Snorri’s Edda, Mythology, and Anglo-Saxon Studies

TIM WILLIAM MACHAN

University of Notre Dame

One of the most influential Anglo-Saxonists of the nineteenth, twentieth,

and twenty-first centuries was unable to speak English and apparently never
set foot in England. So far as we know, he never read Beowulf, never held an
Anglo-Saxon manuscript in his hands, and could not have explained Sec-
ond Fronting. In fact, he was born in twelfth-century Iceland, where, with
the exception of two trips to Norway, he seems to have spent his entire life.1
I mean, of course, Snorri Sturluson, author of the Prose Edda. In its four
parts—a prologue, Gylfaginning (The deluding of Gylfi), Skáldskaparmál
(The language of poetry), and Háttatal (A listing of meters)—the Edda
offers an expansive, coherent vision of Norse poetics and mythology. And
in describing Snorri as an influential Anglo-Saxonist, I mean the influence
of the Edda in particular, which has served, indeed, as an authority on Old
English poetics and cultural history as well as on the continental Germanic
traditions that are presumed to underlie Anglo-Saxon practices. The myths
and mythic culture found in Gylfaginning have been used to explain every-
thing from English place names to folk traditions, while the terminology of
Skáldskaparmál forms the basic vocabulary for discussions of poetic practice
in Old English. In its entirety, Snorra Edda has been both a reference work
for Anglo-Saxon studies and the image of a kind of cultural coherence that
not only shaped early England but linked it to early Iceland. These are all
significant accomplishments for a work composed by someone who talks
about England only in the narratives of Heimskringla and Egilssaga (if he is
in fact the latter’s author). And they are reasons enough to ask how and
why Snorri’s Edda has become so influential in Anglo-Saxon studies and
then whether it merits the status it has.

I would like to thank the two anonymous readers for Modern Philology, whose comments
were extremely valuable to me in rethinking and revising this essay.
1. The only contemporary biographical details of Snorri’s life occur in Sturla Þórðarson’s
Íslendinga Saga, within Sturlunga Saga.

Ó 2016 by The University of Chicago. All rights reserved. 0026-8232/2016/11303-0001$10.00

295

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 296 MODERN PHILOLOGY

While the first complete English rendering of the Edda did not appear
until 1987, interest in Snorri began much earlier. Indeed, the work’s first
partial translation, limited to the prologue, appeared already in Aylett Sam-
mes’s 1676 Britannia Antiqua Illustrata. The translation that effectively intro-
duced an English-speaking audience to medieval Scandinavians in general,
however, was not published until nearly a century later in Bishop Thomas
Percy’s Northern Antiquities (1770, but reissued several times). This actually
contained not an original English translation of the Edda but a translation
of Paul Henri Mallet’s French rendering (limited to Gylfaginning and synop-
ses of Skáldskaparmál and Háttatal ) along with the Latin translation of
Johannes Göransson. Yet as far removed as the Northern Antiquities materials
were from the original texts, the volumes drew emphatic connections
between medieval Scandinavians and their English counterparts. “As to the
Anglo-Saxon, and Icelandic poetry,” Percy says, “these will be allowed to be
in all respects congenial, because of the great affinity between the two lan-
guages, and between the nations who spoke them. They were both Gothic
Tribes, and used two not very different dialects of the same Gothic lan-
guage. Accordingly we find a very strong resemblance in their versification,
phraseology and poetic allusions, &c. the same being in a great measure
common to both nations.”2 It was Percy’s work that inspired Thomas Gray
and other early Scandinavian enthusiasts, with the result that, from the late
eighteenth century on, talking about Norse poetry often amounted to talk-
ing about Old English poetry as well, with much of the primary evidence for
both coming from Snorri’s Edda.
Subsequent to Percy’s Northern Antiquities, partial translations of the Prose
Edda appeared with increasing frequency. Grenville Pigott translated ex-
tracts from Gylfyaginning in his 1839 Manual of Scandinavian Mythology, and
in 1842 George Dasent provided a translation of Gylfyaginning in its entirety
(more or less), along with a small section of Skáldskaparmál. Five years later,
a new translation of Gylfyaginning appeared in a reissue of Northern Antiqui-
ties, and again in 1879 in Rasmus Anderson’s The Younger Edda, Also Called
Snorre’s Edda or the Prose Edda. Arthur Brodeur’s English translation of both
Gylfyaginning and Skáldskaparmál was published in 1916, and Jean Young’s
rendering of the former alone in 1964.3

2. Thomas Percy, trans., Northern Antiquities; or, A Description of the Manners, Customs, Religion
and Laws of the Ancient Danes, a translation of Paul Henri Mallet, Introduction à l’Histoire de Dane-
mark (1755) and Monuments de la mythologie et de la poésie des Celtes (1756), 2 vols. (1770; repr.,
New York: Garland, 1979), 2:194–95. Percy actually used the revised 1763 edition for the Intro-
duction à l’Histoire de Danemark.
3. See further Andrew Wawn, “Early Literature of the North,” in The Oxford History of Literary
Translation in English, vol. 4, 1790–1900, ed. Peter France and Kenneth Haynes (Oxford Uni-
versity Press, 2006), 274–85.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 297

A bibliographical sketch like this tells only part of the story of how and
why the Prose Edda has influenced Anglo-Saxon studies. A second part is
simply the existence of Snorri and his Edda. We have no Anglo-Saxon man-
ual of poetics and myth; we likewise have few myths from other Germanic
traditions, and certainly no mythological handbooks in Old Saxon, Old Fri-
sian, Old High German, or Gothic. For the study not only of Anglo-Saxon
mythology but also of its putative Germanic predecessor, then, Snorri’s
Edda is preeminently convenient. This was recognized as early as 1676, well
before most of the Eddic material had been rendered into English, when
Edward Stillingfleet silently elided any distinction between Snorri’s Iceland
and a hypothetical “Teutonic” mythology. “Edda of Snorro Sturleson,” he says,
“contains the ancient Religion of the Goths.”4
A third part of the story, one well told by Andrew Wawn and others, is
the growth of Anglo-Saxon studies in the early modern period, which was
driven by a conflux of impulses involving history, religion, politics, ethnicity,
linguistics, and documentary culture. The focus of this growth was not the
disinterested recovery of some historical moment but the vindication of a
stable, exemplary, Germanic, and (often) Protestant nation.5 From this per-
spective, arguments about England and its past had an ideological cast, and
Scandinavia provided evidence for those arguments. By the nineteenth cen-
tury, in both popular and learned circles, what might be called a northern
orientation—as opposed to a southern or Roman one—came to prevail in
English philology. While the presence of many French and Latin borrow-
ings was acknowledged, dialect words, cognates, and grammatical parallels
were understood to witness the language’s fundamentally Germanic charac-
ter. More generally, learned groups (like the one that would become the
Viking Society for Northern Research) anchored English ethnicity in the
amalgamation of Anglo-Saxon and Norse traditions. The impact of Snorri’s
Edda and medieval Scandinavia on Anglo-Saxon England underwrote Vic-
torian movements in art, architecture, naturalism, culture, ethnicity, his-
tory, travel, politics, poetry, and, in the efforts of writers like William Morris

4. Edward Stillingfleet, Defence of the Discourse concerning Idolatry (London, 1676), quoted in
Ethel Seaton, Literary Relations of England and Scandinavia in the Seventeenth Century (Oxford:
Clarendon, 1935), 251. Throughout his work Stillingfleet relies on quotations from Gylfagin-
ning. The earliest reference to the Edda in an English work may be Robert Sheringham’s com-
ment that he consulted Peder Resen’s 1665 Latin translation of the Edda (De Anglorum Gentis
Origine Disceptatio [Cambridge, 1670], b1; also see Seaton, Literary Relations of England and Scan-
dinavia in the Seventeenth Century, 265). Sheringham makes extensive use of the translation.
5. See, e.g., Allen J. Frantzen, Desire for Origins: New Language, Old English, and Teaching the
Tradition (New Brunswick, NJ: Rutgers University Press, 1990); Angus Vine, In Defiance of Time:
Antiquarian Writing in Early Modern England (Oxford University Press, 2010); and Timothy Gra-
ham, ed., The Recovery of Old English: Anglo-Saxon Studies in the Sixteenth and Seventeenth Centuries
(Kalamazoo, MI: Medieval Institute Publications, 2000).

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 298 MODERN PHILOLOGY

and W. G. Collingwood, fiction.6 The same historical context that recovered

a usable Anglo-Saxon past also recovered Snorri’s Edda, which allowed that
past to conform to an emergent sense of British identity, one that ran from
preliterate Germanic traditions recorded by Tacitus through Anglo-Saxon
England and Snorri and, finally, to post-Reformation Britain.
But while the formation of these arguments relates to the early modern
formation of an English nation, the arguments themselves have persisted to
the present. It has recently been stated, indeed, that ‘‘Primarily because of
Icelanders’ late conversion, linguistic conservatism and readiness to trans-
mit literature rooted in pre-conversion culture, Scandinavia has provided
the basis for research into all traditional Germanic-speaking cultures.”7
Iceland has the only Germanic literature suitable for “comparative work,”
observes another modern critic. “No other Germanic literature of the mid-
dle ages has the same quantity of pre-Christian material, and no other for-
eign literature has thrown so much light on English heroic poetry.”8 Another
scholar suggests that “‘common Germanic’ tradition viewed the most impor-
tant component of poetry as skillful technique, rather than loftiness of
theme.” Snorri’s role in this approach to Germanic and specifically Old
English literature remains crucial: “The clearest, most systematic and plenti-
ful of this evidence stems from the Old Norse tradition, most notably Snorri
Sturluson’s Edda.”9 As a witness to earlier traditions, Snorri’s Edda even has
been understood to serve as the “Teutonic” matrix out of which J. R. R. Tol-
kien constructed Middle Earth.10
For Anglo-Saxon studies, in particular since the nineteenth century, Snor-
ri’s gods and myths have become increasingly more influential, whether in

6. See further David M. Wilson, “The Viking Age in British Literature and History in the
Eighteenth and Nineteenth Centuries,” in The Waking of Angantyr: The Scandinavian Past in Euro-
pean Culture, ed. Else Rosedahl and Preben Meulengracht Sørensen (Aarhus University Press,
1996), 58–71; Andrew Wawn, ed., Northern Antiquity: The Post-medieval Reception of Edda and Saga
(Enfield Lock: Hislarik, 1994), and The Vikings and the Victorians: Inventing the Old North in Nine-
teenth-Century Britain (Cambridge: Brewer, 2000); Margaret Clunies Ross, The Norse Muse in
Britain, 1750–1820 (Trieste: Parnaso, 1998); and Matthew Townend, The Vikings and Victorian
Lakeland: The Norse Medievalism of W. G. Collingwood and His Contemporaries (Cumberland & West-
morland Archeological & Antiquarian Society, 2009). While such medievalism was a broadly
European phenomenon, it was particularly strong in England. See Michael Alexander, Medie-
valism: The Middle Ages in Modern England (New Haven, CT: Yale University Press, 2007).
7. Alaric Hall, Elves in Anglo-Saxon England: Matters of Belief, Health, Gender and Identity
(Woodbridge: Boydell, 2007), 21.
8. Richard North, Pagan Words and Christian Meanings (Amsterdam: Rodopi, 1991), 1. Also
see his Heathen Gods in Old English Literature (Cambridge University Press, 1997), and The Origins
of Beowulf from Vergil to Wiglaf (Oxford University Press, 2006).
9. Emily Thornberry, “Aldhelm’s Rejection of the Muses and the Mechanics of Poetic
Inspiration in Early Anglo-Saxon England,” Anglo-Saxon England 36 (2007): 81.
10. Tom Shippey, The Road to Middle Earth: How J. R. R. Tolkien Created a New Mythology, rev.
ed. (Boston: Houghton Mifflin, 2003).

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 299

poetic details or in the outlines of larger poetic traditions. Although pagan-

ism evidently survived at least into the tenth century in England, the written
record has left us no compendium of stories, precious few references to
Woden and Þunor, only a handful of place names like Wōdnesdene and Frı̄ge-
dene, and little if any contemporary evidence of cultic veneration or even
popular remembrance.11 What Anglo-Saxon studies does have is compara-
tive philology and the Edda, which can be and has been used to flesh out
the bare bones of an Anglo-Saxon mythology. Yggdrasill thus has been used
to clarify popular Anglo-Saxon beliefs about trees, Grendel’s glof to recall
Þórr’s night in Ymir’s mitten and so sustain parallels between Beowulf and
the Norse god, a passage from Háttatal to explain the meaning of a word in
an Anglo-Saxon will, skaldic poetry to shed light on Anglo-Saxon legal termi-
nology, and the killing of Herebeald in Beowulf to evoke the killing of Baldr
as told by Snorri. From broader perspectives, it has been argued that the
antiquarianism of Gylfaginning substantiates the Beowulf poet’s antiquarian-
ism, that Snorri’s accounts of Óðinn can serve as accounts of Woden, that
Snorri’s description of the Hreiðgotar provides insight on the Hreithgoths of
Widsith, and that Snorri’s “imaginative historiography” in Heimskringla paral-
lels the fictional reworking of a tenth-century Danish raid into Grendel’s
raid on Heorot.12
Snorri’s poetic terminology, specifically kenning and heiti, also has figured
prominently in Anglo-Saxon studies. Snorri defines these terms in Skáldska-
parmál, where in response to Ægir’s question about the three kinds of poetic
language, Bragi says that one is “at nefna hvern hlut sem heitir; ǫnnur grein
er sú er heitir fornǫfn; in þriðja málsgrein er kǫlluð er kenning.”13 Later in

11. See, e.g., David M. Wilson, Anglo-Saxon Paganism (London: Routledge, 1992).
12. See Della Hooke, Trees in Anglo-Saxon England: Literature, Lore and Landscape (Wood-
bridge: Boydell, 2010); Andy Orchard, A Critical Companion to “Beowulf” (Woodbridge: Boydell
& Brewer, 2003), 121–23; Hilda Ellis Davidson, The Sword in Anglo-Saxon England (1962; repr.,
Woodbridge: Boydell, 1994), 188–89; Dorothy Whitelock, ed. and trans., Anglo-Saxon Wills
(1930; repr., Cambridge University Press, 2011), 148–49; John D. Niles, “Myth and History,” in
A Beowulf Handbook, ed. Robert E. Bjork and John D. Niles (Lincoln: University of Nebraska
Press, 1997), 220–21; J. R. R. Tolkien, “Beowulf: The Monsters and the Critics,” in The Monsters
and the Critics and Other Essays, ed. Christopher Tolkien (Boston: Houghton Mifflin, 1984), 22;
Christopher R. Fee, with David A. Leeming, Gods, Heroes, and Kings: The Battle for Mythic Britain
(New York: Oxford University Press, 2001), 20; John D. Niles, Old English Heroic Poems and the
Social Life of Texts (Turnhout: Brepols, 2007), 126; and Helen Damico, “Grendel’s Reign of Ter-
ror: From History to Vernacular Epic,” in Myths, Legends, and Heroes: Essays on Old Norse and Old
English Literature in Honour of John McKinnell, ed. Daniel Anlezark (University of Toronto Press,
2011), 148–66.
13. “One kind is to name something by its own name; another kind is what is called fornafn;
and the third kind is called kenning” (Anthony Faulkes, ed., Edda: Skáldskaparmál, 2 vols. [Lon-
don: Viking Society for Northern Research, 1998], 1:5).

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 300 MODERN PHILOLOGY

the same paragraph Bragi labels kenning as kent heiti. As Snorri subsequently
uses the terms, and as modern Norse scholarship has come to understand
them, the first category consists of heiti (poetic synonyms), the second of
nominal substitutions (both “circumlocutions” and “true descriptions”),
and the third of kenningar (metaphors dependent on the transference of a
noun’s reference).14 Phrased in this manner, Snorri’s terms seem clear and
distinct, and Skáldskaparmál and Háttatal, each after its own fashion, can
be seen as the elaborate development of an argument about the construc-
tion and use of these tropes. Yet as frequently as both Skáldskaparmál and
Háttatal illustrate these terms, and despite much critical commentary, the
precise definitions and the differences among the terms are still unclear.15
Here, however, my concern is not with exactly what Snorri meant but with
how his analysis has influenced Anglo-Saxon studies.
As far as I can tell, the earliest occurrence of kenning as a technical term
in reference to Anglo-Saxon poetry is in Henry Sweet’s “Sketch of the His-
tory of Anglo-Saxon Poetry” in Hazlitt’s 1871 edition of Thomas Warton’s
History of English. Speaking of the characteristics of Anglo-Saxon poetry,
Sweet observes that there are only “half a dozen” similes in Beowulf, all of
them straightforward, “such as comparing a ship to a bird. Indeed, such a
simple comparison as this is almost equivalent to the more usual ‘kenning’
(as it is called in Icelandic), such as ‘brimfugol,’ where, instead of compar-
ing the ship to a bird, the poet simply calls it a sea-bird, preferring the direct
assertion to the indirect comparison. Such elaborate comparisons as are
found in Homer and his Roman imitator are quite foreign to the spirit of
northern poetry.”16 Fifteen years later, Wilhelm Bode states (in German)
that in his discussion of Anglo-Saxon poetry he will use the Old Norse term
kenningar to describe “substantival paraphrases.”17 Sweet and Bode thus
moved kenning laterally (as it were)—from Old Norse to Old English. It was
Albert Cook, in his 1888 edition of Judith, who extended the term backward
from England into pan-Germanic primitive history. “A characteristic orna-
ment of Old English, as well as of early Teutonic poetry in general,” says

14. Phillip Pulsiano et al., eds., Medieval Scandinavia: An Encyclopedia (New York: Garland,
1993), 205–6, 279–80, 351–52.
15. It is particularly puzzling that kent heiti and its antonym ókend heiti occasionally seem to
refer to the same things. See Margaret Clunies Ross, A History of Old Norse Poetry and Poetics
(Cambridge: Brewer, 2005), 242. Also puzzling is the fact that in grammatical handbooks as
well as Háttatal, fornafn has the quite different sense of “pronoun.” See Anthony Faulkes, ed.,
Háttatal (London: Viking Society for Northern Research, 1999), 4.
16. Thomas Warton, History of English Poetry from the Twelfth to the Close of the Sixteenth Century,
ed. W. Carew Hazlitt, 4 vols. (London, 1871), 2:6.
17. Wilhelm Bode, Die Kenningar in der Angelsächsischen Dichtung, mit Ausblicken auf Andere Lit-
teraturen (Darmstadt, 1886), 7.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 301

Cook, “are the kennings.”18 By 1909 Teutonic kennings were well enough
established that J. W. Rankin could describe the impact of the Old Testa-
ment on them prior to their (unattested) disbursement throughout the
Germanic world.19
Sweet indicates that he’s using a foreign word—a comparison is called a
kenning in Icelandic, he says—but in discussions of Anglo-Saxon poetics
since Sweet’s day, kenning has become not only prevalent but naturalized,
losing the scare quotes and Icelandic localization that otherwise mark it as
nonnative. In his monumental edition of Beowulf, Klaeber still surrounds
the term with quotation marks, but Arthur Brodeur, who translated the Prose
Edda and who builds his analysis of the English poem’s art around Snorri’s
terminology, does not. Since the appearance of Brodeur’s book on Beowulf
in 1959, a nativized kenning has been a constant in discussions of Old English
poetics, appearing in prefaces to translations, commentary for editions,
introductions to the Old English language, literary critical interpretations,
and discussions of Old English metrics.20
Snorri’s heiti and the subclasses thereof, such as kent heiti, have also
informed discussions of Anglo-Saxon language and literature, though their
English vitality has been less vigorous. The earliest English occurrence of
heiti is apparently in a 1927 dissertation devoted to the kenning in Anglo-
Saxon and Old Norse poetry.21 But heiti has never achieved an entry in the
OED, and it has failed to become as naturalized as its partner term. While in
Anglo-Saxon criticism kenning is often used by itself, heiti, marked as nonna-
tive word, typically appears only as an accompaniment to kenning.22 Even

18. Albert Cook, ed., Judith, An Old English Epic Fragment (Boston, 1888), lix.
19. J. W. Rankin, “A Study of the Kennings in Anglo-Saxon Poetry,” Journal of English and Ger-
manic Philology 8 (1909): 357–422 and 9 (1910): 49–84. I note here that kenning also occurs in
Gudbrand Vigfusson and F. York Powell’s Corpus Poeticum Boreale: The Poetry of the Old Northern
Tongue from the Earliest Times to the Thirteenth Century, 2 vols. (1883; repr., New York: Russell &
Russell, 1965), 2:448, though there the discussion is narrowly on Old Norse poetry.
20. Howell D. Chickering Jr., “Beowulf ”: A Dual-Language Edition (New York: Anchor, 2006),
6; Craig Williamson, “Beowulf” and Other Old English Poems, with a foreword by Tom Shippey (Phil-
adelphia: University of Pennsylvania Press, 2011), 8–9; R. D. Fulk et al., eds., Klaeber’s Beowulf, 4th
ed. (University of Toronto Press, 2008); Peter S. Baker, Introduction to Old English, 3rd ed. (Mal-
den, MA: Wiley-Blackwell, 2012), 136; Gillian R. Overing, Language, Sign, and Gender in “Beowulf”
(Carbondale: Southern Illinois University Press, 1990), 14–21; Donald G. Scragg, “The Nature
of Old English Verse,” in The Cambridge Companion to Old English Literature, ed. Malcolm Godden
and Michael Lapidge (Cambridge University Press, 1991), 66–67; Haruko Momma, The Compo-
sition of Old English Poetry (Cambridge University Press, 1997), 181; and Jun Terasawa, Old English
Metre: An Introduction (University of Toronto Press, 2011), 75.
21. Hendrik van der Merwe Scholtz, The Kenning in Anglo-Saxon and Old Norse Poetry (Utrecht:
Dekker, Vegt & Leeuwen, 1927), 36.
22. Kemp Malone and Albert C. Baugh, The Middle Ages, vol. 1 of A Literary History of England,
2nd ed. (London: Routledge, 1967), 30; Bjork and Niles, Beowulf Handbook, 92–94.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 302 MODERN PHILOLOGY

Brodeur italicizes heiti, as do others, whether prefacing translations, discuss-

ing Anglo-Saxon literary history, or analyzing Old English literature.23
Much recommends this use of the Edda. Its length, detail, and unique
content do make it far and away the single most important medieval source
on Scandinavian poetics and mythology. As many critics have noted, there
is nothing else like it in any of the medieval northern literatures. As a result,
a significant portion of what we know (or think we know) about mythic cul-
ture anywhere within Germanic traditions depends directly or indirectly on
Snorri. The Prose Edda may be a suspect, compromised authority, but it is
the best authority we have; in this sense it has the flawed but crucial histori-
cal value of the Icelandic family sagas. Only the most foolhardy of critics
would disregard either.
At the same time, it is worth asking about Snorri’s authority for Anglo-
Saxon England. He would have had few opportunities to learn firsthand any-
thing about English poetry that might have sustained or challenged the pos-
sibility of shared, premigration traditions. For all the martial and mercantile
connections between Iceland and Anglo-Saxon England, we have little evi-
dence of literary exchanges from the days of Offa (say) to those of Edgar
and Æðelræd. More extensive contact did occur after the Anglo-Saxon
period, when a copy of abbot Ælfric’s “De Falsis Diis” (ca. 1000), which
underlies one homiletic piece in the fourteenth-century Hauksbók, evidently
made its way to Iceland.24 In this same period, English missionaries brought
other English books to Iceland, where England was regarded as a center
of learning.25 The author of the twelfth-century Icelandic First Grammatical
Treatise knew that English writing used letters from the Latin alphabet and
judged that “vér erum einnar tungu, þó at grǫzk hafi mjǫk ǫnnur tveggja

23. J. E. Cross, “On ‘The Wanderer’ Lines 80–84: A Study of a Figure and A Theme,” Veten-
skaps-Societetens i Lund Årsbok, 1958–59, 92; Stanley B. Greenfield and Daniel G. Calder, A New
Critical History of Old English Literature (New York: NYU Press, 1986), 125; and Dick Ringler, Beo-
wulf: A New Translation for Oral Delivery (Indianapolis: Hackett, 2007), xciv. For the application
of kent heiti to Old English poetry, see Alvin A. Lee, Gold-Hall and Earth-Dragon: “Beowulf ” as Met-
aphor (University of Toronto Press, 1998).
24. See John C. Pope, ed., Homilies of Ælfric: A Supplementary Collection, EETS, o.s., 259, 260
(Oxford University Press, 1967–68), 2:669–70; Eirı́kur Jónsson and Finnur Jónsson, eds., Hauks-
bók, Utgiven efter de Arnamagnænske Håndskrifter no. 371, 544 og 675 4o , 3 vols. (Copenhagen:
Thieles Bogtrykkeri, 1892–96), 1:156–64; and Arnold Taylor, “Hauksbók and Ælfric’s De Falsis
Diis,” Leeds Studies in English 3 (1969): 101–9.
25. The twelfth-century churchmen St. Þorlákr and Páll Jónsson are both said to have gone
to England to study. See Biskupa Sögur, vol. 2, ed. Ásdı́s Egilsdóttir, Íslenzk Fornrit 16 (Reykja-
vı́k: Hið Íslenzka Fornritafélag, 2002), 52, 147, 297–8. Also see Magnús Fjalldal, Anglo-Saxon
England in Icelandic Medieval Texts (University of Toronto Press, 2005), 9–10; and Geraldine
Barnes, “The Medieval Anglophile: England and Its Rulers in Old Norse History and Saga,”
Paregon 10 (1992): 11–25. Barnes describes the various economic and social connections be-
tween England and Scandinavia during this period.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 303

eða nǫkkut báðar.”26 Yet even though English merchants continued to visit
Iceland in the centuries immediately before Snorri lived, Icelandic familiar-
ity with England seems to have been more limited than the sagas sometimes
imply. In England, where the court of King Cnut (ruled 1017–35) flour-
ished as a literary and social focus of Norse activity, significantly more con-
tact between the Norse and the English occurred. In the words of Roberta
Frank, at that time “England rivalled Norway as the centre in the North for
the production and distribution of skaldic poetry.”27 But all this contact
occurred not in the mists of a common, pre-Christian era but in the late
Anglo-Saxon period. And though Snorri speaks about this era in Heims-
kringla and Egilssaga, some of whose skaldic poems date to the tenth or even
ninth centuries, whether he had any specific personal knowledge of this or
the later period is impossible to say. Snorri’s own claims about England
often disagree with Anglo-Saxon sources.28 In any case, a thirteenth-century
critic like Snorri who remembered even Cnut’s era would be remembering
not an English past but a Norse one. All this makes it difficult to talk about
unadulterated Anglo-Saxon practices that Snorri could, or could not, cor-
roborate.
Focusing more narrowly on the construction and purpose of Snorri’s
Edda renders it even more suspect as a source of reliable comparative infor-
mation on Anglo-Saxon England. Just why Snorri wrote the Edda as he did,
and for that matter just what he wrote, cannot be determined with certainty.
Earlier generations of critics saw his motive as largely academic and anti-
quarian: a desire to preserve a pagan past, specifically its mythology, that by
the thirteenth century was rapidly fading away. Others have seen Snorri as
primarily a writer devoted to the preservation of a literary tradition. Still
others have argued that Snorri’s aim was political advancement, and that
mythology and poetics were primarily a historical means toward this con-
temporary goal. To the extent that Snorri perceived familial connections

26. “We are of one tongue with them, even though one of our languages has been greatly
changed or both of them somewhat.” Both original and translation are from Einar Haugen,
First Grammatical Treatise: The Earliest Germanic Phonology: An Edition, Translation, and Commentary,
Language Monograph 25 (Baltimore: Linguistic Society of America, 1950), 12.
27. Roberta Frank, “King Cnut in the Verse of His Skalds,” in The Reign of Cnut: King of
England, Denmark, and Norway, ed. Alexander R. Rumble (London: Leicester University Press,
1994), 107. Also see her “Did Anglo-Saxon Audiences Have a Skaldic Tooth?,” Scandinavian
Studies 59 (1987): 338–55.
28. See Bruce E. Gelsinger, Icelandic Enterprise: Commerce and Economy in the Middle Ages
(Columbia: University of South Carolina Press, 1981), 131; and Anthony Faulkes, “The Sources
of Skáldskaparmál: Snorri’s Intellectual Background,” in Snorri Sturluson: Kolloquium anlässlich
der 750; Wiederkehr seines Todestages, ed. Alois Wolf (Tübingen: Narr, 1993), 71. Magnús Fjalldal
even suggests that “for medieval Islanders Anglo-Saxon England largely existed as a never-never
land,” though this seems truer of the fornaldursǫgur than of the ı́slendingasǫgur (Anglo-Saxon
England in Icelandic Medieval Texts, 124).

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 304 MODERN PHILOLOGY

between himself and the kings of Norway, after all, he also accepted (at least
in principle) that he descended from the gods he described.29
But if Snorri’s motives are unclear, we know with certainty at least some
things about the Prose Edda itself. We know that, whatever motivated Snorri,
he wrote as the Icelandic Commonwealth was beginning to collapse and
that, in the thirteenth century, Icelanders were particularly interested in a
time when they could see themselves as having achieved significant cultural
and political achievements outside of the influence of Norway. Besides
Snorri’s Edda, the thirteenth century produced a large collection of Eddic
poetry—the Codex Regius—and a body of Íslendingasǫgur focused on the
days of settlement and the creation of the commonwealth.
Significantly for the purposes of this essay, all these works have a kind of
nostalgia that is much less evident in Anglo-Saxon England. Writers such as
Gildas and Bede tend to recall premigration days in the aggregate or as his-
torical events, not shared individual myths that unified a people. They do
not frame this memory as a golden age but as a pagan past, the rejection of
which enabled the founding of a Christian civilization in England. And even
this memory was never cultivated as actively as were memories of a precon-
version past in Iceland.30 In the twelfth century, when Anglo-Saxon manu-
scripts still were being copied, historians like William of Newburgh, Henry
of Huntington, and Ordericus Vitalis presented pre-Conquest England in
ways that might mitigate and justify the disruption caused by the Normans,
not co-opt a pagan past. Nostalgia is still more negligible in Snorri’s thir-
teenth-century English peers. Between internal conflicts among the barons
and Continental conflicts with France, England’s political interests were
focused on current political history, and never on pre-Christian beliefs,
poetry, and cultural practice. The thirteenth-century Havelok the Dane, for ex-
ample, bespeaks strong Scandinavian connections and background, but
looks not to Iceland but to Denmark, which it presents in distinctly contem-
porary terms. In short, Snorri’s historical and mythographic impulses were
shared neither by his English contemporaries nor by their predecessors.31

29. See Kevin J. Wanner, Snorri Sturluson and the Edda: The Conversion of Cultural Capital in
Medieval Scandinavia (University of Toronto Press, 2008).
30. See Nicholas Howe, Migration and Myth-Making in Anglo-Saxon England (New Haven, CT:
Yale University Press, 1989).
31. See A. N. Doane and William P. Stoneman, Purloined Letters: The Twelfth-Century Reception
of the Anglo-Saxon Illustrated Hexateuch (British Library, Cotton Claudius B. iv) (Tempe: Arizona
Center for Medieval & Renaissance Studies, 2011); Mary Swan and Elaine M. Treharne, eds.,
Rewriting Old English in the Twelfth Century (Cambridge University Press, 2000); Christine Fran-
zen, The Tremulous Hand of Worcester: A Study of Old English in the Thirteenth Century (Oxford: Clar-
endon, 1991); T. W. Machan, “Language and Society in Twelfth-Century England,” in Placing
Middle English in Context: Selected Papers from the Second Middle English Conference, ed. Irma Taavit-
sainen et al. (Berlin: Mouton de Gruyter, 2000), 43–66, and English in the Middle Ages (Oxford
University Press, 2003), 71–110.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 305

More generally, it is easy to overlook just how peculiar Snorri’s Edda is

within Germanic literary traditions and not to consider how this peculiarity
undermines its value for cultural comparison. As I noted earlier, we have
nothing like it in any other Germanic language, and even among extant
Old Norse texts, the Edda is unique. The First Grammatical Treatise preceded
it by perhaps a century, as did the poem Háttalykill en Forni, which illustrates
a number of skaldic meters. But neither work could be linked generically
with Snorri’s efforts. The Grammatical Treatise concentrates on orthography,
while Háttalykill provides neither mythological narratives like Gylfaginning
nor prose analysis of poetics like Skáldskaparmál and was never as ambitious
as Háttatal.32 The thirteenth-century poetic counterpart to the Prose Edda,
the Poetic Edda, so far diverges from Snorri’s efforts as to leave his Edda peer-
less as a Norse poetic handbook. The Codex Regius is almost all poetry
alone, none of it skaldic, and its stories do not have the kind of scholastic
commentary and critical distancing that appears throughout Snorri’s Edda.
Since Snorri’s Edda was composed relatively late in the Middle Ages, it is
fundamentally Christian in outlook. It remembers pagan gods and practices
but only through the lens of a Christian and following Christian exegetical
traditions of compilation and synthesis. Snorri certainly was aware of Latin
grammatical habits, including specifically doctrinal language; he was perhaps
aware of the twelfth-century Platonists, albeit indirectly. Anthony Faulkes has
asserted that Snorri “does not think like a scholastic theologian, and it is
absurd to think that he read any scholastic writers,” but rough contemporar-
ies such as the anonymous author of Skáldatal and Óláfr Þórðarson, who
wrote the Third Grammatical Treatise, recognized overlaps between skaldic
practice and Latin grammatical traditions, and Snorri evidently did have
familiarity with medieval encyclopedias and rhetorical handbooks.33 In
Snorra Edda, this familiarity appears preeminently in the work’s coherence,
relating the cosmos to its inhabitants, the gods to human beings, the stories
to one another, and the stories to their poetics. As a mythological statement,
the Edda’s most salient characteristic is this same coherence, which emerges

32. On the substantial differences between Háttalykill and Háttatal, see Jón Helgason and
Anne Holtsmark, eds., Háttalykill enn Forni, vol. 1 (Copenhagen: Munksgaard, 1941), 118–34.
33. Faulkes, “Sources of Skáldskaparmál,” 75. Also see Margaret Clunies Ross, Skáldskapar-
mál: Snorri Sturluson’s Ars Poetica and Medieval Theories of Language (Odense University Press,
1987). Guðrún Nordal has argued that even the First Grammatical Treatise and Háttalykill show
evidence of—or at least are consistent with—scholastic discussions of grammatica. See Tools of
Literacy: The Role of Skaldic Verse in Icelandic Textual Culture of the Twelfth and Thirteenth Centuries
(University of Toronto Press, 2001), 19–72 and 199–236. For an even stronger (and not widely
shared) view of Snorri’s dependence on Latin Christian culture, see Ursula Dronke and Peter
Dronke, “The Prologue of the Prose Edda: Explorations of a Latin Background,” in Sjötiu Rit-
gerðir helgaðar Jakobi Benediktssyni, ed. Einar Petursson and Jónas Kristjansson (Reykjavı́k: Mag-
nússonar, 1977), 153–76. Whether Snorri could read Latin, or simply heard about arguments
from Latin, is up for debate.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 306 MODERN PHILOLOGY

in sharp relief when Snorri’s Edda is juxtaposed with the Eddic poems and
the practices depicted in the sagas. The similarities between Snorri’s systemi-
zation and scholastic practice seem to have struck even some of Snorri’s
early readers, such as the compilers of the fourteenth-century Codex Wor-
mianus, which includes the Prose Edda alongside all four Icelandic grammat-
ical treatises.34
The Anglo-Saxon prelates Ælfric and Wulfstan demonized the pagan
gods whom Snorri could look at with a kind of dispassionate academic inter-
est, what Faulkes elsewhere describes as an “almost humanistic detach-
ment.” This detachment and Snorri’s “respect for antiquity,” Faulkes adds,
“make him in fact much more like the Latin mythographers of the Middle
Ages.”35 The stories told in the Eddas retained their value for thirteenth-cen-
tury Icelanders like Snorri precisely because pagan beliefs were denuded of
any cosmic explanatory power but remained what Margaret Clunies Ross
calls “cognitive tools to think with and live by.”36 Part of the uniqueness—
and achievement—of the Edda, then, is the way its coherence is filtered
through a Christian lens. As Faulkes observes, “‘Snorri had an intimate
knowledge of Norse mythology, but he did not reproduce that mythology
unchanged, and his account needs to be handled with care by those who
wish to use it to shed light on heathen religion.”37 As Snorri himself says,
the purpose of at least Skáldskaparmál is to teach the old poetic techniques
to young poets—but not, for Christians, to believe in any of the heathen
gods or myths.38
Here I once more turn to Old Norse and Old English poetics. If we have
no Old English poetic handbook or mythography comparable to Snorri’s
Edda, we also have nothing in cross-references or manuscript catalogs to
indicate that there ever existed either a coherent composite like the Prose
Edda or the panoptic vision that imagined it. It is worth recalling that Snor-

34. Þórir Óskarsson describes the Codex Wormianus “as an interesting source for the medi-
eval literary and rhetorical disciplines, both domestic and international, that Icelandic writers
were familiar with” (“Rhetoric and Style,” in A Companion to Old Norse-Icelandic Literature and Cul-
ture, ed. Rory McTurk [Malden, MA: Blackwell, 2005], 356).
35. Anthony Faulkes, ed., Edda: Prologue and “Gylfaginning” (Oxford: Clarendon, 1982),
xxii.
36. Margaret Clunies Ross, Prolonged Echoes: Old Norse Myths in Medieval Northern Society, vol.
2, The Reception of Norse Myths in Iceland (Odense University Press, 1998), 23. Also see Clunies
Ross, Prolonged Echoes: Old Norse Myths in Medieval Northern Society, vol. 1, The Myths (Odense Uni-
versity Press, 1994).
37. Faulkes, Edda: Prologue and “Gylfaginning,” xxviii. Also see Clunies Ross, Skáldskaparmál.
It is worth adding that where there is no extra-Edda confirmation for one of Snorri’s stories or
interpretations, in a lausavisa for example, it is also always possible that Snorri, driven as he was
to explain, might have engaged in a kind of back-formation: reading a poem, deciding what it
must mean, and then creating a category in which that meaning is possible.
38. Faulkes, Edda: Skáldskaparmál, 1:5.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 307

ri’s thirteenth-century English contemporaries like Robert of Gloucester,

or even slightly earlier counterparts like Henry of Huntingdon and William
of Malmesbury, evince plenty of interest in English social and political his-
tory but, lacking the historical interest that would arise in England only with
works like Richard Verstegan’s 1605 Restitution of Decayed Intelligence, pay
little attention to Anglo-Saxon paganism and poetics. And to the Anglo-
Saxons who lived closer to the times Snorri often is understood to corrobo-
rate, to clerics like Ælfric and Wulfstan, pagan practices were perhaps too
real to be treated dispassionately. In any case, we have no reason outside the
Prose Edda to believe that Old English poetics once were imagined by a syn-
thetic vision like Snorri’s, or that Old English poems and myths ever existed
in even the casual systemization of the Codex Regius or Gylfaginning. Cer-
tainly none of the four great Anglo-Saxon poetic manuscripts suggests such.
In this context, I return to kenningar and heiti, neither of which has a
direct cognate in Old English with a similar sense or usage. For the latter,
the closest comparisons may be to the distantly related hāt, a noun meaning
a “promise” or “vow” that has nothing to do with poetic terminology, or to
the later northern Middle English hight, which is also distantly related, which
also means “promise” or “vow,” and which also has no poetic significance, or
to the Old English verb hātan, meaning “to be called,” again with no poetic
implications. As for kenning, the Old English cenning, despite its ortho-
graphic similarity to the Norse word, derives from cennan (“to bring forth”)
and refers only to childbirth. A semantically related form is not recorded
until the fourteenth-century kenning, meaning “teaching,” “instruction,” or
“cognition.” This word, which the OED treats as a distinctly English deverbal
formation from ken, persists into the modern period, mostly in Scotland and
northern England. For five hundred years, from its fourteenth-century first
citation to the late nineteenth century, kenning is not recorded as a technical
term.39 The OED does list a poetic sense for this deverbal word, but necessar-
ily illustrates it with Victorian quotations, since, as I noted above, the word is
not used as an English term for a poetic technique until 1871.40 The best
one can say about this listing is that it is misleading, in effect naturalizing
kenning and avoiding its actual history by simply adding a putatively medie-
val but fundamentally modern sense to a homonym. Sweet’s kenning rather

39. Joseph Bosworth and T. Northcote Toller, eds., An Anglo-Saxon Dictionary (1898; repr.,
Oxford University Press, 1983), s.vv. cennan, vb., hāt, n., hātan, vb.; Dictionary of Old English: A to G
Online, s.v. cenning, n., http://tapor.library.utoronto.ca/doe/; OED, s.vv. hight, n. 1, kenning,
n. 2, http://www.oed.com; MED, s.vv. hight, n., kenning(e, n., http://quod.lib.umich.edu/m/
med/; Dictionary of the Scots Language, s.v. ken(n)ing, vbl. n., http://www.dsl.ac.uk. The MED also
records kenninge, n. (“mother”) and kenning, n. (“a small ulcer”), both of which correlate with
kenning, n. 3 in the OED and derive, ultimately, from Old English cennan (“to give birth”).
40. OED, s.v. kenning, n. 2, sense 6. The earliest OED citation is not to Sweet but to Vigufsson
and Powell’s 1883 Corpus Poeticum Boreale.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 308 MODERN PHILOLOGY

should be treated as a distinct, modern loan word with no English history

and no semantic or derivational connection to its current headword in the
OED. Kenning was not an English poetic term until Sweet made it so.
It is important to recognize that more than simply characterizing skaldic
poetry, kenningar and heiti are among its requirements. They function as part
of a specifically Norse poetics: in Guðrún Nordal’s words, they provide
the feature that “divides the skilled from the less accomplished.”41 A Victorian
sonnet that lacked metaphors but had an octave and a sestet would be
generically regular; a skaldic poem that lacked kennings would not be. Just
as skaldic kennings can be linked to precursors in the Poetic Edda, so skaldic
poetry itself is a specifically Norse genre, one that depends on syntactic and
metrical patterns that originate not in some premigration “Teutonic” poet-
ics but in developments to the fornyrðislag line that themselves depended
on eighth- and ninth-century phonological changes in Scandinavian.42
For their substance, kenningar and heiti characteristically bespeak historical
events of the kind Snorri remembers in Heimskringla or a mythic world
memorialized in the Prose Edda. They exist, that is, not merely to describe
swords and shields in general but to invoke uniquely Scandinavian experi-
ences, such as the flight of Hrólfr kraki across Fyrisvellir or Þórr’s penchant
for driving a chariot. In the baldest terms, kenningar and heiti are not isolable
phenomena, and Snorri’s Edda exists to explain a kind of poetry that did
not exist in English or, for that matter, any other language. Collapsing Old
English poetic technique into its Old Norse counterpart, however much
resemblance there may be between the two, transforms the character of
both.
By no means is Snorri’s Edda the only influential or even the most fre-
quently translated Norse work; Njálssaga, Egilssaga, and even Friðþjófs saga
hins frœkna have all appeared with considerably more frequency. Similarly,
our knowledge of—or perhaps speculation about—pre-Christian Anglo-
Saxon England does not derive from Snorri alone. But since the late seven-
teenth century, the Prose Edda has occupied an increasingly prominent posi-
tion in the study of Norse literature and culture. In that same time frame,
medieval Scandinavia, in turn, has occupied an increasingly prominent
position in the study of Anglo-Saxon literature and culture: first, by encour-
aging suppositions that there were gaps in what is known about pre-Con-
quest English mythology and social practice, and, second, by filling in those
same gaps. Recalling James Macphersons’s Gaelic poet Ossian, whose rise

41. Nordal, Tools of Literacy, 205.

42. Kari Ellen Gade, The Structure of Old Norse “Dróttkvætt” Poetry (Ithaca, NY: Cornell Univer-
sity Press, 1995); and Heather O’Donoghue, Skaldic Verse and the Poetics of Saga Narrative (Oxford
University Press, 2005). For discussion of skaldic poetry as dependent on a larger cultural
matrix of literary composition, see Clunies Ross, History of Old Norse Poetry and Poetics, 83–113.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).
 Tim William Machan Snorri ’s Edda 309

to prominence coincides with Snorri’s, Snorri would have to have been

invented if he had not lived.43 Ossian was in fact invented and, in some ways,
so was Snorri. It is this multifaceted prominence that has transformed him
from thirteenth-century Icelandic politician and poet into the expert on
Anglo-Saxon poetics and mythology that he is today.
In this context, of course, mythology can be taken two ways. We might
think of it as beliefs about Woden, Þunor, and Frig, of which the Anglo-
Saxon record has preserved, for the most part, only tantalizing glimpses. Or
we might think of mythology as the process of reception that fashioned a
sense of Anglo-Saxon England as fundamentally rooted in earlier “Teutonic”
traditions whose best expression appears in later medieval Scandinavian
works. Given the eighteenth- and nineteenth-century contexts in which this
fashioning initially took place, such mythology had perhaps even greater
ramifications for an emergent sense of Great Britain as itself a fundamentally
Germanic place, despite the historical impact of French and its speakers
and the more recent impact of immigrants from the expanding empire. Of
the first kind of mythology, Snorri may have relatively little to tell us. Of the
second kind, he has much to say.

43. For an account of how Macpherson’s Gaelic primitivism drew on contemporary discus-
sions of medieval Scandinavia, see Margaret Omberg, Scandinavian Themes in English Poetry,
1760–1800 (Uppsala: Almqvist & Wiksell, 1976), 26–33.

This content downloaded from 128.252.067.066 on July 25, 2016 20:38:32 PM

All use subject to University of Chicago Press Terms and Conditions (http://www.journals.uchicago.edu/t-and-c).