Chapter 3
Cædmon’s Hymn and Germanic convention

[Skip to Content]


General Navigation

This chapter

[Beginning of Content]


§ 3.1    The preceding chapters suggest that Bede was impressed by Cædmon’s skill as a poet and interested in demonstrating how unusual he was as a composer and performer. Although few modern scholars have shared Bede’s enthusiasm for the intrinsic aesthetic quality of Cædmon’s work, more have shown an interest in the question of his distinctiveness, arguing either that his poetry developed primarily out of (non-native) Christian traditions, or, more commonly, that his “bold adaptation” of pagan Germanic conventions to Christian ends marks the beginning of a new vernacular tradition in Old English and the other Germanic languages.[1]

§ 3.2    As we shall see there is little evidence to support either claim. Structurally speaking, Cædmon’s Hymn is a very Germanic poem. It shows a deep debt to the same major metrical, stylistic, and formulaic conventions found in other, mostly later, verse texts in Old English and other Germanic languages. Stylistically, there is little evidence to suggest that Cædmon’s verse played the innovative role commonly assigned to it in modern literary histories. Bede nowhere suggests that Cædmon was known for his formulaic innovation, and Cædmon’s use of traditional Germanic epithets to describe the Christian God betrays no self-consciousness about his own supposedly seminal role in the development of native Christian poetry. In fact, the surviving contemporary evidence suggests that Cædmon’s verse was valued primarily for the prosodic and formulaic skill with which it was composed. While Bede’s account does suggest that Cædmon’s training and methods of composition and performance were unusual, there is nothing in the Historia ecclesiastica or the Hymn itself to suggest that the poet’s actual verse was strikingly innovative.


§ 3.3    Metrically, Cædmon’s Hymn is an accomplished but very conservative poem. Its eighteen half-lines show four of the five major stress-patterns identified by Sievers and traditionally used to describe the most conservative Old English verse. Only Type C verses (i.e. verses showing the stress pattern ×‐́‐́×) are not found in the poem (in the following scansion, ‐́ indicates a “heavy” metrical lift, ⏑́‿× a “resolved” lift, ‐̀ a half-lift, × a metrical drop, and (×) a syllable to be ignored in scansion):

Table 1: Scansion of Cædmon’s Hymn.

(A-3: ×××‐́×) Nū scylun hergan      hefaenricaes uard (E: ⏑́‿×‐̀×‐́)
(A-1: ⏑́‿××‐́×) metudaes maecti      end his mōdgidanc (B: ××‐́×‐̀)
(D-2: ‐́‐́(×)‐̀×) uerc uuldụrfadur      suē hē uundra gihuaes (B-2: ××‐́××‐́)
(A-1: ‐́×‐́×) ēci dryctin      ōr āstelidæ (A-1: ‐́×⏑́‿××)
(B-1: ×‐́×‐́) Hē āerist scōp      eordu barnum (A-1: ‐́×‐́×)
(A-1: ⏑́‿××‐́×) heben til hrōfe      hāleg sceppend (A-1: ‐́×‐́×)
(B-1: ×‐́×‐̀) thā middungeard      moncynnæs uard (E: ‐́‐̀×‐́)
(A-1: ‐́×‐́×) ēci dryctin      æfter tīadæ (A-1: ‐́×‐́×)
(A-1: ‐́×‐́×) fīrum foldu      frēa allmectig (D-1: ‐́‐́‐̀×)

§ 3.4    Isaac and Creed have suggested on different grounds that the Hymn shows an unusually high frequency and anomalous distribution of Sievers Type A lines (Creed 1992; Isaac 1997, 223-225).[2] Both claims are true in absolute terms: the Hymn does show an unusually high number of Type A lines in both the on- and off-verse when compared to the corpus as a whole or to large poems such as Beowulf; it also shows an unusually high number of long lines with Type A verses on both sides of the caesura (comparison statistics can be found in Hutcheson 1995, Appendix D).[3] The small size of the sample in question, however, ensures that neither difference is statistically significant: the Hymn shows anomalies for all verse combinations and types (except Sievers Type D in the on-verse) and can be paralleled easily by similarly anomalous distributions of verse types in other samples of similar length from the corpus (examples include the Leiden Riddle and the first nine lines of the West-Saxon Seafarer poem).

§ 3.5    Cædmon’s Hymn agrees well with one particular aspect of the distribution of verse types in the Anglo-Saxon poetic corpus: the use of a light Type A-3 verse for the poem’s initial half-line. As Stanley notes, such light lines are a characteristic choice for the opening of Old English poems, particularly biblical paraphrases, prayers, and translations (Stanley 1987a, 128-129). Parallels include Daniel, Christ and Satan, Andreas, Seafarer, Fortunes of Men, Descent into Hell, Alms Giving, Pharaoh, Aldhelm, all of Cynewulf’s poems, and a large number of Riddles and Boethian Metres.

§ 3.6    In a brief article, Robinson challenges this view, suggesting instead that the Hymn’s first word, the adverb Nu, might have received contrastive stress for rhetorical purposes (Robinson 1993b, 118-119)[4]:

[in Cædmon’s Hymn] means “in this new Christian age.” The familiar framework of the formulas of praise refers to the old heroic age that minted them; their Christian reshapings refer to the new age of Christianity. The introductory establishes the relationship between the two: “In the past we used these terms to praise men or pagan deities; now we must praise God with them.” The freight of meaning borne by the adverb being so heavy, I cannot think that it was submerged phonically in the murmur of a metrical dip. If any word in the poem must have logical and metrical emphasis, it would seem, it is this one.

§ 3.7    While this reading is attractive, it is probably beyond demonstration. In the first place, as Robinson notes, it depends on the assumption that Cædmon’s Hymn represents a bold Christianisation of traditional Germanic encomic poetry, “a poetic form that seems to have been primal in many heroic cultures and that must have been one of the commonest forms practised by Anglo-Saxon scopas during their pre-Christian period” (Robinson 1993b, 117). Although an eulogistic tradition of some type does appear to be implied by classical accounts of continental Germanic customs, the genre itself, as Opland, Bessinger and others have noted, is not actually attested in Old English or other early cognate literatures, leaving us no body of contemporary material against which to compare Cædmon’s poem (Opland 1980, 98; Bessinger 1974, 92; Lehmann 1971, 24-29).

§ 3.8    In the second place, the adverb does not participate in the line’s alliteration. As the alliteration of various forms of the demonstrative pronoun in Beowulf in 197, 790, 806, and 1797 demonstrates, Anglo-Saxon poets could indicate rhetorical emphasis by allowing elements that are normally without metrical stress to participate in alliteration, either as part of the main alliteration, or, as frequently occurs with elsewhere in the corpus, as part of a “transverse” or “crossed” pattern in which each stress in the on-verse alliterates on a different sound in the off-verse, e.g. Beowulf 2053, Nu her þara banena      byre nathwylces (Stanley 1974, 142).[5]

§ 3.9    Finally, as we shall see below (§ 3.30), the interpretation is not supported by Bede’s discussion of the poem’s initial performance and reception, which includes little to suggest that Cædmon was either the first Anglo-Saxon poet to compose Christian Old English verse, or that others found his poetry to be noteworthy for anything other than its exceptional quality, effectiveness, and, to a lesser extent, unusual method of inspiration.

§ 3.10    On the other hand, Robinson’s suggestion may receive some support from Bede’s paraphrase, which, as we shall see, ignores or provides unusual translations for several key terms in Cædmon’s poem and may indicate that Bede found Cædmon’s terminology so innovative as to warrant suppression in much the same way Green has argued Wulfila suppressed similar vocabulary in his Gothic translation of the Bible (see Green 1965, 278-279). Bede’s “suppression” of Cædmon’s epithets is itself quite uneven, however, and, as we shall see, can be explained perhaps better by reference to his desire to eliminate all syntactic apposition from his translation (§§ 3.31-3.35, below).

Alliteration and rhyme

§ 3.11    Alliteration or initial rhyme is the most certain aspect of Old English poetic style (Hoover 1985, 75, and the statistics in n. 1). The Hymn agrees absolutely with standard expectations for the distribution and phonological characteristics of alliterative syllables in conservative Old English poetry. Consonantal alliteration in the poem always involves like consonants; vocalic “alliteration” pairs unlike vowels in every instance except 5 of the Northumbrian aelda recension. No line in the poem alliterates on palatal or velar /g/ or any of the /h/ or /s/ clusters, two tests often used for dating Old English verse (see Amos 1980, 93-96).

§ 3.12    The distribution of alliterating sounds in the Hymn also corresponds to standard Old English poetic practice. In the off-verse, alliteration is always found on the first stressed syllable and never on the second. In the on-verse alliteration falls on the first stressed syllable only or both the first and the second. There are no examples of crossed alliteration and relatively few lines showing double alliteration (in which both stresses in the on-verse alliterate with the first stress of the off-verse): in five lines, alliteration in the on-verse involves the first metrical stress only; in only four does it involve both. As Amos notes, this relatively low percentage of lines showing double alliteration is characteristic of all three externally datable Northumbrian poems, Cædmon’s Hymn, Bede’s Death Song, and the Leiden Riddle. It probably represents a regional stylistic feature (Amos 1980, 95-96).

§ 3.13    One verse-pair in the poem is tied together by rhyme as well as alliteration: tha middingard, moncynnæs uard, 7 (text: P). That the device plays no structural importance in the poem (or line), however, is suggested by the Moore scribe’s spelling of the same set of verses, in which the rhyme is partially obscured: tha middungeard, moncynnæs uard (see Stanley 1988, 23, and n. 10). The rhyme (or its West-Saxon equivalent, middangeard... ƿeard) is preserved in all other copies of all other recensions of the poem except West-Saxon ylda recension, where it is obscured by the use of an inflected form, middangearde, in the on-verse.

§ 3.14    Howlett proposes an additional and complex series of rhymes linking various sections of the poem together (text adapted from Howlett 1997, 268-269; rhymes indicated by italic type[6]):

    Nu scilun herga      Hefenricæs Uard,

    Metudæs Mehti      and His Modgithanc

    uerc Uuldụrfadur.      Sue He uundra gihuæs

    Eci Dryctin      or asteli.

5    He ærist scop      ælda barnum

    hefen to hrofæ      Halig Sceppend.

    Tha middingard      Moncynnæs Uard,

    Eci Dryctin,      æfter tia

    firum foldu      Frea Allmehtig.

§ 3.15    The very complexity of this system argues against its deliberateness. The rhymes indicated by Howlett’s typography seem to involve a series of almost random correspondences. Howlett marks no rhymes in 1a, 2b, 3a-b, 5a-b, or 9a; the rhymes he does record vary greatly in significance and proximity (all spellings from Howlett’s text): Uard, 1b and 7b, rhymes with middingard, 7a; Mehti, 2a, rhymes with Allmehtig, 9b; Eci Dryctin shows an identical rhyme in 4a and 8a; hefen and Sceppend show double rhyme in 6a and 6b; and the inflectional ending in astelidæ, 4b, rhymes with that of tiadæ, 8b (in light of the feminine rhyme in this last example, it is not clear why Howlett does not include two more examples: herga and Tha, 1a and 7a, and barnum and firum, 5b and 9a). While such correspondences may help give the text as a whole a sense of coherence (cf. the discussion of sound patterning in Kubla Khan in Crystal 1987, 74), so few of the rhymes appear at regular intervals or involve necessarily connected ideas that it is difficult to see how an Anglo-Saxon auditor would be able to perceive them as being of structural importance.


§ 3.16    Stylistically, Cædmon’s Hymn is probably most remarkable for its heavy use of ornamental poetic variation, particularly in the poem’s last five lines. The device has been formally defined in Robinson 1993a (73) as consisting of

syntactically parallel words or word-groups which share a common referent and which occur within a single clause (or, in the instance of sentence-variation, within contiguous clauses).

Informally, the term “variation” is often extended to cover the repetition of similar words or the restatement of similar ideas or concepts, regardless of whether or not they stand in strict syntactic apposition (see Robinson 1993a, 71-73). As Robinson notes, this stylistic feature is so prominent in Old English verse that it attracts “the attention of even the most casual observer”.

§ 3.17    Regardless of the precise definition used, Cædmon’s Hymn shows a great amount of poetic variation. Rhetorically, as Schwab has noted, epithets for God occur in the Hymn at what approaches the maximum possible frequency: a noun, adjective, or pronoun referring to God the Father is found in every single line of the poem (Schwab 1983b, 23). Structurally, the poem can be divided into two major sections, each containing two clauses, of which the second can be read largely as an expansion of material in the first.[7] The poem’s first four lines are used to state the theme, that men ought to worship God, and explain in general terms why this worship is due: sue he uundra gihuaes, / eci dryctin, or astelidæ, “as he, the eternal Lord, created the beginning of each of wonders” (3b-4); the last five lines expand on this point by recounting the actual order in which the physical manifestation of these wonders was completed: first God created the heavens, then the earth for men. As Huppé, Bloomfield, and Schwab have noted, this two-part structure echoes contemporary ideas about the dual nature of creation: a “first” creation in which God produced the eternal and spiritual ideal and a “second” in which the temporal and physical aspects of this original ideal were realised (Huppé 1968, 129-131; Bloomfield 1962, 42-43; Schwab 1983b, 23).

§ 3.18    In Robinson’s stricter sense of the term, variation is confined almost entirely to the last five lines of the poem. As I have argued elsewhere (O’Donnell 2004), Cædmon’s Hymn changes tone significantly at the end of 4b. In the first four lines, the poem’s theme and rationale are stated in a relatively spare and straightforward fashion: in the traditional reading (see § 3.17 and n., above), the two clauses making up the first section show an adverbial relationship (the second clause explains why we should do as the first commands), have different subjects, and employ verbs that differ in sense, number, tense, and perhaps person. Even the numerous references to God in these lines are not strictly speaking syntactically apposite, but vary either in case or in their association with different aspects of the Godhead: the first epithet in the poem, uard, “Guardian,” 1b, is accusative singular and refers to God the Father; the second, third, and fourth references, metudæs, “Creator,” 2a, his, 2b, and uuldurfadur, “Father of glory,” 3a, are all genitive singular but modify nouns describing different parts of God’s person and work. Indeed the only true example of syntactically apposite variation in the first four lines of the poem is found in 3b-4a, where the nominative singular pronoun he, 3b, and the nominative singular noun phrase eci dryctin, “eternal Lord,” 4a, have the same referent and serve as subjects for the same singular finite verb (in the following citation, loosely appositive forms are in italic type; strictly appositive terms are in bold; superscript letters are used to group apposite elements):

    Nu scylun hergan      hefaenricaes uardA

    metudæsA maecti      end hisA modgidanc

    uerc uuldurfadurA      sue heB uundra gihuaes,

4    eci dryctinB,      or astelidæ!

§ 3.19    In the Hymn’s last five lines, however, Cædmon makes far greater use of appositive variation in Robinson’s strictest sense. In contrast to the clauses in 1 through 4, the two clauses in 5 through 9 are syntactically equivalent: they are loosely linked by the time adverb tha, “then,” 7a, and employ verbs, scop, “created,” and tiadæ, “appointed, established,” 5a and 8b, that are approximately synonymous, refer to the same subject, and agree in tense, person, and number. Above all, however, epithets for God used in each clause in this section are all in strict syntactic apposition: all five are nominative singular and refer to the subject of their respective and parallel verbs (in the following citation, loosely appositive forms are in italic type; strictly appositive terms are in bold; superscript letters are used to group apposite elements)[8]:

5    HeA aerist scopB      eorduC?/aelda barnumD

    heben til hrofe,      haleg sceppendA;

    thaE middungeardC,      moncynnæsD uardA,

    eci dryctinA,      æfterE tiadæB

    firumD folduC,      freaA allmectig.

This pattern of appositive repetition has important implications for our understanding both of what Bede saw as being the most significant features of Cædmon’s art (see below, §§ 3.31-3.35) and the relationship between Bede’s paraphrase and the Old English poem (see Note B: Cædmon’s Hymn and Bede’s paraphrase).

Poetic diction and formulae

§ 3.20    In addition to being metrically regular, following the expected rules for alliteration, and, in its second half in particular, showing heavy use of ornamental appositional variation, Cædmon’s Hymn also corresponds to the norms of traditional Old English verse in two other respects: its use of “poetic” diction and its heavy reliance on poetic formulae. At least six of the poem’s forty-two words belong to the Anglo-Saxon “poetic vocabulary” (the precise number varies according to the recension examined and criteria used for establishing this vocabulary): metudæs, “of the creator,” 2a; uuldurfadur, “of the father of glory,” 3a; or, “point, origin,” 4b; firum, “for men,” 9a; foldu, “(of) earth,” 9a; and frea, “lord,” 9b.[9] A number of other forms, uard and dryctin among them, are very common elements in poetic formulae, although they are also found frequently in prose.

§ 3.21    Depending on the specific recension chosen and definition used, moreover, between nine and fifteen of the poem’s eighteen half-lines can be paralleled closely elsewhere in the Old English poetic corpus: verbatim in eleven half-lines: 1b, 2a, 2b, 4a, 4b, 5a, 5b (Northumbrian aelda and West-Saxon ylda recensions), 7b, 8a, 9a (Northumbrian eordu and West-Saxon ylda recensions), and 9b; as representative of a clear formulaic system in another three: 3b, 5b (Northumbrian eordu and West-Saxon eorðan recensions), and 6b.[10] In some, more controversial, formulaic analyses, as much as 100% of the poem is said to be paralleled by poetic formulae or systems found elsewhere in the Old English corpus (see Fry 1975 and Fry 1979 but cf. the objections in Miletich 1983). As Fulk indicates, this highly formulaic structure and use of poetic language go together: many formulae make use of poetic words, and participation in common formulae is no doubt responsible for the preservation of otherwise archaic terminology in poetry long after such words have fallen out of colloquial use in prose (Fulk 1992, 265-268).

§ 3.22    The heavy use of poetic and formulaic diction in Cædmon’s Hymn is significant for two reasons. In the first place, the poem is among the earliest recorded examples of sustained vernacular verse in any Germanic language (Stanley 1995, 139). While the evidence of later texts suggests that much of the poem’s language belongs to a traditional poetic register, this traditional register is, in the case of several words and most if not all formulae, first attested in the Hymn itself.[11] Thus while most scholars assume that Cædmon is drawing on and adapting a pre-existing Anglo-Saxon poetic tradition in his work, his poem is also very often our earliest evidence of this tradition’s existence.[12]

§ 3.23    But the traditional poetic vocabulary and formulae of Cædmon’s Hymn are significant for another reason as well: as Bede’s account suggests and as the internal evidence of its Germanic vocabulary makes clear, Cædmon’s poem, while one of the earliest attested examples of traditional Germanic poetry, is not itself directly representative of the earliest Germanic tradition. As a Christian poem, Cædmon’s Hymn by definition adapts, rather than seamlessly continues, the original pre-Christian tradition from which it presumably draws its vocabulary and formulae. This can be seen nowhere more clearly than in the poem’s epithets for God. Two of the four terms for God in the last five lines of the poem, dryctin and frea, have their origins in pre-Christian Germanic lordship terminology (Stanley 1995; Green 1965); a third, uard, is commonly used in a non-religious sense to describe a “keeper” or “guardian.” One of these terms, dryctin, is attested for the first time in any Germanic language in Cædmon’s Hymn (Green 1965, 287); a second, frea, is found earlier only in cognate form in Wulfila’s Gothic translation of the Bible (the implication of Green 1965, 503).

§ 3.24    This means that our earliest knowledge of these terms, and hence the formulae in which they are found, is from their secondary, metaphorical, usage in Christian poetry. In describing God as a uard, dryctin, and frea, Cædmon is at best comparing the Christian deity to concepts traditionally applied to Germanic heroes and leaders; at worst, his terminology may consist of little more than dead metaphor, tags already so conventional as to be more or less drained of any symbolic force. Either way, neither the Hymn nor most of the subsequent Old English poetic tradition gives us direct access to the precise connotations this pre-Christian vocabulary may have had for early Anglo-Saxon audiences. As Stanley points out, only twenty-two of the approximately 15,500 occurrences of dryctin in the Old English corpus are intended in the secular sense of “lord, ruler, chief” (Stanley 1995, 138); a similarly effective Christianisation is also seen in the other Germanic languages. Green notes that most examples of traditional Germanic lordship and comitatis terms such as dryctin and frea in the cognate languages survive in Christian contexts, in most cases, indeed, in contexts that can be traced to centres of Anglo-Saxon missionary work or influence (Green 1965, 286-287).

§ 3.25    Traditionally, this Christianisation of traditional Germanic poetic language and formulae has been attributed directly to Cædmon and his “school” (e.g. Stanley 1995, 139-140; Robinson 1985, esp. 30 n. 9). Most accounts of the poem’s composition and reception assume, therefore, that Cædmon’s initial audiences would have found his use of pre-Christian terms such as frea, dryctin, and uard daringly novel. As Smith 1978 suggests (15):

...[I]n Cædmon’s time when Northumbria had been converted to Christianity for only half a century these phrases belonging to Christian poetry could scarcely have become conventional, as they certainly were in later Old English; on the contrary, the poem represents the beginnings of such a diction and its freshness and originality must have been felt a generation or more after its composition; no mere assembling of clichés would have called for inspiration, divine or otherwise.

§ 3.26    Whether Cædmon’s verse really had this effect on contemporary audiences is open to doubt, however. While the poem does contain some of the first recorded examples of presumably traditional Germanic terminology in the new non-traditional Christian sense, nothing in Bede’s account of its initial reception, or indeed in the use of this terminology in the poem itself, suggests that Cædmon’s text was valued primarily for the novelty of its diction. As we shall see below (§ 3.30), Bede’s own praise of the poem concentrates on its sweetness, order, and propagandistic value; the praise he puts more indirectly in the mouths of Hild and her counsellors is primarily concerned with its smoothness, propriety, and faithfulness to its sources. No one in Bede’s chapter comments on Cædmon’s role in the Christianisation of previously pagan poetic diction; the Hymn itself applies its supposedly novel vocabulary without any obvious recognition of the freshness of its invocation of God as a traditional Germanic lord.[13] Opland has suggested that Cædmon’s Hymn is important because it demonstrates how an Anglo-Saxon scop can be not “the praise-poet of a lord, but... a praise-poet of the Lord” (Opland 1980, 116). It is difficult, however, to imagine a less vehement vehicle for such a message.

§ 3.27    The strongest evidence against the traditional reading of Cædmon’s poem as a bold Christianisation of an earlier pagan Germanic tradition lies in the Hymn itself, which seems to show no recognition of its supposedly seminal function. Cædmon’s Hymn is not the poem to turn to to discover what it is like to feel the loss of a recent pagan heritage or to exult in the recognition that the new religion can provide its adherents with heroes as significant and dynamic as those celebrated in the old songs. If Cædmon is responsible for introducing terms like dryctin and frea in a religious sense, then he manages to rededicate this vocabulary with remarkable ease and lack of self-consciousness. He may describe God in terms appropriate to the Germanic hero, but nothing in his poem suggests that he is using these words in anything other than the stereotypical sense they acquire in later Old English Christian translations and prayers. His use of uard in 1b and 7b does contain some sense of “protector” or “guardian.” His description of God the Father as dryctin and frea in 4a, 8a, and 9b, however, shows no evidence that Cædmon actually saw Him in terms of a traditional Germanic leader.[14]

§ 3.28    Just how little emphasis Cædmon places on the pre-Christian and Germanic connotations of the epithets he chooses can be seen when the poem is compared to two passages of approximately similar length from poems in which tension between pre-Christian and Christian world views is at issue. The Song of Creation in Beowulf (ed. Kiernan 1999 [Edition]; trans. Bradley 1982), a frequently mentioned analogue to Cædmon’s Hymn (see above, § 2.31), for example, uses the ambiguity inherent in Old English poetic diction and formulae to emphasise the distance between the full Christian knowledge of world history shared by the poet and his audience and the far more limited world view available to the pagan characters of his poem. This is particularly true of the Song’s double ending: after a relatively orthodox account of creation in 92-98, Hroðgar’s scop goes on to what a Christian audience would initially understand as a brief resume of prelapsarian life and the Miltonesque introduction of the feond responsible for bringing “death into the world, and all our woe”:

         ....         Þær wæs hearpan sweg,

90    swutol sang scopes.      Sægde, se þe cuþe,

    frumsceaft fira      feorran reccan,

    cwæð þæt se Ælmihtiga      eorðan worh[te],

    wlitebeorhtne wang,      swa wæter bebugeð,

    gesette sigehreþig      sunnan ⁊ monan,

95    leoman to leohte      landbuendum.

    ⁊ gefrætwade      foldan sceatas,

    leomum ⁊ leafum,      lif eac gesceop

    cynna gehwylcum      þara ðe cwice hwyrfaþ.

    Swa ða drihtguman      dreamum lifdon

100    eadiglice,      oð ðæt an ongan

    fyrene fre[m]man,      feond on helle.

[T]he lyre’s music was there, and the clear singing of the poet. He who was skilled in recounting the creation of men in time distant declared that the Almighty made the earth, a plain radiant to look upon which water encircles; he, taking delight in his achievement, established the sun and the moon, those luminaries, as light for those living in the world; he embellished the earth’s surfaces with branches and with leaves; life too he created in each of those species which go their vital ways.

So the men of that community lived happily, blessedly, until one being, a fiend in torment, began to perpetrate outrages.

As the next line of the poem, Wæs se grimma gæst Grendel haten, “[t]hat savage visitor was called Grendel,” indicates, however, this Christian view of the song’s conclusion is, in the context of the poem’s narrative at least, incorrect. The identification of Grendel as the grimma gæst makes it clear that the scop’s song actually ended in 98, not 101, and that the happy retainers and the feond on helle of 99 through 101 are in fact part of the framing narrative. In forcing his audience to re-evaluate its understanding of the precise moment at which the song of Hroðgar’s scop ends, the Beowulf poet emphasises the distance between contemporary Christian understanding of world history, in which the significance of the underlying biblical history is recognised, and that of his pagan characters, who are doomed to repeat the original error unknowingly. As Robinson has argued in relation to other passages of the poem, this dual vision is part of a calculated effect whereby the poet allows his “devoutly Christian audience” to admire his pre-Christian heroes “while remaining fully aware of their [i.e. the heroes’] hopeless paganism” (Robinson 1985, 11).

§ 3.29    A second example of the possibilities inherent in the pre-Christian tradition but ignored by Cædmon can be seen in the passage describing Christ’s ascension to the cross in the Dream of the Rood (ed. Swanton 1970; trans. Bradley 1982):

         ....         Geseah ic þa Frean mancynnes

    efstan elne mycle      þæt he me wolde on gestigan.

35    Þær ic þa ne dorste      ofer Dryhtnes word

    bugan oððe berstan,      þa ic bifian geseah

    eorðan sceatas.      Ealle ic mihte

    feondas gefyllan,      hwæðre ic fæste stod.

    Ongyrede hine þa geong hæleð,      (þæt wæs God ælmihtig),

40    strang ond stiðmod;      gestah he on gealgan heanne,

    modig on manigra gesyhðe,      þa he wolde mancyn lysan.

    Bifode ic þa me se beorn ymbclypte;      ne dorste ic hwæðre bugan to eorðan,

    feallan to foldan sceatum.      Ac ic sceolde fæste standan.

    Rod wæs ic aræred.      Ahof ic ricne Cyning,

45    heofona Hlaford;      hyldan me ne dorste.

Then I saw the Lord of mankind hasten with much fortitude, for he meant to climb upon me. I did not dare then, against the word of the Lord, to give way there or to break when I saw the earth’s surfaces quake. All the enemies I could have felled; nonetheless I stood firm. The young man, who was almighty God, stripped himself, strong and unflinching. He climbed upon the despised gallows, courageous under the scrutiny of many, since he willed to redeem mankind. I quaked then, when the man embraced me; nonetheless I did not dare to collapse to the ground and fall to the surfaces of the earth, but I had to stand fast. I was reared up as a cross; I raised up the powerful King, Lord of the heavens. I did not dare to topple over.

In this passage, only slightly longer than Cædmon’s Hymn and perhaps based on an originally contemporaneous text, Christ is portrayed in terms strongly reminiscent of those used by the Whitby poet for God the Father: he is described as the frean mancynnes, “Lord of mankind,” 33b; dryhten, “Lord,” 35b; geong hæleð, “young man/hero,” 39a; ricne cyning, “powerful King,” 44b; and heofona Hlaford, “Lord of the heavens,” 45a. But while the terminology is similar, the force of the language is completely different: the power of the passage in the Dream of the Rood depends on our recognition of the double meanings inherent in the traditional poetic terms. Where Cædmon’s frea, dryctin, and uard have little more metaphoric force than the relatively bland equivalents for dominus used by the Old English Psalter poet (see Griffith 1991), the Dream of the Rood poet’s God is at once both dominus and miles, a new Christian type of hero whose triumph lies in his ability to conquer by embracing apparent defeat. As Swanton argues (Swanton 1970, 71):

Christ... [in the Dream of the Rood] is not led... [to the place of crucifixion] by a jeering mob; he is stripped by no mocking soldiers. Instead, as in Ravennate iconography or as seen by Ambrose or Cynewulf, he is a young and confident champion striding from afar, efstan elne mycle þæt he me wolde on gestigan (34). Vigorous and single-minded, he strips himself for battle and a kingly victory. The action is entirely his, an eager sacrifice; there is no question at this point of his being nailed to the cross. Instead he climbs to embrace it (40-2). It is pre-eminently an act of dominant free will by a prince confident of victory. With the agony transferred to the cross, Christ can sensibly be seen to rule from the gallows.

§ 3.30    The seeming lack of interest in the metaphorical force of the Cædmon’s vocabulary is echoed in Bede’s paraphrase and account of its reception in the Historia ecclesiastica (IV. 24 [ed. Colgrave and Mynors 1969]). At no point in his chapter does Bede imply that Cædmon’s verse was noteworthy for anything other than its metrical quality, organisation, and propagandistic success (see also Abraham 1992, 334). The chapter’s opening lines suggest that Bede saw Cædmon’s poetry as praiseworthy primarily because it was well constructed, scripturally based, never frivolous or impious, and very effective in leading men to turn their thoughts to heaven. His description of Hild and her counsellors likewise, portrays them as being struck more than anything by the poet’s musicality and ability to translate accurately into well-formed verse. When they give him their first commission, it is for a translation of scripture or religious doctrine into modulationem carminis, “metrical form.” After they accept him into the monastery and set him on his programme of translation, they appear to be most impressed by his poetry’s “sweetness”:

[S]usceptumque in monasterium cum omnibus suis fratrum cohorti adsociauit, iussitque illum seriem sacrae historiae doceri. At ipse cuncta, quae audiendo discere poterat, rememorando secum et quasi mundum animal ruminando, in carmen dulcissimum conuertebat, suauiusque resonando doctores suos uicissim auditores sui faciebat.

She and all her people received him into the community of the brothers and ordered that he should be instructed in the whole course of sacred history. He learned all he could by listening to them and then, memorizing it and ruminating over it, like some clean animal chewing the cud, he turned it into the most melodious verse: and it sounded so sweet as he recited it that his teachers became in turn his audience.

§ 3.31    Slightly more ambiguous evidence is provided by Bede’s paraphrase. As we have seen above (§§ 3.20-3.24), the epithets Cædmon uses in his original Old English are all relatively common in a religious sense in subsequent vernacular poetry. They are in the case of dryctin in particular, moreover, associated with stereotypical Latin translations. In Bede’s paraphrase, however, the majority of these “traditional” epithets are either omitted or translated using Latin words carrying no sense of the original connotations. Thus dryctin, “the standing term for dominus” in the Germanic languages (Green 1965, 298), is not translated in Bede’s paraphrase using the expected equivalent. Instead it is paraphrased using Deus, “God,” on its first occurrence in 4a and omitted altogether in the translation of 8a. The pre-Christian lordship term, frea, 9b, similarly, is omitted in Bede’s translation, which nevertheless retains the second word in the half-line, allmectig (translated by omnipotens)—an adjective which, Robinson demonstrates, was common in both pagan and Christian contexts (Robinson 1985, 34-38). Only in the case of uard does Bede translate Cædmon’s traditional Germanic terminology using a Latin word with equivalent connotations, albeit even then on only one of the noun’s two occurrences: in 7b he translates moncynnæs uard with Custos humani generis, a paraphrase that retains the connotations of the original vernacular text quite closely; in 1b, however, Bede’s translation of hefaenricaes uard by auctorem regni caelestis abandons the sense of guardianship implicit in Cædmon’s original formulation.

§ 3.32    There are several reasons why Bede might fail to translate Cædmon’s supposedly most novel epithets with either their stereotypical Latin equivalents or forms showing similarly striking connotations. One, implied by the traditional understanding of Cædmon’s pioneering role in the Christianisation of Germanic poetic diction, is simply that the epithets themselves were too new, too unusual, and perhaps even too daring at the time Bede set to work (see above, § 3.25). In this view Bede avoids translating dryctin or frea with a Latin term showing a similar sense of lordship either because the traditional equivalency between these words and Latin dominus had yet to be established or, because, like Bishop Wulfila who consistently avoids the cognate of dryctin (but not, interestingly, that of frea) in his Gothic translation of the Bible, he found the comparison too extreme for his intended readership (on Wulfila, see Green 1965, esp. 265-269).

§ 3.33    A second possibility, however, is simply that Bede did not consider Cædmon’s specific choice of epithets to be all that significant. In this view, Bede’s failure to use what later material suggests to be the usual Latin translation for Cædmon’s vocabulary comes not from any qualms about its novelty or suitability but rather Bede’s sense that this diction was not a crucial aspect of the vernacular poem’s success or structure.

§ 3.34    The first piece of evidence suggesting that this second possibility might be true lies in his treatment of uard in 1b and 7b. The fact that Bede uses a Latin equivalent with similar connotations to Cædmon’s original on one of the word’s two occurrences demonstrates that he did not consider this form to be too bold for his intended audience; the fact he does not use similar Latin terms both times, however, suggests that he did not see its repetition in Cædmon’s Old English to be particularly important to the poem’s form or meaning. In this light, the omission of an equivalent for frea and the translation of dryctin as Deus rather than Dominus might as easily be a sign that Bede did not find Cædmon’s specific diction to be very important as evidence that he found it too bold.

§ 3.35    The second piece of evidence suggesting that Bede was not overly concerned with Cædmon’s original vocabulary involves his treatment of poetic variation in the paraphrase. As I have argued elsewhere (O’Donnell 2004), Bede’s translation, while treating Cædmon’s supposedly novel diction in an inconsistent fashion, pays extremely close attention to his use of appositive variation: in paraphrasing the text, Bede removes or rephrases every single example of repetition among syntactically apposite elements in the poem.[15] This policy has its greatest effect in the last five lines, where Cædmon’s vernacular shows the most ornamental repetition (see § 3.18, above). The original text’s two, more-or-less synonymous verbs, scop and tiadæ, are collapsed into a single Latin form, creauit. The five syntactically apposite nominative pronouns and noun phrases, likewise, are reduced to two-and-a-half: in translating 5 to 9, Bede retains the first nominative reference in each clause (he [translated as qui], 5a; uard [translated as Custos], 7b) but eliminates their poetic variants, haleg sceppend, 6b, eci dryctin, 8a, and frea, 9b. In contrast, Bede’s translation of the first four lines of the poem renders Cædmon’s relatively spare Old English nearly word-for-word. The only exception comes in the equivalent to 3b through 4a, where, significantly, Bede recasts the vernacular text in order to eliminate the only example of ornamental syntactic apposition in the Old English poem’s opening lines.


§ 3.36    Bede’s account of Cædmon is by far the most detailed contemporary non-fiction account of poetic practice to survive the Anglo-Saxon period. As Bede was not an eyewitness to the events he was recording, however, its greatest value to modern scholars lies in what it tells us about how Cædmon was understood by his near-contemporaries. Taken together, the evidence of the first two chapters suggests that, for Bede at least, Cædmon was a superbly accomplished vernacular poet who learned to sing in an unusual way. The evidence of this chapter, however, suggests that he was not in practice particularly aesthetically innovative. The evidence of the rest of the poetic corpus and cognate languages suggests that Cædmon was drawing on traditional Germanic language and forms in composing his Hymn. But while the Hymn represents a secondary development of this (presumably originally pagan) tradition, there is no evidence to suggest that Cædmon was himself responsible for the initial adaptation of pagan terminology to Christian ends. Cædmon’s verse shows very little self-consciousness about its supposedly novel formulations for God; Bede does not devote any special attention to Cædmon’s use of formulae or poetic diction in either his translation or his account of the poem’s initial reception. What Bede’s account does suggest, however, is that Cædmon was a good technician. Bede’s paraphrase of the original vernacular pays close if negative attention to the rendition of Cædmon’s use of ornamental apposition; his description of the poem’s initial composition and reception suggests that Cædmon was valued by his contemporaries for his skill as a versifier.


[15]Cædmon’s significance to the development of (Christian) Old English vernacular poetry is a critical commonplace; the wider importance of his “school” among the Germanic speaking peoples is discussed in Green 1965, 286-300. For the suggestion that the poet’s primary debt is to non-Germanic poetic or compositional traditions see, among others, N.F. Blake 1962, esp. 245; Howlett 1974, 10.

[3]Isaac 1997, 224, incorrectly scans 9b as Type A-1 (Sx/Sx in the notation of Russom 1987). Since “eall- is stressed” in compounds (Campbell 1991, § 86), the line is Type D-1 (in Pope’s restatement of Sievers’s notation), or S/Ssx (in Russom’s notation). See the scansion in Pope and Fulk 2001, 141; also Creed 1992, 37.

[4]Hutcheson’s statistics are not precisely comparable, since his tables admit a sixth verse type (“Type 3”). The small number of “Type 3” verses and the large statistical uncertainty attached to Cædmon’s Hymn, however, ensure that this anomaly does not invalidate the comparison. I am grateful to Dr. Jonathan Seldin of the Department of Mathematics and Computer Science at the University of Lethbridge for his help in interpreting the statistical significance of verse distribution and combinations in the Hymn.

[5]A similar suggestion is implicit in the discussion of “archetypal verses” in Creed 1992 (see particularly the scansion of 1a on 41, fig. 1).

[6]Robinson notes Stanley’s qualification, but argues that the participation of in transverse alliteration in Type A-3 lines in fact indicates that “this is... one of those areas of Old English prosody where stress can fall where it ordinarily would not fall when there is a special, contextual reason for stress” (116). See Momma 1997, esp. 142, for a discussion of the placement of unstressed in Old English poetry.

[7]Howlett’s citation also uses underlining to indicate alliterative syllables (not reproduced here).

[8]This reading assumes sue in 3b is a conjunction; for an alternative, asymmetrical, reading of the poem, in which sue is an adverb and 3b-4b associated with 5a-9b instead of 1a-3a as here, see Howlett 1974, 6-7; Howlett’s reading is discussed below, § C.9.

[9]In the following lines, eordu may be construed as accusative singular (in which case it is parallel to middungeard and foldu, both of which can also be understood as accusative singular) or genitive singular (in which case it may be parallel to foldu, which might be construed as a genitive). See § 5.16 and note.

[10]This list is based on the critical text of the poem in this edition and, with the exception of uuldurfadur, the vocabulary printed in Griffith 1991, 183-185. Griffith’s method of collection excludes numerous compounds that, while only found in poetic contexts, contain simplices that are also found in prose (see 168). uuldurfadur is identified as poetic in Clark Hall 1984 (“wuldorfæder”). The Northumbrian aelda and West-Saxon ylda recensions contain at least one other poetic word, aelda, “of men” (5b).

Wrenn 1947 claims a total of “at least nine words which must have then belonged to the aristocratic heroic-poetic tradition—words of a diction which would not be in use among the peasants” (287): five of the six listed here (he doesn’t mention or) and eordu, tiadæ, modgidanc, and middungeard. Neither Griffith nor Clark Hall identify any of these last four words as belonging to the poetic vocabulary, and, indeed, all four are common in prose.

[11]All figures drawn from Fry 1975, 50-61. Although Fry claims to have found parallels for “every line” in the Hymn in the form either of verbatim repetitions or membership in an attested formulaic system (50), some of his “systems” seem more strained than others. In some cases, indeed, they seem little more than an observation that the same words are found elsewhere in Old English poetry. It is hard to see, for example, how the following examples, which Fry claims suggest the existence of a formulaic “system” “(X') teo'n,” can be anything other than evidence that forms of tēon are occasionally found as the second stressed element in Old English half-lines: on flet teon, eftsiðas teah, ond næs togen, swa hine oxa ne teah, ond to ham tyhð, and hwilum ut tyhð (Fry 1975, 57). See also Fry 1979 and, for objections to Fry’s method, Miletich 1983.

[12]“Attested for the first time” in this context means “earliest appearance in the surviving written record.” Other Old English texts showing these terms may be as old as or older than Cædmon’s Hymn, although they are found in later (usually much later) witnesses. Thus Green cites Widsith (known from the late tenth-century Exeter Book) for frea (Green 1965, 503). frea and dryctin are also found in the late tenth-century Vercelli Book text of the Dream of the Rood (but not in the related Ruthwell Cross Inscription, a perhaps eighth-century runic text).

[13]Careful statements of the consensus position can be found in Robinson 1985, 29-59, esp. 30 and n. 9; Stanley 1995, 139-140. Opland 1980, 95-97, argues that Anglo-Saxon traditional poetic diction may in fact date only from the seventh and eighth centuries.

[14]This lack of attention to Cædmon’s role in the supposed conversion of Germanic tradition is all the more surprising given Bede’s emphasis elsewhere in the Historia ecclesiastica on the process by which pagan monuments, rituals, and feasts were adapted to Christian use or overthrown. See, for example, his account of Gregory’s letter to Mellitus (I. 30), and the account of the conversion of King Edwin in which the chief priest Coifi is shown destroying pagan temples (II. 13). In Bede’s introduction to Gregory’s letter, he suggests that the advice to take advantage of pagan monuments and customs demonstrated the Pope’s eagerness erga saluationem nostrae gentis, “for the salvation of our race.”

[15]In addition to the evidence from Anglo-Saxon England cited here and below, it is useful to compare Cædmon’s Hymn to the admittedly much longer 1970’s musical Jesus Christ Superstar (Lloyd Webber and Rice 1993), a work in which similarly striking (and “vernacular”) diction is used to refer to God. In contrast to Cædmon’s Hymn, where traditional Germanic terminology is used in what can at best be described as an ornamental fashion, however, the “contemporary” language of Jesus Christ Superstar is central to the musical’s conceit.

[16]Cavill 2002 argues that Bede also removes “structural alliteration” from the poem (11). As Cavill notes, however, Bede is less consequential in his treatment of this aspect of Old English poetic practice: while Bede’s paraphrase does not use alliteration as a structural principle, it does reproduce some alliteration for what Cavill describes as “decorative purposes” (12).