The Basic Metre of Beowulf

Zlatomir Fung

The history of alliterative English poetry begins with the oldest known poems written in Old English. The metrical systems that govern the verse of Old English poetry are part of a larger Germanic alliterative tradition that includes Old Norse and Old High German texts. The earliest recorded Germanic verse, written in a form of early Norse, exemplified the basic tenets of a rich tradition which lasted over 1000 years.[1] Richard Wagner believed that early Germanic texts (specifically the Poetic Edda) exhibited a “primal” human expression, and he sought to emulate it in the libretto of his mythological opera cycle Der Ring des Nibelungen[2]; yet, through careful analysis of the metrical elements of alliterative poetry, it is clear that the Germanic alliterative verse is crafted upon a deeply sophisticated metre. This paper aims to provide a basic summary of Old English metrical principles as they appear in Beowulf, and briefly analyze how knowledge of the metre influences translation of the poem from Old English into Modern English.

Beowulf holds primacy in the canon of Old English verse. Of some 30,000 extant lines of Old English poetry,[3] the 3,000-odd lines of Beowulf comprise the longest single work of Old English literature and about 1/10th of the corpus. The verses of Beowulf contain a remarkably large vocabulary, including words which only appear in Beowulf and in no other extant works in Old English. It is considered to be a poem of great thematic interest, in part owing to J.R.R. Tolkien’s important 1936 lecture-essay Beowulf: The Monsters and the Critics, which favorably altered the course of critical and cultural opinion. The Beowulf poet’s evocative use of kennings and the imaginative employment of metre are other reasons why Beowulf has been considered the apotheosis of early English literature.[4]

Nonetheless, very little about Beowulf aside from its artistic and historical importance is a matter of agreement among scholars. The manuscript of Beowulf, housed in the British Library, exists without a title or indication of authorship and is thought to be a copy of an earlier version of the poem, written down around the turn of the 11th century.[5] The date of composition of the poem is heavily debated; various arguments employing a number of metrical, linguistic, historical, and stylistic interpretations have yet to achieve consensus. Historians and metrists do agree[6],[7],[8] that there is a degree of scribal error in the extant Beowulf manuscript but do not agree on the nature or extent of the errors. One of the most controversial aspects of Beowulf is its metrical organization, which has been analyzed and debated extensively for over 100 years.

Basic Metrical Principles of Old English Verse

Each line in Old English poetry can be divided into two parts, called half-lines or verses; between these verses there exists a strong metrical caesura, which is made clear by the relative regularity of the stresses in Old English and the patterns of alliteration. The first half-line is called the a-verse, and the second the b-verse:

(a-verse)                 (b-verse)

Oft Scyld Scefing    sceaþena þreatum                            (Beowulf[9], 4)

The extant manuscripts of some Old English poems, including Beowulf, do not mark the positions of the lines or verses, but other manuscripts have small scribal marks indicating the caesurae or line breaks, which has greatly assisted metrical analysis.

One of the first important efforts at metrical scansion of Old English poetry came from George Hickes, an early 18th-century metrist, who, through his study of Old English religious poetry, postulated that Old English metre was quantitative, like Latin and Ancient Greek metre.[10]The accentual-syllabic metres of modern English poetry, like iambic pentameter, rely, in part, on a regular syllable count per verse, which Hickes did not find in the Old English poems he examined. His “quantitative” hypothesis was that Old English differentiated words with the same spelling based on vowel length (e.g. God [modern English “God”] vs gōd [modern English “good”]), which meant that longer vowels could claim the same metrical duration as multiple smaller syllables. His first supposition was correct, but not the second; Old English does distinguish phonologically on the basis of vowel length, but not as the primary element of the metre.

Eduard Sievers was the first metrist in the modern tradition whose observations about Old English metre continue to exert a lasting influence on the field. His pioneering research was conducted in the late 19th century, and classified five basic types of half-lines, with some variations to each line. Variations were based on the presence of four basic metrical positions per verse, and the different types were based on the positioning of three levels of lexical stress:

 A         Sx|Sx:      S     x     S   x

                            gomban  gyldan,  “to  pay  tribute” (11a)

 B         xS|xS:     x    S        x  S

   hīe  wyrd  forswēop,  “events  swept  them” (477b)

 C        xS|Sx:     x    S        S    x    

               oft  Scyld  Scēfing, “Scyld, son of Scef, often” (4a)

 D        1) S|Ssx:  S     S   s    x

    wīs  wēlþungen,  “wise,  accomplished”  (1927a)

            2) S|Sxs  S       S     x  s      

    Fyrst forð gewat, “time passed” (210a)

 E         Ssx|S:      S    s     x      S

    glēomannes  gyd,  “musician’s  song”  (1160a)[11]

In the examples above, an “S” indicates a strong (primary) stress (otherwise known as a “lift”), and an “x” a weak (tertiary) stress (otherwise known as a “dip”). An “s” indicates a half (secondary) stress, which only occurs in the D and E types. The D type has two permutations, depending on the placement of the syllable with secondary stress. In Old English, as in Modern English, the strength of a word’s stress depended mostly on the grammatical category of the word: adjectives, nouns, infinitive verbs, and other central words were either always or often strongly stressed, whereas conjunctions, prepositions, prefixes, and other smaller linguistic units received less stress. Some linguistic categories, like pronouns, prepositions and articles, were promotable to higher levels of stress depending on the needs of the verse.[12]

The symbol “|” indicates the half-point in the organization of the verse, creating two subgroups known as “feet”, which typically consist of two syllables or their metrical equivalent. Each foot of verse possesses its own hierarchy of stress, and is not tied to the other foot’s metrical hierarchy. Sievers argued that each foot contained one strong stress, with two strong stresses total per verse. This tradition of metrical analysis has been dubbed the “accentual paradigm” by Ian Cornelius.[13] It is an approach to the metre based on the perceived organization of lexical stress. The Five Types were classified according to their frequency in Beowulf, beginning with Type A.

The essential decorative element is alliteration, the consonance of sounds at the beginning of words or morphemes. The presence of alliteration highlights the rhythmic stress (and therefore the metre) in a line, but it is merely an ornamental element of the metre, not a functional one. Consider the following complete lines from the opening of Beowulf, and how the poet connects the two verses with alliteration in various ways:

monegum mægþum,     meodosetla ofteah              (Beowulf, 5)

egsode eorlas.               Syððan ærest wearð          (Beowulf, 6)

Him ða Scyld gewat      to gescæphwile                   (Beowulf, 26)

Oft Scyld Scefing           sceaþena þreatum              (Beowulf, 4)

geong in geardum,        þone god sende                  (Beowulf, 13)

The bold sounds both alliterate and initiate the lifts. In Beo, 5, there is a simple alliteration off the consonant “m.” In Beo, 6, the alliterating sounds are vowels; all vowels in Old English may alliterate with all other vowels.[14] In Beo, 26, the alliterating consonant in the b-verse does not begin the word, but is preceded by a prefix which has fluid metrical significance; this type of alliteration is common and permitted.

In Beo, 4, the sound “Sc” alliterates three times. Consonants like “sc” and “st” could not alliterate together, even though they were written with the same starting letter. (Note that in the same line, “þ” occurs twice in the b-verse. This does not have any rhythmic or metrical significance because neither syllable containing that sound is strongly stressed.) In Beo, 13, the initial sounds of  “geong,” ”geardum,” and “god” alliterate, although by the time of the Beowulf manuscript, the “g” had diverged into both palatalized and velar sounds depending on the vowel that succeeded it (a velar g in the case of “geong” and “geardum,” and a palatalized g in the case of “god”).[15] There are only two constraints on the placement of alliteration within a line: 1) In the b-verse, only the first stressed syllable may alliterate--in other words, double alliteration (i.e. alliteration on both stressed syllables) may not occur in the b-verse, and 2) in the a-verse, the first and second stress may alliterate together or independently with the first stress of the b-verse. Occasionally, in lines such as

S  |  S  s   x   (D1)                     x    S    | x      S       (B)

swiðhicgende                           on  sele  þam hean             (Beo, 1016)[16]

(The bold sounds alliterate with one another and the italicised sounds alliterate with one another.)

there is a transverse alliteration. In this verse, there are two separate alliterative units, in the pattern abab (the pattern abba also occurs, though more infrequently). This technique is seldom employed, and typically only at moments of intense lyricism or melody in the poetry[17]: this particular line occurs in a passage describing the celebratory feast following Beowulf’s dismembering of Grendel.

Despite the importance and utility of Sievers’ classification, the majority of the verses of Beowulf cannot be simply categorized as one of Sievers’ Five Types. Only about 10% of Beowulf contains eight syllables per line[18] (in other words, a one-to-one correspondence of the syllables to the proposed metrical stresses), but a majority of the remainder can be explained with the theoretical addition of three metrical phenomena: 1) expansion of unstressed syllables, 2) resolution of an unstressed syllable, and 3) anacrusis. The first two concepts were described and explained by Sievers, but the third, anacrusis, is a controversial addition to the existing framework, and one of the reasons for the recent surge in criticism of Sievers’ model.

The expansion of unstressed syllables may occur in the A, B, and C types. If we are to again return to the abstraction of the types, we get the following addition to the form:

Type A: Sx(x…)|Sx

Type B: x(x…)S|xS

Type C: x(x…)S|Sx

In the following examples, the extra unstressed syllable(s) may count simply as an extension of the unstressed region, such that the unstressed syllables merge together in the metrical count:

S        x  x  |  S    x      (Type A)        x      x      S  |  S  x    (Type C)        x   x     S | x    S  (Type B)

Weox under wolcnum (Beo, 8a)      þæt wæs god cyning (11b)             Syððan ærest wearð (6b)

These dips with multiple unstressed syllables may be called “strong dips,” and occur only in the first dip of the verse (i.e. a SxSx(x…) is not a possible variation on the A type). This allowed the Beowulf poet much more metrical flexibility: there can be up to four unstressed syllables in the strong dip in Type A, and up to five in Types B and C.[19]

Resolution is the process by which a short stressed syllable followed by a short unstressed syllable is metrically counted as one long, stressed syllable. Important work in historical linguistics by Max Kaluza in the late 19th century led to an understanding of the phonological distinction between long and short vowels in Old English,[20] which became a key feature of its poetry. Consider the following line:

            S--x    S   s     x

            Wudu bundenne  (216b)

The syllable “du” is counted as a resolved syllable of “wu,” allowing this line to be metrically analyzed as a simple D1 type, whereby the first two syllables amount to one single, long-vowel stressed syllable. For resolution to occur, the first stressed syllable must be followed by only one consonant,[21] and then a short, unstressed vowel. Most importantly, however, the initial syllable must contain a short vowel.

Anacrusis is the insertion of an unstressed syllable before a verse type that would otherwise begin with strong stress (a lift). The presence of anacrusis has been a challenge to the logic of Sievers’ system. Consider that in the B-type verses (xSxS), an initial unstressed syllable is counted as an integral part of the verse’s character; without it, the verse would lack the necessary four metrical positions. However, in Type A with an anacrusis (xSxSx), this initial unstressed syllable is uncounted. An influential interpretation of anacrusis was proposed by A.J. Bliss, who argued that the anacrusis occurring in the A and D types was a natural poetic technique to offset the relative metrical weight of the second foot of the verse, which often had stronger stress than the first foot.[22] (For instance, in type D2, S|Sxs, the second foot has more metrical weight with three positions, so the anacrusis balances out the weight of the two feet. This also explains why anacrusis never occurs in Type E, Ssx|S, in which the first foot has a greater metrical weight than the second foot.) Later research by Donoghue and Cable discovered that “...that the syllables in anacrusis tended to be unstressed prefixes or the negative proclitic ne and, secondly, that these morphemes exhibit similar behavior interior to the verse.”[23] Nikolay Yakovlev deemed this finding the “prefix license,” but still argued against the logic of this license in the context of an accentual paradigm.[24]

One of the most fundamental ideas of Sievers’ accentual paradigm is that the stresses of each of the five types contain an inherent natural logic of pronunciation; in other words, an alternation of strongly and less strongly stressed syllables. The biggest challenge to this paradigm is the presence of clashing stress in the C and D types, in which two seemingly equal syllables are uttered in direct succession. This led Thomas Cable, in the early 1970s, to propose a refinement to Seivers’ system involving C and D type lines, whereby the first strong stress in the verse would be even stronger than the second strong stress[25]:

                                 (stronger) (weaker)

 C-type   xS|Sx:     oft Scyld     Scēfing      (italics indicate strong stress) (Beowulf, 4a)

This idea attempts a solution for the problem by introducing yet another level of rhythmic stress which has little linguistic justification.[26] Moreover, various types of verses that appear in Beowulf defy simple categorization, like type A3 (waes min faeder, 262a, xxSx), which possesses only one strong stress, and type D*, (side saenaessas, 223a, SxSsx), which seems to contain an extra unstressed syllable which cannot be explained by resolution because the first syllable has a long vowel.

The Sieversian tradition of classification of the lines of Beowulf peaked with A.J. Bliss’ influential work The Metre of “Beowulf,” in which he classified every line in Beowulf as one of 130 different metrical subtypes, all falling under the larger umbrella of Sievers’ five types.[27] Modern theorists began to notice a fundamental flaw in the Sieversian tradition: by so extensively cataloguing the various contours of the verse, metrists had lost sight of a metre as being able to fundamentally represent a deeper layer of linguistic truth. Additionally, metrists were increasingly seeking basic metrical principles which could explain the evolution of the alliterative tradition from Old English poetry into Middle English poetry.[28] The most important reformist was Thomas Cable, who criticized the cataloguing of the Sieversian types and their subtypes. He hoped establish a metrical theory that would function as a basic canvas upon which all the existing verses of Beowulf could be constructed.[29] He proposed successful simplifications and additional rules (for instance, the “antepenultimate syllable rule for resolution”) which manage to explain around 99% of Beowulf’s verses.[30]

R.D. Fulk, in his important 1992 work A History of Old English Metre, had hinted at the fact that the accentual paradigm may have been given undue importance to rhythmic stress in the analysis of Old English metre. In 2008, a brilliant unpublished dissertation by Nikolay Yakovlev fully fleshed out the idea and shifted thinking about old English metre. Yakovlev proposed a metre of Old English based on morphology rather than the stress as the organizing principle of the metre.[31] As Cornelius notes, “Stress accent establishes the place of alliteration within the line, but it does not follow that the meter itself is stress-based, for alliteration may be only a regularized highlighting of stress peaks that are themselves the output of metrical organization located in a deeper linguistic layer.”[32]

Drawing heavily on the Sieversian tradition, Yakovlev preserves the idea of four basic metrical positions, the presence of a lift and a dip (strong and weak metrical stress), and expansion of a dip, and resolution. However, Yakovlev classifies the metrical position of a syllable or word according to its morphological function.[33] Strong words, like “roots, suffixes, and stressed prefixes of open-class words”[34] will always receive metrical stress. Therefore, a verse like

lēof  lēodcyning        (Beo, 54a)

is annotated as SSSS, four strong metrical (not lexical) stresses, and one of eight possible permutations.[35] Yakovlev’s theory explains why Sievers’ Type A is the most common verse type, as it may occur from four different morphological permutations (SxSx, SSSx, SxSS, and SSSS), more than any other type.[36] Yakovlev’s theory also explains why the presence of anacrusis (the “prefix license”) in types A and D is allowed: prefixes and the negative “ne” would naturally be given little to no morphological (and therefore metrical) weight. Much research on the morphological paradigm remains to be done, but Yakovlev’s ideas are a promising expansion upon the traditional model.[37]

Modern Translations and Old English Metre

In 1940, J.R.R. Tolkien published an important essay titled “On Translating Beowulf,” in which he discussed some of the key considerations of the translator approaching Beowulf. In it, Tolkien offers an excerpt of a possible alliterative translation of the famous passage beginning in line 210. (Tolkien ultimately abandoned a complete alliterative translation of the poem, instead opting for a rhythmic prose translation, published and edited by his son Christopher.) Seamus Heaney’s 1989 translation of Beowulf into Modern English is an important and high-quality translation of a very different character. It succeeds in part because of its adherence to Beowulf’s original metre in a degree that is neither irreverent nor overly faithful; it strikes an exceptionally natural medium which lends the translation a great natural rhythm when recited. In defending his choice not to alliterate in the strictest sense, Heaney wrote “...I prefer to let the natural ‘sound of sense’ prevail over the demands of the convention…”[38]

Compare the two translations of the following excerpt from Beowulf, lines 210-217 (proper alliteration in bold, approximate alliteration italicized):

Regarding the original verse, two metrical elements are relevant:

1. Metrically, these lines are conservative. None are hypermetrical (a rare verse type containing more than two stresses), contain anacrusis, or employ expansion of a dip lasting more than two syllables. This “tightness” in the poetic construction is contrasted greatly by the evocative vocabulary and concision with which the Beowulf poet describes the seafaring journey of the Geats. The abundance of the Type-A (SxSx) gives the passage a strong, regular rhythm. The result is a dense, sensory, and heightened form of poetic expression.

2. Six of the seven lines contain double alliteration in the a-verse, which is highly ornamental.

The challenge of the translator is whether these formal elements of the poetry are as essential as the tone or emotional sweep of the verse. Tolkien’s translation is eminently faithful to the metrical style of the original, so much so that Tolkien was able to assign Sievers’ line types to his own translation.[41] He is able to successfully alliterate according to the rules of alliteration, and even preserve a great deal of the a-verse double alliteration in the original. Heaney’s translation, while retaining the metrical principle of four strong stresses per line and the general verse structure, only properly alliterates on five out of seven lines, never uses double alliteration, and even corrupts the verse arrangement between lines 211 and 212.

On the other hand, Tolkien’s translation is cryptic and lacks colloquial syntax. Heaney, approaching the original as if it were eminently colloquial in its own time, beautifully interprets the lyricism and grandeur of the original with its simplicity and imagery, but his translation is admittedly looser. If we are to adopt the spirit of Yakovlev’s morphological metre, then the evolution of such a metre from Old English to Modern English would likely not retain the same verse types that Sievers identified. In this sense, Heaney captures the spirit of the metre, since his verse retains approximately the morphological density of the original. A qualitative judgment about these two different translations is irrelevant--the metrical analysis above only serves to underline that metre is a fundamental consideration for translation. The rich alliterative nature of Beowulf’s verse and the new horizons of metrical analysis ensure that more comparative research is left to be done in this vein, specifically the application of Yakovlev’s morphological theory to the translation of Beowulf.

[1] Geoffrey Russom, Beowulf and Old Germanic Metre (Cambridge: Cambridge University Press, 2008), Introduction.

[2] Spencer, Stewart. 2001 "Stabreim." Grove Music Online. 4 May. 2019.

[3] Scragg, Donald G. “The nature of Old English verse” in The Cambridge Companion to Old English Literature (Cambridge: Cambridge University Press, 2008), 58.

[4] Eric Weiskott, English Alliterative Verse (Cambridge: Cambridge University Press, 2016), 23.

[5] Leonard Neidorf. The Transmission of Beowulf: Language, Culture, and Scribal Behavior. (Cornell University Press, 2017), 21

[6] Goering, Nelson. "Metrics, Scribes, and Beowulf: A Response to Neidorf (2017), The Transmission of Beowulf." Neophilologus 103, no. 1 (2019): 115-127.

[7] Terasawa, Jun. Old English Metre: An Introduction. (University of Toronto Press, 2011)

[8] J.R.R. Tolkien, “On Translating Beowulf” in The Monsters and the Critics and Other Essays (HarperCollins, 1997).

[9]All examples in the original Old English will henceforth be taken from Beowulf, but the basic metrical principles and problems generally apply to other Old English poetry of the same period.

[10] Ian Cornelius. "The Accentual Paradigm in Early English Metrics." The Journal of English and Germanic Philology 114, no. 4 (2015), 463.

[11] Examples taken from Cornelius, Ian. "The Accentual Paradigm in Early English Metrics."

[12] Thomas Cable, The English Alliterative Tradition (University of Pennsylvania Press, 1991), 22

[13] Cornelius, 462

[14] Jun Terasawa. Old English Metre: An Introduction. (University of Toronto Press, 2011), 2.1 Consonantal and Vocalic Alliteration.

[15] Jun Terasawa. 2.1: “Consonantal and Vocalic Alliteration”

[16] Example from Scragg, Donald G. “The nature of Old English verse” in The Cambridge Companion to Old English Literature (Cambridge: Cambridge University Press, 2008), 62

[17] Scragg, 63

[18] Chris Golston, and Thomas Riad. Scansion and alliteration in Beowulf. 2003. Web, http://zimmer.csufresno.edu/~chrisg/index_files/Scansion.pdf

[19] Thomas Cable, The English Alliterative Tradition (University of Pennsylvania Press, 1991), 12

[20] Seiichi Suzuki. "Preference Conditions for Resolution in the Meter of "Beowulf": Kaluza's Law Reconsidered." Modern Philology 93, no. 3 (1996): 281-283

[21] Paraphrased from Alaric Hall, and Sheryl McDonald. "A beginner’s guide (hopefully) to Old English metre (version 1.5, September 26th 2016), 4-5.

[22] Rudolph Willard, and Elinor D. Clemons. "Bliss's Light Verses in the "Beowulf"." The Journal of English and Germanic Philology 66, no. 2 (1967): 230-44.

[23] Ian Cornelius, 475

[24] Nikolay Yakovlev. "The development of alliterative metre from Old to Middle English." PhD diss., (University of Oxford, 2008), 80: “...it is very hard to understand how a syllable may be completely omitted from the metrical count: the prominence of the syllable may be low, but it will still be part of the intonational contour of the verse.”

[25] Thomas Cable, The English Alliterative Tradition, 26-37

[26] Nikolay Yakovlev, 72

[27] Ian Cornelius, 460

[28] See Yakovlev, “The Development....” and Cable, The English Alliterative Tradition

[29] Thomas Cable, 6-9

[30] Nikolay Yakovlev, 31

[31] Nikolay Yakovlev, 74

[32] Quoted from Ian Cornelius, 59

[33] See Yakovlev, 74-78, for a complete definition of his morphological paradigm, which is outside the scope of this paper.

[34] See Nikolay Yakovlev, 75, for a more detailed description.

[35] The eight permutations of metrical stress are SxSx, xSxS, xSSx, SSSx, SSxS, SxSS, xSSS, and SSSS.

[36] Eric Weiskott, English Alliterative Verse, 24

[37] Yakovlev’s ideas are still fresh and are only recently gaining adoption: see Cornelius, 2015 and Weiskott, 2016 (in the bibliography) for contemporary applications.

[38] Seamus Heaney, Beowulf: An Illustrated Edition (New York: W.W. Norton and Company, 2008), Introduction, xxiii

[39] Seamus Heaney, 17

[40] J.R.R. Tolkien, “On Translating Beowulf” from The Monsters and the Critics and Other Essays, 72

[41] J.R.R. Tolkien, “On Translating Beowulf” from The Monsters and the Critics and Other Essays, 72

Works Cited 

Cable, Thomas. "Clashing Stress in the Metre of “Beowulf”." Neuphilologische Mitteilungen 72, no. 1 (1971): 42-50.

Cable, Thomas. "Metrical Simplicity and Sievers' Five Types." Studies in Philology 69, no. 3 (1972): 280-88.

Thomas Cable, The English Alliterative Tradition (University of Pennsylvania Press, 1991)

Cornelius, Ian. "The Accentual Paradigm in Early English Metrics." The Journal of English and Germanic Philology 114, no. 4 (2015)

Godden, Malcolm, and Michael Lapidge. The Cambridge Companion to Old English Literature. (Cambridge: Cambridge University Press, 1991).

Goering, Nelson. "Metrics, Scribes, and Beowulf: A Response to Neidorf (2017), The Transmission of Beowulf." Neophilologus 103, no. 1 (2019): 115-127

Golston, Chris and Thomas Riad. Scansion and alliteration in Beowulf. 2003. Web, http://zimmer.csufresno.edu/~chrisg/index_files/Scansion.pdf

Hall, Alaric and Sheryl McDonald. "A beginner’s guide (hopefully) to Old English metre (version 1.5, September 26th 2016)

Heaney, Seamus. Beowulf: An Illustrated Edition (New York: W.W. Norton and Company, 2008)

Neidorf, Leonard. The Transmission of Beowulf: Language, Culture, and Scribal Behavior. (Ithaca: Cornell University Press, 2017)

Russom, Geoffrey, Beowulf and Old Germanic Metre (Cambridge: Cambridge University Press, 2008), Introduction.

Spencer, Stewart. 2001 "Stabreim." Grove Music Online. 4 May. 2019

Suzuki, Seiichi. "Preference Conditions for Resolution in the Meter of "Beowulf": Kaluza's Law Reconsidered." Modern Philology 93, no. 3 (1996): 281-283

Terasawa, Jun. Old English Metre: An Introduction. (Toronto: University of Toronto Press, 2011).

Tolkien, J.R.R. “On Translating Beowulf” in The Monsters and the Critics and Other Essays (New York: HarperCollins, 1997).

Weiskott, Eric. English Alliterative Verse (Cambridge: Cambridge University Press, 2016)

Willard, Rudolph, and Elinor D. Clemons. "Bliss's Light Verses in the "Beowulf"." The Journal of English and Germanic Philology 66, no. 2 (1967): 230-44.

Yakovlev, Nicolay. "The development of alliterative metre from Old to Middle English." PhD diss., University of Oxford, 2008.

Copyright © 2020 Zlatomir Fung