ORAL-FORMULAIC CHARACTER OF ANGLO-SAXON NARRATIVE POETRY! By FRANCIS P. MAGOUN, JR Iw the course of the last quarter-century much has been dis techniques emmlayed by sinlettered cingere in their eomnosition af narrative verse. Whereas a lettered poct of any time or place, composing (as he does and aust) with the aid of writing materials and with deliberation, creates his own language as he proceeds, the unlettered singer, ordinarily composing rapidly and ‘extempore before a live audience, must and does call upon ready-made language, upon a vast reservoir of formulas filling just measures of verse. ‘These formulas develop over a long period of time: they are the creation of countless generations of singers and can express all the ideas a ginger will ned in order ta tell his story. itself usually traditional. ‘This progress is primarily due to the work of two men, the late Milman Parry! and his former pupil and successor in this field, Professor Albert Bates Lord of Harvard Universty.* First in connection with Homeric Iunguage, Inter us a result of eld-work in Yugosiavia, chleny among unietered Muslim singers, Parry, aided by Lord, demonstrated that the eharacterstie fea ture of all orally composed poetry is its totally formulaic character. From this « second point emerged. namelv. that the recurrence in a given poem of an an~ “This poner in mith revisions, esentally the second (The Formlais Character of Anglo Saxon ‘Nerative Poetry) of three Special Univenity Lectures (erieotitle ‘rat Formlake Tradii ‘Anglo Sezon Petry), delivers et the invitation ofthe Unive of London inthe Senate Hou om 10, 1% wal 24 Jose 1005) nl nae sists Mass 3098 Te Bs two pe ta gree De sete tn it re The Art nd Ca of Oral Pty, il he lt ew og sini get the third (Some Problems of the Futur) Charts T snd TT are revisions of mimeorapbed counter i (1948), 43-4, To of Party's papers tay be specially noted representing the full development of hia thought: ‘Stndies in the Epie Tecbniue of Oral VerseMaking, I: Homer and Homerie Stl Horan Studie iu Clatial Phibogy, 213 (1980), 73- 1, esp. pp. 18-18) for charts exposing the formulaic character of I 1-28 of the Iliad and the iar respetively; asl "TE: The Homeric Langunge a tke Langange of Orel Poetry tidy Xu (1982), 1-50, exp. pp. 1217 ("The Art of Oral Poetry). These pepers ae cited here ax Pary T and IT nd by page. * Parcy in th summer of 1088, and Parry and Lord in the ytce 1096 85, sted Ube production eh ort ep style rn Yuosivin and cllected some 1800 tes, "The Parry Colleton of South Slavic Texts,’ now deposited in the Harvard Coleg Iirary. Following Parry's lad aod working with this opulent material Lord submitted in 190» Ph.D, thesis (Harvard unpublished), “The Singer of Tales: A Study in the Process of Yugotlav, Gtec, and Gormanie Oral Posty.’ Lond revisited egonavin 1089 ant 1901, fo» beset om the wag ip of L4o0 ee “Yugi Pipe Fee Port’ Journel ofthe Intarnational Pole Muniz Court (1851), 87-81. Hin then, revised aad ex ‘panded, willbe poblshed by Uhe Harvard Univenity Peas a The Singer of Tale inthe eres“ ‘van Studies in Comparative Literate.” Tha won of Bory eed Toe sry tant to Sir Coal Marie Bowra and utilized by him in hia Hers Poetry (London: Macmillan, 1952). Tia be little songs of the order anza 60) gs about z hy Ger. rasize the tradition of the art metrical- int to the s identical ‘erman — ‘way and metrical ys one is be Anglo- of Anglo- Oral-Formulaie Character of Anglo-Saxon Poetry 449 Saxon poetry before the introduction of the art of reading and writing by Chris- tian missionaries from Rome and from Iona in the Hebrides; indeed, we have no poetical text which can in exactly the form preserved be thought of as having been put together very early at all. Consequently, it has been natural to think of the precerved poems ao composed as we compose postr. io. hy lattarad per. sons making use of writing materials, and until the time of Parry and Lord ‘there was no available technique permitting one to decide on the basis of internal evidence alone to which tradition a given text might belong — to the oral or to the lettered. The recurrence of verses and verse-pairs m Anglo-Saxon poetry, ‘the ‘Parallelstellen’ of German scholars, has been much noted and commented upon, and cross-references accumulated and often cited by editors of individual [poams, with the main conehision drown from this phenemenan Keing that those pparallels might constitute evidence of the direct influence of one poem upon another (see p. 461, below). But with the discovery of the dominant réle of the formula in the composition of oral poetry and of the non-existence of metrical formulas in the poetry of lettered authors, we have suddenly acquired a touchstone ‘with which itis now possible to determine to which of the two great categories of poetry a recorded text belongs — to the oral or to the lettered tradition Aes Bret test T hase analyvod the fret twenty-five ines or. better. the frst fit verses or twenty-five typographical lines of Béowulf, chosen because they deal with highly specialized thematic material not represented elsewhere in the poetry, for the presentation of which in verse one might suppose that a poet Would need to create his own language If he would ever have wo do s0. ‘The for- rulaie character of the verse is demonstrated by Chart I (pp. 404-465, below).* A word-group of any size or importance which appears elsewhere in Béoioulf for other Anglo-Saxon poems unchanged or virtually unchanged is marked with solid underlining snd is @ formula according to Parry's definition that. a formula is “a group of words which is regularly employed under the same metrical conditions to express a given essential idea.”™ A word-group marked with solid and broken underlining, or with broken undertining only, amy be called a foriua- Iaie phrase or system; such groups are of the same type and conform to the same verbal and grammatical pattern as the various other verses associated with them and cited in the supporting evidence. For verses which are unmarked I have found no supporting evidence, Following the marked text on the chart comes the ‘supporting evidence asserobled under numbers answering to the a and b parts of the respective typographical fines. Tookinng at Chast Tue wvtes fiat Wael uf dae Mitty verses vuly svuie tistecis or twenty-six per eent, are not matched wholly or in part elsewhere in Anglo- ‘Saxon poetry. In a word, despite the relatively limited corpus of some $0,000 + ucla mid Tnwa fev fea Dbl ars nl oe Bey Kaaba hind ition with Pint ‘and Second Supplements (Boston, 1800), those fram Judith ou the edition of Benno J. Ttomer (Loa on: Methuen, 1088); ll thers on The Angl-Sazon Pads Records (New York: Columbia University Pras, 1981-12), Spellings ate normalized on the basis of early WS. an ot forth in Las Langues mo-450 Ural-Formulaic Uharacter of Anglo-Saxon Poetry lines — about the same as the two Homerie poems — in which to find eorrespond- ing phraooo, come ooventy per cont of the tent of this pamage dos occa eloe= where, Were the surviving corpus, say, twice as big and if, above al, we had other songs of any extent dealing with anything like the seme thematic material, there well might be almost nothing in the language here used that could not be demonstrated as traditional ‘Though usefulness rather than mere repetition, is what makes a formula, itis instructive to look at the repeated formulas frst, since it is easier Lo recognize a formula ao exch when it oocurs a cceond or third timey* and from this regular use {in various songs one readily sces how it helps this and that singer to compose his verses, Verses 1b, 3a, 3b, 4b, Ba, Sb, Sa, 10b, 11b, 18a, 14a, 15a, 162, 17a, 238, and 25a are of this sort. They occur exactly the same elsewhere or with only some significant. change in inflection about which a singer would searecly have to Aevote conscious thought in order to fit. them into some different context or slightly different grammatical situation. ‘The very fact of their recurrence in and/or auteide of thie paccage boure witnooo to their woefulnces not only to the singer of Béowulf but to singers of many other songs dealing with quite different themes. A number of these formulas are something more than mere repeats and form part of larger formulate systems used to express the samme, of almost the same, ‘dea or used to ft some larger thythmical-grammaticel pattern. As Perry ob- serves of such formulas in Homer, ‘any group of two or more such like formulas rakes up a system, and the eystom my he dafined in tur ao a group of phrases which have the same metrical value and which are enough alike in thought and ‘words to leave no doubt that the poet who used them knew them not only as a single formula, but also as formulas of a certain type.”? Here belong verses 1b, ‘9b, 11D, 10a, ahd 1a Tb. on géar-dajum is one phase of a system on x-dagum used to express the idea ‘long ago’ and occurs twice elsewhere in Béowulf and in other poems as well. Either alone or with one ar tw preceding nnatroesed words i forme « complete Coverse. With the substitution for géar, with the sense “of yore," of dr, eald, or fyr, the formula remains unchanged in meaning and meter, though the variant first clements of the compound are patently more then useful in meeting the eigeuties of alteration, a restrictive and technical problem with whieh neither Homeric uor Yugostav verse, for instance, have to contend. ‘The degree of thrift that marks the use of formulas in Homeric verse is searcely conceivable in the construction of the much more restrictive alliterative German worse 8b. eibban éredt wearp shows us three words repeated as « formula ia Bésoulf where it serves to express the general idea ‘after something or other has hap- pened’; it must have often been used by singers to express this same idea in a complete D-veree. Dul sippun erest (or furdum) can be followed by any mono- syllabic verb-form in the psst tense and in the recorded instance with awéar expresses a closely related idea. * Pary LAS and Pay Tae Ub. pe sentucetiy Stylistiea serve as where wil tion. The formulas 6a. la: uk is de Blowulf 6 smitde sul here allite measure ¢ Va. Se se doce system, © syllabic p involving favorite k Tess come ‘The pr Iined aa f whole ver systems 0 ges6bas (2 bein bbe mere r Dotan oh eymingy the limits one way 0 the Danis telets? of: tional and Waldor-ey Foie oynin for Frit litle to t gmbsite » Rorfurtrvesprousle recur else- had other material, ald not be mule, surging @ cegular use spose his Ya, 238, only some y have to context or aly to the e different and form the same, Peny ob- > formes of phrases ought and only asa verses Ib, xpress the one ae wall complete { dr, eal, he variant ceeling the Oral-Formulaic Character of Anglo-Sazon Poetry 451 1b, put wee gS oyningt is @ formule that mey well have eome into being f connection with encomiastie verse, of which we hear so much and have so little. Stylistically this and related formulas stop the narrative for a moment and thus serve as a kind of emphatic punctuation. Tt is used twice in Béowulf, and else- ‘here with unfavorable adjectives it serves as a parallel phrase of cisapproba- tion, The system is bat wae (i) x eying. There are other more distantly related formulas noted in the supporting evidence, all referring to persons. 16a, lange hells fo part of « lange ayetom expressing the ies ‘fr « lang time! and is closely related to a similar system with frége, equally popular with the ‘Bowulf singer. This formula or formulaic system occurs with ealle, de, and miele substituting for lange, alternates which affect neither sense nor meter; htere alliteration must dictate the singer's choice. Whether he uses mile or Prage is surely a matter of accident or indifference, since both words fill the same ‘measure of verse and here will not enter into the alliteration. Oa. Seieldeesafora fe not repented sleowhors in the postin corps for nowhere clse does the need exist to uso this particular patronymic, The value of this system, whereby an A-verse can be constructed with the genitive of any mono- ayllabie personal name, is obvious from the supporting evidence. For patronymics involving the numerous dithematie names 1t may be observed that sum is the favorite keyword and sutomatically forms a D- or E-verse, as do the somewhat, less common maga and magu. “Tho present pessoye inches three nominal compounds which T have under- lined as formulas not merely because they are repeated elsewhere to make up ‘hole verses but because their second elements constitute the core of many stall systems of formulas. These are Séod-oyninges (22), ymbsitiendra (9b), and wil gesibas (23a). If these words did not make up entire verses, one might perlaps be inclined to view them merely as repeated words, and just as formulas need not be mere repetitions, so mere repetitions need not constitute a formula. ‘Péed-ryningn (a) i one of & larer number of compounds with inflected forms of eyning, usually in the genitive singular, which express the idea ‘king’ within the limits of a D3-verse. In most cases the first clement merely emphasizes in ‘one way or another the importance of the king or kings in question, as here where the Vanish péod-eyningas are tacitly opposed, as i were, to smdkurcunyur telets' of ancient Scandinavia, Occasionally the first element will be more func- tional and will define or locate a king. In the on-verse postion Béowulf 2705 has Waltor-eyningn and in the offoverse position eorb-. héah-. béod-ovninaes, also Friseyninge and s4-cyninga; of the same general order is weorold-eyninga. Except for Fris-, used to place geographically Daxghreetn’s overlord, the first clementsadd little to the thought and were presumably chosen for alliterative convenience."* ymbrizendra (9b), 8 compound present part. forming % Di-veraes presents a ° Por further instances of word of scilar structure, and thus with sinilar chythmical pattern Barwa ote John Colin Pope, The Ry of Blowul (New Haven, Con, 1942), pp. 200, 858 (type DI, No, 1) and #48 (type Al, No a), Hxamples (rom other Poets anc with ter AI elecnenes CA ‘be found in Chistian W. Mf. Grein—Verd, Holthausen—J. J. Kohler, Spraokchats dr angeatehs- ‘hen Disker Heidelberg, 1912) "Seo further sid, p. 108, col 1, under ning.452 Oral-Formulaie Character of Anglo-Saxon Poctry somewhat similar aituntions it handily expresses the idea of ‘porsons residing round about.’ Very close is ymbstandendra. In a broader way ymbsitendra is to be associated with a large number of verses consisting of a compound present participle, of which there are many in Béowdf, which tend in tum to break down into various semantic systems such as the idea of ‘sea-farer’ expressed by brim and sd-liSende in Béowulf, and in other songs with the substitution of éa-, meré and wdg- asthe first element but with no change in thought. Nl geattor (@8a) ic but ono of w largish formulaic ayetom oentoring om geotP to express in & complete A-verse the idea of ‘fllower(s) ‘retainer(s),’ the large variety of available first elements being highly useful to the singers in connection with alliteration, Thus are found compounds with dryht-, eald- folo-,wfary and seynn- ‘Within the first Gfty verses of Béowulf ovcur three so-called kennings, two Christian: Lé-fréa (1b), varied by wuldres Wealdend (17a), and one non- Chsatian: hramordde (We) Rosorvi in connection with the special diction of the Christian songs (pp. 456 ff. below), ‘we may examine here the formulaic character of the C-verse afer hran-réde and some closely related expressions by the aid of which the singers were able to place people on the sea oF to get them over it. Much has been wntten about Anglo- ‘Saxon kennings by themsefves and as part of Old-Germanie poetical technique, ‘but there is one particular aspect of this problem which ean probably support farther thanght and investigation, namely. the formulaic charactor af the kone ning. Like the rest of the language of oral poetry kennings must have developed over a long period of time and must be traditional and formulaic. An examina- tion of the phrase ofor hran-rdde will tend to bear out this view. The feminine accusative singular hran-rade, combmed with the prepositions geond, ofr, and on, forms a complete C-verse, whose repeated use marks it as formulaic. Yet ‘more than that, in that itis also one phrase of « formulaic system on (ofer, orond) xara. where for x one oan substitute any appropriate manoasliahie fest clement. With the substitution of swan one finds ofer swan-rdde in Béowulf and Elene, on swan-rdde in Juliana, while on seg-rdde appears in Béowoulf with little (or no real difference in meaning, and none in meter, from the other combinations. ‘The singers are presumably concerned not primarily with some refinement of imagery produced by varying the first elements hran, zeg, and swan — something for which an oral singer could scarcely have time — but with recalling a formule expressing the fundamental idea in question with availability for different al- literative situations. Itis hard to believe that they had much eoncern with possi- ble connotative effects produced by passing mention of sails, swans, or whales. ‘There is another aspect of this general problem that semantically at least is related Uke ufor frunrdde verse in Bérwuif. Now uals particular forwuls nad related formulaic systems were obviously useful to Anglo-Saxon singers and provided them with e C-verse with the aid of which they could get their charac- ters onto ot across the sea. Nevertheless this system imposed certain limitations, 1 Thi, p08, under get. ¢ Including tt yet the need ‘way. A fair ceusative sini in this gram pattern othe important al Dar, fas Ofer bap-ves ‘pressing on Ireeab ofer b alliterating * serve their} rental cons Ue vee, wb probably # s Jaran on flo Fce-aneg, in on hea-1eg alliterative verb in a sir lw eral the: [shall cot of Béowulf the singer ¢ formulas. HE important m Ieeat plus a the singer 6 toa Bor and a few ir all instances by two poin familiarity ¥ fet that he ‘at tmoe od in preserved second. Te verbal idea but the és fore proceec in what apy ee Dorot 1 parrimonst for purposes olresiding dra is to (present: to break “essed by ution of on gest the large snneetion ofa-, and ngs, two >ne ‘non Iehelow), -réde and 2 to place ae Anglo sehnique, 1 support the ken- leveloped examina- feminine eat oy Yet it is on (efor labie first neuf and ith litle sinations, recat of vomething formula erent al- ith possi ve whales. at Teast is ule ond agers and sr charac nitations, Oral-Formulaie Character of Anglo-Saxon Poetry 453 including the fact that a verce based on thie farmmnla eannot well contain a verb: {yet the need for composing such verses was felt and was met in more than one ‘way. A fair example centers on a parallel to rd, 4, namely, reg, m. In the ac- cusative singular of weg there will be no endings hence any compound of wey in this grammatical case, where ending a verse, must be Heted tnvo a metrical pattern other than C, one in which there will be place for a verb or some other ‘important alliterating word at the beginning. The pertinent compounds of weg are Bop, fed , fot, nod bevel, of which hrhoreg is the mort frequent combination. Ofer bap-wep occurs three times, always with some form of breean in the sense ‘pressing on across the sea’: thus, brecan ofer bapoeg (And 228, le 248) and ‘recap ofer bap-sceq (And 513), where the phrase ofer ba}-s2eg combines with the alliterating verb to make a formula, The two J-compounds, dd- aul for-wey, serve their purpose in combination with faran. Fiéd-veg appears in an instru- mental construction foron fléd-wwege (E70 106) ‘the sailors] journeyed on oF over the con's while in the arrnsative plural there is For fléd-weaas (Bid $6, 9) “it, probably a ship] traversed the seas.” With on the combination flt-any appears in ‘faran on flot-weg (HoBE 42) “Iwas fated) to journey on the sea.’ Finally comes ‘wate, in meaning identical with hran-rdd of Buf 10a and occurring in huetep an hot woe (Sea 63a) “impels on [C02] the whales rouse" Deside vileving vativus alliterative alternates this cluster of wey-formulas permits the inclusion of a verb in a single D-verse, an opportunity of which thesingers were obviously glad Ao avail themselves, T shall conclude this discussion of the formulaie character of the first 50 verses ‘of Béowulf with a brief word on the first five verses (1a-Sa) of the poem, where the singer appears to have sjusted, combined, and recombined a number of formulas, He begins with a formula siucl used W start svugs or te introduce an ‘important new section of a song, a formuls built around the weakly exclamatory ‘wat plus a personal pronoun. ‘Tis isin effect a sort of filer-in, something to let the singer got going: the phrase, ordinarily metrically unstressed, opens the way toa B. or C-verse, The total system, embracing all personal pronouns in the nom. and a few in oblique cases, is vast and cannot be presented here, but looking at all instances of the subvariety Hivat, wé (1a), eolleeted on Chart I, one is struck Dy two polis: (1) Una in exch ease the singe includes his audience in acouming familiarity with the thematic material of his song,"* and (2), more important, the fact that he is saying ‘we have all heard or learned about something or other,” at times adding that the events took place long ago. Htcran is the verb favored in preserved song, with frignan of Béowulf running (perbaps by chance), a poor second. It will be noticed that the singers ordinarily work in the important ‘verbal idea ‘hearing about,’ Tearning about” in the course of the first to verses, Bur the Béowudp singer iutinluved mention of the Epear Danco (Gér Dena) be- fore proceeding farther. This apparently spoiled his chance of getting in a verb in what appears to be the favored or ordinary position in the first verse, Com- ee Derethy Whitey The edie of Bll (Ontord Unianity ram. O40) pp SLL and 1, posin o euencefnailarity in rors and dati with the Blowal/ stores aed subtorieaintrodared fer purpoos of embellishment; the latter are notin aay ordinary sense ‘ireasions”454 Oral-Formulaie Character of Anglo-Saron Poetry arable. Cyuewult in Bte 77, be might in some tashion have worked in a suit able verb in 1b, had there been such a one capable of g-alliteration, but at all vents he next called upon one of the several available formulas expressing the idea ‘long ago.’ already discussed (p. 450. ahove) under an pfardagwm (1h). Thus gefrugnon. is put off to the fourth verse (2h), while the Axi of the total phrase wé gefrugnon Ini has to wait for the fifth (8a). The basic formula is all there and the singer has used every scrap of it, though not in what would appear to be the ‘woual may. Oue ight Interpret this exeeptional treatment 28 an example of & first-rate singer coping quickly and deftly with an almost awkward situation into which he had got himself, even though the resulting order of wards is per- haps not quite natural. To sugzest that this order af wards ix any ennsa ‘Titersey? is virtually to deny oral technique in the composition of the poem, a technique demonstrated in the preceding analysis of the first fifty verses of the poem, ‘The traditional character of the recorded text is further born out by the fact that, st Kao! fifteen yer veut f the verses of the poem are to all intents and purposes repeated within the poem,” a phenomenon unthinkable in lettered tradition. In the opening lines of Béowulf are two formulas which must be called Chris- tian: L{f-fréa (16b) and wuldres Wealdend (Via). Neither of these sowallad kan ‘nings could well refer to anything but the Deity and hence could not have formed part of the traditional language of pre-Christian poetry. They must be relatively, young and their presence in Béowulf raises the general question of the relation of the language of Claisiau ustrauive poctsy —by far the largest genre in the ‘corpus — to the older traditional poetic language. There are no means of knowing when first a singer or singers started making songs based on such novel thems material as that found in the Old Testament, Apocrvpha. saints’ lives, and ho: ‘but it cannot well have been before the arrival of Augustine in Kent in 597 and of Paulinus in York in 625, an influence fortified by the settlement of Aidan on Lindisfarne (Holy Island) off the Northumberland coast in 635. Yet some- where in the neighborhood of O76 Ot lla was quite ponstbly singing religious ‘verse, interspersed among diverting secular songs, in public at Malmesbury in Wiltshire in order to get the local populace to stay on after mass for the ser- ‘mon, and sometime between 658 and 680, the vears during which Hild ruled as " Comune orally by Mr Robe P, Crd f Sith Calls, who is prsaly studying the ora sya in Pal. "Reported by Wiliam of Malebury (119) n bis De Gee Ponta Angora (cd NB & A Mamitan Bal Sue, Woo Landen 19), Bla) Pe Chile seal goes he Ch ‘Aled the Gras ot Tendo (Wan Mowe, ot pp. 39 389), “Litera od plenum tren nativee uti nae non nla cara ado, tate ve Bete parla iin, Pec gin one ten tir adi ago aa Aiconm fan. Popul colangomsoniceas yore ivi semmonib intent, statin eantatis nis dace castare slum, Te sacs vn ‘eps pontem quire ot rte contin bets oppose gus an cnt po fess. Bo plus quam sezelfavorem et concurvam emeritom, Hoe commento, sensia iater ldo ‘vers Septurarum insert, ever ad saitatetn rons “And thus fully natracted in [Latin trate be alo didnot neglect the songs of his antive tongue, that, acoarding to Alles book of mhich [spoke above at no ine was anybody ever his equal ablces of the mone ccomposin ‘more abo probably the backs Tn talk, Poems on ‘conservat formance it is clear recitation lay circle: Creation ave beet to lay auc from the’ stronghol entire fab innity ar Now, ¢ slope ve only with familiar.’ the tradi themes of reason th measures ld forme He was able Alfred reas monly oe icons, Tly man st the country Aloe thins ‘ion people Ie ed the r Te mayb oct ned othingi Browne, 8. Tine rom, sure, 3 White Tae)tin a suit- but at all essing the 1b). Thus phrase tob there and to be the imple of a situation relatively relation are in the Ocnoming thematic und homi- i eae oe his eral. Orut-Furmul jc Character of Anglo Saxon Poctry ABB abbess of Whitby in the North Riding, the unlettered Ciedman, farm-hand on the monastic estate, is said on first-rate authority"? to have been successfully ‘composing all sorta of songs based on Christian story. There is no way of learning tore about Aidan’s compositions but, as I hope to show elsewhere, Credman was probably the father of nothing but his own songs and composed these against the backgrounl uf « develyyed tradition. “Ia talking or thinking about the chronology, real or relative, of Anglo-Saxon ‘poems one is notoriously treading on very swampy ground, but if one adopts the conservative view that a Béowulf song in form fairly close to the preserved per~ formance had come into being not far from, sey, 780 or even somewhat Tater, it is clear that by that time Christian poetry was a commonplace and that its recitation waa a familiar form of entertainment not only in monasteries but in Ty circies, Were Usie uot the caocy the rositation = Fiona af a semg abont the Creation (Buf 90-98) would, as Dr Whitelock has recently pointed out, ‘surely have been incongruous, or even ludicrous, if minstrels never sang on such themes tolay audiences." As itis, the Creation song seems to enjoy a status no different fom thet of tongs sung about Sigemund and fitela oF the wagedy of Finn's stronghold in the same hall on another occasion. Indeed, apart from this, the entire fabric of Béowulf is shot through with the language and thought of Chris- ly aid nist be viewed ae « Christian pom thongh of an unusual sort ‘Now, aa Parry emphasizes, the traditional language of unlettered singers de- vvelops very, very slowly and over a long petiod of time and is created to deal only with traditional themes with which singers and audiences are in the main familiar, On his visits to Yugoslavia in 1950 and 1M01 Professor Tord noted that the traditional singers were proving unable to cope with such radically new themes ofa social-political nature as Marxism and related matters, for the simple Season that they Incl] formulae nonessary to express these new ideas in just Ineasures of verse. Except for rather obvious substitutions of key-words in an old formula (eg, enjla Dryhten for corla dryhter), no one singer ever creates many He was abe to make Eaglith poetry, compose a metoay, ad propel ei ut wets man: Fety ‘Abed rears that AVdhelm composed alight song which way sl (Le, in Alfred's dap) beng com- ‘Bonly sung, The people, at that tim [about 678] semibarbaroot and to ite intent on divine ‘Sacouren, weve inthe habit of hurrying to thet homes after masses had been sung. Therefore, the iy mau sro sel swage forse Avo wi enneste he tom fof Malmo] ne the countryside as an obotece to thowe ging vay, at Unough rofexing the ato song. After he had done this sveel times [tore than ebe'} be pine the good-wil and the attendance af the com- ‘som people. By thi device, gradually working a words of Ue Seriptaret among entertaining word, ‘emmy bo wemasked Ut the Scriptural words introduced inthe coune ofthe recitation of secular poco ned wot have ben in verse though thi sa reasonable inference. Ttshould alo be noted that athing bad about witng despite the rndeting “rite x poem’ (Poesim fee) of George E. ements Bt Addis: His Lite and Timus (London, 3908), p79 ‘ST eter nt merely to Boe hee! but to the tradition ofthe Whithy community on wanen Bede ier, surely compleely rlable in tht local mater, nlet one assumes a monstros conspireey of {eleifcation. 3 Whitelock, op. ip. 9s.om pp. O11 Whitelck i on the vege of suggesting what I suggest eve, 9 Fem, pp. S-4; Klaber, ed ly Px456 Oral-Formulaic Character of Anglo-Saxon Poetry new formulas and most of them never create any at all. Thus, standing on the Uurestild, 60 to speak, of the year 600, one might well have wondered whether and how Anglo-Saxon singers would be able to meet the challenge of ada their traditional verses to the needs of singing about themes so different as Chis. tian material wonld seem to ho. Tm actual fast they did rise to this vevasivu and ‘often magnificently. A glance of Chart II (pp, 466-67) analyzing I. 612-885 of Christ and Salan, @ poem of appreciably later date than Béowulf and mainly telling a atory of ‘Cluiat's harrowing of helt, exhibits plainly the formulaic character of the lan~ suage. If not as many verses are underlined as in Chart I, this can, in the ease of the unmarked verses, only mean that the surviving corpus of Anglo-Saxon poetry does not happen ta ramtsin verses whish furnioh supporting evidence, ‘that is, either exactly similar verses or, equally significant, verses constructed on elosely similar formulaic patterns. Te will be unnecessary to take up the text of this chart in detail, for the sup- Porting evidence will now be telling sts own story. There are, however, two matters, quite diferent from one another, which the present passage brings to ‘one’s attention. The frst concerns the ‘Christianity? of the language of this and Perhaps any other Christian nae. while the coaond concerns the possibility of occasionally making use of an understanding of the nature and function of the formula in textual criticism, i terest in the sample of verse analyzed on Chart II lies sm a Christian poem, At 1s 8 passage trealing a most ‘central event in Christian belief, the Ascension of Jesus Christ, and in that sense at least could scarcely be more Christian as opposed to the opening verses of Béowulf. What, a8 far a8 the language is eonoerned, ie Christian about it? Very largely references to God, specifically Jesus Christ. This passage of forty-six verses includes thirteen such references, more than one for every four verses: ‘wuldres Weard (§12a), Meotod mann-cynnes (6182), Drylten God (6144), enjla Deyhion (G10b), Gat (6221), Goes Sunu (o20b}, Sunu Meoloder (S27b), we Ha (880b), Péodine (882), Scieppend enfla (685b), and Drykten (6852). These ar all in one way or another diferent from one another. In addition there are ten other ‘Christian’ words, that is, words which swonldl normelly uly appear fn a Christian context: Galilee is mentioned twice (62a, 520a), Simon called Peter twice (621b, 686b); there is one reference to the Holy Spirit (525b), two to the disciples (520b, 529b), and three to angels (518b, 520a, 683b), of which two ‘econ ao parte of Kennings designating Use Delty. Zn all these forty-six verses include twenty-three Christian words, or words used in a Christan way; thus there is one Christian word for every other verse or one for each typographical Tine. Tt might be hard to find a more ‘Christian’ passage. and for theee wwe and formulas used m a Christian way only giengran lacks supporting evidence. This is no doubt due to the limits of the surviving corpus and, had the singer happened to have preferred formulas with the much more frequent equivalent of ‘disciple,’ namely fon, it would probably be possible Uv uullect no tittle supporting evi dence. In thio « Anglo-Saxc justed trad In the first singers ap] extraordint center And ‘of Anglo &« striking an names of F bbe and wer manic prop sccented o1 offered few awkwardly five times i complete A Christian ® the singers in words at ready to ta rooult of ne pounds of + ‘the mumbe: ‘odds the la land hence “angel isn by the old he old we ‘prosperou: tian doctrit tion predor tion, The D ignorance ¢ ‘ideas. The Levenson Ascension narrative.” Soe Head Nimeen. 19 Konipberger (6D “Chri see of Cyewull” ves (iesind Satan, a story of f the lane athe case glo-Saxon evidence, onseructed + the sup- ever. two brings to this and possibility fou of the art TI Ties 1g & most that sense verses of Lit? Very onty-aix 40), engla )s ae Bea ese are all are ten opear in a Hed Petor wo to the hich two ay; thus >araphical words and Save. The Tbappened “disciple, veting evie Oral-Formutaic Churucter of Anglo-Saxon Posty 457 In this so very Christian passage there may be & hint and more as to how Angle Sevon cingors were able. anparently from early on, to sing in a slightly ad- jnted traditional language songs based on these novel and untraditonal theres. {In the fist place and stated in most general terms, the Christian themes that the singers apparently liked best to sing about are in the main stories involving Cxteaordinary and exciting adventures aud creits, cuch ao the otorios on which center Andreas, Azariaa, Danial, Elene, Ezodus, Judith, and Juliana, To the ear of Anglo-Saxons not yet fully initiated inthis new development most immediately Steiving and strange were no doubt the presence of non-Germanie proper nouns, ‘ames of persons such as Simon Peter and places such a8 Galilee. ‘These could be and were, however, readily fitted into older formulas created to embody Ger- imanie proper names, and since these strange new names were all but invariably Recented on the Ast syllable, regardlcas of the stress in the original tongae, they ‘offered few, if any, metrical problems to the singer. Some of them must have been awkwardly long and more than queer sounding, such as Nabockodonossar, tsed five times in Daniel (48, 411, 407, 618, 668) and once in Azarias (183) to form a ‘complete A-type on-verse, yet the singers made do with then. Aside from the pre- Christian word God and elsewhere Ort, to be viewed as ordinary personal narnes, the singers had available from pre-Christian tradition, already evidently rich in words and Kenuluye weapreas the idea of ‘ruler’ large number of expressions reaily to take off the rack, available as substitutory epithets for the Deity. As @ result of new formations on the analogy of the old, e4., the weaving into com- yronnds of such characterstically Christian word-lements as heofor and wuldor, The number of epithets for the Deity was increased to a point where ths by all ‘cls the largest single group of kennings in the poetical eorpus.* The frequency fand hence importance of this group can scarcely be overestimated. The concept angel” rw aa is Ure lvaurword eno am iden also onphle of being expreseed by the old word gést. The Latin titles Sanctus and Beatus were easily handled by the old words hdlig (originally “inviolate’) and éadig (favored by fortune,’ ‘prosperous'). Expression of general conceptions of theology, dogma, and Chris- dian doctrine is notably rare in the Christian songs," as it1s m Beowudy, where ac- tion predominates, and even ia that most beautiful song of meditation or devo tion, The Dream ofthe Rood. This Jack is surely due nether to mere accident nor to Ipuctauve ur hndifesencey but to @ lacks of formulas cepable af adaptation to such ideas. he lyrically keyed poem on the Advent. (Christ J) and the song on the ‘Ascension (Chri I), based on the latter part of Pope Gregory the Great's ‘Ascension homily, are both traditional in diction and adhere pretty strictly to narrative.* 4 See Hendrik van der Merve Schult The Kenning in Alo Sano ed O1¢ Nore Posy (Utrecht [Nijonge. 102), pp 92-98 and Hertha Marquardt, Die engliosen Kenning, ete. Schrten der napster eelcrin Geet rll 200) wp a0 209, te Sn RN. at, {GD "Chratiche Bete’ 1 Gp. Kleber ipl, nit 1 £5. Mr Robert B Diamond presently engaged the study of “The Diton ofthe Sign Poems at Opn torvard the ooparatin) falls me 30 Aor 106) thst 80 per cent of the S196 eau Ue, 2508 nuabeved typos Hint ofthe ets) inthe signed poens of Cynewal are r=458 Oral-Formulaic Character of Anglo-Saon Poetry Tt would be wrong to sugzest that the ailaptation of the tredtioual laugunge of the ancient poetry to this new and differeat thematic material did not take doing on the part of the singers or to withhold from them full credit for the successful exercise of what at the outset particularly must have called for skill ‘and ingenuity. It io, however; far 1 Point out, m view of the obviously traditicnal Tanguage of the Christian poems —a matter that in essence has long since been noticed and stressed — that the singers did not make things unnecessarily hard for themselves by attempting to sing ahont matters for the expression uf which the old diction would have been inadequate. As it was singers and audience Drobably felt little difference between the general style and narrative tech nique of, say, Béowulf and Christ and Satan, to mention two poems of very jlifferont thematic backgrounds. This marked unuformness oF unity of style i largely to be accounted for by the continuity of the traditional formulaic language of the Anglo-Saxon singers, a continuity that seems to live until the Norman Con, ‘Many factors, political and social as well as linguistic, probably contributed to the death of the traditional poetry after the Conquest, and one must also reckon with the diffculties, probably insuperable, which the relatively. swift [iiretaetion of idone aut activities Ineldental to the advent of the feudal age brought in their train, ideas which eould not easily be sidestepped by singers ‘trying to sing in the old tradition and for which they had no formulas, Quite by chance the present passage from Christ nnd Satan offers an oppure tunity to consider the general possibilty of the use of an understanding of the role of the formula in occasional matters of textual criticism, Verse 513b, with the manuscript reading dr on morgen (A), ‘early in the morning’ lechnially vio Infos « basio principle of alliteration in that the inst down-beat or lotus fa tag ‘olf-verse does not here allterate with the preceding on-verse where the alliters, ton is m. Acting on « suggestion of Professor Holthausen, Professor M.D. Clabh cmended tis verse in his edition of 1925 to reac am morgen so, thao producing Rich he nghtly described as a normal (B) verse. Nevertheless, in the light of the supporting evidence which demonstrates the existence of x formula tr on morgen, taken together with the phrase on ér-morgen, with which inay also be compared mii dr-dage of cimilac sucauiny, une may wonder whether the did not himself violate the usual procedures of 8 formula that he needed, a formula or system in which dr preceded the word it modified. Consequently, one might do well. not only here ut $9 other seven ftuations, to test such alliteratively defective verses for their formulaic character before embarking on a course of emendstion, however rach better emendaticg ray make, or may seem to make, matters. If given time to think his vere ever; in & word to comps at a more lfoucly pace, a singer might well agree with what a modern editor was proposing todo; on the other hand, such an enowdavon er rit preposing to dos on the other hand, such an e Seeee hare 80 pe oat epeated vere, 6. per ent of verc of which ura Bp etary ‘eeureacocliewhere, demonstrate themselves tobe forme, On aight produce technical defer Tt this dise lett regardless the ease of mi cemendation te by certain ear ‘Tho oupportic or of, meaning always except tomb. Now it venient to ret from a surfac ‘from’; itis si in Christ and . holy if not © that Our Lord long te ine of the preposi not support ‘The futare aspects of An these are, or @ ofall let it be, is to be condut construction 0: lecting of supp poup of waree More sampl seems doubtfu # For aay comp forthe OlbNowe (ce 27, belo). “Baise chat Coop of Carell nln ond ape the Slips thee No eomconacceaL language 4id not take adit for the led for skill "traditional ‘since been ssarily hard a of wale «d audience ‘ative tech- sontributed nically vio~ cctus in the hho allitora- LD. Club + producing :he Tight of mula ér on ray also be the singer ake use of the word it hor similar ce character -mendation verse over, agree with mendation Sndoding the ‘Stic sarples saxon poetical by virtue of Oral-Formulaic Character of Anglo-Saxon Poetry 459 sight prodhior a secuence of words which would strike him as stranger than the technical defect in versfication. Tf this discussion of manuscript ér on morgen suggests that it should be Jeft regardless of the technical imperfection that its use and retention produces, the case of manuscript un jut feelonne (E100) would ecom to eponk in favor of temendation to of bam fastonn, “trom, out of the tomb,’ an emendation adopted by eertain earlier editors, though not by Clubb or Krupp, last to edit the poem ‘The mupporting evidence on Chart II exhibits two expressions, one with fram or of, meaning from or out ofthe prison, stronghold or tomb,’ the other with om, flways except here with the obvious meaning ‘in the prison, stronghold or tomb’ Now it is true that Old-English uses expressions with on which are con enlent wW renter by ‘hrvas? generally in conncotion with romaving something from a surface on which the object in question is lying or reposing (see B.T. ‘2, ‘on,’ M1, 2). From the Anglo-Saxon point of view on is in these eases entirely ‘apropriate, though the approach to the act ig different from ours. It is as if one said ‘he took the pencil on the table,’ that is, ‘he took the penest which was Tying fon the table,’ in the sense thet he took it from the table. When Grendel assails ‘Béowul, itis said that the troll name. ..rinc on reste (I, 746-47), ‘took the war- Flor frum lia seating place.” hia iy however, far from saying that OTR om means ‘from’; it is simply to say that the image of the action is different, In the verse jn Chriet and Satan such an image would in the nature of things be highly wn- Tikely if not out of the question altogether. The singer must be trying to say that Our Lord went out of the tomb and thus it s all but certain that te man seript on does not go back to the words of the singer or to anybody who was giving attention to the thought but to» miscopying by scribe somembere un Une Inc of written tranomiosion. If thie eso, then in the small verbal matter of the preposition, manuscript on, the supporting evidence involving on’s does not support the manuscript reading, but rejects it rathe® ‘The future is full of many problems involving a reappraisal of ‘certain aapeots of Anglo-Saxon poetical style and compositional vechnique aul ‘these are, or at present seem likely to be, can here be merely adumbrated. First of all let it be said that, if further study of the formulaic character of the poetry fs to be conducted in « thoroughgsing way, the fra and most erving need is the construction of a concordance of the entre poetical corpus; without tbis the eol- leeting of supporting evidence to test the formulaic character of « given-verse ot ‘group of verses will prove to be incredibly laborious and often uncertain.” More sample analyses of narrative verso are certainly desirable, Uugle it seems doubtful that any narrative poem will be found to be non-traditional in ‘For any comparative sndy of Ok Germanic formulsc dition oncordanos are equally needed a OM do type eee (eed af Sod. I! sla) al fo the Old Saxon cOrDo8 (dee 127, below, ict techngbe fr concordancesmaking have bee worked out by Profoutor Berit Tne Cooper of Cornell University andar ct forth n considerable detail in “The Making and the Use of 1 Verte Connon Sree Ra x7 (010), 18-208, ep. 101-185, repented in his Tsiution ond Repentance (Thee, N.Y 1080), ep pp. 2488. Sesto ie Tstructions er preparing the ipa thre pegs inserted in A Conerdane to the Wark of Hace (Warhiogton, D.C 101), No consordance should eve be ataspted without consulting these wigs.160 Oral-Formulaic Churucter uf Anylo-Sueon Poetry language. Particularly interesting will be a study from this point of view of the diction of the rather small body of Iyrical- Reg, Klacher, ad cit, pp. cx-eti, For a competent nurvey of thought on the testimony of the Parallels ce Clas Schaar, Crital Studien ix the Cyne) Group (Lund Studies ta Engh, xvte ‘and, 140), pp, 288. Over sixty years ago J. Kai, ‘Uber de Paralleltllen in der anglaise ben Foie: Anglia, x1 (1859-00) 81-80. wan leary nner ght shen he iced to kovee fa ae of OlaSaxon poetry a stare Was tne by Eduard Sievers i bin 0 Hiiend alle, 1879) c sense represt force learn f verbal sia singers have measure of ¥ ture. One of more strikin, to make use get to know eis neue individual = broad archit traditional 1 stylistic, an nov to all Tack of selative age something © suffered. cot din the Anglo- 89-1008 of some of the in the sup- dales. often the verses in), notably verween let- a Christian ss the story expounded easily have + how could vy means of fg for Use ssting to be 1 the tradi- Jictating to ‘procedure reontradic- 2 or lack uf doems than ‘smay ina the uni si ‘comp, show of recurence 4a gid ere) sie. Does thin by « clas of ‘al timer? One feweare whl eye Pe ogi, x0, tehtchriachen| a the ease ‘ate, 1678) Oral-Formulaie Character of Anglo-Saxon Poetry 461 sense represent borrowing from one poem to another, for traditional singers per- force learn from other singers. But one verbal similarity or even a number of verbal similarities in themselves prove nothing heyond suggesting that given ngcca have found the came formulas weaful to oxprese x pertain iden in » eimilar Ieasure of verse. To quote Parry, ‘Plagiarista is not possible ture, One oral poet is better than another not because he has by himself found a ‘more striking way of expressing his thought, but because he has been better able to make use of the tradition.» When by the aid of a concordance we gradually get to know what the Anglo-Saxon formulas are and what, indeed, constitute ‘their dimensions® and the like, it will perhaps be possible to begin to detect Individual styles. Apart from general aver-all organization of material. the broad architectonics of a given poem, a singer's individuality will, as in other traditional poetry, presumably emerge in rather small matters, verbal and stylistio, and will not be revealed by the large and rather obvious components, Known to all or almost all singers. ‘Lack of truly eat]y material will preclude our ever knowing much about the relative age of the formulas encountered in the preserved poems, but perhaps fomething non ha dane with verses containing words which in earlier times had suffered contraction, either from the simple contraction of two vowels (as din being doe, As: Impets toa revaluation ofthe expanded verse has eently bees given by Benno J. Time, “Rapended Eine ie OMEagl Puery” Nenpuopes, SOXY UdSI), GAZE. On this to be the seribes sue sin when they me schools. That ‘typographical ‘the onset of a time with foo musical comp ‘Haxvano Uretry hhave hung on to atraction-process ay singers and fs and accepted tably be further rmulas. 80 is the poems there are tch Anglo-Saxon sat is required, Y offers at least ‘th the aid of a ‘Ta, 2942-450); aig uf 970, igin and special veasual survey ccems to he for- organization of 5 somewhat less a not available + further study problems of the test and solve, of the physical al background, the vernacular 1 parchment as hard to believe "aformala leds to setons cf th Amer >og iid woe 1 babe may. Toere (ine aboenine Benno J. Timmer, Oral-Formulaie Character of Anglo-Sazon Poetry 468 this to be the case and suspect it was written as prose merely because neither setibes nor singers understood in a formal sense the metries of the verse, even when they may have had an understanding of Latin verse studied in monastic schools. That tenth-leventh century senibes at times separate verses (not our typographical lines) by dots may merely reflect a feeling for the basic rhythm, the onset of a down-beat, comparable to « inusically unschooled person's tapping time with font ar Ringer thenigh Imawing nathing of the writing of music ar of misical composition. Hanvanp Untvgesiry