘struggle’, see below, p. 91, and cf. Bo, Lay IV, lines 55b–6, where the ‘hit nu eall winð on þam yðum þisse worulde’ of the prose (p. 10, line 27) appears as ‘nu hi on monegum her / worulde yðum wynnað 7 swincað.’

¹⁰⁰ Offrung: Ps(P) 10 × and CP 7 ×; lac: Ps(P) 1 × and CP 19 ×.

¹⁰¹ Earfoþ: Ps(P) 54 ×, CP 15 ×, Bo 17 × in prose and 2 × in verse; earfoþnes: CP 3 ×, Bo 1 × in prose. For geswinc, see above, n. 99. Gedrefednes occurs in both Ps(P) (1 ×, for conturbatio) and the accepted works of Alfred (CP 5 ×, Bo 9 × in prose and Solil 2 ×, with gedrefnes 2 × in Bo verse, Solil 1 X). Gedeorf is not found in either corpus.

¹⁰² Ps(P) 5 ×, CP 1 × and Bo 6 × in prose.

¹⁰³ For Ps(P)'s figures, see above, p. 74. Ps(P) also has one instance of hweorfan ymb where the Gallican text of the vulgate psalms has the verb circumdare, however, Ps(A), like the ‘Hebrew’ text, here uses the verb circuire, a word rendered by hweorfan ymb elsewhere in Ps(P). The usage of the accepted works of Alfred is CP behringan, besittan, ymbbringan and ymbsettan (each 1 ×) and ymbsittan (3 ×); and Bo ymbstandan (1 × in prose and 1 × in verse). Ps(P) also has ofsittan (2 ×), rendering Lat obsedere; ofsittan also occurs in CP (4 ×), Bo (3 ×) and Solil (1 ×), in the sense ‘to beset’, ‘to oppress’.

¹⁰⁴ I do not include an entry t o correspond t o the (h) of the first section, since equivalents of confirmare are absent from the accepted works of Alfred; however, gestrangian occurs three times in CP for confortare and roborare, while getrymian or getrymman is found sixteen times in CP and once each in Bo prose and Solil and trymman four times in CP and once in Solil. Similarly I omit any reference to (j) since I have found no equivalent for deficere in the accepted works of Alfred. However, Bo has geteorian in the sense ‘to get tired’. For Ps(P)'s usage, see above, p. 74

¹⁰⁵ Gaderian: Ps(P) 4 ×, CP 11 ×, Bo 7 × in prose and Solil 2 ×; gegaderian: Ps(P) 6 ×, CP 5 ×, Bo 26 × in prose and 1 × in verse and Solil 1 ×; gesamnian: CP 1 × and Bo 2 × in prose and 5 × in verse (the presence of the larger number of forms in the verse being quite possibly significant).

¹⁰⁶ Anbidian: Ps(P) 7 × an d Solil 1 ×; geanbidian: Bo 2 × in prose; gebidan: Ps(P) 3 ×, CP 1 ×, Bo 4 × in prose and 2 × in verse and Solil 1 ×; bidan. CP 4 ×. For hopian, see above, p. 83.

¹⁰⁷ Fagnian: Ps(P) 23 ×, CP 22 ×, Bo 19 × in prose and 2 × in verse and Solil 5 ×; blissian: Ps(P) 19 ×, CP 10 ×; gefeon: Ps(P) 2 × and CP 6 ×. For wynsumian (Ps(P) 2 ×), see above, p. 84. Th e tw o instances of gefeon in Ps(P) correspond to the Gallican psalter, supergaudere, Roman insultare; see also Ps(P) xxxiv. 19, where blissian corresponds to supergaudere in the Gallican, laetari in the Hebrew and insultare in the Roman.

¹⁰⁸ (Ge)bu(i)an: Ps(P) 8 ×, CP 1 × and Bo 7 × in prose; eardian: Ps(P) 5 ×, CP 4 ×, Bo 3 × in prose and 2 × in verse and Solil 2 ×. Wunian is used also for ‘to dwell’ as well as ‘to remain’, ‘to continue’, both in Ps(P) and the accepted works of Alfred.

¹⁰⁹ Gieldan: Ps(P) 13 ×, CP 6 × and Bo 5 × in prose and 1 × in verse; forgieldan: Ps(P) 3 ×, CP 4× and Bo prose and Solil each 1 ×.

¹¹⁰ Ps(P) 2 ×, CP 3 ×. Cf. arleasnes: CP 1 × and Bo 1 × in prose; and arleast: Bo 2 × in verse.

¹¹¹ Ps(P) 1 × and wel la Bo 2 × in prose, Solil 1 ×.

¹¹² For Ps(P)'s usage, see above, p. 76. In CP faran outnumbers feran more than 5:1, while Bo has faran some twenty-two times in prose and ten times in verse, gefaran three times in prose and feran three times in prose and once in verse and Solil has faran six times. Gan occurs in Ps(P) nine times, CP thirty-four times and Bo fourteen times in prose and once in verse; gangan is found in Ps(P) five times, CP ten times and Bo three times in verse, with eode the only form in the preterite. Ps(P)'s examples of gangan are all first person singular, a form never found in Bo.

¹¹³ Goderian occurs once, gegaderian thirty-three times and gesamnian five times.

¹¹⁴ Bu(i)an occurs five times in the section dealing with Ohthere's report to King Alfred, with gebud once and gebun twice in this section and gebudon once and gebun twice elsewhere in the work. Wunian is used for both ‘to dwell’ and ‘to remain’, Eardian is found three times (including two instances in Ohthere's report).

¹¹⁵ Twice, in the same sentence. Or also uses both earfoþ and geswinc (each 4 ×) and mettrumnes and untruntnes (each 1 ×).

¹¹⁶ Or has nearly 200 examples of intransitive faran, with gefaran ‘to go’ thirty-four times and faran six times. Gan occurs three times, beside gangan three times and gegan once.

¹¹⁷ In the accepted works of Alfred blotan occurs only once (in CP).

¹¹⁸ Ymbsittan occurs four times and besittan twenty-eight times. Cf. also the range of words used to describe surrounding by water: behabban, belicgan, ymblicgan, ymbhabban, ymbfon and licgan ymbutan; and see also uton ymbfaran.

¹¹⁹ Bidan and gebidan each occurs five times and onbidan occurs once. Hopian, which Ps(P) occasionally uses for expectare, is not found in Or.

¹²⁰ Gegaderian is found three times and gesamnian once, while the figures for the following section, 891–900, are 6:1. The single instance of eardian is in annal 491 in MSS A and E, MSS B, C and F reading waron, while bu(i)an occurs also in the 891–900 section. For the distribution of faran (35 × in MS A) and feran (7 ×), see Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, pp. 111Google Scholar, n. 1, and 127; gan occurs twice, with nine instances in the ‘genealogical preface’ alongside two instances of agan.

¹²¹ See Kleinere Angelsächsische Denkmäler 1, Bibliothek der Angelsächsischen Prosa 6, ed. Leonhardi, Günther (Hamburg, 1905).Google Scholar

¹²² I include here usages which word studies suggest may be generally restricted to Anglian dialects in the later period but which occur also in early West Saxon texts.

¹²³ In the sense ‘power’, ‘strength’, ‘might’; cf. Gretsch, , Regula, pp. 347–8Google Scholar, and Theodulfi Capitula, ed. Sauer, pp. 252–3.Google Scholar It should be noted that, while crœft, like mœgen, is found in Ps(P), C P and Bo, miht is confined to Bo (3 × in prose and 14 × in verse). See, further, the discussions of these words by Käsmann and Gneuss, cited by Sauer, Ibid. p. 253.

¹²⁴ Ps(P) 5 ×, CP 15 ×, Bo 4 × in prose and 2 × in verse and Solil 1 ×; beside bliss (Ps(P) 11 ×, CP 7 × and Bo 8 × in prose and 1 × in verse) and bliþnes (Bo 2 × in prose). For details of late West Saxon usage and for Ostheeren's conclusion that gefea is the central word for Freude in early West Saxon, see Gretsch, , Regula, pp. 335–6Google Scholar, and Theodulfi Capitula, ed. Sauer, p. 228;Google Scholar see also The West-Saxon Gospels, ed. Grünberg, p. 322.Google Scholar

¹²⁵ Ofermetto: Ps(P) 2 ×, CP 35 ×, Bo 7 × in prose and 3 × in verse and Solil 1 ×; ofermod: Ps(P) 5 ×, CP 15 × and Bo 5 × in prose and 3 × in verse; ofermodlic(e): Ps(P) 1 ×, CP 3 × and Bo 1 × in prose. Cf. ofermodnes (CP 4 × and Bo 1 × in prose) and ofermedu, ofermodgung and ofermodig (each C P 1 ×); beside oferhygd (CP and Bo prose each 1 ×) and oferhygdig (CP 1 ×). The usages o f Ps(A) is oferhygd. See, further, Schabram, Superbia, passim, and Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 444, 455 and 456.Google Scholar

¹²⁶ Ps(P) 8 ×, CP 17 ×, Bo 2 × in prose and 1 × in verse and Solil 8 ×. Cf. hyht (CP 1 ×) and hyhtlic (Bo 1 × in verse) and (to)hopian beside gehybtan, for which, see, further, above, p. 83, and see Theodulfi Capitula, ed. Sauer, p. 263.Google Scholar The normal Old English word for Lat spes is hyht.

¹²⁷ Ps(P) and CP also use uncyst, while Bo has leahtor (2 × in verse), a word which in Ps(P) occurs alongside leahtrung in the sense ‘opprobrium’, ‘reproach’; see, further, Gretsch, , Regula, pp. 343–4Google Scholar, and Theodulfi Capitula, ed. Sauer, pp. 249–50.Google Scholar

¹²⁸ Ps(P) 1 × and CP 4 ×, for clamare; beside (ge)clipian (Ps(P) 32 ×, CP 24 ×, Bo 6 × in prose and 2 × in verse and Solil 11 ×); see, further, Jordan, , Eigentümlichkeiten, p. 93.Google ScholarCigan is found also in the Laws of Alfred.

¹²⁹ Ps(P) 16 ×, CP 10 ×, Bo 7 × in prose and 1 × in verse and Solil 4 ×. Bo also has feon (1 × each in prose and verse). Cf. feoung Ps(P) 1 ×, CP 4 × and Bo prose 2 ×), beside ‘ normal’ West Saxon hete; and see Theodulfi Capitula, ed. Sauer, p. 238.Google Scholar The usage of Ps(A) is feon and feoung.

¹³⁰ See Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 448, 451, 455 and 458Google Scholar, n. 200. For eac swelce in Or, see below, n. 133.

¹³¹ This usage has caused a great deal of trouble to Anglo-Saxon scribes as well as to modern editors and commentators. All five instances in Ps(P) are obscured either by emendation by Bright or by scribal error, with confusion of s and r (a common error in Anglo-Saxon minuscule and one found elsewhere in Ps(P); e.g., xlviii. 19 thorn;œs for adverb þœr): ×, Introduction swa þes spearuwa and xxi.5 swa þes wyrm (described, The Paris Psalter, ed. Colgrave et al., p. 16Google Scholar as ‘a curious use of the demonstrative þes in similes’) beside xxi.12 swa þoelig;r weax, xxxvi.19 swa ðer smec and xlv.3 swa þer muntas (emended by Bright to swa þœt weax, swa swa smec and swa þa muntas); cf. CP, p. 90, line 19 swa ðer bieme MS Cii sua ðer, with variant ðœr in C12, ðes in H and oþer in R5 and 12). For Horgan, Dorothy (‘The Relationship between the OE MSS of King Alfred's Translation of Gregory's Pastoral Care’, Anglia 91 (1973), 153–69, at 156)Google Scholar H's reading ðes ‘seems to be better than that in the other manuscripts’. ‘It seems likely’, she says, ‘that there was some difficulty in reading a primary source in which the letter s may have been confused with the letter r. The reading of H indicates that the scribe has resolved satisfactorily the difficulty presented by the spelling ðer … T and U [= Carlson's R5 and I2], both inheriting r instead of s, have a form which has been created to fit the context; J [ = Carlson's Ju] and Cii appear to have kept the misreading, whilst the scribe of CC [= Carlson's C12] substitutes another word.’ For Carlson (The Pastoral Care 1, 150), on the other hand, Horgan's ‘general line of argumentation cannot be accepted without reservations’. His tentative suggestion is that the original reading was sua oðer bieme and he compares the construction swilce oðer, described by Tengstrand, E., ‘A Special Use of Old English oþer after swilce’, SN 37 (1965), 382–92.Google Scholar However, if we add to the instances of swa þer and swa þes in Ps(P) and CP the four instances of swaþer or swœþer in this context in Bo (3 x in prose and 1 × in verse, with two instances altered in MS B to swa þœt and swa þe respectively) and a further instance in the ‘890 Chronicle ’ (473 A swa þœr fyr, with þœr subsequently erased), there seems no doubt at all that, whatever its ultimate origin and affiliations, the construction swa þer or swaþer is not only genuine but also associated particularly with texts of the time of Alfred.

¹³² Beside crœft it has both mœgen and miht; beside gefea (3 ×) it has bliþnes (1 ×). The other figures are ofermetto 7 ×, tobopa 1 × and unþeaw 2 ×. Ps(A) differentiates between miht for potestas and potentia and mœgen for virtus and vis.

¹³³ For ofermodig (4 ×), see Schabram, Superbia, passim, and Bately, , ‘King Alfred and the Old English Translation of Orosius’, p. 444;Google Scholar for eac swelce (3 ×), see Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, p. 126Google Scholar and n. 3, and Theodulfi Capitula, ed. Sauer, pp. 136–7.Google ScholarClipian occurs twice. The concepts ‘to hate’ and ‘proudly’ are not found, but cf. hete (2 ×).

¹³⁴ See above, n. 131.

¹³⁵ See, further, the discussion of the distribution of wisdom and snyttro, See bold, , ‘Sapiens und prudens ’, pp. 291–33.Google Scholar

¹³⁶ Used intransitively in the sense ‘to appear’.

¹³⁷ One of these has been altered to œtiewan in MS A. The first ‘original’ instance of œtiewan in this manuscript is s.a. 892. For Lat ostendere, ostendi and (ap)parere Ps(A) has oteawan, while Bede's normal usage appears to have been œteawan (common MS variant œtywan).

¹³⁸ See Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 450–3, 455 and 457.Google Scholar

¹³⁹ Cf. Bede, where the order is feohtan, winnan, with campion in third place, followed by gtfeohlan and gewinnan; and see, further, Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, pp. 104, 105, 107, 114 and 122–3.Google Scholar

¹⁴⁰ See, further, Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 445, 451, 453, 455 and 456Google Scholar, and ‘The Compilation of the Anglo-Saxon Chronicle’, p. 119, n. 2. Ps(A) normally uses gefultumian, with fultumian once. Cf. Or's use of acwencean as well as the Alfredian adwœscan (Ps(P) adwœscan and gedwœscan), for which, see Theodulfi Capitula, ed. Sauer, pp. 218–19.Google Scholar

¹⁴¹ Ps(A) differentiates between mildheortnes for misericordia and milds for miseratio and propitiatio.

¹⁴² See, further, Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, pp. 104, 107, 113Google Scholar, n. 4, 114 and 123, n. 4. Ps(A) has only adrifan

¹⁴³ That Ps(P) shows an apparent preference for hattt over is (ge)haten, while the converse is true of CP and Bo, is no argument against common authorship. In both Ps(P) and CP hatte is the normal usage with personal names, the only exceptions being one instance of is gehaten in Ps(P) (as the variant to the hatte immediately preceding in the same verse) and three instances of is (ge)haten in CP (one of these as a variant to a preceding hatte). In Bo the figures for the prose are hatte with personal name eight times and is (ge)haten nine times (once as a variant to a preceding hatte and four times in a duo nomine or tria nomina construction with several names for the same person, a context not occurring in Ps(P)). It should be noted that in Ps(P) the longest run of instances of hatte uninterrupted by an alternative construction is two, the norm also in Bo.

¹⁴⁴ Hatan is also used frequently in first and second person singular and second and third person plural constructions in Bo. At first sight the usage of the Lœceboc seems very similar to that of Ps(P) and the accepted works of Alfred, with an order of preference hatte (22 ×), we/men hataþ (5 ×); however, the majority of instances of hatte occur in the paratactic construction ‘× hatte wyrt’ (e.g., ‘ nim weax 7 hemlic hatte wyrt’ (Kleinere Denkmäler, ed. Leonhardi, p. 45Google Scholar, line 31), with ‘þa wyrt þe hatte ×’ in second place and ‘þa wyrt þe × hatte’ (the norm in Ps(P) and the accepted works of Alfred) in third place.

¹⁴⁵ Only instances where these verbs are interchangeable are included in the statistics that follow; the past participle gebaten is taken to be part of the verb hatan and gecweden of cweþan.

¹⁴⁶ In the form gecweden and limited to the introductions to pss. 11, iv and v.

¹⁴⁷ A second instance of cweþan (in the form gecweden) is found only in MSS C, Cii and C12, the other manuscripts reading genemned (H) and gehaten (R5 and 12).

¹⁴⁸ See, further, Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, pp. 104, 107, 114 and 123.Google Scholar The construction þœs nama is is also found frequently in Mart (beside se is on naman) and Bede, while GD has a preference for þœm is nama. Mart and Bede also share a marked preference for constructions of the type is (ge)nemned over (þe) man nemneþ, the latter construction being totally absent from GD. In the case of Bede the frequency of these constructions is to a large extent determined by the usage of the Latin source.

¹⁴⁹ See, further, Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 444, 451 and 457Google Scholar, and, for the usage in manuscripts of WS, The West-Saxon Gospels, ed. Grünberg, p. 321.Google Scholar

¹⁵⁰ One instance of gylt in CP occurs in the phrase ðara scyldegena gyltas, presumably to avoid repetition of the root scyld-. See, further, Bately, , ‘King Alfred and the Old English Translation of Orosius’, pp. 443, 451–3, 455 and 457Google Scholar, and Büchner, G., Vier altenglischc Bezeichnungen für Vergehen und Verbrechen (firen, gylt, man, scyld) (Berlin, 1968).Google Scholar

¹⁵¹ Two of the three instances of ben in CP occur in the same sentence, with MS H reading bed for the second of them (p. 399, line 31). In Ps(A) gebed is reserved for operatio and oratio and ben for deprecatio, petitio and prex. Ps(P), however, uses gebed indiscriminately. For Ps(P) healsung, see above, n. 82.

¹⁵² It frequently renders prex and prices by the pairs gebed 7 ben, bena 7 gebeda etc. GD also has a large number of instances of ben beside gebed in a ratio of approx. 1:3.

¹⁵³ BT also records the construction hit is cyn þœt in the Laws of Ine. Ps(P) and the accepted works of Alfred also use gemetlic, ‘fitting’, ‘suitable’, but never in a construction of this type.

¹⁵⁴ For full details of the distribution of these words in Ps(P), the accepted works of Alfred and Or and for a number of other instances of differences, see Bately, ‘King Alfred and the Old English Translation of Orosius’,passim.

¹⁵⁵ The ‘890 Chronicle ’ agrees with Ps(P) and the accepted works of Alfred in using eaþmod.

¹⁵⁶ Ps(P) and the accepted works of Alfred also show the same preferences in their choice of conjunctions to introduce subordinate clauses; e.g., Ps(P), CP and Bo œr œr, a form which I have not noted in Bede, GD or Or, and a marked preference for þam (þam) over for þœm þe, for þœm þœat etc. and over forþon. Indeed given the special relationship between Ps(P) and its Latin source there appears to be no significant difference in usage between Ps(P) and the accepted works of Alfred in respect of the constructions studied by Elizabeth Liggins (see her ‘The Authorship of the Old English Orosius’, pp. 57–90).

¹⁵⁷ For the dangers inherent in ignoring the wording of the underlying Latin sources, see Bately, , ‘The Compilation of the Anglo-Saxon Chronicle’, p. 118Google Scholar and n. 3, and ‘World History in the Anglo-Saxon Chronicle: its Sources and its Separateness from the Old English Orosius’, ASE 8 (1979), 177–94Google Scholar, at 190 and n. 1. In contrast, Bromwich's arguments for Alfred's authorship of Ps(P) would have carried far greater conviction had his lists of correspondences included the Latin equivalents where available.

¹⁵⁸ We may compare word pairs in Bede, such as mod 7 magen for vires, geon 7 geomertung for gemitus and snyttro 7 wisdom for both prudentia and sapientia, none of which is found in Ps(P) or the accepted works of Alfred. See, further, Koskenniemi, Inna, Repetitive Word Pairs in Old and Early Middle English Prose, Annales Universitatis Turkuensis 107 (Turku, 1968).Google Scholar

¹⁵⁹ For Latin inutiles: e.g., Ps(P) xiii.4 and xxv.4; beside CP, p. 271, lines 7–8, and Bo, p. 68, line 29. Similar constructions are found in Bede and Ælfric's works.

¹⁶⁰ Ps(P) ix.14 and xxx.21 and 22; beside Bo, p. 131, line 6. CP has gebal 7 ophealdan. See also GD, p. 98, line 20 geheold 7 gehœl.

¹⁶¹ Ps(P) xvii.31 and 37 and xxxvii.10; beside CP, p. 41, line 11, p. 163, line 8, and p. 465, line 5, and Bo, p. 72, line 11, and p. 108, line 28, etc.

¹⁶² Ps(P) xvii.19, xx.i, xxxvii.8 and xlv.1; beside CP, p. 199, lines 15 and 17, Bo, p. 51, line 8, and p. 107, line 31, and Solil, p. 57, line 11, and p. 63, line 15, etc. So also Ælfric, beside Or swa swiþlice.

¹⁶³ Ps(P) 6 ×; beside CP, p. 43, line 17, and Bo p. 140, line 28, etc.

¹⁶⁴ Ps(P) some 23 ×, occasionally coupled with heorte. The normal rendering of cor and anima is, of course, heorte and sawol. See, e.g., CP, p. 219, line 2, p. 283, line 11, and p. 306, line 12.

¹⁶⁵ Ps(P) vii.17 and Bo, p. 141, line 7. See also CP, p. 389, line 20 ‘ðin sio swiðre and The Life of St Chad, ed. Vleeskruyer, p. 48.

¹⁶⁶ ‘Who was the Translator?’, p. 296, n. 2.

¹⁶⁷ See Bromwich's comment that ‘speculation about the degree of help given to the king by his different assistants cannot be very profitable until there is an absolutely complete Old English Dictionary. Even then the material from the works of King Alfred's reign would have to be transferred t o modern card-indexes with electrical sorting machinery before their statistical pattern could be computed. With so many different persons possibly involved, it is doubtful in the extreme if any unanimity in interpreting these statistics could be achieved, so it is likely that the matter must rest where it is’ (Ibid. p. 303, n. 1). I would add that a similar analysis of the usage of, say, the Ælfric canon would be necessary as a control.

¹⁶⁸ I have heard it suggested that King Alfred's involvement in ‘his’ translations was possibly purely nominal. Lexical studies cannot, of course, either prove or disprove that theory. But, if Alfred's authorship were to be rejected on non-linguistic grounds, the linguistic evidence is still that overall responsibility rested with one man.