Sagareading in an Icelandic Farmer's House.
H.Aug.G. Schiott, (1823-1895).
National Museum of Iceland.



EINAR HAUGEN, (1906-1994)

Professor emeritus of Scandinavian Studies, Harvard University.

(Originally published in "Proceedings of the Eighth Viking Congress, Arhus. 24-31 August 1977". Edited by Hans Bekker-Nielsen, Peter Foote, Olaf Olsen. Odense University Press. 1981)

As the only representative of Vinland who has been admitted to this distinguished company, I feel a certain obligation to comment on the location of Vinland. The Vinland problem is a happy one, since it is one on which archeologists and philologists can fruitfully cooperate. Others have contributed: botanists and zoologists, astronomers and explorers, historians and journalists. But primarily it is a problem of literary texts and their philological interpretation, which only the archeological excavations can definitely confirm or disconfirm. The remarkable discoveries made by Anne-Stine and Helge Ingstad on the northern tip of Newfoundland since 1960 have put an entirely new face on the problem, which students of the sagas have to take into account.

The most I can hope to do in this brief paper is to sketch an approach and hint at a possible conclusion. The needed research has not been done, in spite of the mountain of writings on the Vinland sagas, and I have not had the time, the energy, or the resources to do so. But I have re-read the sagas in the light of some of the best research done by such eminent Viking scholars as Sven B. F. Jansson and Jón Jóhannesson.(2)

The crux of the problem lies in the relation between the two widely deviant sagas on the one hand and their common relationship to possible written and oral sources on the other. On this point one can adopt either a maximum or a minimum approach. The maximum approach assumes that everything which is not contrary to common sense and historical probability is true, and that one can get a satisfactory account by harmonizing the two major versions and smoothing out their contradictions. I confess to having once tried this approach in a charmingly illustrated popular book on the subject.(3) For the general reader there is much to be said for this approach, which brings him all the vivid excitement of saga narrative without the confusing necessity of bringing two images into a single focus.

In view of the present consensus that the sagas are essentially literary narratives, based on a mélange of oral and written historical lore, it is time to adopt a minimum approach to the Vinland sagas as well.(4) There have always been writers who decried their historicity, the most eminent having been Fridtjof Nansen, who considered the description of Vinland to be merely a reflection of the old European myth of the Blessed Isles.(5) Without endorsing this extreme minimum, I shall grant that on re-reading the texts and making a detailed comparison of their treatment of parallel events, I was more impressed by their literary excellence than by their historical acumen.

The groundwork for a critical approach to the Vinland sagas was laid by Jansson in his line-by-line comparison of the two versions of the Eiríks Saga. His wholly convincing conclusion was that the Skálholtsbók version was badly copied but textually faithful, while the Hauksbók version was heavily edited by Haukr Erlendsson. No comparable critique has been made of the relation between the proto- version of the Eiríks saga and the so-called Greenlanders' saga, which is in reality just a sequence of 'ţćttir' inserted into the Greater Saga of Olaf Tryggvason.(6) Jón Jóhannesson reversed all previous scholarship by demonstrating that the Greenlander's Saga was older than 1200 and could be based on the authority of (if not written by) Thorfinn Karlsefni's great-great-grandson, Bishop Brandr Sćmundsson (1163-1201). The saga cites Thorfinn as its ultimate source, and Jón sees no reason to doubt it. The Eiríks Saga, on the other hand, can be no older than 1263, the year in which that other descendant of Thorfinn, Brandr Jónsson, became a bishop.

When Jón goes on to suggest that the author of Eiríks Saga used the Greenlanders' Saga as one of his sources, I am not convinced. The massive deviation in content and ordering and the almost total absence of verbal echoes either in narrative or dialogue make it highly questionable that the author of Eiríks Saga had ever seen the Greenlanders' Saga. Jansson's research has confirmed the view that Eiríks Saga was written on Snćfellsnes in western Iceland, from which Gudrid and her family emigrated, perhaps at the monastery of Helgafell.(7) Since Guđbrandur Vígfússon it has been generally held that the Greenlanders' Saga stems from northern Iceland, where Thorfinn and Gudrid settled down on their return, and where their episcopal descendants had their see at Hólar (as pointed out by Matthías Ţórđarson).(8)

A close reading of the two rival versions makes it clear that they have some very basic features in common. Both are highly clerical and Christian in tone. The clerics who must have written them were more than usual in the sagas concerned about the conflict between paganism and Christianity at the turn of the millennium. Both sagas are edifying and moralizing in tone, exempla showing how the good Leif finds new lands while his pagan father Eirik fails, and why the good Gudrid has a blessed progeny while that of the evil Freydis is cursed for all time by her own brother. The role of Gudrid is particularly interesting: one could even say that these sagas are strongly woman- oriented. The reason appears to be that Gudrid was the link between the families of Eirik raudi and Thorfinn Karlsefni: she was Eirik's daughter-in-law and Thorfinn's wife. In the supposedly more primitive and reliable Greenlanders' Saga she is even sent on a highly anachronistic pilgrimage to Rome and ends her days as a nun, a woman's highest destiny in good Catholic doctrine. But not before she has become the mother of at least three children, from whom are descended four bishops plus lawman Haukr Erlendsson, who personally copied her saga into his great anthology.(9) Finally, the sagas have in common what can only be described as remarkably imaginative and skillful authors, who did not hesitate to invent episodes and vivid dialog for the entertainment of their readers and the greater glory of God and the descendants of his faithful servant, Gudrid Thordardottir.

Perhaps the best evidence of this literary independence is the scene which Jón uses to prove the dependence of the Eiríks Saga on the Greenlanders' Saga. I refer to the death of Gudrid's first husband, Thorsteinn Eiriksson. The common element is that Thorsteinn had to die so that Gudrid could be free to marry Thorfinn. They agree that he died in Lysufjord in the Western Settlement in the household of one Thorsteinn whose wife also died in the same epidemic, and that Gudrid's husband prophesied Gudrid's future after he himself had died. But beyond this the two authors have embroidered the story in entirely different ways. They disagree on whether Gudrid had been married before; on whether Thorsteinn had made an abortive effort to reach Vinland; on how they got to Lysufjord and why; on the name of the other Thorstein's wife; on how she died and what she said; on the order of events; on how Thorsteinn got his wife's attention; on what he said when he died; and on what happened after his death.

In short, a common set of facts which could have been derived from an oral or written source has been freely and independently embroidered by two imaginative and not impartial authors. It surprises me that more has not been done by students of saga creativity with this remarkable case of literary embroidery. The problem is not unlike that faced by theologians who try to harmonize the disagreements of the synoptic gospels and find what they call "the historical Jesus". A distinguished Norwegian theologian, Jacob Jervell, has astutely pointed out that the gospels are not intended to be history: they are an interpretation of history which is intended to edify the worshipper.(10) Similarly, we have to recognize that the purpose of the Vinland sagas was not to report on the discovery of America, but to glorify the descendants of Gudrid, and incidentally to make of it a good story. This is even more blatant in the Eiríks Saga, where she is provided with an ancestry right back to the landnám. But it is also strong in the Greenlanders' Saga. Both of them end with tell- tale geneaIogies, which may be implied but are not stated in other sagas.

What is the bearing of all this on the location of Vinland? Simply that both versions agree on the finding of grapes as a central feature in the naming of Vinland. Even with a minimum theory of the validity of the sagas and a maximum emphasis on their authors' skill as fabulators, I cannot accept the proposition now being advanced that L'Anse-aux-Meadows on the northern tip of Newfoundland is the Vinland of the sagas. That cold and casual bay, right opposite Greenland and at the shortest sailing distance, is an obvious way-station to something greater and more interesting. The harsh northern winters, the rivulet running down to a shallow bay, the barren surroundings: what has become of the mild climate, the lake and river large enough to float a viking ship, the huge population of hostile natives? The sagas report explorations southwestward from their first landfall. One amateur archeologist of the 1880s, the Harvard chemistry professor Eben Norton Horsford, was so convinced of the New England location of Vinland that he pointed out a site on the Charles River, placed a plaque on the spot, dedicated a statue of Leif Ericson on Commonwealth Avenue in Boston, and built a tower in Watertown to commemorate Thorfinn Karlsefni's settlement.(11) Whatever we may think of his findings, there can be no doubt that New England fulfills the conditions described in the sagas better than does Newfoundland.

Part of the argument hinges on the meaning of the first syllable of Vinland. Sverre Marstrander has written an excellent survey of the Ingstad findings in which he said: "On a philological basis it can hardly be determined whether the first member is to be interpreted as 'vine', as most have done, or as 'pasture, meadow'."(l2) There is a certain plausibility in this statement, in view of a forceful attack on the standard interpretation launched by Professor Sven Söderberg in 1898. There is no denying the ingenuity and brilliance of Söderberg's reasoning, which first became known in wider circles when his lecture was printed in 1910, though obscurely and posthumously.(l3) His theory was accepted by the Finnish geographer Vainö Tanner in 1941 and by the Swedish philologist Valter Jansson in his classic 1951 dissertation on the vin-names of Scandinavia.(14)

If I venture to differ with these scholars, it is in part because of the testimony of the sagas that I have already suggested. But I believe there is a strong linguistic and philological argument as well.(15) I shall not bore you with a detailed presentation of the evidence here, but remind you first of all that the word vin 'pasture' is totally absent from the rich Old Norwegian and Old Icelandic literature beginning around 1150, where its function has been taken over by the word eng (or tún for enclosed pastures). Even the archaic skaldic language does not know it, except possibly in a kenning for Sjaelland, viney, where we have no means of knowing exactly what it implies. In the laws of Norway and Denmark there are two terms for levied taxes, vinjarspann 'a tub of butter' and vinjartoddi 'a leg of beef'; in these words vin clearly means 'farm', not 'meadow'.(16)

Farm-names containing the element vin were popular in the grassy areas of Norway and in the Swedish province of Västergötland.(17) The materials presented in Valter Jansson's study reveal that nearly all of these have vin as a second member, not as a first. They also show that such names are unknown in Iceland and the Faroes, as well as Greenland, Denmark, Ireland, England, and other places where the Northmen lived or travelled, aside from two or three names in the Orkneys and Hebrides. The period when most of these names were given was from AD 1OO to 600, well before the Viking Age, when we may presume that the word was still in active use. By the time Greenland and Vinland were discovered, however, it appears certain that these names had lost their v (=w). An original *Bjorg-win had become Bjorgyn, with a final syllable that was no longer identifiable as the word vin.(l8) Jansson finds that especially in western Norway some place names have survived with v, and on this basis contends that new names could be formed with vin as late as 1200.(19) On this argument is based his belief that the word could have been known to the Icelanders of Leif Eiriksson's day.(20)

At best this proves that the new names could have been created on the pattern of older names. It suggests that the independent word could have been known, but does not provide an argument for its existence as an element comparable with the well-known words hella in Helluland and mark in Markland. Even if it were still identifiable as a suffix applicable to farms, it could hardly have signified 'pasture' at this time. The farms to which it applied had long since ceased to be pastures; they were a rather prosperous group of farmsteads. Just as in the words cited, the only meaning which an Old Norse speaker without an etymological dictionary at hand could deduce would be 'farm' rather than pasture. In fact, the only basis we have for thinking that it ever meant 'pasture' is its etymology, which shows that it is cognate with Gothic winja and High German winne, both of which meant 'pasture'.(21) In Jansson's dissertation it is tacitly assumed that it continued to mean 'pasture' for a thousand years; he nowhere discusses its semantic development, which must have been inevitable as the word became less common and was primarily attached to farm-names.

I shall not try to go into the problem further, since it involves discussing the far-fetched assumption that the whole story of grapes is based on a folk-etymology produced by Adam of Bremen.(22) In conclusion I will say that I do not venture to propose any single location for the Vinland of the sagas. I am happy to recognize Newfoundland as its beginning and congratulate the Ingstads on their discoveries. But Newfoundland cannot be the end of Vinland. And I suspect that if techniques had been as refined in the 1880s as now, and the soil had been equally undisturbed, we might have been able to confirm some of Horsford's hunches.

As it is, I shall conclude with an invitation. If this Congress should ever decide to meet in Vinland, I shall be happy to serve its members a delicious repast of those Concord grapes which have delighted visitors and residents alike from Leif Eiriksson to, shall we say, Ralph Waldo Emerson.


1. For the Ingstad explorations and discoveries see Helge Ingstad, Vesterveg til Vinland (Oslo, 1965), Eng. tr. Westward to Vinland (New York, 1969); Anne Stine Ingstad, Det nye land med de gronne enger (Oslo, 1975); Anne Stine Ingstad, "The Norse Settlement at L'Anse aux Meadows, Newfoundland. A Preliminary Report from the Excavations 1961-1968", Acta Archaeologica, (1971), XLI, 109-154. Anne Stine Ingstad, The Discovery of a Norse Settlement in America (New York, 1977).

2. Sven B. F. Jansson, Sagorna om Vinland I. Handskrifterna till Erik den Rodes saga (Lund, 1944); Jón Jóhannesson, "The Date of the Composition of the Saga of the Greenlanders", tr. by Tryggvi Oleson, Saga-Book XVI 54-66 (1962). Jansson's text of the Eiríks Saga has been used here; for the Greenland Saga the text in A. M. Reeves, The Finding of Wineland the Good (London, 1890) has been used.

3. Einar Haugen, Voyages to Vinland: The First American Saga, ill. by Frederick T. Chapman (Chicago, 1941); rev. general edition (New York, 1942). For a current translation of both sagas, with a valuable (if traditional) introduction, see M. Magnusson and H. Pálsson, The Vinland Sagas (New York, 1966).

4. For a view of saga research today see Lars Lönnroth: Njál's Saga (Berkeley, 1976).

5. Fridtjof Nansen, In Northern Mists: Arctic Exploration in Early Times (London, 1911).

6. Sven B. F. Jansson refers to an unprinted essay on the subject, which he hoped to develop into a volume at the time of publishing his 1944 dissertation; but so far this has not appeared.

7. Jansson, op. cit., pp. 263-272; see also his article in Kulturhistorisk leksikon for nordisk middelalder, s. v. Eiriks saga.

8. Matthías Ţórđarson, Vinlandsferdirnar (Reykjavik, 1929), Eng. tr. The Vinland Voyages (New York, 1930); Islenzk fornrit, IV (Reykjavik, 1935), see esp. p. LXXXI f. Gudbrand Vigfusson, ed. Sturlunga saga, I (Oxford, 1878), p. lix.

9. For the evidence see Jansson, op. cit., pp. 100-101.

1O. Jacob Jervell, Den historisk Jesus (2. ed.; Oslo, 1969).

11. See especially Eben Norton Horsford, The Landfall of Leif Erikson AD 1000 and the Site of His Houses in Vineland (Boston, 1892). For a report on an excavation of his sites see the partial excerpts in Cornelia Horsford, "Vinland and Its Ruins," Appleton's Popular Science Monthly, December, 1899. The excavations were by the Icelandic scholars Valtyr Guđmundsson and Thorsteinn Erlingsson (dated July, 1896); they granted that the sites resembled Icelandic farmsteads, but they could not confirm it since the sites also contained glazed pottery and bricks of later manufacture.

12. Sverre Marstrander, "Arkeologiske funn bekrefter sagaens Vinlandsberetninger," Forskningsnytt, XIX:3 (1974), 2-11.

13. "Professor Sven Söderberg om Vinland," Sydsvenska Dagbladet Snallposten, Nr. 295, Oct. 30, 1910.

14. Vainö Tanner, De gamla nordbornas Helluland, Markland och Vinland (Abo, 1941); see also the critique by Sigurđur Thorarinsson, "Vinlandsproblemet," Ymer, LXII (1952), 39-46; Valter Jansson, Nordiska Vin-namn: En ortnamnstyp och dess historia (Lund, 1951) .

15. I call attention to a vigorous defense of vín in articles by Erik Wahlgren: "Fact and Fancy in the Vinland Sagas," Old Norse Literature and Mythology, ed. E. C. Polomé (Austin and London, 1969), 19-80; "Ordet och begreppet 'Vinland'," Gardar, V (1974), 16-42.

16. L. Heggstad, Finn Hodnebo, Erik Simensen, 'Norron ordbok', 3. ed. (Oslo, 1975). See also discussion in V. Jansson, op cit.

17. See maps in V. Jansson, op cit.

18. For examples see the lists in V. Jansson, op cit., chap. 11.

19. Examples and discussion in V. Jansson, op. cit., chap. 10.

20. As stated in his Nordiska Vin-namn, p 422.

21. On the etymology see e.g. Jorn Sandnes and Ola Stemshaug, Norsk Stadnamnleksikon (Oslo, 1976), s.v. vin.

22. See especially Söderberg's article, above n. 13. We need hardly take seriously the Vinland Map, since the announcement by the Yale Press that the ink has been shown to be modern: see "The Strange Case of The Vinland Map," The Geographical Journal, CXL (1974), 183-214.

Maintained by