¹ See, for example, Cook, A. S., ‘An Irish Parallel to the Beowulf Story,’ Archiv für das Studium der neueren Sprachen und Literaturen 103 (1899) 154–56; Gerould, G. H., ‘Offa and Labhraidh Maen,’ Modern Language Notes 17 (1902) 201-3; Olsen, O. L., ‘Beowulf and “The Feast of Bricriu,”’ Modern Philology 11 (1913/14) 407-27; Whitbread, L., ‘Grendel's Abode: an niustrative Note,’ English Studies 22 (1940) 64-66; Donahue, C., ‘Grendel and the Clanna Cain,’ The Journal of Celtic Studies 1 (1949/50) 167-75; Carney, James, Studies in Irish Literature and History (Dublin 1955; rev. imp. 1979) 77-128; Puhvel, Martin, ‘Beowulf and the Celtic Under-water Adventure,’ Folklore 76 (1965) 254-61, and ‘The Deicidal Otherworld Weapon in Celtic and Germanic Otherworld Tradition,’ Folklore 83 (1972) 210-19. I have not seen Puhvel's book, Beowulf and Celtic Tradition (Waterloo, Ontario 1979).Google Scholar

² This paper was first prepared for a colloquium on the date of Beowulf which met under the auspices of the Centre for Medieval Studies of the University of Toronto in April 1980. Most of the papers will appear in The Date of Beowulf, ed. Chase, Colin (Toronto 1981). I am obliged to many members of the colloquium for discussion and comment; specific debts are acknowledged in the relevant footnotes, but I should like to record special thanks to Dr. Ashley Amos both for helpful discussion and for pre-publication access to her important book, to which frequent reference is made in this article.Google Scholar

³ There is no general survey available. See Blair, Peter Hunter, The World of Bede (London 1970); Mayr-Harting, Henry, The Coming of Christianity to Anglo-Saxon England (London 1972); Hughes, K., ‘Evidence for Contacts Between the Churches of the Irish and English from the Synod of Whitby to the Viking Age,’ in England before the Conquest: Studies in Primary Sources Presented to Dorothy Whitelock, edd. Clemoes, Peter & Hughes, K. (Cambridge 1971) 49-67; and, for the ninth and tenth centuries, my O'Donnell Lectures cited in n. 4. Google Scholar

⁴ For full discussion of Anglo-Welsh contacts, see Dumville, D. N., Wales in the Early Middle Ages (Oxford 1982). For links with Strathclyde, Cornwall, and Brittany, see my forthcoming volume of O'Donnell Lectures, England and the Celtic World in the Ninth and Tenth Centuries. There is no satisfactory comprehensive survey of Anglo-Brittonic links in the seventh and eighth centuries; for the fifth and sixth centuries see Kenneth H. Jackson, Language and History in Early Britain (Edinburgh 1953) 194-261.Google Scholar

⁵ See Ionae Vitae Sanctorum Columbani Vedastis Iohannis, ed. Krusch, Bruno (Hannover 1905), or in MGH, Scriptores rerum Merovingicarum IV. For a translation (by Munro, D. C.), see Peters, Edward (ed.), Monks, Bishops and Pagans: Christian Culture in Gaul and Italy, 500-700, Sources in Translation (Philadelphia 1975) 75-113. Google Scholar

⁶ HE 4.13. Bede makes the same point about Augustine and his monks in their early days at Canterbury: HE 1.26.Google Scholar

⁷ We know that the Christian mission in Northumbria was continued during the difficult years 633-35 by James the Deacon (HE 2.16 and 20; 3.25); but it is not to be supposed — I think — that the Roman mission had left any substantial body of Christians with whom the Irish would have found it easy to communicate.Google Scholar

⁸ HE 3.1 and 3.Google Scholar

⁹ HE 3.3; of Aidán Bede says ‘qui Anglorum linguam perfecte non nouerat’!.Google Scholar

¹⁰ See Adomnán's Life of Columba, edd. & tr. Anderson, A. O. & Anderson, M. O. (Edinburgh 1961) 485–88 (3.10). The source is, of course, a century later than the events of which it speaks; it is therefore possible that the presence of an Englishman reflects conditions at Iona in Adomnán's time rather than Columba's.Google Scholar

¹¹ HE 1.25.Google Scholar

¹² HE 1.2. See also The Earliest Life of Gregory the Great, ed. & tr. Colgrave, Bertram (Lawrence, Kan. 1968) 90–91.Google Scholar

¹³ Gregory, , Epistola ad Candidum (ca. 595), in MGH, Epistolae I (Gregorii I Papae Registrum epistolarum, t. I libri 1-8), edd. Ewald, Paul & Hartmann, L. M. (Berlin 1887-91) 388–89. Compare the behaviour of Bishop Aidán, reported by Bede, HE 3.5: ‘Denique multos, quos pretio dato redemerat, redemtos postmodum suos discipulos fecit, atque ad sacerdotalem usque gradum erudiendo atque instituendo prouexit.’ Google Scholar

¹⁴ On this aspect of the Roman mission see Jones, P. F., ‘The Gregorian Mission and English Education,’ Speculum 3 (1928) 335–48.CrossRefGoogle Scholar

¹⁶ HE 3.14. Ithamar was a native of Kent. His predecessor, Paulinus, died on October 10, 644. The next English bishop was Thomas, who succeeded Felix in East Anglia in 648 (HE 3.20).Google Scholar

¹⁶ HE 3.15 and 21. He had a brother, Adda, who was also a priest (3.21).Google Scholar

¹⁷ HE 3.26, discussing the aftermath of the Synod of Whitby in 664. Eata became bishop of Bernicia ‘non multo post’: fourteen years in fact, in 678.Google Scholar

¹⁸ HE 3.26-28. In 664 Tuda became bishop of Northumbria. Bede describes him as ‘apud Scottos austrinos eruditus atque ordinatus episcopus’ (HE 3.26); he had arrived in Northumbria, already a bishop, during Colmán's episcopate (661-64). Colgrave, Bertram & Mynors, R. A. B., Bede's Ecclesiastical History of the English People (Oxford 1969) 308 n. 3, take him to be an Irishman, but I can see no justification for this. Unless his name is an Old English hypocoristic from an Irish name such as Tóthal (Old Irish Tuathal), which seems most unlikely, he must be an Englishman: for others called Tuda or Tudda see Searle, William G., Onomasticon Anglo-Saxonicum (Cambridge 1897) 460; they are perhaps hypocoristics of Old English Theod- names. Later in 664, Wilfrid and Ceadda also became bishops in Northumbria (HE 3.28).Google Scholar

¹⁹ He was the first abbot of Gilling in Deira (HE 3.24), founded after (postmodum : HE 3.14) 651; as a relative of King Oswine (3.24) he was presumably a Deiran. He was educated by the Irish (though probably in England): ‘edoctus et ordinatus a Scottis.’ The precise date of his succession to the bishopric is uncertain, but it seems to have been after (and probably shortly after) King Wulfhere's accession in 658.Google Scholar

²⁰ HE 3.22: after 653, but not necessarily long after. On the chronological problems see Plummer, Charles (ed.), Venerabilis Baedae opera historica (Oxford 1896) II 176–79.Google Scholar

²¹ The circumstances are not directly comparable, of course. King Oswald and some of his nobles were already Christian on his assumption of power in Northumbria in 634, whereas Augustine enjoyed no such advantage in Kent. Also, it is difficult to be clear about the extent to which Paulinus' mission during the period 625-33 laid reliable foundations on which Aidán could build; cf. n. 7.Google Scholar

²² HE 3.21. This should be noted, for it gives the lie to any notion that Irish episcopi uagantes were roaming the area in any numbers.Google Scholar

²³ HE 1.34. Bede concludes, ‘Neque ex eo tempore quisquam regum Scottorum in Brittannia aduersus gentem Anglorum usque ad hanc diem in proelium uenire audebat.’ Google Scholar

²⁴ We hear of this in St. Patrick's Epistle to Coroticus (§ 14), for example: see L. Bieler's text and commentary in Classica et Mediaevalia 11 (1950) 98, and 12 (1951) 203, and the work of J. N. Garvin to which he refers.Google Scholar

²⁵ HE 3.3 and 5.Google Scholar

²⁶ An important recent case-study of such matters is by Miller, M., ‘Eanfrith's Pictish Son,’ Northern History 14 (1978) 47–66.CrossRefGoogle Scholar

²⁷ Ammianus Marcellinus is the principal source: see the recent discussion by Tomlin, R., ‘The Date of the “Barbarian Conspiracy,”’ Britannia 5 (1974) 303–9.CrossRefGoogle Scholar

²⁸ We still lack a comprehensive discussion of these secular court-functionaries of Anglo-Saxon kings; however, Hermann Moisl (in papers now at press) and Hollowell, I. M., ‘Unferσ the pyle in Beowulf,’ Studies in Philology 73 (1976) 239–65, and ‘Scop and woÐbora in OE Poetry,’ Journal of English and Germanic Philology 11 (1978) 317-29, are doing their best to remedy this deficiency.Google Scholar

²⁹ Miller, M., ‘The Dates of Deira,’ Anglo-Saxon England 8 (1979) 35–61 (especially 60).CrossRefGoogle Scholar

³⁰ Vita anonyma S. Cudbercti , ed. & tr. Colgrave, Bertram, Two Lives of Saint Cuthbert (Cambridge 1940) § 28; and Vita auctore Beda, prose § 24, verse § 21. I see no reason to follow Plummer, Baedae opera historica II 263-64, in believing the later mediaeval Irish equation of Aldfrith with Fland Fína. As far as I know, Fland Fína was an abbot of Clonmacnoise who died in 733 (see the Annals of Ulster, s.a. 732).Google Scholar

³¹ See the Annals of Ulster, s.a. 730 (= 731): ‘Clericatus Echdach fili [MS filius] Cuidini, rex Saxan, et constringitur.’ Google Scholar

³² See Bruce-Mitford, R. L. S., The Sutton Hoo Ship-Burial II (London 1978).Google Scholar

³³ HE 3.19.Google Scholar

³⁴ HE 3.22: King Sigeberht was baptised by Bishop Fínán in Northumbria. Fínán appointed the Irish-trained Northumbrian Cedd as bishop; Cedd's methods were those of the Irish ecclesiastic, as Campbell, J., ‘The First Century of Christianity in England,’ Ampleforth Journal 76 (1971) 12–29 (especially 23-24), has pointed out.Google Scholar

³⁵ HE 4.13.Google Scholar

³⁶ If the letter about Bishop Dagán can be so interpreted: HE 2.4. We might note also Aldhelm's portrayal of Archbishop Theodore of Canterbury (668-90) as ‘hemmed in by a mass of Irish students’: Lapidge, Michael & Herren, M. (tr.), Aldhelm: The Prose Works (Ipswich 1979) 144, 163.Google Scholar

³⁷ HE 3.7, but Bede says nothing of Irishmen in Wessex. Bede's reference to Malmesbury (HE 5.18) as Maildubi Urbem has often been taken to indicate an Irish foundation, but there is no basis for the academic romance which this view has generated: see the salutary remarks of Lapidge & Herren, Aldhelm: The Prose Works 6–7, 9, 138, 147.Google Scholar

³⁸ HE 3.27: ‘multi nobilium simul et mediocrium de gente Anglorum.’ He is referring to the time of Bishops Fínán and Colmán of Lindisfarne (651-64) in particular. Bede also mentions (passim) a good many other individuals as having sojourned in Ireland.Google Scholar

³⁹ Lapidge, & Herren, , Aldhelm: The Prose Works, especially 143-46, 160–64. Cf. also Aldhelm's letter to Wihtfrith, ibid. 139-40, 154-55.Google Scholar

⁴⁰ HE 4.4 (and Plummer's note, Baedae opera historica II 210-11).Google Scholar

⁴¹ See Hughes, K., art. cit. (n. 3).Google Scholar

⁴² See Dumville, D. N., ‘Antecedents of Ecclesiastical Reform in Tenth-Century Northumbrian,’ Journal of Theological Studies, forthcoming.Google Scholar

⁴³ Some aspects of this question are discussed by Kahl, H.-D., ‘Papst Gregor der Grosse und die christliche Terminologie der Angelsachsen,’ Zeitschrift für Missionswissenschaft und Religionswissenschaft 40 (1956) 93–111, 190-200.Google Scholar

⁴⁴ The standard work is that of Gneuss, Helmut, Lehnbildungen und Lehnbedeutungen im Altenglischen (Berlin 1955), which gives a full bibliography of earlier studies. See also Funke, O., ‘Altenglische Wortgeographie (Eine bibliographische Überschau),’ in Anglistische Studien: Festschrift … Friedrich Wild, edd. Brunner, Karl et al. (Vienna 1958) 39-51. For details of some recent work see Gneuss, H., ‘The Origin of Standard Old English and Æthelwold's School at Winchester,’ Anglo-Saxon England 1 (1972) 75 and n. 4. I have not seen the dissertation of Manfred Scheler, Altenglische Lehnsyntax: Die syntaktischen Latinismen im Altenglischen (Berlin 1961).Google Scholar

⁴⁵ See, for example, Betz, Werner, Deutsch und Lateinisch: Die Lehnbildungen der althochdeutschen Benediktinerregel (Bonn 1949; 2nd ed. 1965). See also Reiffenstein, Ingo, Das Althochdeutsche und die irische Mission im oberdeutschen Raum (Innsbruck 1958). There is also much of general value in the fundamental work of Dennis H. Green, The Carolingian Lord (Cambridge 1965).Google Scholar

⁴⁶ On this point, see further below, pp. 127-29, 139, 143-44, 160.Google Scholar

⁴⁷ HE 3.5.Google Scholar

⁴⁸ HE 3.3-6, 14-17, 25 (especially cc. 5, 17, and 25).CrossRefGoogle Scholar

⁴⁹ HE 1.27.Google Scholar

⁵⁰ This is a considerable subject, with innumerable and complex ramifications. The best starting-point is still Kenney, James F., Sources for the Early History of Ireland: Ecclesiastical (New York 1929; rev. imp. by L. Bieler, 1966).Google Scholar

⁵¹ See, for example, Hughes, Kathleen, The Church in Early Irish Society (London 1966) 79–156 (especially 103-90).Google Scholar

⁵² For a striking reminder of this, see Sharpe, R., ‘Hiberno-Latin laicus, Irish láech and the Devil's Men,’ Ériu 30 (1979) 75–92.Google Scholar

⁵³ Again, there is nothing approaching an adequate survey of this considerable literature. It must be pieced together from the following: Grosjean, P., ‘Sur quelques exégètes irlandais du vii^e siècle,’ Sacris Erudiri 7 (1955) 67–98; Byrne, F. J., ‘Seventh-Century Documents,’ Irish Ecclesiastical Record, 5th series 108 (1967) 164-82; Hughes, Kathleen, Early Christian Ireland: Introduction to the Sources (London 1972); Díaz y Díaz, Manuel C. (ed. & tr.), Liber de ordine creaturarum (Santiago de Compostela 1972); McNally, Robert E. (ed.), Scriptores Hiberniae minores (CCL 108-B; Turnhout 1973); Bieler, Ludwig & Kelly, F. (edd. & tr.), The Patrician Texts in the Book of Armagh (Dublin 1979); Law, Vivien, The Insular Latin Grammarians (Ipswich 1981).CrossRefGoogle Scholar

⁵⁴ A convenient starting-point is Hughes, Early Christian Ireland 41–64.Google Scholar

⁵⁵ Thesaurus Palaeohibernicus, edd. & tr. Stokes, Whitley & Strachan, J. (rpt. in 2 vols.; Dublin 1975), is still the standard repertory of this gloss-material.Google Scholar

⁵⁶ See, for example, Jones, Charles W., (ed.), Bedae opera de temporibus (Cambridge, Mass. 1943); Blair, Hunter, The World of Bede; and works cited in n. 53 above, especially McNally, Scriptores pp. v-xvii, and Díaz, y Díaz, , Liber. Google Scholar

⁵⁷ It is studied by Grosjean, , ‘Sur quelques exégètes,’ and by Simonetti, M., ‘De mirabilibus sacrae scripturae: Un tratatto irlandese sui miracoli della Sacra Scritura,’ Romanobarbarica 4 (1979) 225–51. The critical edition is that of McGinty, F. P., ‘Augustinus Hibernicus, De mirabilibus sacrae scripturae’ (unpublished National University of Ireland Ph.D. diss., University College Dublin 1971).Google Scholar

⁵⁸ ‘Beowulf, Ireland, and the Natural Good,’ Traditio 7 (1949-51) 263–77.Google Scholar

⁵⁹ See further Donahue's paper, ‘Beowulf and Christian Tradition: A Reconsideration from a Celtic Stance,’ Traditio 21 (1965) 55–116. For a measured assessment of Donahue's contribution, see the important paper by Wormald, C. P., ‘Bede, “Beowulf,” and the Conversion of the Anglo-Saxon Aristocracy,’ in Bede and Anglo-Saxon England , ed. Farrell, Robert T. (Oxford 1978) 32-95 (especially 49-50, 64, 65, 66). In his 1965 paper (pp. 63-64) Donahue develops the notion, which may prove a useful concept for the seventh to ninth centuries (and perhaps later), of an ‘Insular Mode’ of Christianity; it will, however, need a great deal of testing. Much less useful — in fact, potentially very misleading — is the notion of an ‘Insular Church,’ a phrase apparently coined by Professor R. T. Farrell. It is far from clear what this means, but it seems to embody a recognition that (the now discredited concept of) ‘the Celtic Church’ had its reflexes outside the Celtic-speaking countries: see Farrell in Review of English Studies n.s. 20 (1969) 405; his excellent edition Daniel and Azarias (London 1974); and Bede and Anglo-Saxon England p. [viii]. Google Scholar

⁶⁰ Campbell, Alistair, Old English Grammar (Oxford 1959; rev. imp. 1962) 1–29 passim. Google Scholar

⁶¹ This is the date of the earliest English charters and the first works of Aldhelm. English names would have been used in correspondence before that, but we have no reliable evidence as to the system(s) used, for all such documents appear only in later copies or incorporated in later sources.Google Scholar

⁶² Meroney, H., ‘Irish in the Old English Charms,’ Speculum 20 (1945) 172–82.CrossRefGoogle Scholar

⁶³ Brown, C. F., ‘Irish-Latin Influence in Cynewulfian Texts,’ Englische Studien 40 (1908/9) 1–29, whose conclusions seem, in general, sound.Google Scholar

⁶⁴ Ker, Neil R., Catalogue of Manuscripts Containing Anglo-Saxon (Oxford 1957) 43–45, and Grant, Raymond J. S., Cambridge, Corpus Christi 41: The Loricas and the Missal (Amsterdam 1978).Google Scholar

⁶⁵ Studies in Irish Literature and History 77–128 (especially 114-28).Google Scholar

⁶⁶ See the chapters by Weisweiler, J. & Betz, W., ‘Deutsche Frühzeit,’ and especially Betz, W., ‘Lehnwörter und Lehnprägungen im Vor- und Frühdeutschen,’ in Deutsche Wortgeschichte, edd. Maurer, Heinrich & Rupp, H. (3rd ed.; Berlin 1974-78), I 55-133, 135-63. CrossRefGoogle Scholar

⁶⁷ The Middle Irish period is conventionally defined as ca. 900 - ca. 1200, the Early Modern Irish period as ca. 1200 - ca. 1650.Google Scholar

⁶⁸ The classic presentation of texts edited with these considerations in mind is by Murphy, Gerard, Early Irish Lyrics (Oxford 1956). He discusses the syllabic system fully in Early Irish Metrics (Dublin 1961). Murphy's view was that the syllabic system derived from Latin models (Metrics 11-12), but for an argument for an Indo-European origin see Watkins, C., ‘Indo-European Metrics and Archaic Irish Verse,’ Celtica 6 (1963) 194-249; Watkins' view is itself now vigorously disputed (see below, n. 70).Google Scholar

⁶⁹ Expressed by James Carney in various papers, but especially ‘Notes on Early Irish Verse,’ Éigse: A Journal of Irish Studies 13 (1969/70) 291–312.Google Scholar

⁷⁰ See Carney, J., ‘Three Old Irish Accentual Poems,’ Ériu 22 (1971) 23–80.Google Scholar

⁷¹ As against a normal figure in early Irish poetry of about 40%: Ériu 22 (1971) 71 n. 1.Google Scholar

⁷² Ériu 22 (1971) 53–80, and see also Carney's, ‘Aspects of Archaic Irish,’ Éigse 17 (1977-79) 417-35.Google Scholar

⁷³ Ériu 22 (1971) 53–80 (passim). Google Scholar

⁷⁴ Éigse 17 (1977-79) 426–27.Google Scholar

⁷⁵ Éigse 13 (1969/70) 291–312, especially 292-96.Google Scholar

⁷⁶ Compare the modern Albanian evidence discussed briefly by Hamp, E. P., ‘On Dating and Archaism in the Pedeir Keinc,’ Transactions of the Honourable Society of Cymmrodorion (1972/3) 95–103 (at 98-99).Google Scholar

⁷⁷ See, for example, Thurneysen, R., (ed.), ‘Mittelirische Verslehren,’ apud Stokes, Whitley & Windisch, E. (edd.), Irische Texte mit Übersetzungen und Wörterbuch III (Leipzig 1891-97) 1–182. The Early Modern Irish texts are edited by Bergin, O. J., ‘Irish Grammatical Tracts’ (Dublin 1915-55), issued as supplements to Ériu 8-10,14, and 17.Google Scholar

⁷⁸ This system has generally not been followed in the edition of Old Welsh texts, arguably to the considerable detriment of the discipline: cf. the remarks of Greene, David, ‘Linguistic Considerations in the Dating of Early Welsh Verse,’ Studia Celtica 6 (1971) 1–11 (especially 3).Google Scholar

⁷⁹ In spite of the major study by Bliss, Alan, The Metre of Beowulf (Oxford 1958), the new appearance of Pope, John C., The Rhythm of Beowulf (2nd ed.; New Haven 1964), and various other works on Old English metrics, there has been no direct assault on these particular topics.Google Scholar

It is to be hoped that a new departure is indicated by the publication of Amos, Ashley C., Linguistic Means of Determining the Dates of Old English Literary Texts (Cambridge, Mass. 1980).Google Scholar

⁸⁰ Wrenn, C. L. (rev. Bolton, W. F.), Beowulf with the Finnesburg Fragment (3rd ed.; London 1973) 6.Google Scholar

⁸¹ Wrenn, C. L., (ed.), Beowulf with the Finnesburg Fragment (London 1953) 31–32.Google Scholar

⁸² As pointed out in the review (of the third edition) by Robinson, F. C., Speculum 52 (1977) 188–92.Google Scholar

⁸³ First edition 21-22, 31-32.Google Scholar

⁸⁴ See below, pp. 143–44.Google Scholar

⁸⁵ See below, pp. 137–55.Google Scholar

⁸⁶ Kevin Kiernan is now bravely making such an attempt, arguing in fact for a date of composition in the reign of Cnut (1016-35): Beowulf and the Beowulf Manuscript (New Brunswick, N.J. 1981).Google Scholar

⁸⁷ Holthausen, F., (ed.), Beowulf nebst den kleineren Denkmälern der Heldensage (2nd ed.; Heidelberg 1908) I 103. Cf. also n. 112 below. For Sievers' (unpublished) reconstruction see Westphalen, Tilman, Beowulf 3150-55: Textkritik und Editionsgeschichte (Munich 1967) I 124-32 (and plates iii-iv).Google Scholar

⁸⁸ Cook, Albert S., (ed.), Judith: An Old English Epic Fragment (2nd ed.; Boston, Mass. [1889]) 75–85. The date of Judith has been a much debated question: see also Foster, T. G., Judith: Studies in Metre, Language and Style (Strassburg 1892).Google Scholar

⁸⁹ Cook, A. S., ‘Metrical Observations on a Northumbrianized Version of the Old English Judith,’ Transactions of the American Philological Association 20 (1889) 172–74.Google Scholar

⁹⁰ A thorough study, based on those few poems which survive in more than a single copy, of the variants arising in Old English verse in the course of manuscript-transmission is much to be desired as an essential preliminary to further appreciation of the limits of our understanding of the interrelationship of scribes and metrical practice: a start was made by Ricci, A., ‘The Chronology of Anglo-Saxon Poetry,’ Review of English Studies 5 (1929) 257–66 (especially 261-63).Google Scholar

⁹¹ Campbell, , Old English Grammar, §§ 341-54, 394: ‘before the period of the oldest texts, but after i-umlaut’ (p. 161); cf. Bülbring, Karl D., Altenglisches Elementarbuch I (Heidelberg 1902) §§ 433, 438–39 (in ‘Urenglisch’).Google Scholar

⁹² Campbell, §§ 363-64, 395, who pronounces it older than smoothing in Anglian; cf. Bülbring §§ 440-47 Luick, Karl, Historische Grammatik der englischen Sprache (1914-40; rev. imp. Oxford 1964) 1297 (§ 320), took the view that the orthographic evidence of the early glossaries indicated partial analogical re-creation of (spoken) forms without parasiting. Campbell (§ 363) suggests that early written forms not indicating parasiting arose in imitation of the inflected members of the declension: he does not specify the existence of spoken forms without parasiting, and that he goes on to mention as an alternative ‘scribal neglect of the parasite vowel’ may as well indicate that he was thinking of purely orthographic conventions. I labour the point because Amos (op. cit. 70-71) has read Campbell through Luickian spectacles, and has used Luick's view of the orthographic evidence to mount an assault on the relevance of this sound-change for dating Old English verse. Luick's view seems to me to be unnecessarily complicated; but, to the extent that the metrical test arising from the sound change is called into question by his view, the test must now be judged more by its results than by its theoretical strength — all the more reason, then, both for its rigid application and for its use as a proportional method. For another approach, see Lehmann, W. P., ‘Post-consonantal l m n r and Metrical Practice in Beowulf,’ in Nordica et Anglica: Studies in Honor of Stefán Einarsson , ed. Orrick, Allan H. (The Hague 1968) 148-67, and Amos' objections (op. cit. 72-76), not all of which seem to me equally well founded. Google Scholar

⁹³ Campbell, §§ 231, 241.2a, 242, who says that it occurred later than breaking, but generally earlier than smoothing; cf. Bülbring § 529 (by ca. 700).Google Scholar

⁹⁴ Campbell, §§ 234-39, 461.Google Scholar

⁹⁵ Campbell, (§ 255) has established the following sequence of sound-changes, all of which are to be placed before ca. 700: (i) i-mutation; (ii) back-mutation; (iii) smoothing; (iv) reduction of hiatus and compensatory lengthening. These are the last major prehistoric changes in accented vowels. The changes with which we are concerned are related to this scheme as follows: (a) syncope of medial vowels is placed after i-mutation; (b) development of parasite-vowels is placed before smoothing; (c) loss of h between r and a vowel occurred partly before smoothing and partly when smoothing had begun (see § 231).Google Scholar

⁹⁶ Campbell, §§ 236.3, 237.1, 238.2 and 3, 242, 363, for example.Google Scholar

⁹⁷ If we were to generalise all the archaic forms in Beowulf we should argue that parasiting had not taken place; if the text were originally Anglian, Anglian smoothing would therefore not have occurred either. If the text were not originally Anglian, then reduction of hiatus is the next major change not to have taken place.Google Scholar

⁹⁸ Some vocabulary-evidence has been used to support the idea of Anglian composition. However, if the poem is earlier in date than the Alfredian period, we lack many of the controls necessary to establish beyond doubt the dialect-distribution of particular words. For some of the difficulties, see Sisam, Kenneth, Studies in the History of Old English Literature (Oxford 1953) 119–39; but cf. n. 224 below.Google Scholar

⁹⁹ This would require that syncope of medial vowels and the development of parasiting took place after back-mutation but before smoothing and reduction of hiatus. It is perfectly possible, however, that these both occurred before back-mutation: that ‘there is no certain case of a parasite vowel causing that change’ (Cambell § 395) might perhaps argue against it, however, though the parasiting with which we are concerned arises after long stems where back-mutation would not in any case be operative.Google Scholar

¹⁰⁰ The only sound-change uncertainly located (with reference to i-mutation) in which we are interested here is the development of parasite-vowels, which preceded Anglian smoothing.Google Scholar

¹⁰¹ Syncope of medial vowels seems to have been accomplished before the period of the extant texts: Campbell §§ 341-54, 394.Google Scholar

¹⁰² This argument has been able to be made with conviction from phonological evidence in The Battle of Brunanburh because that poem has a secure terminus post quern of a.d. 937: see Campbell, Alistair (ed.), The Battle of Brunanburh (London 1938) 13. A promising new line of enquiry (exploiting a point scarcely submerged in the work of Sievers and Bliss) has been suggested by Jacobs, N., ‘Anglo-Danish Relations, Poetic Archaism and the Date of Beowulf: A Reconsideration of the Evidence,’ Poetica (Tokyo) 8 (1978) 23-43 (at 25-26): syllabic nasals and liquids, uncontracted vowel-groups, and some aspects of the use of weak adjectives, often reckoned as old or archaising features, show a marked tendency to recur in certain types of metrical patterns; their use may therefore be formulaic. This line of enquiry seems well worth following up. Much relevant material (though almost certainly not a complete collection) is gathered in the charts published by Amos, op. cit.Google Scholar

¹⁰³ Cf. Klaeber, F., (ed.), Beowulf and the Fight at Finnsburg (3rd ed.; Boston, Mass. 1936; 2nd rev. imp. 1950) xcii and cviii, and the references given there. Amos, op. cit. 110-24, has discussed this feature and finds it lamentably wanting as a dating criterion.Google Scholar

¹⁰⁴ Wrenn, , 1st edition 83. The case is sometimes argued, as with the Welsh Gododdin — cf Jackson, K. H., Studia Celtica 8/9 (1973/4) 1–18 (at 8)— that the presence of a good many ‘archaic’ (or unintelligible) words is an indicator of antiquity. But otherwise unattested forms are no guarantee of antiquity, as difficult eleventh-century Middle Irish poems attest, or as the tenth-century Battle of Brunanburh (ed. Campbell 41) indicates for Old English. Amos, op. cit. 146, has discussed this phenomenon in Old English verse, with especial reference to Funke, , ‘Altenglische Wortgeographie’ 49.Google Scholar

¹⁰⁵ See, for example, Woolf, Rosemary, (ed.), Juliana (London 1955) 4, 5–6; Brooks, Kenneth R. (ed.), Andreas and the Fates of the Apostles (Oxford 1961) xxi-xxii.Google Scholar

¹⁰⁶ Ed. Campbell, 18, 29.CrossRefGoogle Scholar

¹⁰⁷ Sisam, , Studies 3. Sisam prefers the ninth century (ibid. 7, 27–28): see n. 108 below.Google Scholar

¹⁰⁸ Sisam, , Studies 6. I cannot see that Sisam has established convincing grounds for ruling out the second half of the eighth century for Cynewulf's floruit. See further Dumville, D. N., ‘The Date of Cynewulf,’ forthcoming.Google Scholar

¹⁰⁹ A start has been made in the general reassessment offered by Amos, op. cit.Google Scholar

¹¹⁰ One must ask, for example, about the implications of scribal substitution of non-alliterating synonyms (as for example Beowulf lines 965a, 1073b, 2298b, 3056a: Pope, , Rhythm of Beowulf xxxii).Google Scholar

¹¹¹ Amos, , op. cit. 35-36, discusses the reworking of poetic formulae to fit new pronunciations; cf. Chadwick, H. M., The Heroic Age (Cambridge 1912) 46.Google Scholar

¹¹² Professor Stanley tells me that such rewriting, especially of prose texts, has been a minor industry in Oxford for some time. As in the case of most cottage-industries, its products have regrettably not achieved widespread distribution.Google Scholar

¹¹³ Transactions of the American Philological Association 20 (1889) 172.Google Scholar

¹¹⁴ Studies 44.Google Scholar

¹¹⁵ Sisam, , ibid. Only if that manuscript could be shown to be in some sense authorial (as Kiernan would argue for Beowulf) could that be the case.Google Scholar

¹¹⁶ See Ridgeway, W., ‘The Date of the First Shaping of the Cuchulainn Saga,’ Proceedings of the British Academy 2 (1905/6) 135–68.Google Scholar

¹¹⁷ The classic exposition in this style is by O'Curry, Eugene, Lectures on the Manuscript Materials of Ancient Irish History (Dublin 1861).Google Scholar

¹¹⁸ Hamp, As, ‘On Dating and Archaism’ 100, points out.Google Scholar

¹¹⁹ See, for a full exposition, Thurneysen, Rudolf, Die irische Helden- und Königsage bis zum siebzehnten Jahrhundert (Halle 1921). The early version of Táin Bó Cúailnge, together with its associated texts, may be found conveniently in translation by Kinsella, Thomas, The Táin (2nd ed.; Oxford 1970).Google Scholar

¹²⁰ For discussion of all these matters, and more, see Jackson, Kenneth H., The Oldest Irish Tradition: A Window on the Iron Age (Cambridge 1964); important modifications are suggested by Mac Cana, P., ‘Conservation and Innovation in Early Celtic Literature,’ Études celtiques 13 (1972/3) 61-119.CrossRefGoogle Scholar

¹²¹ Carney, , Ériu 22 (1971) 73–80.Google Scholar

¹²² Mac Cana, Proinsias, The Mabinogi (Cardiff 1977), gives a helpful discussion. There is a translation by Gwyn, and Jones, Thomas, The Mabinogion (2nd ed.; London 1949 and subsequent revisions).Google Scholar

¹²³ The main historical commentators have been: Whitelock, Dorothy, The Audience of Beowulf (Oxford 1951); John, E., ‘Beowulf and the Margins of Literacy,’ Bulletin of the John Rylands Library 56 (1973/4) 388-422; Wormald, art. cit.CrossRefGoogle Scholar

¹²⁴ Donahue, , Traditio 7 (1949-51) 263–77. For other aspects of the poet's consciousness that the society he was depicting was removed in time and attitudes from his own, see Klaeber, , Beowulf cxxiii n. 4; Hollowell, , ‘Unferσ’ (n. 28 above); Spolsky, E., ‘Old English Kinship Terms and Beowulf,’ Neuphilologische Mitteilungen 78 (1977) 233-38. Schücking, L. L., ‘Die Beowulfdatierung: Eine Replik,’ Beiträge zur Geschichte der deutschen Sprache und Literatur 47 (1923) 293-311, under attack by Felix Liebermann, took the view that no accurate portrayal of the poet's own society need have been intended. For the reasons given above, I think that Wormald's endorsement (art. cit. 35) of Marc Bloch's view that ‘in every literature, society contemplates its own image’ (Feudal Society [London 1961] 102) is, in this context, prejudicial.Google Scholar

¹²⁵ See below, pp. 137–38, 144-46.Google Scholar

¹²⁶ See, for example, Klaeber's edition, Beowulf xxx–xlv.Google Scholar

¹²⁷ See Klaeber's extraordinarily generous assessment, ibid, xxix–xxx.Google Scholar

¹²⁸ Chadwick, H. M. & N. K., The Growth of Literature (Cambridge 1932-40); Bowra, C. M., The Meaning of a Heroic Age (Newcastle-upon-Tyne 1957), and Heroic Poetry (London 1952) Vansina, Jan, Oral Tradition: A Study in Historical Methodology (London 1965) is worth consulting Henige, David P., The Chronology of Oral Tradition: Quest for a Chimaera (Oxford 1974) is now compulsory reading. For a recent sociological assessment, see Finnegan, Ruth, Oral Poetry: Its Nature, Significance and Social Context (Cambridge 1977).Google Scholar

¹²⁹ But Walter Goffart has now argued (in the volume cited in n. 2 above) that the Liber historiae Francorum, completed in 727, is a source of Beowulf, and an historically disreputable one at that.Google Scholar

¹³⁰ See, for example, Dillon, Myles, The Cycles of the Kings (London 1946).Google Scholar

¹³¹ Fingal Rónáin and Other Stories, ed. Greene, David (Dublin 1955) 1–15.Google Scholar

¹³² Tales about Mongán are edited and translated by Meyer, Kuno, The Voyage of Bran (London 1895). See also the study by Mac Cana, P., ‘Mongán mac Fiachna and Immram Brain,’ Ériu 23 (1972) 102-42.Google Scholar

¹³³ Scéla Cano meic Gartnáin, ed. Binchy, D. A. (Dublin 1963). See also the study by Carney, J., ‘The So-Called “Lament of Créidhe,”’ Éigse 13 (1969/70) 227-42.Google Scholar

¹³⁴ Carney, , Éigse 13 (1969/70) 231, 236.Google Scholar

¹³⁵ Carney, , ibid. 234, 236.Google Scholar

¹³⁶ Kiernan, , Beowulf and the Beowulf Manuscript, argues, however, that the poem in its extant form had no existence before the writing of the sole surviving manuscript.Google Scholar

¹³⁷ See pp. 121–125.Google Scholar

¹³⁸ See p. 32.Google Scholar

¹³⁹ Klaeber, F., ‘A Few Beowulf Notes,’ Modern Language Notes 16 (1901) 14–18 (at 17). Much the same sentiment was expressed in aggressive terms by Kock, E. A., Anglia 27 (1904) 220 n. 2.CrossRefGoogle Scholar

¹⁴⁰ Studies 119-39; Sisam ‘popularised’ a theory already expressed by Collitz, H., ‘The Home of the Heliand,’ Publications of the Modern Language Association 16 (1901) 123–40 (especially 132-36), with especial reference to Old Saxon and Old High German poetry, and extended overtly to English (with acknowledgment to Collitz) by Jespersen, Otto, Growth and Structure of the English Language (Leipzig 1905 and successive editions) § 53.CrossRefGoogle Scholar

¹⁴¹ That is not to say, of course, that an earlier poet writing (or orally composing) verse would not have worked within a tradition where models were influential or even carefully studied. What is implied in Campbell's arguments about Brunanburh is that the specific models (probably written) were the author's poetic tradition.Google Scholar

¹⁴² See Campbell's edition passim. Google Scholar

¹⁴³ On the phonology, cf. Campbell, , Grammar §§ 256-64 for example.Google Scholar

¹⁴⁴ Cf. Sisam, , Studies 65–96.Google Scholar

¹⁴⁵ We may note that in England in the ninth and tenth centuries the practice of giving old ‘heroic’ names to one's children seems to have ceased: Chadwick, , The Heroic Age 64–66, and Wormald, , art. cit. 94-95.Google Scholar

¹⁴⁶ A project by the University of Toronto Dictionary of Old English staff to examine ‘The Language of Beowulf’ (see the volume cited in n. 2 above) has, by concentrating on consistent features of the transmitted text, suggested that the medley of eccentric forms which has been taken to indicate a long and complex text-history for the poem has perhaps been allowed to create an illusion. It is probably fair to say that, while the tentative conclusions currently emerging from this important project will prove controversial, the invaluable collections of materials, and the expert and sensitive handling to which Professor Cameron and his colleagues have subjected them, will almost certainly provide the starting-points for any further treatments of the language of the poem.Google Scholar

¹⁴⁷ The description ‘rather out-of-date’ implies a date of writing in the first quarter of the eleventh century rather than the last quarter of the tenth, and requires an assumption that the two scribes collaborated.Google Scholar

¹⁴⁸ Art. cit. 47-48.Google Scholar

¹⁴⁹ Studies 65–67; cf. Whitelock, , Audience 50-53.Google Scholar

¹⁵⁰ The reformers were, or became, enthusiastic supporters and practitioners of vernacular literature, but not of this sort. Their interests seem to have been exclusively ecclesiastical.Google Scholar

¹⁵¹ Asser's Life of King Alfred, ed. Stevenson, W. H. (Oxford 1904; rev. imp. by D. Whitelock 1959) § 98 (p. 85).Google Scholar

¹⁵² Review of English Studies n.s. 11 (1960) 419-21, reviewing Celia, & Sisam, Kenneth (edd.), The Salisbury Psalter Edited from Salisbury Cathedral MS. 150 (London 1959), with especial reference to p. 13 § 26. For the difficulties in localising this manuscript, see Stroud, D. I., ‘The Provenance of the Salisbury Psalter,’ The Library 6th Series 1 (1979) 225–35.Google Scholar

¹⁵³ By a.d. 1000 only the following sees had had no monastic bishop: Dorchester-on-Thames, Elmham, Hereford, Lichfield, Rochester. By 1016 only Hereford, Lichfield, and Rochester remain in this category.Google Scholar

¹⁵⁴ Klaeber, , Beowulf lxxxviii–xc. We could associate the Kentish element with this period, too; it is not impossible to imagine West Saxon and Kentish being written by two collaborating scribes at, say, Canterbury at any time after 825.Google Scholar

¹⁵⁵ A notable attempt to do just this, however, was that, a generation ago, of Andrew, S. O., Postcript on Beowulf (Cambridge 1948) 133–52. One must also mention the interesting paper of Prokosch, E., ‘Two Types of Scribal Errors in the Beowulf MS,’ in Studies in English Philology: A Miscellany in Honor of Frederick Klaeber, edd. Malone, Kemp & Ruud, M. B. (Minneapolis 1929) 196-207. Google Scholar

¹⁵⁶ Wrenn, C. L., Transactions of the Philological Society (1943) 18. There is also perhaps room for the suspicion that an antecedent copy used u rather than Ϸ (= w): cf Andrew, , Postscript 151-52, on line 3005b (and perhaps also on 2298b). On u/uu/Ϸ , see Ker, , Catalogue xxxii, and Sisam, K., Proceedings of the British Academy 39 (1953) 310-11.Google Scholar

¹⁵⁷ Wrenn, , loc. cit.; cf. Ker, , Catalogue xxxi.Google Scholar

¹⁵⁸ In particular the work of Pope, Lehmann, Bliss, Cable, and Stevick.Google Scholar

¹⁵⁹ Klaeber's edition, which first appeared in 1922, has not been revised since the 1936 edition; it now impresses more by its comprehensiveness than its scholarly judgment. Dobbie's appeared in 1953 and shares the faults of the series of which it is a part — uncertainty of aim being the principal difficulty. Wrenn's, also of 1953, is an eccentric and erratic piece of work; its latest revision has rendered it colourless but no less erratic. The latest editions of Sedgefield (third, 1935) and Wyatt-Chambers (second, 1920) have perhaps slipped further from view than is their due. The principal difficulty is of course that the three decades since 1950 have seen a vast outpouring of Beowulf-scholarship, leaving the subject in many respects in a very different position: Swanton, M. J. (ed. & tr.), Beowulf (Manchester 1978) p. [v], estimates a total of about 3,000 articles; we shall see the full horror of this in the new Robinson & Greenfield Bibliography. Google Scholar

¹⁶⁰ See especially Pope, , The Rhythm of Beowulf , and Bliss, , The Metre of Beowulf. Also valuable is Roberts, J., ‘A Metrical Examination of the Poems Guthlac A and Guthlac B,’ Proceedings of the Royal Irish Academy 71 C (1971) 91–137.Google Scholar

¹⁶¹ The project described above (in n. 146) is likely to form the basis for such an advance.Google Scholar

¹⁶² See Sisam, , Studies 126–32, 138, and his references to Menner's work.Google Scholar

¹⁶⁸ It is in the editions of Wrenn-Bolton and Swanton, for example. Dobbie, Elliott (ed.), Beowulf and Judith (New York 1953) 193, sums up the position in restrained, judicious language: ‘but if this explanation [wundini] of the MS. reading is correct, it is odd that we have no other endings of this sort in the poem. The emendation to wundnum puts less of a strain on probability.’ Google Scholar

¹⁶⁴ See Craigie, W. A., ‘Interpolations and Omissions in Anglo-Saxon Poetic Texts,’ Philologica 2 (1923/4) 5–19.Google Scholar

¹⁶⁵ Kiernan, , Beowulf and the Beowulf Manuscript, now offers us a remarkable reassessment: even if only a proportion of his findings be accepted, a major rethinking of attitudes to the poem and its history will be necessary.Google Scholar

¹⁶⁶ Catalogue 282.Google Scholar

¹⁶⁷ Oskamp, H. P. A., ‘Notes on the History of Lebor na hUidre,’ Proceedings of the Royal Irish Academy 65 C (1966/7) 117–37. For further refinements of the collation, see Powell, R., Ériu 21 (1969) 99-102; Oskamp, , Ériu 25 (1974) 147-56; D. N. Dumville, Éigse 16 (1975/6) 24-28.Google Scholar

¹⁶⁸ Bonjour, Adrien, The Digressions in Beowulf (Oxford 1950).Google Scholar

¹⁶⁹ Art. cit. (n. 164 above).Google Scholar

¹⁷⁰ An implication of Kiernan's work, Beowulf and the Beowulf Manuscript, is that in the extant manuscript we have two separate poems newly stitched together or one shorter poem (part one) newly extended.Google Scholar

¹⁷¹ Ed. Meid, Wolfgang (Dublin 1967).Google Scholar

¹⁷² For sources, see Campbell, A., ‘The Use in Beowulf of Earlier Heroic Verse,’ in England Before the Conquest, edd Clemoes, & Hughes, , 283–92. On the diptych-image, see Donahue, , Traditio 21 (1965) 55.Google Scholar

¹⁷³ See pp. 132–37.Google Scholar

¹⁷⁴ For the phrase, and some discussion of recent work, see Lapidge, M., ‘Aldhelm's Latin Poetry and Old English Verse,’ Comparative Literature 31 (1979) 209–31.CrossRefGoogle Scholar

¹⁷⁵ See pp. 121–32.Google Scholar

¹⁷⁶ See p. 142.Google Scholar

¹⁷⁷ HE 4.24 (22). For full discussion, see Plummer, , Baedae opera historica II 248-58. See also the valuable article by Magoun, F. P. Jr., ‘Bede's Story of Cædmon: The Case-History of an Anglo-Saxon Oral Singer,’ Speculum 30 (1955) 49–63, a paper which is none the poorer for its obvious ideological bias.Google Scholar

¹⁷⁸ The best studies are by Wrenn, C. L., ‘The Poetry of Cædmon,’ Proceedings of the British Academy 32 (1946) 277–95, and Schwab, Ute (ed.), Caedmon (Messina 1972).Google Scholar

¹⁷⁹ The classic, if heavily interpretative, study is that of Mac Neill, E., ‘A Pioneer of Nations,’ Studies (Dublin) 11 (1922) 13–28 and 435-46. According to the Annals of Ulster, Cennfaelad died (as a cleric, as in this context the title sapiens implies) in 679.Google Scholar

¹⁸⁰ The principal text is the preface to Bretha Étgid : see Hancock, W. N. et al. (edd. & tr.), Ancient Laws of Ireland III (Dublin 1873) 86–89 (cf. 550). For discussion see Mac Cana, P., ‘The Three Languages and the Three Laws,’ Studia Celtica 5 (1970) 62-78.Google Scholar

¹⁸¹ Though many scholars have made him say that, including Whitelock, Audience 21, and Swanton, Beowulf 3. In other words I agree enthusiastically here with Goldsmith, Margaret, The Mode and Meaning of 'Beowulf (London 1970) 17.Google Scholar

¹⁸² Plummer, , Baedae opera historica, II 251–52; Dobbie, Elliott, The Manuscripts of Cædmon's Hymn and Bede's Death Song (New York 1937); Blair, Peter Hunter (ed.), The Moore Bede (Copenhagen 1959).Google Scholar

¹⁸³ This was, I find, the opinion of Wülker, Richard, Grundriss zur Geschichte der angelsächsischen Literatur (Leipzig 1885) 117–20.Google Scholar

¹⁸⁴ HE 4.25 (23).Google Scholar

¹⁸⁵ Gesta pontificum Anglorum Bk 5 § 190 (ed. Hamilton, N. E. S. A. [London 1870] 336).Google Scholar

¹⁸⁶ The evidence is provided by the early copies (of 737 and 746) of Cædmon's Hymn; by Bede's Death-Song (of 735, but the earliest MS is of the ninth century); and by the verse proverb from the collection of Boniface's letters (in an anonymous letter, perhaps of 757 × 786). If the inscription on the Ruthwell Cross does indeed date from ca. 700, then it offers the earliest specimen of Old English Christian verse.Google Scholar

¹⁸⁷ Williams, Ifor, (ed.), Canu Aneirin (Cardiff 1938); Jackson, Kenneth H. (tr.), The Gododdin: The Oldest Scottish Poem (Edinburgh 1969).Google Scholar

¹⁸⁸ For recent discussion, see Mac Cana, P., Celtica 9 (1971) 316–29 (reviewing Jackson); Greene, D., Studia Celtica 6 (1971) 1-11; Jackson, K. H., Studia Celtica 8/9 (1973/4) 1-18; Miller, M., ‘Historicity and the Pedigrees of the Northcountrymen,’ Bulletin of the Board of Celtic Studies 26 (1974-76) 255-80; Dumville, D. N., ‘Palaeographical Considerations in the Dating of Early Welsh Verse,’ ibid. 27 (1976-78) 246-51.Google Scholar

¹⁸⁹ For all this, see Jackson's concordance, Gododdin p. 159, and his introduction 41-67.Google Scholar

¹⁹⁰ This seems the only conclusion to be drawn from the evidence presented by Jackson, , Gododdin, pp. 47–48 (on B1 = A78), and perhaps 48-49 (B2 = A52) and 49-50 (B5-6 = A40-41).Google Scholar

¹⁹¹ Ibid. 86–91. The following quotations are from Jackson.Google Scholar

¹⁹² A useful brief discussion of this question by Gerard Murphy may be found in Knott, Eleanor & Murphy, G. (ed. Carney, J.), Early Irish Literature (London 1966), especially 191–93. See also his ‘On the Dates of Two Sources Used in Thurneysen's Heldensage,’ Ériu 16 (1952) 145-56 (at 151-52).Google Scholar

¹⁹³ It is discussed briefly above, pp. 126–28.Google Scholar

¹⁹⁴ See, for example, Vansina, , Oral Tradition (n. 128 above).Google Scholar

¹⁹⁵ See pp. 155–59.Google Scholar

¹⁹⁶ Bonjour, , Digressions; Jacobs, , Poetica 8 (1978) 41, is unconvinced.Google Scholar

¹⁹⁷ Edd. Dickins, B. & Ross, A. S. C. (3rd ed.; London 1954; rev. imp. 1963); or ed. Swanton, Michael (Manchester 1970). Google Scholar

¹⁹⁸ Other aspects of its history have already been alluded to above, pp. 133–34.Google Scholar

¹⁹⁹ The stemma is a form of that offered by O'Rahilly, , Stowe Version (see n. 200 below) xxx, modified by her remarks in Recension I (vide ibid.) xx, xxii.Google Scholar

²⁰⁰ On all the preceding, see the editions and translations by O'Rahilly, Cecile, The Stowe Version … (Dublin 1961) (= Recension IIb); The Book of Leinster Version … (Dublin & London 1967) (= Recension IIa); Recension I (Dublin 1976). For the establishment of the relationships, see Thurneysen, , Heldensage 99–118, 235-41, as modified by O'Rahilly, , Stowe Version vii-xxxi, and Book of Leinster Version xiv-lv; for the dates, see Thurneysen, R., Zeitschrift für celtische Philologie 19 (1931-33) 193-209; cf Jackson, , Oldest Irish Tradition 52. On early witnesses to the story, see the papers by J. Carney cited in nn. 70, 72 above. On the earliest manuscript, see Concheanainn, T. Ó, Éigse 15 (1973/74) 277-88. The fragmentary Recension III does not exist in an up-to-date edition: it is edited from its two manuscript-witnesses separately: Thurneysen, R., ‘Táin Bó Cúailghni nach H.2.17,’ Zeitschrift für celtische Philologie 8 (1910-12) 525-54; Nettlau, M., ‘The Fragment of the Tain Bó Cuailnge in MS. Egerton 93 (ff. 26a1-35b2),’ Revue celtique 14 (1893) 254-66 and 15 (1894) 62-78, 198-208.Google Scholar

²⁰¹ Much ground has been cleared by Bliss, The Metre of Beowulf 44ff., and by Pope, , The Rhythm of Beowulf 232-38 and xxiii-xxxiv, making a new examination of this question both possible and highly desirable.Google Scholar

²⁰² It is in the nature of past academic work on the subject that such approaches will be primarily phonological. Any results obtained on a specific point will likely be able to be expanded with the aid of the evidence already amassed for the dialectal colouring introduced by successive scribes (but cf. n. 146 above).Google Scholar

²⁰³ To which one could add the search for noticeable variations in syntax (or vocabulary) as between different parts of the poem.Google Scholar

²⁰⁴ Audience 4–23, 99-105 (especially 19-20).Google Scholar

²⁰⁵ See, however, Goldsmith, , Mode and Meaning 10–12, and Wormald, , art. cit., especially 38-39 and n. 28 (pp. 74-75). (I cannot, however, imagine what evidence would lead Wormald, 38, to speak of ‘extensive lay literacy’ in early mediaeval Ireland.) One can accept Wormald's point (35) that Beowulf is ‘literature about, for, and even by, the Anglo-Saxon aristocracy,’ but one must at the same time remember the large number of aristocrats who became ecclesiastics.Google Scholar

²⁰⁶ Audience, especially 5, 19–20.Google Scholar

²⁰⁷ The proposition has repeatedly been urged that Beowulf is intended for a royal patron as a sort of speculum regale; I am not aware, however, that any detailed or convincing arguments have been advanced to support this view.Google Scholar

²⁰⁸ Wormald, art. cit. 42-58. There is no reason why this need be restricted (from either end) to the eighth century.Google Scholar

²⁰⁹ See below, pp. 157–58.Google Scholar

²¹⁰ Audience 19–20.Google Scholar

²¹¹ Aldfrith, (685-704) and Ceolwulf (729-37) of Northumbria, and Ælfwald of East Anglia (713-49) are all kings of the ‘age of Bede’ who had Latin literary compositions dedicated to them — by Aldhelm, Bede, and Felix, respectively.Google Scholar

²¹² Art. cit. 75 n. 28.Google Scholar

²¹³ See above, pp. 130–32.Google Scholar

²¹⁴ In his letter to the bishop of Lindisfarne, ed. Duemmler, Ernst, MGH, Epistolae IV (Epistolae Karolini Aevi 2; Berlin 1895) 181–84 (no. 124).Google Scholar

²¹⁵ The most important recent work is that of Carney, J., Ériu 22 (1971) 53–80.Google Scholar

²¹⁶ For some discussion of these matters, see Byrne, F. J., Zeitschrift für celtische Philologie 29 (1962-64) 381–85, and his article cited above, n. 53.Google Scholar

²¹⁷ Williams, J. E. C., ‘The Court-Poet in Medieval Ireland,’ Proceedings of the British Academy 57 (1971) 85–135.Google Scholar

²¹⁸ See his articles cited above, nn. 58-59.Google Scholar

²¹⁹ Audience 28.Google Scholar

²²⁰ And on these grounds, as on many others, the absurd arguments of Cook, A. S., ‘The Possible Begetter of the Old English Beowulf and Widsith,’ Transactions of the Connecticut Academy of Arts and Sciences 25 (1922) 281–346, for its composition at the court of King Aldfrith of Northumbria should be laid to rest; it seems extraordinary that scholars still feel the need to remind us of Cook's lapse from sound judgment, and incredible that any scholar of judgment could quote it with approval. Whether or not that time and place is likely for Beowulf, Cook's arguments do nothing to advance its claims.Google Scholar

²²¹ Cited in n. 102 above.Google Scholar

²²² A recent example I have noted is by Farrell, R. T., Bede and Anglo-Saxon England 116 n. 43.Google Scholar

²²³ The first half of this requirement, an examination of the metrical tests in the light of the results they offer, has now been completed by Amos, op. cit.Google Scholar

²²⁴ Sisam, , Studies 131: ‘In the present state of the investigation, vocabulary remains unsatisfactory or inconclusive as evidence of the original dialect of poems presumed to be early.’ Sisam has been quoted repeatedly over thirty years, while there is no sign of an attempt to advance the ‘state of the investigation’ Google Scholar

²²⁵ Brook, G. L., ‘The Relation Between the Textual and the Linguistic Study of Old English,’ in The Anglo-Saxons: Studies in Some Aspects of their History and Culture Presented to Bruce Dickins, ed. Clemoes, Peter (London 1959) 280–91, discusses this practice, which he refers to as ‘normalisation.’ His opposition seems to be based on the fear that such practice would conceal the manuscript-evidence. However, his concern is surely inappropriate in the Old English poetic context, where editions displaying what the manuscripts say are readily available. Unwillingness to ‘normalise’ could itself be said to be concealing from students the evidence for and (especially) against the theories of their fellow-workers.Google Scholar