• 1 This article is based on the keynote lecture I gave at the II^e congrès de l’Association p (...)
• 2 On the relationship between manuscript and print in Iceland, and the reasons for the (...)

1Iceland – nowadays probably best known for its glaciers and volcanos, its financial crisis and its lively music scene – also has a long history as a hotbed of literary activity, starting with the Eddas and sagas of the medieval period and extending to the crime fiction of the present day. For most of Iceland’s history, this literature circulated predominantly in handwritten form, as manuscripts, initially on vellum, later on paper. Indeed, for certain types of literature at least, manuscript transmission remained the norm in Iceland well into the modern era, long after the arrival of the first printing press in the mid-sixteenth century2.

• 3 A preliminary list of Icelandic translations and adaptations from French is fou (...)
• 4 The selection of texts treated here has not been made on any scientific basis; rather I h (...)

2The range of types of literature preserved is vast, from sombre works of religious devotion to romping romances and tales of derring-do. Although Latin learning was certainly not unknown in Iceland, the bulk of this literature was in Icelandic, and was the product of the indigenous Icelandic imagination. But many works were translated, too, both from Latin and the various European vernaculars, not least French3. The Icelanders’ appetite for good stories was voracious, and it is clear that pretty much anything that was known on the continent was likely eventually to find its way to Iceland. In the present essay, I will examine a selection of the French literary works that survive in Icelandic translations, primarily handwritten but some also in print, the routes by which they travelled there and their later fates in Iceland once they were there4.

3The culture of the book came to the north with Christianity, and among the first things to be translated were works which served the needs of the Church, such as saints’ lives, homilies and so on, the bulk of which were translated from Latin. As the native literature began to emerge in the second half of the twelfth century, so too did translations of secular works begin to appear, both from Latin and, increasingly as time went on, vernacular languages.

• 5 ‘Old Norse-Icelandic’ denotes the common language of Norway and Iceland (and th (...)
• 6 The same designation is applied to slightly younger indigenous Icelandic works which were (...)
• 7 On these translated romances see e.g. Maclean, 1972, Kalinke, 1985, Barnes, 1993, Thorleifsdóttir, (...)

4About twenty Old Norse-Icelandic5 translations of European courtly romance, most of them from French, survive from the medieval period. These translations, known as riddarasögur (lit. sagas of knights)6, are for the most part thought to have been made in Norway, but nearly all survive chiefly or exclusively in Icelandic copies7.

• 9 Schach, 1975. The suggestion that these translations were the work of one and the same ma (...)

7Owing to his name – and his obvious familiarity with Norman French – it has been suggested that “Bróðir Róbert” may have himself been a Norman, who had however become fluent enough in Old Norse to be able to translate into that language. Elis saga ok Rósamundu, a translation of the French romance Elie de Saint Gille, is similarly attributed to a “Róbert ábóti” (Abbot Robert), presumably the same man now further advanced in his ecclesiastical career9.

• 10 For a survey of this, see Barnes, 2011.
• 11 Rubow, 1936, p. 21.
• 12 Schlauch, 1934, p. 151.
• 13 Einarsson, 1961, p. 163.
• 14 Finlay, 2004.

8Tristrams saga has received a good deal of scholarly attention, both as it is the most complete representative of Thomas’s Tristan extant and also owing to its assumed position as the earliest of the Old Norse translations of French romances10. It is also thought to have exerted considerable influence on Old Norse-Icelandic narrative literature. The theory has had few supporters, but it was even suggested, by the Danish scholar Paul Rubow in the early part of the last century, that it was Brother Robert’s Tristrams saga that kick-started Old Norse-Icelandic saga writing in general: “Der burde et Sted oprejses ham en Statue, thi han er efter al Sandsynlighed Grundlægger af den oldnordiske Underholdningslitteratur i Prose” (There ought to be erected a statue to him somewhere, for he is in all probability the founding father of Old Norse prose fiction)11. Motifs ostensibly deriving from Tristrams saga have been identified in Rémundar saga keisarasonar, Haralds saga Hringsbana – which Margaret Schlauch thought “seem[ed] to have been constructed as a deliberate reply to the French romance”12 – Jarlmanns saga ok Hermanns and several other romances, and also in “Spesar þáttr”, the final section of Grettis saga. Similarly, alleged influence from the Tristan story has been detected in the four so-called skáldasögur, or “poets’ sagas”, Kormáks saga in particular, to such a degree that Bjarni Einarsson, for example, could refer to Kormákr as “Tristan Íslands” (the Icelandic Tristan)13. Recently, however, a number of scholars, Alison Finlay foremost among them14, have rejected the idea that there are any narrative correspondences between Tristrams saga and the poets’ sagas, and the whole question of the saga’s position and influence may be said to be ripe for reassessment.

• 15 The part of the saga dealing exclusively with Gawain was at some point separated from the (...)
• 16 See e.g. Leach, 1921, p. 149-157.
• 17 The scholarly literature on these translations is extensive; for surveys see Bo (...)

9From the same period, the mid-thirteenth century, are translations of three of the five romances attributed to Chrétien de Troyes (late twelfth century): Erex saga (Érec et Énide), Ívens saga (Yvain, le Chevalier au Lion) and Parcevals saga, with Valvens þáttur15 (Perceval, le Conte du Graal). The translations are all thought to date from the mid-thirteenth century, and all, like Tristrams saga, are believed to have been made at the behest of King Hákon Hákonarson16; in the case of Ívens saga, this is explicitly stated in the epilogue. As with most material of this sort, texts of these translations survive only in copies made in Iceland. Ívens saga and Parcevals saga are both found in fifteenth-century vellums, but these are defective, so the only complete texts are younger, and Erex saga survives only in paper copies from the seventeenth century and later17.

• 18 Kalinke, 2011, p. 32-37.

10All of these translations suffer from textual attrition vis-à-vis their French sources, on average a reduction of 40-60%. It is difficult to know at what stage this attrition took place, i.e. whether at the time of translation or later in the course of transmission, though, on the whole, the latter seems more likely18. Typically, descriptions of equipment, clothes and horses are reduced or removed, as are expressions of inner conflict by the protagonists on the nature of love and honour and any first person commentary by the narrator – in short, much of what may be said to characterise Chrétien’s romances.

• 19 Discussions of Möttuls saga can be found in Kalinke, 1981, Driscoll, 1991, Larrington, 20 (...)

11Apparently also from the time of King Hákon Hákonarson is Möttuls saga (The saga of the mantel), a translation of La mantel mautaillé (also known as Le Lai du cort mantel), a somewhat frivolous tale of a chastity testing mantle sent to King Arthur’s court, dating from the end of the twelfth century or the beginning of the thirteenth (there are vellum manuscripts from the fourteenth and fifteenth centuries)19.

• 20 Kalinke, 1987, p. lviii-lxi. The suggestion that Möttuls saga was the first romance (...)

12It has been suggested that Möttuls saga may have been, after Tristrams saga ok Ísöndar, the earliest of the Arthurian romances to be translated into Old Norse-Icelandic. Among the evidence adduced for this is the presence in the saga of a long introductory section describing King Arthur in the most glorious terms. This section has no parallel in the French original – and indeed would have been quite out of place there – but, it has been argued, would have been necessary to bring a Norwegian audience unfamiliar with the figure of King Arthur up to speed. The misnaming for example of Parceval as Paternas and Yvain/Íven as Meon in the saga has also been cited as evidence for an early date20.

• 21 On rímur in general see Hughes, 2005, and references there.
• 22 Skikkjurímur are discussed in Driscoll, 1991, Larrington, 2011, p. 90-93.

14In Iceland, Möttuls saga, in turn, became the basis for Skikkjurímur (Mantle verses), composed probably in the fifteenth century, the only Icelandic rímur to deal with an Arthurian theme (fig. 2). Rímur – long metrical romances in complex metres making use of both alliteration and end-rhyme – were almost invariably based on pre-existing prose narratives21. While most rímur follow their sources quite closely, even slavishly, Skikkjurímur are unusual in that, although the basic plot of Möttuls saga remains unchanged, there are both lengthy additions and significant omissions. The effect of these is to increase even further the burlesque element of the story, at the expense of any serious critique of courtoisie which underlies the humour of La mantel mautaillé22.

• 23 Reichert, 1986.

15A reference to Arthur’s kringlótt sess, “round seat” – not mentioned in Möttuls saga – indicates that the poet had some notion of the existence of the famous round table, first described in Wace’s Roman de Brut (1155), otherwise scarcely mentioned in early Icelandic sources23.

• 24 Leach, 1921, p. 184.
• 25 Schach, 1957-1961, p. 118.
• 26 Schach, 1964, p. 281; this view was first put forward in Schach, 1960, and pursued in Sch (...)
• 27 Kalinke, 1981, p. 199-202.

17This saga has also been the object of a good deal of scholarly attention, much of it focusing on its relation to its apparent source, the older Tristrams saga. Henry Goddard Leach, writing in 1921, thought it to be a vulgarisation, written for a less refined audience: “a boorish account of Tristram’s noble passion”24. Paul Schach initially deemed it to have been “based on an imperfect recollection of the longer saga, which has been further distorted by the addition of names and situations from other sources”25, but later decided, “upon closer scrutiny”, that it was “a deliberate caricature of the translated romance”26; Marianne Kalinke went even further, seeing it as a parody not just of Tristrams saga but of Arthurian romance in general27.

• 28 See in particular Barnes, 1999, Kramarz-Bein, 2000, Francini, 2005.

18I tend, I confess, to be rather sceptical of attempts to see irony everywhere, or the assumption that any two works dealing with vaguely similar subjects must be directly related to each other, and only to each other, ideally by one being a parody of the other. When I first read Saga af Tristram ok Ísodd it struck me as being a fairly typical late medieval Icelandic romance, no more distorted, exaggerated or vulgar than any other. In the last two decades or so a number of scholars28 have argued for the necessity of viewing the younger saga in the literary and cultural context of late mediaeval Iceland, resulting in a far more nuanced view of this highly interesting work.

• 29 The standard work on Icelandic ballads is Ólason, 1982.
• 30 On the Icelandic ballad of Tristram see Driscoll, 2011, p. 169-71 and references there.

19The story of Tristan also gave rise to an Icelandic ballad. Although the ballad genre was never as popular in Iceland as it was in mainland Scandinavia or the British Isles, some fine examples have been preserved29, perhaps none finer than Tristrams kvæði, thought to have been composed in the fifteenth century. Unlike many of the Icelandic ballads, Tristrams kvæði has no parallels among the ballads of other traditions and there appears to be no reason to see it as anything other than an original Icelandic composition. Its simple refrain, “Þeim var ekki skapað nema (að) skilja” (They were fated only to be parted), seems to capture so beautifully the pathos of the Tristan story30.

• 31 This little known but highly interesting manifestation of the Tristan story is discussed (...)

20Several centuries later, a Danish sentimental novel loosely based on the story of Tristan, En Tragœdisk Historie om den ædle og tappre Tistrand, Hertugens Søn af Borgundien, og den skiønne Jndiana, den store Mogul Keyserens Datter af Indien (A tragic story of the noble and valiant Tistrand, son of the duke of Burgundy, and the fair Indiana, daughter of the great mogul, emperor of India), first published in Christiania (present-day Oslo) in 1775 and reprinted at least nineteen times over the next century, also found its way to Iceland, where it was translated into Icelandic six times, both into prose and verse. In this version of the story, a great many changes have been made: the scene of the action has shifted from Cornwall, Ireland and Brittany to Spain, France and India, and, rather more fundamentally, the two protagonists, Tistrand and Indiana, though fated forever to love only each other, choose not to enter into an illicit sexual liaison. Instead, they agree to limit themselves to simple expressions of affection: Tistrand may kiss her hand, while she will allow herself to stroke him on the cheek31.

• 32 See Kjær, 1996, Driscoll, 2011, p. 188-191.

21By far the best known of the Icelandic treatments of this story are the Rímur af Tístrani og Indíönu by Sigurður Breiðfjörð (1798–1846), the pre-eminent rímur-poet of the nineteenth century. Breiðfjörð composed the rímur in the autumn of 1828 while staying at Helgafell on Snæfellsnes with his uncle, the clergyman Jóhann Bjarnason, who asked his nephew to compose a set of rímur for the entertainment of his household. Finding a copy of the Danish book on his uncle’s bookshelf, he decided to use it as the basis for his rímur. Rímur, as was mentioned, are almost invariably based on previously existing prose sources, which the poets tend, by and large, to follow quite closely. Breiðfjörð was rather free with his source, however, skipping over some sections and amplifying others as he saw fit. Not surprisingly, he tended to dwell on the types of scenes generally favoured by rímur-poets: descriptions of sea voyages, feelings of passion and, not least, battles32.

• 33 Driscoll, 2011, p. 189-190.

22Breiðfjörð’s Rímur af Tístrani og Indíönu were printed in 1831 – the first of his rímur to appear in print – but are also found in seven manuscripts, including an autograph (which in fact post-dates the printed text). They are nowadays chiefly known because of an extremely negative review of them written by the naturalist and romantic poet Jónas Hallgrímsson (1807-1845) in the third number of the journal Fjölnir (1837). Jónas takes issue with both the substance and style of Tístrans rímur, but his attack is directed not so much at them in particular as at the entire rímur genre, which he dismisses as simply bad poetry. Although Jónas’s review is generally credited with having dealt the rímur their death blow, their popularity continued unabated throughout the nineteenth century33.

• 34 Traditionally, most Icelanders do not have surnames, but rather a patronym, ending (...)
• 35 On Magnús and his scribal activity see Driscoll, 2012.

23The youngest of the Icelandic prose versions of the story, called simply Sagan af Tístran og Indíönu, is found in four paper manuscripts, the earliest from 1875, the latest from 1913, all written by the same man, Magnús Jónsson í Tjaldanesi34 (1835-1922). Magnús, an ordinary farmer with no formal education, developed early on an interest in collecting and copying manuscripts, and copies, generally more than one, of nearly two hundred sagas are preserved in his hand35. Unlike the other prose versions, Magnús’s Sagan af Tístran og Indíönu is not derived directly from the Danish novel, but is in fact a prose retelling of Sigurður Breiðfjörð’s rímur. The phenomenon of recasting in prose rímur which were themselves based on prose sagas was commonplace, and may perhaps be compared to the relationship between, say, the books and films dealing with Harry Potter. Both versions tell essentially – but not necessarily entirely – the same story, only the medium is different.

• 36 For more information on Sigurður Breiðfjörð and Núma rímur, see the introductio (...)

25Sigurður Breiðfjörð composed another set of rímur several years later based on material which also has its roots in France. These are his Rímur af Núma Pompilssyni, based on the novel Numa Pompilius, second roi de Rome by Jean-Pierre Claris de Florian (1755-1794), published in Paris in 1786. Scarcely known today, Florian was a popular author of novels, comedies and, in particular, fables, which were translated into all the languages of Europe. His historical novel Numa Pompilius appeared in Danish translation in 1792 as Numa Pompilius, Roms anden Konge. Núma rímur, composed in 1833, are based on this Danish translation. Breiðfjörð had studied cooperage in Copenhagen and worked as a cooper in Greenland for a time. It was while in Greenland, suffering from scurvy, that he composed his Núma rímur, which are widely regarded as the best of Breiðfjörð’s rímur36.

• 37 26 sets of rímur by Sigurður Breiðfjörð have survived, of which ten were printed during h (...)

27Núma rímur are found in an autograph manuscript, Reykjavík, National and University Library of Iceland, ÍB 304 4to (fig. 3), which was used as the basis for the first printed edition, which appeared in 1835. The rímur have been printed a further three times, in 1903, 1937 and 1963, something of an achievement even by Breiðfjörð’s standards37.

• 38 Sigurjónsson, 1937, p. xxiii.

29Unlike the prose Sagan af Tístran og Indíönu mentioned above, this version is not a “prosification” of Sigurður Breiðfjörð’s rímur, but a translation of the Danish translation of Florian’s novel. It must, however, post-date the composition of the rímur, certainly if it is true that the translation is by Kristján Ívarsson, who was born in 1830 and would have been only three years old when the rímur were composed. The impetus for translating the saga presumably came from the great popularity of the rímur38.

• 39 On Hallgrímur Scheving see Guðmundsson, 1969.

31“Sélico” was translated into Icelandic by Hallgrímur Scheving (1781-1861), teacher and headmaster at the Latin school at Bessastaðir (from 1846 in Reykjavík), as “Selikó (Afrikanisk frásaga)”, and printed in the second volume of Margvíslegt gaman og alvara (Assorted jocularity and seriousness), which appeared in 1818, twenty years after the first volume, which had been published in 1798 (fig. 5). Hallgrímur Scheving was one of the most learned Icelanders of his day, an inspiring teacher, a great promoter of the Icelandic language and one of the precursors of the Icelandic national-romantic movement of the nineteenth century39. It is not clear why he should have chosen precisely this story to translate, but presumably he found it edifying and thought it would appeal to his countrymen. If so, he was right.

• 40 There has been no full study of these translations, but they are treated br (...)

35Although not French as such, the collection of oriental tales translated and adapted by Jean-Antoine Galland (1646-1715) as Les Mille et une nuits made this wealth of material available to the European reading public through the medium of French. First published in twelve volumes in Paris between the years 1704 and 1711, Les Mille et une nuits was quickly translated into most other European languages. Although not initially printed, translations of the collection also found their way to Iceland, circulating in manuscript. There are in fact three separate Icelandic translations of all or part of the collection, all based on the Danish translation of 174640.

• 41 On this material see Steingrímsson, 1980.

38A similar collection of Persian tales was translated by François Pétis de la Croix (1653-1713) and published in five volumes as Les Mille et un jours, contes persans in 1710-1712. Although less well known today than the 1001 Nights it was highly popular in its day – one of its tales, the fourth one, that of Calaf, served as the source for Puccini’s opera Turandot. Given its popularity, the collection also inevitably found its way to Iceland, where it was translated into Icelandic in whole or in part five times41. The first of these, made sometime before 1770, was from the English translation of Ambrose Phillips, published in 1714; although only found in a manuscript from 1840, this translation must have been made before 1770, as it served as the basis for a set of Rímur af Refsímu by Páll Sveinsson, which were composed in that year and are based on the nineteenth and final tale in the collection, which tells of the virtuous Repsima, falsely accused of adultery by her brother-in-law, whose advances she had spurned. The other four translations were all from the Danish translation of 1759. The five translations are preserved in fourteen manuscripts altogether. The last of them, which is only of the fourth tale, the source of Turandot, was made by Steingrímur Thorsteinsson, who had also translated the entire 1001 Nights collection. This tale, uniquely among these translations, also appeared in print – in fact twice, first in 1933 and again in 1947.

• 42 The Icelandic translation of Zadig is dealt with in Driscoll, 1997, p. 117- (...)

39The vogue for these contes orientaux had a profound effect on eighteenth-century Europe. Among those influenced by them was François-Marie Arouet, better known as Voltaire (1694-1778), whose Zadig, ou la destinée, histoire orientale, borrows much from them. First published in Paris in 1748 and quickly translated into all the major languages of Europe, Zadig is found in an Icelandic translation by the clergyman and poet Jón Oddsson Hjaltalín (1749-1835) in Reykjavík, National and University Library of Iceland, Lbs 893 8vo, a miscellany from around the end of the eighteenth century (fig. 8). The translation is based on the Danish translation of Frederik Christian Eilschov (1725-1750), Zadig eller Skæbnen: En østerlandsk Historie, published in Copenhagen in 1750 (the first of two Danish translations, the other, by Jens Schelderup Sneedorff, appearing in 1759). It is a reasonably faithful, if slightly abridged, rendering of the Danish text (which is itself somewhat simplified vis-à-vis the French original). There are a couple of significant changes, however: a long passage in chapter VII in which the growing passion felt by Zadig and the Queen for each other is greatly reduced, and a significant omission is found in chapter XI, where an animated debate on the various religions practised by a group of people assembled at the market town of Balzora – which ends with Zadig explaining that whatever they may worship separately they all recognise a supreme being and are therefore all of one mind – is reduced to a few sentences, and the question of religion avoided entirely. Presumably the reason for this omission is that the translator, a minister in the Lutheran church, disagreed with its message, that Christianity is no better than any other religion. Given that this is one of basic tenets of deism, a philosophy espoused by Voltaire, this is indeed a significant change. Ironically, this changes the nature of Voltaire’s work, removing, as it were, the philosophie from his conte philosophique and bringing it more into line with the contes orientaux of which Zadig is in some ways a pastiche42.

• 43 Pétur Gunnarsson’s translation of Du côté de chez Swann (Í leit að glötuðum tíma: leiðin (...)

43It is striking how much reached Iceland, demonstrating that despite its geographical location on the edge of the inhabitable world, Iceland was far from isolated. It is also striking how quickly things reached Iceland, often within a generation. Voltaire’s Zadig first appeared in French in 1748; it was translated into Danish just two years later, and into Icelandic by the end of the century. And Florian’s story “Sélico” appeared in print in Iceland just 26 years after its original publication in France. The wait has been longer for Proust’s À la recherche du temps perdu – 84 years (and that’s just the first volume)43.