"Noticia de los Kenningar": Borges, Kennings, and the Spanish Tradition

In 1932, in the periodical Sur, Borges published "Noticia de los Kenningar," his first essay on Scandinavian literature, which comprised a "report" on the type of metaphorical figures characteristic of Old Norse (and Old English) poetic diction known in English as kennings. Already by 1933 this short piece had been extensively expanded and revised and was republished in the form of a 26-page booklet Las Kenningar, which in turn was incorporated into the collection Historia de la Eternidad (1936) under the same title.1 Almost twenty years later the essay reappeared in a modified version as "La poesía de los escaldos" (The Poetry of the Skalds) in the Scandinavian literature section of Borges' introduction to medieval Germanic literature Antiguas literaturas germánicas (with Delia Ingenieros, 1951),2 and once again in a reworking of that last title as Literaturas germánicas medievales.3 Borges' appraisal of the kennings of Icelandic poetry in the 1932 article was quite disparaging and, although the indignant tone he adopts may well be tongue-in-cheek, his report presents kennings as empty formulaic phrases, clever perhaps, but with little literary merit. Over the years as Borges' appreciation and knowledge of Old Norse literature increased, and the original essay was revised and corrected, some of the more acerbic criticisms were modified or removed.4 However, despite these alterations, the core content of the original essay has remained surprisingly unaltered, especially the list of kennings which was central to his original argument. Moreover, a detailed examination of Borges' inventory demonstrates that his opinions were based on a very imperfect and often utterly erroneous understanding of the nature of kennings and their function in Old Norse poetry; it also reveals a clear tendency to manipulate the material to suit his own agenda, which was to advance the view that the kennings represented the first conscious verbal play of a primitive, "instinctive" literature, as opposed to a long-established, sophisticated and learned tradition. While a number of studies focus on the question of Borges and kennings, with the exception of Lavender's article on La Alucinación de Gylfi, there has been almost no critical analysis of the sources for Borges' list of kennings or his translations of them.5 Indeed, the tendency has rather been to celebrate Borges' achievement in introducing the concept of the kenning to a Spanish-speaking audience.6 In fact, he had been anticipated in this respect 150 years earlier by the Spanish priest Juan Andrés y Morell, and well before the eighteenth century there already existed a substantial body of literature in Spanish about the culture and history of the "Septentrional" lands which, although he was almost certainly unaware of it, shares interesting parallels with some of Borges' own attitudes and concerns.7 The first part of this paper aims to provide an overview of the literature in Spanish which treats of the medieval north and its literature, prior to the publication of "Noticia de los Kenningar," with a view to situating Borges within this tradition. The second section provides a detailed survey of Borges' initial list of kennings, analysing the extent to which his translations diverge from the original Old Norse texts, as well as the implications of and reasons for these differences.

The Reception of the Medieval North in Spanish pre-1932

While it would be an exaggeration to talk of an unbroken tradition of Old Norse scholarship in the Spanish language, at least before the 1980s, there certainly exists an extensive literature that concerns itself with the countries and cultures of the north, which stretches as far back as the early medieval period.8 The first allusions to Scandinavia, or rather Scandinavians, in Iberian sources occur in the Latin and Arabic chronicles and histories that record the Viking raids on the peninsula in the ninth and eleventh centuries. Not unexpectedly, neither the Muslim nor Christian writers demonstrate much curiosity as to the precise nature or culture of the Vikings, who are almost without exception presented as an "extremely cruel" people who come to Spanish shores to ravage, plunder and lay waste with "slaughter and fire," yet it was these histories which generated the first modern scholarly publications on Hispano-Nordic relations in the nineteenth and early twentieth centuries.9 Another episode which generated attention among Spanish historians was the marriage of Princess Christina of Norway to Philip of Castile (brother of King Alfonso X the Wise) in 1258, an alliance which may have led to important cultural interchanges between Spain and Norway and possibly prompted a renewed interest in the gothic origin of the Spanish monarchy in the thirteenth century; at any rate, it was during Alfonso's reign that this idea was promoted and included as a fundamental part in the great historical chronicles of the time, e.g. la Estoria de España (1271-1274). The question of Spain's gothic heritage became a trope monarchs and diplomats frequently had recourse to over the following centuries, such as Carlos V, who in 1542 famously referred to the common origin of the Swedish and Spanish in the words: sumus et nos de gente Gothorum.10 It is a tradition that Borges was well aware of, and to which he evidently subscribed, as we can see in his introduction to La Alucinación de Gylfi:

Los visigodos que ocuparon España a principios del siglo quinto creían que su estirpe era escandinava. Así lo afirma el historiador Jordanes en la obra De Rebus Geticis; así lo reitera Diego de Saavedra Fajardo en su Corona Gótica. Aunque profesasen la fe cristiana, recordarían algunas de las fábulas mágicas que ahora vuelven a España en este libro.
[The Visigoths who conquered Spain at the beginning of the fifth century believed that their lineage was Scandinavian. This is affirmed by the historian Jordanes in the work De Rebus Geticis; and Diego de Saavedra Fajardo reiterates it in his Corona Gótica. Although they professed the Christian faith, they would have remembered some of the magical fables which now return to Spain in this book.]11

In the sixteenth century, due to the discovery and conquest of the Americas, Spain saw a flourishing of interest in distant and exotic lands, and this trend, combined with the appearance of works such as Olaus Magnus' Carta marina et descriptio septentrionalium terrarum (1539), his complementary Historia de Gentibus Septentrionalibus (1555), and his brother Johannes' Historia de omnibus Gothorum Sueonumque regibus (1554), generated something of a craze in Europe for the Septentrional lands, Spain included.12 Thus, information from medieval Scandinavian sources, especially Saxo Grammaticus' Gesta Danorum, gradually filtered into Spanish miscellanies, histories, and even fictional works, such as Miguel de Cervantes' Los trabajos de Persiles y Sigismunda, historia setentrional (1617), in which part of the action is set in "Ultima Thule," Iceland. These publications were more concerned with entertaining and provoking a sensation among their readers than any pretensions to accuracy, but more serious treatments of Nordic material also appeared, such as Bernardino de Rebolledo's Selvas dánicas (1655), an extended poetic work in the Gongorismo style of the Golden Age, written during his time as Spanish ambassador to the Danish court and against the background of the thirty years war. In the section celebrating the royal lineage of his patroness Queen Sophie Amalie, Rebolledo drew extensively from Saxo's Gesta Danorum, which he also used to advance the Spanish cause by emphasizing the Roman Catholic faith of the ancient Danish kings.13

It was not until the late eighteenth century, however, that the first extended description of Old Norse literature in Spanish was published. This was the work of Juan Andrés y Morell, a Jesuit priest from Valencia who, following his expulsion from Spain in 1767, settled in Italy where he wrote his famous Of the Origin, Progress, and Current Situation of All Literature, a seven-volume comparative study of the literatures of the entire world. Originally published in Italian and later translated into Spanish as Del origen, progresos y estado actual de toda la literatura (1784-1806), the third volume of Juan Andrés' work contains three sections dedicated to Old Norse poetry entitled "Septentrional ò Scalda," "El Edda" and "Gusto de la poesía de los Scaldros."14 Although highly acclaimed as a scholar and humanist, Juan Andrés was no expert in Scandinavian literature, and his knowledge of Old Norse poetry was entirely derived from French and Latin sources.15 Therefore, although some parts of his description are surprisingly accurate and scholarly, such as his detailed account of the metrical requirements of the complex drottquade (dróttkvætt) metre (pp. 168-70), the work contains many misunderstandings. For example, Juan Andrés misinterprets a debate around the authorship of the Poetic and Prose Eddas as a dispute as to which of these is the "authentic" Edda. 16 Concluding that the argument for Snorri Sturluson's Edda is more convincing, he proceeds to give a description of that work under the section entitled "El Edda," where the first two parts, The Prologue and Gylfaginning, "an extract about the mythology of the ancients" are called "Demisaghas," Skáldskaparmál, "a poetic treasury," is called (curiously) "Kenningar," and Háttatal, "an Icelandic prosody," is called "Liodsgrientr ... which signifies 'distinction of sounds'" (pp. 167-68).

Using a tone remarkably similar to that of "Noticia de los Kenningar,"17 Juan Andrés' response to the poetry of the "Scaldros" reflects a mixture of amusement, fascination and disgust, as he details the manner in which these poets went out of their way to make their works "enigmatic and unintelligible" and "convoluted and obscure," by means of their syntax, "an extremely strange transposition of words," and their use of kennings, as here:

¿Pero quién podrá entender ciertas metonimias y perifrasis, de que hacian singular ostentacion aquellos Poëtas? Para decir Yo pongo el anillo en mi dedo, dice un Poëta citado en el Edda: yo sujeto la serpiente golpeada toda al rededor abriendo la garganta à la punta del puente del Francolin à la horca del escudo de Odin. La horca del escudo de Odin es el brazo, sobre el qual estriva el escudo; el puente del Francolin es el puño, donde el Alconero pone el Alcon (tomándose el Francolin por el Alcon por licencia de variar las especies comun entre aquellos Poëtas); y la punta de aquel puente es el dedo, la serpiente &c., es el anillo. Los chinches se llamaban habitantes de las murallas; y à otro insecto aun mas asqueroso se le daba el pomposo nombre de Soerkelefant, elefante de camisa. Antonomasias, metáforas, hipérboles, y expresiones atrevidas y obscuras constituian los ornamentos mas apreciables de aquella poesía; sin embargo Troil dice, que no pueden leerse sin mucho placer el Bjarkamal de Lodbrok cercano à la muerte,18 y varios otros de aquellos poëmas. Goce él en hora buena este sumo placer, que yo ciertamente no se lo envidio, y de buena gana le cedo la plena posesion sin pretender la mas mínima parte, y sin la mas leve sombra de zelos ni competencia.19
[But who can understand certain metonyms and periphrases which those poets flaunted so singularly? To say, "I put the ring on my finger," one poet quoted in the Edda says: "I hold the snake beaten all around opening the throat to the point of the bridge of the partridge to the gallows of the shield of Odin." "The gallows of the shield of Odin" is the arm, on which the shield rests; "the bridge of the partridge" is the fist where the Falconer puts the falcon (taking partridge for falcon due to the licence to interchange common species among those poets): and "the point" of that bridge is the finger, the "serpent" etc. is the ring. Bedbugs are called "inhabitants of the walls," and to another insect more disgusting was given the pompous name Soerkelefant, "elephant of shirt." Antonomasias, metaphors, hyperboles, and audacious and obscure expressions constitute the most significant ornaments of that poetry: nevertheless, Troil says that one cannot read without pleasure Lodbrok's Bjarkamal (when he was close to death) and many other of those poems. May he happily enjoy this great pleasure, that I certainly do not envy, and to which with goodwill I concede to him the full possession without a desire for the smallest part, and without the slightest shadow of jealousy nor rivalry.]

Juan Andrés' method of presentation and his choice of examples are obviously intended primarily to titillate and entertain his readers. However, amusing as it may be to believe that "inhabitants of the walls" refers to a bedbug, in fact the original kenning viðbjorn aldinna veggja "wood-bear of old walls" refers to a mouse. I have not been able to locate the source for Soerkelefant, but "elephant of the shirt" does not make any sense as a kenning for "louse," and is obviously a misreading in one of Juan Andrés' sources. The other quotation, although thoroughly mangled, is recognizable as part of a stanza attributed to the tenth-century poet Egill Skalla-Grímsson, which indeed describes the placing of a gold ring on his arm.20 As far as I have been able to ascertain, this represents the first rendering of dróttkvætt verse in Spanish, although of course not translated directly from the Old Norse. The first complete (or almost complete) Old Norse poem to appear in Spanish, Hávamál, is also a translation of a translation. It was included (alongside other information about Old Norse mythology and literature) under the entry for Iceland in a volume dedicated to modern geography of the Encyclopédie méthodique, translated into Spanish by Juan Arribas y Soria y Juan de Velasco in 1792.21

Moving into the nineteenth century, some works worthy of note include: an essay published by the Venezuelan diplomat and scholar Andrés Bello, "Influencia de la poesía germánica en el romance," which contains a brief description about the characteristics of skaldic poetry and the use of kennings; Antonio Bachiller y Morales' Antigüedades Americanas: Noticias que tuvieron los europeos de la América antes del descubrimiento de Colón (1845), a collection of material (including a translation of Carl Christian Rafn's Antiquitates Americanae, 1837), which introduced the Hispanophone world to the notion that Europeans (Vikings) had discovered America five centuries before Christopher Columbus; and Angel de los Ríos' Los Eddas (1856),22 translated from Rosalie de Puget's French work Les Eddas (1846), comprising The Prologue, Gylfaginning and some excerpts of Skáldskaparmál from Snorri's Prose Edda, as well as the poems from the Poetic Edda. Unlike de Puget, de los Ríos did not attempt a verse translation, and his rendering of the Poetic Edda consists in reductive prose paraphrases which preserve little of the original poetic diction. Nevertheless, the reader can see some examples of kennings for poetry in the excerpt from Skáldskaparmál, "Coloquio de Brage con Aeger," which recounts the origin myth of the mead of poetic inspiration:

Hé aquí por qué la poesía se designa con las espresiones siguientes: Sangre de Qvaser, bebida de los enanos, licor de Odraerer, de Bodn, ó de Son, navío de los enanos (pues que este hidromiel los salvó de los arrecifes donde estaban abandonados), hidromiel de Suttung y licor de Nittberg.23
[This is why poetry is designated by the following expressions: blood of Kvasir, drink of the dwarves, liquor of Óðreyrir, of Boðn, or of Són, ship of the dwarves (because this mead saved them from the reefs where they were abandoned), mead of Suttung and liquor of Hnitbjǫrg.]

While de los Ríos' translation of the poetry follows de Puget's text closely, his book also includes original additions in the form of a lengthy introduction and an appendix of "cognate" words, where he strives to emphasize a correspondence between the Old Norse and medieval Spanish languages and literatures. In the introduction, for example, he observes that at the time the "skaldas" of the north were singing of the misfortunes of that Spanish princess, who was also sung of in the Nibelungenlied "the Iliad of Germany," the first strains of the deeds of El Cid begin to be sung in "our language."24 Indeed for de los Ríos, Sigurd's voice is so reminiscent of the eponymous Spanish hero of that poem that he refers to him as "el Cid del Norte," an analogy which is drawn further in his appendix comprising a list of words in the "Old Scandinavian Language" which "more or less disfigured can be found in Spanish."25 Many of these are of dubious etymology, but indeed a number do represent words which share a common Germanic origin, such as "hard" (presumably ON harðr, "hard") for Sp. árdido "brave," a word derived from Frankish *hardjan "harden" or Go. hardus "hard," which occurs in a line from Poema del Cid: "Esto lidiare á tod el mas ardido."26

Borges does not appear ever to have known of the existence of either de Puget's or de los Ríos' translations, and his La alucinación de Gylfi was based on Arthur Gilchrist Brodeur's English translation.27 However, one of these (most probably de Puget's) does seem to have made it into the hands of another Argentinian writer, José Hernández. In an exhaustive study, Emilio Carilla convincingly demonstrates how a whole canto of Hernández's famous epic El Gaucho Martín Fierro, is based almost entirely on extracts from Hávamál "The Sayings of the High One" and other texts from the Poetic Edda.28 Borges was fascinated by Martín Fierro, and wrote about it several times, but never seems to have noticed that one whole section, "XXXII de la Vuelta" better known as "Los consejos de Martín Fierro" or "The Counsels of Martín Fierro," essentially comprised Old Norse wisdom poetry in Spanish guise. In fact, he found this section to be one of the less appealing parts of the epic:

Otro recurso para descuidar el poema, lo ofrecen los proverbios. Esas lástimas —según las califica definitivamente Lugones— han sido consideradas más de una vez, parte sustantiva del libro. Inferir la ética del Martín Fierro, no de los destinos que presenta, sino de los mecánicos dicharachos hereditarios que estorban su decurso, o de las moralidades foráneas que lo epilogan, es una distracción que sólo la reverencia de lo tradicional pudo recomendar. Prefiero ver en esas prédicas, meras verosimilitudes o marcas del estilo directo.29
[Another pretext for neglecting the poem are the proverbs. These poor unfortunates—as Lugones definitively classifies them—have been considered more than once as a substantive part of the book. To deduce the ethos of Martín Fierro, not from the fates and fortunes it features, but rather from the mechanical vulgarities which hamper its development, or from the strange/foreign moralities which sum it up, is a distraction which only a reverence for the traditional could recommend. I prefer to see in this preaching mere truisms or traces of direct speech.]

Another writer we should mention in the context of the reception of Old Norse literature in Spanish is the Peruvian-Bolivian poet Ricardo Jaimes Freyre, whose collection Castalia bárbara (1899), also drew heavily on the mythological and heroic content of the Poetic Edda.30 According to Borges, it was only later in life that he became aware of this Nordic connection in Freyre's work: Cuando yo escribí tantos, y acaso demasiados, poemas sobre mitología escandinava pensé que era el primero que lo hacía en lengua castellana. Luego recordé a Castalia bárbara. "When I wrote so many, and perhaps too many, poems about Scandinavian mythology I thought that I was the first that did so in the Castilian language. Then I remembered Castalia bárbara."31

This overview includes just some of the more significant texts which reflect an interest in the literature and culture of Medieval Scandinavia in Spanish prior to Borges' work, and by no means constitutes a comprehensive bibliography. We can observe certain micro-trends, such as the craze for the exotic north in Golden-Age Spain, or a fashion for the Poetic Edda among nineteenth-century South American poets. However, we cannot talk about a continuous tradition of Old Norse scholarship in Spanish prior to 1932. We can also note certain recurrent tendencies in the reception of this literature in Spanish which mirror historical events, ideological movements and literary trends, especially those linked to the rise and fall of the Spanish Empire. Thus, the Scandinavian peoples were variously depicted in the Hispanophone world as cruel barbarians, gothic cousins, exotic "others," composers of baroque and ridiculous verse, and full circle back to Viking explorers of the new world whose "ethics" can be transposed onto the iconic gaucho of the Argentinian plains, or to the poetics of modernist poetry. Very few of these accounts were born from a direct contact with the Old Norse language; the vast majority are derived second- or third-hand, mostly from French or Latin sources which Spanish writers used, with varying degrees of success, to paint a picture of Septentrional culture and literature to a Spanish-speaking readership. However, with some exceptions, such as Juan Andrés (whose motives appear to have been purely scholarly), it is clear that in the main the chief motivation in the treatment of medieval Scandinavian texts in these works was rather to advance certain political or literary agendas, than to enlighten their audiences.

As we have noted above, Borges appears to have been mostly unaware of this tradition in Spanish. His contact with Scandinavian literature was by way of English and, to a lesser extent, German sources. Nonetheless it is clear that Borges' response to the literature of the medieval North, as exhibited in "Noticia de los Kenningar" and later writings, shared many of the factors which characterized the work of his predecessors.

Borges Writes about Kennings

Una de las más frías aberraciones que las historias literarias registran, son las menciones enigmáticas o kenningar de la poesía de Islandia.32
[One of the most frigid aberrations recorded in literary histories are the enigmatic phrases or kennings of the poetry of Iceland.]

It is clear from this rather sensationalist opening sentence that one of Borges' primary aims in "Noticia de los Kenningar" was to dazzle and entertain his readership. The first phrase used to describe kennings as más frías aberraciones is itself rather enigmatic and could be interpreted as anything from "mechanical anomalies" to "unemotional monstrosities." Either way, this somewhat hyperbolic language sets the tone for the rest of the article in which kennings are described as penosas ecuaciones sintácticas "tortuous syntactic equations," desfallecidas flores retóricas "wilted rhetorical flowers," inexactas menciones "inaccurate phrases," metáfora salvaje "barbarous metaphor," and so on.33 Borges proceeds to analyse some examples to support his perspective starting with the "most insidious," (the kennings are indicated in italics):

El héroe mató al hijo de Mak
hubo tempestad de espadas y alimento de cuervos.

Borges' source text was George Ainslie Hight's The Story of Grettir the Strong: Story of the Eleventh Century, and his translation follows the English text very literally:34

The hero slew the son of Mak;
there was storm of swords and raven's food.
Skuf and Bjarni he also felled;
gladly he bathed his hands in blood.

While this rendering gives the gist of the meaning of the Old Norse text, the form could hardly be further removed from the original stanza, which was composed in the highly complex and prestigious dróttkvætt "court metre." Like most Germanic metres, in dróttkvætt lines had to be linked in pairs by alliteration (indicated below in bold), assonance was required between the penultimate syllable and another syllable in every line (underlined), not to mention other stipulations regarding stress patterns, all of which had to be achieved in six syllables per line.35 It was these strict metrical rules which resulted in the infamously unnatural syntax noted by Juan Andrés, and illustrated in the word-for-word translation below:

The convoluted phrasing makes these texts almost impenetrable to non-experts, so that editors usually include re-ordering of the words into more natural prose as well as explanations of the kennings, and a modern language paraphrase. The source text for Hight's translation, Richard Boer's German edition, gives the prose word order of the stanza as follows:

Hǫlþr lét goldet kapp Máks syne meþ hepne. þá gerþesk sverþa hríþ. Hrafn gat hrátt hold at slíta. Vágs viggríþande varþ enn síþan at víge Skúfs ok Bjarna. Hann bar gjarna greipr at gunne.36
[The hero (man) repaid the son of Mák for his belligerence with success. Then storm of sword > battle arose. The raven got raw meat to tear at. The rider of (the horse of the wave > ship) > seafarer/man afterwards was to slay Skúfr and Bjarni. He eagerly used hands in battle.]

As we can see, compared to the original, in Hight's translation the complexity of the poetic diction has been greatly reduced, although where one kenning has been omitted, another has been introduced. "Raven's food," a reworking of "the raven got raw meat to tear at," is based on the common "beasts of battle" motif whereby wolves and birds of prey feast on the corpses of the slain on the battlefield. In this instance the kenning has been devised by the translator, but many such kennings are attested in the corpus, such as verðr hrafns ("raven's meal" > corpse). The other kenning in this stanza, "storm of sword," is also one of a very common type grounded in the metaphor of battle conceived of as a ferocious tempest in which swords and shields flash like lightning and clash like thunder, as in "blizzard of the light of terror" (where "light of terror" is the sword), or storm where missiles fall like rain or hail, as in "rain of the spear-cloud" (where "cloud of spears" stands for shield). Due to the numerous poetical synonyms for both the base-word (storm/rain/blizzard) and the determinant (sword/spear/shield), as well as the possibility of substituting either by another kenning, these poetical periphrases rarely recur in exactly the same form. However, while a kenning might be ingenious, or even original in the sense that it had not been used before in those precise terms, in their compositions the skalds rarely strayed outside the confines of the conventional poetical language and the established metaphors of their tradition (although, at some point, they must have been original). The ingenuity of the poet did not lie in his or her ability to invent new metaphors, but rather in the facility with which the poet could take a basic statement or image such as "the man killed the man," or "there was battle," and reinvent it in poetical language appropriate to the context and within the appropriate metrical constraints. For the skalds and their audiences, a battle is a storm, or a ship is an animal, functioned as lexicalized metaphors and any kennings based on them would have been immediately recognizable once the verse was deciphered. Thus, Borges' comment that the two metaphors somehow take advantage of the reader (engaña ventajosamente al lector, pp. 202-03), only makes sense in the context of an audience which is ignorant of the norms of skaldic diction.

The next examples Borges analyses occur in one of a pair of stanzas from George Webbe Dasent's The Story of Burnt Njal, placed, as Borges points out, in the "abominable plutonic mouth of Steinvora, mother of Ref the Skald."37 In the verses below, Steinvora, a pagan who opposes the new faith, tells the preaching priest þangbrandr that the destruction of his ship was the work of Thor and the gods:

Compared to the previous example this stanza is much richer in kennings, and of the six, four consist in various expressions of the underlying metaphor a ship is a (large) animal, based on the correspondence between the movement of the ship as it speeds over the waves and a powerful animal running across the land. In skaldic diction the majority of ship-kennings are based on this model where the base-word is most commonly horse, but also reindeer, wolf, bison, or even elephant, and the determinant is a term for the sea (or some other aspect or attributes of the sea or seafaring). In the first example, bisonte ("bison," vísund) is determined by la pradera de la gaviota ("mew-field," mástallr), itself a kenning for sea, because the sea is the "support" or "seat" (stallr) of the seagull (már), giving "bison of the sea" > ship. In the next kenning Dasent (followed by Borges), renders val strandar as "falcon of the strand," although editors now generally prefer to read valr as the mythological name for a horse rather than falcon, thus "horse of the strand" > ship. "Courser of the causeway" (el caballo que corre por arrecifes), also appears to contain a misinterpretation of the original malmfeta varrar, ("metal-pacer (> horse) of oar-pull"). The final kenning in the stanza, "Gylfi's hart," which Borges does not translate, contains an allusion to the mythological sea-king Gylfi, and Gylfa hreinn "reindeer of Gylfi" is again the ship. The message of this stanza is simple: "Thor destroyed the priest's ship, Christ did nothing to protect it." However, the poetic diction (especially in the form of the kennings) evokes a multiplicity of fantastic images of the ship as a bison, horse and reindeer galloping over the waves and being smashed by the mighty giant-killer Thor and the gods, while the Christian priest stands by helplessly, guarding his bell. Although Dasent's interpretation deviates in some details from the original, he is careful to maintain the variety and something of the poetic nature of the kennings. Borges' translation, while faithful to the source text, is flat and prosaic by comparison. As Borges points out, this is partly because the Spanish language is less accommodating than English or German when it comes to translating compound words or kennings.38 Nevertheless, it is hard to believe that the proficient translator and poet is not being deliberately reductive in order to substantiate the point reiterated throughout the essay, that kennings represent some kind of literary cheap trick.

Borges goes on to provide a select list of approximately fifty more examples of kennings, which he prefaces with the following statement:

Predomina el carácter funcional en los kenningar. Definen los objetos por su figura menos que por su empleo. Suelen animar lo que tocan, sin perjuicio de invertir el procedimiento cuando su tema es vivo. Fueron legión y están suficientemente olvidados: hecho que me ha inducido a compilar un índice parcial de sus desfallecidas flores retóricas.
[The functional character of the kennings predominates. They define objects by their usage rather than by their form. They usually animate what they touch, without prejudice to reversing the procedure when their subject is alive. They were legion and are pretty much forgotten; a fact which has induced me to compile a partial catalogue of their wilted rhetorical flowers.]

His observation that kennings can bring an inanimate object to life is an accurate one as we have seen in the stanza above, where a ship can become a horse or a bison. The opposite is also true, as for example in peñasco de los hombros "crag of the shoulders," which defines the head as a "crag" in the "landscape" of the body. However, the idea that kennings predominantly define objects by their usage as opposed to form is not true even of the examples on Borges' list. Numerous kennings are based on a correspondence of shape or other physical aspect between two completely different things, such as "fish of battle" for sword (both are shiny), or "moons of the forehead" for eyes (both are round and bright). Furthermore, a great many kennings allude neither to form nor usage of the referent but rather to certain legends, myths or mythological characters, such as the gold kennings on Borges' list. As Margrét Jónsdóttir has pointed out, in his selection Borges deliberately avoided kennings with mythological content, preferring those based on universal metaphors.39 However, although a few examples on Borges' list could be said to be universally comprehensible, such as bosque de la quijada "forest of the cheek" for "beard" or sacudidor del freno "shaker of the reins" for "horse," the majority are rather based on contextual metaphors which require specific cultural knowledge to be understood. It is these contextual kennings which, when extracted from their cultural and poetic context, seem the most humorous and ridiculous, especially when presented in a simplified form.

Borges' comment that kennings have been relatively forgotten is typical of the cavalier approach which characterizes his writing on Old Norse poetry, and ignores the fact that most kennings survive in poetry preserved in the Icelandic sagas and in the Prose and Poetic Eddas, works which were continuously disseminated through new editions and translations from the nineteenth century onwards. The early twentieth century also saw the publication of several key scholarly works on skaldic poetry and kennings, many of which remain standard today. These include Finnur Jónsson's monumental Danish edition of skaldic poetry, Den norsk-islandske skjaldedigtning, and his Old Norse-Danish dictionary of poetic language.40 Another landmark publication was Rudolf Meissner's Die Kenningar der Skalden, a treatise and systematic inventory of kennings categorised according to their referents, loosely modelled on Snorri's schema in Skáldskaparmál.41 In his 1933 booklet Las Kenningar, Borges explicitly references Snorri's Edda Prosaica as a source for his list, and Brodeur's English and Neckel and Niedner's German translations of this work were included in the bibliography.42 However, none of these texts is mentioned in "Noticia de los Kenningar," although the comment that kennings "were studied in lists by the skalds' apprentices," is surely a direct allusion to Snorri's description of Skaldskaparmál as a guide for young poets.43

The first source Borges credits for his list of kennings is Dasent's The Story of Burnt Njal. As we have seen in our analysis of the stanza from this work above, Dasent's translations of the kennings are relatively accurate, although in his attempt to maintain alliteration, metre and rhyme, some of the poetic renderings reflect the spirit rather than the letter of the original. As Borges translated these forms uncritically into Spanish, the resulting list includes many examples which are simplified, reduced or half-invented, and hence do not provide a faithful representation of skaldic diction. Thus "Lord of rings, a periphrasis for a chief, that is, Mord." becomes señor de anillos, which according to Borges, is a kenning for príncipe "prince."44 In fact, Dasent's phrase misinterprets the original kenning hrings álmr "elm of sword" > warrior. While riacho de los lobos for "rill of wolf — stream of blood" is a very precise rendering of the original vargs bekkr "wolf's brook," the gold kennings fuego del Rhin and fuego del mar for "Rhine's fire" and "Ocean's fire" sound monotonous compared to hyrs haf "embers of sea" and Rínar leygr "flame of Rhine."45 Ciervo del mar, although not as majestic as "sea-stag," does not stray too far from the original concept of elgr unnar "elk of the waves" > ship, but patín de agua and patín de viking, although faithful to Dasent's "water-skates" and "viking's snow-shoe," sound almost comical when taken out of their original context as part of the extended kennings Ekkils vallar andrs Ullr "god of the ski of Ekkill's land" > seafarer/man and skíð Atals grundar "ski of Atli's land" > ship.46 A similarly unfortunate term is baño del cisne for "sea" taken from Dasent's "swanbath's beams — periphrasis for gold," which in the original is svanteigs eldr "fire of swan-field." Ditto roedor de yelmos for "helmgnawer — the sword that bites helmets," where Dasent seems to take hjalmgagarr "helmet-hound" rather literally.47 The underlying metaphor is indeed based on the comparison of a dog's and a sword's biting action, but obviously a sword 'bites' a helmet in a slicing or chopping motion, it does not chew; furthermore, Borges cannot but have been aware that for a Spanish speaker the primary meaning of roedor is "rodent."

While the above samples were translated directly from Dasent's periphrases, a smaller number appear to be of Borges' own invention, although derived from the poetic texts. For example,"'Forge which foams with song,' the poet's head, in which songs are forged, and gush forth like foaming mead" is reworked as fragua del canto "forge of song" with the referent la cabeza del skald "the skald's head."48 While the image of poetic inspiration gushing forth in the form of the mythical mead is a common one in skaldic diction, the location of poetic inspiration (as well as emotion and thought) was the chest, not the head, and in fact the original phrase, bœnar smiðju, denotes the chest as the "smithy of prayers" not "song." More instances of poetic licence can be seen in two of Borges' kennings for axe: querido alimentador de los lobos "wolf's beloved feeder," and ogra de la batalla "ogress of war." The first in the original refers to a halberd, atgeirr, not an axe, which gives the wolf his fill, úlfa fylli, and from which Gunnar vows never to part, and the second is merely the proper name of Skarphedinn's axe, although axe-kennings of the type "trollwoman of battle" do occur elsewhere.49 Finally, custodia de las joyas "guardian of the jewels" may be derived from the phrase "He that hoardeth ocean's fire," although as a kenning for "prince" it does not make sense;50 in skaldic poetry when lords are defined in terms of gold, it is always with regard to their propensity to distribute treasure, not to guard it.

Borges' second named source was Wilhelm Ranisch's Eddalieder, a handbook in German comprising an introduction to Old Norse grammar, a bilingual translation of extracts from five poems in the Poetic Edda, as well as detailed explanatory notes to the texts.51 In Ranisch's edition Borges comes directly in contact with the Old Norse language, and, although he obviously relied heavily on the German notes, the nine examples selected from this text (eight from Atlakviða, one from Brot at Sigurðarkviðu) are generally much closer to the original than those taken from The Story of Burnt Njal. Thus árbol de asiento accurately reflects sess-meiðr (Sitzbaum, p. 124: "seat-beam/tree > bench); sacudidor del freno is a literal translation of skókr bituls (der Erschütterer des Gebisses, p. 131: "shaker of the bridle/bit" > horse); distribuidor de las espadas is sverða deilir (Spender der Schwerter, p. 134: "giver/distributor of swords" > lord). Two sword-kennings also appear: espina de la batalla from rógþorn (Kampfdorn, p. 130: "thorn of the battle"), and rama de las heridas "branch of the wounds" for ben-vöndr (Wundengerte, "wound-stick"), this last from Brot af Sigurðarkviðu (p. 113). Inaccuracies occur, however, in the warrior-kenning árbol del yelmo "tree of the helmet" translating Helmbaum ("helmet-tree," p. 128), for kumblameiðr — all other editions have kumblasmið "wound-smith."52 The translations of other examples from the source are less than exact. In another warrior-kenning, dólg-rögnir "battle-Rǫgnir," which Ranisch correctly explains as "battle-Odin," "battle-god" because Rǫgnir is one of the names of Odin (p. 131), Borges removes the mythological allusion, reducing the kenning to señor de la pelea "lord of the fight." Borges renders Rógmalmr "strife-metal," a kenning for gold which refers to the role of treasure as the cause of discord in the legend of the Nibelungs, as bronce de las discordias "bronze of strife," based on Ranisch's explanation: "Kampferz, Metall das Kämpfeveranlaßt," (p. 129); although indeed Erz can mean "bronze," here the primary meaning "ore" is obviously intended: "battle-ore, metal that causes battle." Finally, one of Borges' battle-kennings, encuentro de las fuentes "meeting of the fountains/springs," which makes absolutely no sense in the context of skaldic diction, is explicable only as a misunderstanding of Ranisch's translation of apaldr brynþings "Apfelbaum (Baum) der Brünnenversammlung" (p. 12), "apple-tree of byrnie-assembly" > warrior, where Borges reads Brünnen in its more usual sense of "well/fountain," although Ranisch here is referring to "mailcoat" or "byrnie," whose assembly is the battle.

The remaining examples on the list must be attributed to Borges' third source, Raimundo Lida, an emerging Spanish philologist and managing editor of Sur, to whose "generosity of scholarship" Borges owes "twenty or so" examples (p. 204). Lida's authority for these examples is not entirely clear. Although they could well be taken from the Snorra Edda, a number of the interpretations point instead to Meissner's catalogue "Das System der Kenningar" in his book, where kennings are explained and listed according to referent, although unlike the Snorra Edda they are not quoted within their poetic context.53 While some of these reflect authentic kennings in Old Norse poetry, the majority are translations of the archetypes (or original creations based on those archetypes) outlined by Snorri and Meissner. The kennings which have an equivalent, or near equivalent, form in the original are: bosque de la quijada "forest of the jaw" for kinnskógr ("cheek-forest" > beard), peñasco de los hombros for klett herða ("crag of the shoulders" > head), juego de los filos for eggja leikr ("game of blades" > battle), perro de cadaveres "dog of corpses" for hrægagarr ("corpse-hound" > sword); and camino de las velas "path of sails" which is a loose translation of slóðir hafskíða ("paths of ocean-skis" > sea), teñidor de espadas "colourer of swords" for rjóðr branda ("reddener of swords" > warrior), and ruta de la ballena "route of the whale" for slóðar vagna ("track of the killer-whale" > sea).

Many of the terms listed are not derived directly from attested kennings: borrachera de las espadas "drunken-binge of swords" for battle, which may be a misunderstanding for the "assembly/meeting of weapons" type. Gansos de la batalla "geese of battle" for flechas "arrows," probably reflects arrow kennings such as bryngagl "gosling of mailcoat" and bengagl "wound-gosling," which are quoted in Meissner as examples of how weapons are compared to birds such as geese (p. 145). When determined by "battle" as Borges does here, the kenning rather suggests a bird from the 'beasts of battle' motif such as a raven or eagle. Pescado de la batalla "fish/catch of battle" for sword, obviously alludes to the metaphor a sword is a fish which underlies a great number of sword kennings, and is based on a comparison between the "slim, narrow shape" of this animal and the blade of the sword.54 Borges' choice here of pescado "catch" as opposed to the live animal pez, would suggest that he does not understand the premise of the metaphor. Borges lists these kennings for blood: marea de la matanza "tide of slaughter," rocío del muerto "dew of the dead one," sudor de la guerra "sweat of war," which appear to be approximations of examples quoted in Meissner under the referent "blood," such as hræflóð "corpse-tide," valdǫgg "slain-dew," sveita sœfis "sweat of sword."55 Finally, agua de la espada "water of the sword" would seem to be derived from Meissner's observation that blood is determined through weapons and the base-word is any term for water.

Other specimens on Borges' list which reflect generic kenning types rather than attested phrases include: fuego de las olas del mar referring to the concept of gold as "fire of the sea" (or any term related to water), but while "fire of the waves" does reflect an authentic kenning, del mar "of the sea" would never occur, as wave already functions as a metonym for sea. Tesoro del dragón alludes to the legend of Fafnir the dragon who lay on a bed of gold, and a great many kennings are based on this model. Distribuidor de los tesoros "giver of the treasures" is a common type for "lord" príncipe; eyes, on the other hand, are never referred to as joyas de la cabeza "jewels of the head," but rather as "stones" of the eyelash or brow or another part of the face. Castillo del cuerpo "castle of the body" does not appear in the corpus, although borg "castle/fortress" does occur in some kennings for head, e.g. borg breiðra brúna "fortress of broad brows" and borgir heila "fortresses of brains."56 As we have seen above, in the ship is a horse type of kennings, it is crucial that the determinant refers to some aspect of the sea or sea-faring; thus apart from the fact that it is not authentic, caballo del pirata "horse of the pirate" does not work because this phrase could literally be interpreted as a pirate's horse. Similarly, campo del vikingo "field of the viking" for sea is too ambiguous, as a viking's domain is not only the sea. Of the remaining four kennings, two, viaje de las olas "journey of the waves" for la inundación "flood" and yelmo de la noche "helmet of the night" for la sombra "shadow," do not appear anywhere in extant Old Norse poetry as kennings or referents. Tesoro del pecho "treasure of the chest" for pensamiento "thought" may be based on a confusion regarding heart kennings of the type stone of the chest but kennings for thought are always based on a quite different model; e.g. "wind of troll-woman." Finally, yelmo del aire "helmet of the air" is not a kenning for la neblina "mist," although there is a kenning lopthjálmr ("air-helmet" > sky/heaven) which occurs in an extended kenning for the Christian God, "King of the air-helmet."

As this survey of Borges' examples and their sources shows, although the author obviously had a good understanding of the basic theory of kennings, his ability to appreciate their polysemic and metonymic character, as well as the intricacies of the dróttkvætt metre was limited by his lack of knowledge of the Old Norse language and its poetical tradition. This is apparent in some of the more risible errors ("meeting of the fountains/wells" > battle) and in comments such as: su misma bastedad - peces de la batalla: espadas - puede responder a un antiguo humour, a chascos de hiperbóreso hombrones "their very crudeness - fish of battle: swords - may reflect an ancient humour, the jokes of hyperborean he-men."57 However, it is evident that the consistent tendency to reduce the complexity and variety of kennings is not only the result of a lack of expertise, but reflects a very deliberate policy to portray them as inherently primitive in nature. Borges does not explain his selection criteria for the specific examples on his list, but although the impression that he gives is that it is somewhat eclectic, the omissions of some major categories such as kennings for women and kennings with mythological content would appear to be calculated. As we have seen above, the examples in his lists which contain mythological allusions in the original have been modified.

Conclusion

Borges' principal intention in writing "Noticia de los Kenningar" was most probably to compose an entertaining short piece for the readers of Sur introducing them to the phenomenon of the kenning, an exotic anomaly of literary history which he felt had been "sufficiently forgotten." By featuring the most absurd-sounding alongside the most mundane of these rhetorical devices, he also aimed to demystify what he felt to be something in between a joke and a trick. He considered the complexity of these figures an illusion and, once unravelled, the "partial knots" that made up extended kennings revealed nothing new to him, and only left him with a sense of disappointment.. Like his antecedents, Borges' first foray into the literature of the North reflected a fascination with the exotic, an "othering" of the medieval Scandinavian barbarian poets whose "enormous ineptitude amazed the red men of the volcanic deserts and the fjords, just like the strong beer and the fights of rearing horses," and an unashamed manipulation of the sources to suit his own ends.58 Undeterred by his lack of expertise and his reliance on secondhand sources, he was outspoken in his condemnation of kennings, although, even in this early piece, one can detect a grudging admiration behind the façade of his strident reproaches.

That Borges did not expect a wide audience for "Noticia de los Kenningar" or its later incarnation Las Kenningar is clear from his comment in the preface to the Historia de la Eternidad: "the unlikely or practically nonexistent reader of Las Kenningar can investigate the handbook Literaturas germánicas medievales which I wrote with María Esther Vásquez."59 This view was echoed many years later by Enrique Bérnardez who pointed out that neither publication served to "palliate the utter lack of knowledge of Old Norse literature in Spanish," although he was keen to emphasize Borges' pioneering role.60 In fact, it was Bernárdez who should rightly be credited with the beginning of a professional tradition of Old Norse scholarship in Spanish in the early 1980s, beginning with his article on the translation of kennings, followed by his Sagas Islandesas, Snorri Sturluson: Textos mitológicos de las Eddas, Saga de Egil Skallagrimmson and Saga de Nial.61 In one of his last literary acts, Jorge Luis Borges included Bernárdez' translation of the Saga de Egil Skallagrimmson, which contains the poetic corpus of its eponymous protagonist and arguably the best of the Icelandic skalds, in his Biblioteca personal, and by so doing he probably did more to promote skaldic verse among the general Spanish-speaking public than any of his other works.62

University of Castilla — La Mancha