Chapter VII: Part II: Discussion (1)

I. NATURE OF THE EVIDENCE

In order to judge what historical value should be assigned to the narrative here translated, it is necessary for the reader to have a clear idea of the nature of saga literature, and some notion of the process by which such stories were transmitted from the time of their occurrence to the period, more than three centuries later, when they assumed the form which is now known to us. In view of the fact, which must be at once conceded, that we are dependent upon an interval of oral tradition before any written account of the Wineland voyages can have come into existence, we must first of all consider how the special characteristics of story-telling in Iceland affect the reliability of such tradition; next we should look for any early corroboration bearing upon the questions involved; and finally we must consider the manuscripts which form the basis of the story, and inquire into any circumstances which may make one source preferable to another.

_Oral Tradition in Iceland._

None of the Wineland voyages which form the subject of our inquiry can have taken place later than—say—A.D. 1030, and the earliest would appear to date from as early as 986. Until the inconvenient runic alphabet, suited only to short inscriptions, was superseded by something better adapted to the requirements of fluent literary composition, the history of such events could be preserved only by word of mouth. This change did not occur till at any rate nearly a century had elapsed from the time of the occurrences with which we are dealing. Oral tradition, however, may, under favourable conditions, show a fidelity to the actual facts which is at first sight surprising. Mention might be made in this connexion of the Scottish Highlands, where, in spite of the Celtic imagination, the ‘shenachies’ or prose annalists attached to the more important families have been found to have transmitted historical facts which have been most exactly confirmed by subsequent investigation of documentary evidence. A little consideration will show that this is not so extraordinary as one might superficially be disposed to imagine. The distinction, recognized by our law, between libel and slander is partly at any rate based upon a consideration which should be borne in mind in this connexion. The written word remains, even though contradicted and disproved; nay, it may not infrequently survive its contradiction. The verbal narrator of contemporary events, however, is always liable to have among his audience those who are as thoroughly conversant with the facts described as he is himself. An inaccuracy may be suddenly and unpleasantly brought to book; the lie is no sooner uttered than it is denounced and exposed. We find a good illustration of the embarrassing predicament in which a story-teller might find himself placed (though the hero in this instance came out of the ordeal with credit) in the episode of the Icelandic saga-man at the court of King Harald Haardraade which is reproduced among the excerpts in Vigfusson and Powell’s _Icelandic Reader_ (p. 141). This young man, we are told, was taken in at court for the purpose of entertaining the bodyguard with his sagas. About Christmas time he began to grow melancholy, and on the cause being investigated it was found that he had used up all his stories but one just at the time—the Yuletide festivities—when his accomplishment was most in demand. This remaining story he hesitated to recite, for it was the saga of the king’s own travels. Encouraged, however, by Harald himself, he ventured upon his embarrassing task, the hero of the exploits described being present among the audience. The story was told, and the days passed by, but the Icelander evinced no curiosity to know how his rendering had pleased the person who had first-hand knowledge of the facts. ‘I am afraid about it’ was his reply, when the king drew his attention to this omission on his part. Harald reassured him, however, saying that his version was perfectly correct, and inquired the source from which so accurate a report had been obtained. On learning that one Halldor Snorrison was the person originally responsible, the king said that he was no longer surprised at the accuracy of the tale, and offered the narrator the hospitality of his court on any future occasion when he might wish to come there.

Another instance, where the consequences were not so satisfactory to the story-teller, occurs in Njál’s Saga, where Gunnar Lambison is requested by King Sigtrygg in the Orkneys to give an account of the burning of Njál in his house, to which he had been a party. He starts telling the story in an unfair and inaccurate manner, stating among other things that Skarphedinn, Njál’s son, had wept as the danger closed round him. Upon this Kári, who has been listening at the door, dashes in with a drawn sword, and cuts off the head of the untruthful historian. Flosi, another of the burners, defends and justifies Kári’s action, and thereupon tells the story himself, and as he favours neither one side nor the other unduly in his narration we are told that his story was believed.

Now the conditions of this art of story-telling in Iceland were unusually favourable to the maintenance of an accurate tradition. In the first place, as may be seen from the instance cited, the practice was to all intents and purposes contemporaneous with the occurrence of the events described. In the second place, a point which will fall to be developed later on, it is evident that the taste of the Icelandic audience was intensely practical and unimaginative. Superstitions no doubt there were, in Iceland as throughout the whole world of this and indeed far later periods, but even their ghosts and supernatural occurrences are treated by this people, far more than by any other with whose works I am conversant, as something all in the day’s work. The Icelander did not want, like the Celt or the later Romancers, to surround his heroes with an atmosphere of picturesque mythology; his principal desire was to learn in the utmost detail exactly how everything was done, with the dates, genealogies, and circumstances relevant to the story to which he was listening.

I have mentioned the word genealogies, and this brings me to the last factor which operated in favour of the accuracy of oral tradition in Iceland. The colony was from its very nature composed of a great number of more or less connected families, equal in social status, and known to each other to consist of men of like passions with themselves. There was no king, no outstanding heroic personality, round whose unapproachable majesty the flattering tongues of courtiers could weave their myths and fictions. The saga-teller moreover was not, like the bards and shenachies of the Scottish Highlands, the appanage of a single family. He moved from place to place, whiling away the monotony of the Arctic winter with his histories, and the hero of one locality was in another an ancestor or a member of a family in no way superior to the persons who were gathered to hear the tale. Each listener was deeply versed in genealogy, a subject which was clearly regarded as of primary importance. Most great families, by dint of intermarriage, were connected with at all events some of the characters which were introduced into almost any saga, and the necessity of reciting correctly before the most critical of audiences the intricate ramifications of all the family trees occurring in the course of the narrative must have been the best possible discipline to produce a school where accuracy was placed above every other consideration.

From the circumstance, too, that the story had to satisfy the inhabitants and the visitors of a number of different settlements, with an equal social status but with frequently conflicting interests, arose the characteristic which has often been noticed by students of Icelandic literature, that both or all sides of a question are stated fairly, the author or reciter being, as Vigfusson has put it, ‘a heathen with the heathen, a wrathful man with the avenger, and a sorrowful man with the mourner, as his style reflects the varied feelings of his _dramatis personae_’.[64]

We have therefore the best of grounds for imagining that the exploits of those who fought, litigated, or explored in the tenth and eleventh centuries were carried with truth, impartiality, and accuracy over the brief interval which separated them from the age of written history, which dawned with Ari the Learned.

_Ari the Learned._

This pioneer of Icelandic history and of the age of writing was born, as we learn from the Icelandic Annals, in the year following the Norman conquest of England (1067). His grandfather, Gelli, was a contemporary of Karlsefni, and was in fact his second cousin. (See Genealogical Table, p. 20.) We are expressly told by Ari that his uncle, Thorkel Gellison, supplied him with information relating to Eric the Red, which he had obtained from direct speech with one of the latter’s companions. The events with which we are concerned thus fall within a period bridged by one human memory from the time of occurrence to the period when they could be recorded in writing, and when written history, as superseding oral tradition, may be said to begin.

It is moreover worthy of note in passing that the most important explorer with whom these sagas deal, Thorfin Karlsefni, was of the same stock as Ari, and must almost necessarily have been personally known to one of his informants, his uncle Thorkel.

It should also be remarked that one of the persons for whom Ari expressly tells us that he composed his Íslendíngabók, and to whom he showed it, was Bishop Thorlak, the grandson of that Snorri who, as we are told in the saga, was born to Karlsefni in Wineland.

To the truthful and conscientious work of Ari the Learned a well-known introductory passage in the history of the kings of Norway known as Heimskringla bears eloquent witness. The author of this book was greatly indebted to the researches of Ari; in fact, though the latter’s original work on the subject of the Norse kings no longer exists in its intact and primitive form, we know that such a book was among his literary achievements, and was in all probability followed closely by subsequent compilers of stories relating to the earlier history of Norway. Unfortunately, however, greatly as later writers were indebted to Ari, of his original work only one book remains, and this in a highly condensed and summarized form. This is the Íslendíngabók, or history of the Icelanders. We know from the author’s own statement that this book was originally written in a different and probably more extended form, of which no copies now remain, but the little book now extant contains, besides a genealogy of Karlsefni, one passage valuable to us in dealing with the present subject, from the early corroboration which it affords of the essential outlines of our story. This passage, which will also be found in the Appendix of Supplementary Passages, p. 74, may be rendered as follows:

‘The country which is called Greenland was discovered and colonized
from Iceland. It was a man called Eric the Red from Breidafjord who
went out thither from this country, and he took land in the place
which was afterwards called Ericsfjord: he named the country and
called it Greenland, saying that the fact that the country had a
good name would attract men to journey thither. They found there,
both in the east and the west of the country, dwellings of men,
and fragments of canoes, and stone implements of a kind from which
one could tell that a race had come (_farit_) there _of the kind
that inhabited_ (_bygt_) _Wineland, and whom the Greenlanders call
Skrælings_. Now the date when the settlement of that country was
started was from fourteen to fifteen winters before Christianity
came here to Iceland, according to an account given to Thorkel
Gellison in Greenland by one who himself accompanied Eric the Red
out.’

This casual reference would appear to afford the strongest confirmation both of the known and recognized existence of Wineland, and, in particular, of the episodes described in the sagas relating to the savages or ‘skrælings’.

It furnishes besides, in the present writer’s opinion, proof positive that a land inhabited by savages had been visited by the Norsemen at a time when no such people had actually been met with in Greenland itself. The Eskimo of Greenland, it will be observed, had, so far as Ari’s information went, come and gone before the Norse occupation (_farit_), and their existence was only inferred from the traces above described, while the natives of Wineland had at the same date ‘a local habitation (_bygt_) and a name’. ‘Skrælings’ was not therefore a title transferred from known inhabitants of Greenland to savages figuring in tales of Wineland; the reverse was the case.

This point will be developed later, and certain objections which have been raised to this interpretation of the passage will be fully dealt with, but it will at once be seen that it is of considerable importance in its bearing upon the accuracy of the saga and the fact of the Norse discovery.

_The Landnámabók._

Another work of high authority, in which it is certain that the conscientious hand of Ari played a large part, is the Landnámabók or history of the settlement of Iceland. Hauk Erlendson, in his edition of this classic, expressly acknowledges the authorship of the master, saying that it is ‘according to that which first priest Ari the Learned, Thorgil’s son, has written, and Kolskegg the Wise’. Kolskegg was a contemporary of Ari’s, and Vigfusson[65] thinks that his share in the collaboration was confined to supplying the genealogies of the Eastern district. Judging from its uniformity of style, this great authority[66] has no hesitation in ascribing the sole authorship of the Landnámabók to Ari and Kolskegg. The authoritative character of this work has a direct bearing upon our subject, for it is evident that the writers of both versions of the story drew largely from its pages, indeed both versions contain a great deal of absolutely literal quotation.

As regards Wineland itself, however, the Landnámabók has but little to say. It was in fact foreign to the purpose of a book whose whole scope was confined to Iceland, and we ought not therefore to expect more than we actually find. The only reliable mention of the place is in the passage relating to Ari Marsson, who is there said to have been cast upon Hvítramannaland, ‘which some call Ireland the Great, it lies westward in the sea near Wineland the Good’. The importance attaching to this passage is that Wineland is casually mentioned as a well-known locality from which the position of Hvítramannaland could be approximately fixed, without the necessity of further explanation. Another passage, relating to ‘Karlsefni who found Wineland the Good’, is of less value, as it is in all probability an interpolation by Hauk, which consequently affords no independent corroboration of the discovery.

_Adam of Bremen._

It has therefore been established so far that at the time when writing superseded oral tradition the fact of the discovery of a ‘Wineland’ by the Norsemen was perfectly well known, that it lay to the west (vide Landnámabók), and contained savages. The name moreover affords some corroboration in itself of the details given in the sagas with reference to the discovery of grapes there. A further confirmation of the facts recorded as to the principal products of the country must now be dealt with. This dates from an even earlier period, and comes from an independent source, the _Descriptio_ of the ‘islands’ or countries of the North which was written by Adam of Bremen. This worthy became director of the cathedral school in Bremen in or about the year of Ari’s birth (1067), and derived, as he tells us, the information upon which his description is based from Svein Estridson, King of the Danes, who died in 1076.

Knowledge obtained from such a source brings us practically to the lifetime of Karlsefni’s contemporaries, and well within that of many who might remember him or his associates. In the geographical work referred to, Adam inserts the following reference to Wineland:

‘He (King Svein) told me of yet another island besides, discovered by many in that Ocean, which is called ‘Wineland’, from the fact that there vines grow naturally, producing the best wine. Moreover that corn abounds there without sowing we have ascertained, not from fabulous conjecture, but from the reliable (_certa_) report of the Danes.’

_Prima facie_, therefore, we have here the most controversial part of the whole story—the existence of the wild corn and vines—substantiated by an authority based on a Scandinavian source, almost within the lifetime of the explorers themselves. In view of a contention which will be dealt with more fully later, that the accounts of vines and wild corn occurring in the sagas are derived from references to the Fortunate Islands in Isidore Hispalensis and classical works, it may be important to note here the emphasis laid by the writer on the source of his information.

Adam of Bremen, a learned continental _magister_, must have been already familiar with the numerous legends relating to these Fortunate Islands, references to which are frequent in many classical authorities, and he appears to be anticipating the criticism which has in fact been made, when he draws, as he does, a careful distinction between _fabulosa opinio_ and _certa relatio Danorum_. He seems in fact to be saying,—‘Of course you think that this is another story based on classical legends which are familiar to you, but it is nothing of the sort: when I was in Denmark I had the opportunity of questioning the Danes whose information I have recorded, and I find it impossible to conclude that this is merely a case of the Insulae Fortunatae at second hand.’

_Date of the Existing Manuscripts._

We may now pass on to consider the sources from which the present translation is drawn. The existing manuscripts, it will be found, are none of them earlier than the fourteenth century, but it may be well to point out that this fact is not so damaging to their credit as might be supposed.

The day of oral tradition was long over, the day of documentary history had been long established, and the compilers of those versions which we now possess must have worked in the main not from oral tradition, but from earlier written sagas which had then attained to a large extent the form in which we have them. A well-known passage in the Sturlunga Saga is not without a bearing on this point. ‘Nearly all stories.’ it says, ‘which had been made in Iceland before Bishop Brand Sæmundson died (A.D. 1201) had been committed to writing; but stories of things which have taken place since were hardly committed to writing at all before the skald Sturla Thordson dictated the Iceland Sagas.’ Now while we may admit, with Vigfusson, that this passage has reference primarily to the three sagas which have at this point been incorporated in Sturlunga, it is clear that ‘nearly all stories’ cannot be a statement confined to three, and must have a general reference to the condition of all the stories known at that date. It follows that any events which took place before 1201 had in all probability assumed a more or less fixed written form before Sturla (born _c._ 1217) started to write down the later occurrences.

The contributions of later scribes would appear to have been confined for the most part to bringing the genealogies down to their own day; the fashion of romanticizing the earlier material to any great extent did not become general till a later date than those which we have to consider.

That Eric’s Saga had assumed a written form before the Flatey Book version was compiled is evident from the reference to it in the opening chapter of that story: ‘Thence arose the quarrels and fights between Eric and Thorgest which are related in Eric’s Saga.’ How far the saga of Eric known to the compilers of the Flatey Book corresponded with any work which now bears the same name is a question which cannot be adequately discussed till we have considered further the nature and authenticity of the versions from which the translation has been derived.[67]

_Hauk’s Book and the Saga of Eric the Red._

Our knowledge of the Wineland voyages is obtained, as the careful reader of the translation will discover, from two apparently independent sources, which may for convenience be described as Hauk’s version and that of the Flatey Book. The story as known to Hauk is found in two manuscripts: one contained in Hauk’s Book and partly written by his own hand; the other, in an early fifteenth-century hand, is No. 557 4to in the collection of Arne Magnusson, and is most conveniently designated—according to its actual title—as the Saga of Eric the Red.

This last-named manuscript, while it was undoubtedly written long after Hauk’s Book, probably embodies the earlier and better text of this version. It is certainly not a free rendering of the story, but a literal transcript of some earlier manuscript, for it contains a number of typical copyist’s errors. There are, for example, words repeated twice in succession, and passages which as they stand are meaningless, and require some simple emendation. It is equally certain that the text followed was not that of Hauk, for the language differs slightly throughout, and there are sentences in each version neither occurring in the other nor arising from it by necessary implication. The theory that the Saga of Eric the Red embodies an earlier text than that of Hauk is deduced by experts from the greater simplicity of the language in the former version. To the lay mind the most convincing proof is to be derived from the genealogy at the end of the saga. As has already been stated, it was the practice of transcribers to bring such pedigrees down to their own day. Hauk follows this practice, tracing the line of Karlsefni down to himself. The Saga of Eric stops short at Bishop Brand the first, several generations earlier. Hauk, according to his account, was the great-great-great-grandson of Bishop Brand’s second cousin. (See Genealogical Table, p. 20.) The fact, however, that Bishop Brand is described as ‘the first’ shows conclusively that the text copied in Eric’s Saga was not completed till the ordination of the second bishop of that name, which took place in 1263.

Of course, as far as this goes, it is not inconsistent with the writers of these two versions having worked from the same manuscript, which Hauk altered and edited, while the other scribe contented himself with a literal copy. While, however, the sense of Hauk’s version follows approximately that of the rival manuscript, the language is rarely identical for many words together. Had both been working from the same manuscript, this is not what one would expect to find: it is so much simpler to transcribe a passage _verbatim_, when the meaning which it is intended to convey is as adequately given by such a method. And Hauk’s text occasionally gives us information which cannot be explained as a mere intelligent amplification of the other.

We are consequently justified in all probability in imagining that the common origin of the two versions must be assigned to a period considerably earlier than either. Finnur Jónsson, an excellent critic of Icelandic styles, considers that we may give the common archetype as early a date as 1200. As regards the date of the extant manuscripts, to which, for reasons already given, too much importance should not be attached, it is sufficient to state that Hauk died in 1334, and as his own hand concludes the saga it must have been written some time before that date. The clue given by the mention of Bishop Brand ‘the first’, noticed above, is common to both manuscripts, and fixes the period before which neither manuscript was completed at 1263. In the case of Hauk’s Book these limits are further narrowed by the mention of Hallbera with her title as Abbess of Reynisness. We know that this lady attained this position in 1299, so that Hauk’s Book cannot have been completed before this date.

_Hauk’s Personal Authority._

Mr. W. H. Babcock, in his clear and valuable treatise on the subject,[68] lays considerable stress on the fact that Hauk was a descendant of Karlsefni, as enhancing the authority of this version of the narrative. To some extent this is a good point, but it may be doubted whether Hauk’s knowledge of his ancestors was sufficient to check the written records accessible in his day. He follows the demonstrable error of Landnámabók in making Thorbjörn Vifilson the son of Aud’s freedman, which a close examination of the chronological data shows to be an altogether untenable theory. (See Genealogical Table, p. 20.) He was separated from Karlsefni by no fewer than eight generations, and any reader who takes the trouble to consider how much he knows of the achievements of so distant an ancestor will no doubt form the conclusion that Hauk was not in a position to throw much additional light on the subject, though it was naturally of peculiar interest to him. All we can say is that he regarded the saga as historical and not romantic, and his wide experience of Icelandic literature, quite apart from his family connexions, made him a good judge. That he had no special private sources of information is clear from the fact that he transcribed the saga practically as it stood. It cannot be sustained that he discarded the Flatey version, or preferred the alternative; it seems much more likely that the editors of the Flatey Book tapped sources to which he never had access. Hauk, had he deliberately compared the two authorities, would for example inevitably have selected the Flatey version of the stranded-whale episode, as this tallies much better than his own text with the older verses incorporated. (Cf. next chapter, p. 132.) Hauk, in fact, merely copied, with more or less intelligence, the only version of the story which he knew, and his manuscript, therefore, stands on exactly the same footing as the Saga of Eric the Red: coming from a common archetype they of course afford no independent corroboration of one another.

_Independence of the Flatey Version._

That such corroboration is, however, afforded by the version contained in the Flatey Book is, I think, clear to demonstration. But for the attitude of some modern writers on the subject, the independence of this account might be said to be beyond dispute, whatever its relative value as an authority might be. Some commentators have, however, attempted to establish that the Flatey Book is but an embroidery based on the rival text. Thus Mr. Juul Dieserud, in the Bulletin of the American Geographical Society (1901), states boldly that the Flatey Book ‘borrowed incidents and descriptions from the story of Thorfin’. He adds: ‘This may seem to be a hazardous conjecture, but ... the only way out of it is to regard the saga of Thorfin as the result of a similar process.’

The alternative, however, with which Mr. Dieserud here considers himself to be faced, is by no means the only one. The depositions of two witnesses to a matter of fact may show many points of agreement as well as discrepancies without any collusion or borrowing whatsoever. So, too, different authors may treat of a question of history or tradition without having consulted each other’s works. Again, if I and a friend go through some experience together—suppose, for instance, that we serve in the same unit during the war—the accounts which we transmit to our respective descendants may be quite independent of one another. A charge of plagiarism, under such conditions, needs to be established by definite and cogent evidence.

Now what does Mr. Dieserud put forward as proof or support of his contention? He says, for example, ‘an incident related of the stalwart Freydis and the short mention of some quarrels caused by the women during the last winter in Straumsfjord sets somebody’s imagination working till we get a gruesome tale of her separate expedition to Wineland in company with the brothers Helgi and Finnbogi’. The quarrels over (not otherwise caused by) the women in the Saga of Eric the Red are of a purely sexual character. The bachelors, we are told, coveted the wives of the married men. This situation, though hardly unique, might well provide an imaginative mind with a plot like that of a modern problem novel. But where is anything of the kind to be traced in the Flatey Book story of Freydis? There is no quarrel about women; in fact, feminine charm was hardly Freydis’s strong point. There is a purely mercenary dispute about the ownership of a boat, in which a person who is incidentally a woman plays the principal part. In short, there is no sort of connexion between the two plots; it might as well be said that the story of Jezebel and Naboth was a plagiarism from that of David and Bathsheba.

In the same way, the alleged development of Bjarni Herjulfson from Bjarni Grimolfson, which is also asserted by Joseph Fischer,[69] rests upon no more solid foundation than the coincidence of a name by no means uncommon in Icelandic literature. Storm, more correctly, recognizes the Bjarni of the Flatey Book as ‘en ellers ganske ubekjendt person’ (a person otherwise quite unknown), and Neckel’s _Erste Entdeckung Amerikas_ makes use of an identical expression. Would anyone, desiring to make up a good story about Bjarni Grimolfson, neglect the dramatic episode of his death in the worm-eaten ship, as given in the saga of Eric? Why, as Neckel says, not let him land and find the vines and corn, if the object was to give him a credit not his due? Apart from their first names, Bjarni Grimolfson and Bjarni Herjulfson have nothing whatever in common. When Fischer says, ‘Only in this way (i.e. by inventing the Flatey Book story) could the priest (John Thordson, one of the scribes of the Flatey Book) ascribe the honour of the discovery of Wineland to his hero Bjarni, who was really only one of the band who accompanied Karlsefni on his later expedition’, one is disposed to ask, Who treats Bjarni as a hero? He gets no credit for the discovery which accident threw in his way; Leif is here, as elsewhere, treated as the discoverer of Wineland: nay, we are told that Bjarni was severely criticized for lack of enterprise in not pursuing his investigations further. Moreover, if Bjarni Grimolfson was John Thordson’s hero, why change his surname altogether?

The third parallel suggested by Mr. Dieserud is between Tyrker in the Flatey Book and Hake and Hekja in Eric’s Saga. Hake and Hekja, one would think, make a more picturesque appeal to an imaginative writer than Tyrker. They are at least as good material for a story. But they are Scots or Celts while Tyrker is a German, they are two while he is one; in fact, they show few points of resemblance. A better case could be made out for a comparison between Tyrker and Thorhall the Hunter, though even this would be pretty remote. These are the three instances most prominently put forward to substantiate a charge of plagiarism.

When we look for points in one version which must inevitably have been included in the other if the two accounts were interdependent, we are only struck by the dissimilarity. The wild corn, so prominent in Eric’s Saga and in the popular accounts which reached Adam of Bremen, is not mentioned anywhere in the Flatey Book. The stranded whale, evidently a fact, as shown by Thorhall’s verses, is referred to, but the whole point of the story, as a story, is destroyed by too literal adherence to what appears to be the simple truth.

On the other hand, numerous statements of a circumstantial nature are made in the Flatey version which find no place in the rival account. The important ‘eyktarstad’ observation (see Chapter V) is a good instance of this. The Flatey Book gives the south-westerly course which the necessities of the case, as known to us, demand, but we look in vain for such a course in Eric’s Saga or Hauk’s Book, which follow the current ideas of Icelandic geographers in reporting a uniform progress to the south. Is it suggested that the greater accuracy of the Flatey Book in this particular is a freak of a vivid but uninstructed imagination? The savages, sleeping under their boats, as Jacques Cartier found them centuries later, are also mentioned in the Flatey Book alone. It is true that the authors of this version, coming to the conclusion that all the explorers made the same landfall, have felt at liberty to draw the description of Leif’s camp from what appears to be a report of Karlsefni’s Hóp, but, assuming the latter place to have been actually discovered by Karlsefni, there is no evidence in this that another saga was consulted at all. In short, I can find no evidence whatever that the compilers of the Flatey Book version had any knowledge of the rival account known to us. It is true that Finnur Jónsson[70] considers that the reference to ‘Eric’s saga’ in the introductory matter quoted from Landnáma is to the document known to us by that name; but, with all respect to the views of so fine an Icelandic scholar, such a theory seems to me untenable. In the first place, in the passage in question the author must be alluding to a story so well known to his audience that he can refer them to it without hesitation. _A fortiori_ a story known to himself. Yet no one who had more than the haziest recollection of our Eric’s Saga could possibly make the wide departures from it which are characteristic of the Flatey version. Secondly, the reference to the ‘quarrels and fights’ between Eric and Thorgest suggests a detailed account of the whole dispute. Yet the matter omitted in the Flatey Book from that supplied by Landnáma, which is the source quoted by all our authorities at this stage, amounts to no more than a bare mention of the battle which brought about Eric’s banishment, and that on his return to Iceland which was the prelude to reconciliation. The omissions are in fact hardly longer than the explanation which the author inserts. The object of the reference being clearly to effect a saving of time or space, one must suppose that the allusion is to some fuller account. But even if the reference were to our Eric’s Saga, it would not disprove the independence of the Flatey version as a whole, since at this point the compiler has not reached the stage where he incorporates new matter, but is copying practically _verbatim_ an abridgement from Landnáma which is to be found in other texts of the Saga of Olaf Tryggvason. The reference to ‘Eric’s saga’ is part of a quotation, rather than an original observation. In fact, as Neckel puts it, ‘the (Flatey Book) narrative makes pretty strong departures from the Saga of Eric the Red. It knows on the one hand more, on the other less; above all, the same occurrences appear in quite different order and connexion’ ... ‘Between both accounts runs the remarkable relationship that while clearly harmonious in the main features they are widely separated from one another in details. The use of the older narrative by the younger is accordingly excluded.’

The motive apparently suggested by Mr. Dieserud and those who agree with him for the tone adopted in the Flatey Book is the glorification of the family of Eric the Red. The introduction of a prior discoverer to Leif does not seem likely to conduce to such a result, and one feels that a member of Eric’s family would hardly regard the story of Freydis with pride or pleasure. But let that pass. Those who adopt this position seem to be faced with a dilemma. No one outside Greenland had any interest in attempting such a task, while if—as I myself believe (see next chapter, p. 139)—this version comes in the main from a Greenland source, it is far more likely that it represents an independent tradition than that compilers in so inaccessible a country had access to the version current in Iceland. For these reasons we need have no hesitation in accepting the independence of the Flatey version, and in concluding with Vigfusson that ‘the correspondence of these distinct versions throws great light on the vitality and faithfulness of tradition, and is a strong confirmation of the credibility in main points of a saga which is especially important for historic reasons’.[71]

_Date of the Flatey Book._

The date and circumstances of composition of the Flatey Book are known to us from the invaluable researches of Vigfusson, who transcribed the entire manuscript for publication. From this source we learn that it was compiled for one John Haakonson, who was born in 1350; the date of its commencement can therefore hardly have been earlier than some twenty years later (_c._ 1370). As originally planned it commenced with the mythical tale of Eric the Far-travelled, a fact which is plain from the words of the text, ‘He that wrote this book set this story first’. It continues in the same hand to set down a long saga of Olaf Tryggvason, King of Norway, followed by the saga of King Olaf the Holy. At this point the first scribe, John Thordson, lays down his pen, and the book is carried on by one Magnus, terminating with some Annals, which it was intended to keep up to date by additions from time to time. When therefore Magnus found himself in possession of some additional matter, which it was thought desirable to incorporate in the volume, he added a few leaves at the _beginning_ of the work, leaving the blank pages at the end for the continuation of the Annals. Towards the end of the newly incorporated matter comes the statement that it was written in the year 1387. Magnus then added a title-page with a list of the contents, and continued to add to the Annals from time to time till 1394. The story of the Wineland voyages given in the Flatey Book consists of two ‘thættir’ or episodes, interpolated after the manner of the time in the saga of Olaf Tryggvason, which is the second piece of literature included in the original volume. It follows therefore that, so far as we are concerned, the manuscript dates from some time after 1370, when the owner came to man’s estate, and before 1387. Considering the time which must have been occupied in writing a book of such gigantic proportions, we may fairly ascribe the Wineland parts of the book to a date considerably earlier than the year last mentioned.

The manuscript at present extant is therefore of a later date than that of Hauk’s Book. In admitting this we should, I think, for the reasons given earlier, be chary of attaching too much importance to the fact. Evidence is not wanting that the sources followed compare favourably in age with the rival version. Two such proofs are mentioned by Reeves, though only one of these seems to me of real importance. This is the fact that, unlike the rival version, the Flatey Book refers to Bishop Brand without the distinguishing title ‘the first’, which would in all probability have been added by anyone composing the archetype used by John Thordson at a date subsequent to the second Bishop’s ordination. The other point mentioned by Reeves is the reference to Eric’s landfall in Greenland by its original name of Midjökul, as well as by the later designation of Bláserk, which latter is given alone in Hauk’s version. A reference to the Landnámabók, however, shows that both names are there preserved, and as the part of both versions where the name occurs is obviously founded on Landnáma, the omission of a word of the matter copied by Hauk appears to me devoid of significance.

Turning to the contents of the rival productions of Hauk and the Flatey Book, though the two stories are obviously the same, we are at once confronted by certain striking dissimilarities. Bjarni Herjulfson and his adventure are recorded in the Flatey Book, and nowhere else in literature. Leif’s voyage is represented by the same version alone as being deliberately undertaken as a result of Bjarni’s discoveries; elsewhere it is accidental, an episode of a different voyage. A separate voyage of Thorvald Ericson, terminating in his death, is detailed in the same account, whereas in the Saga of Eric the Red no such person is mentioned at all till the episode of his death, and in Hauk’s Book and the companion manuscript he is represented as sailing and meeting his death under the auspices of Karlsefni’s expedition. Finally, after Karlsefni’s return, we have in the Flatey Book alone the story of Freydis’s second visit to the newly discovered country. With these discrepancies, and the attitude of modern criticism towards them, it will be necessary to deal in a separate chapter.

II. THE DISCREPANCIES OF THE FLATEY BOOK

The earlier writers on the subject of the Wineland voyages based their theories very largely on the Flatey version, and indeed accepted its authority as in every way preferable to the alternative rendering of the story. Laing, for example, in his preface to the Heimskringla, laments the fact that any other document besides the Flatey Book should come into the discussion at all: and Hauk’s version is dismissed by a writer in the _Cornhill Magazine_ for 1872 (vol. xxvi) as ‘a later manuscript ... full of the most marvellous impossibilities’. In a slashing and sceptical paper on the subject in vol. VII of the _Proceedings of the Royal Geographical Society_, by R. G. Haliburton, the same view is emphasized. This writer had but little faith in any of the stories, but he treated the Flatey account as at all events preferable to that of the Saga of Eric the Red.

Perhaps none of the writers cited above can be considered as of very high authority, but their attitude is typical of the older school of thought, and the Flatey Book has as great a critic as Vigfusson on its side. They are quoted to show how widely the opinions of students can vary. For since Gustav Storm in 1887 published his _Studier over Vinlandsreiserne_[72], his views, which have found very general acceptance and still hold the field, have completely reversed the relative status of the different versions. To-day it is the Flatey Book which is criticized, and on all points where it joins issue with the rival version the evidence of the latter is preferred. With great deference to those whose learning has contributed to such a result, it seems to me that such criticism has gone a great deal too far. Let us endeavour impartially to consider the main points wherein there is variance, and thus form our own conclusion as to which story is the more correct.[73]

_Bjarni Herjulfson._

Herjulf, Bjarni’s father, was undoubtedly a real person, whose name and pedigree occur in Landnáma, and it appears to be historically established that he was one of Eric’s companions when Greenland was colonized in A.D. 985 or 986. A well-known headland in Greenland was named after him, and in fact no one hitherto has ventured to question Herjulf’s existence, or his emigration to Greenland.

We start then from the certain fact that Herjulf, Bjarni’s father, has sailed to Greenland about the summer of 986. If he had a sailor son, absent in Norway on a trading voyage, that son on his return to Iceland would almost certainly endeavour to rejoin his parent in the new colony. All the best available pilots are gone, neither Bjarni nor his crew have any clear knowledge of the seas they will have to traverse, and it is with a knowledge of their risk, clearly stated, that they start sailing west in the direction of Greenland, separated from them by a distance imperfectly known, and also, if there is the slightest deviation to the south of Cape Farewell, in the direction of America. To America we are accordingly informed that they came, driven thither by suitable winds and weather. From America, without landing, without any information to impart as to these strange countries, they returned to Greenland, and Iceland saw no more of Bjarni thenceforward. As fiction, it is a pointless and barren narrative, whatever may be its historical interest to persons of a post-Columbian age. It was evidently disappointing to those who heard and to those who subsequently wrote the story. So far from being treated as a hero, as Professor Fischer would have us believe, we are told that Bjarni received nothing but blame for his lack of enterprise and curiosity on the occasion which chance and unsuccessful navigation had thrown in his way. These were not circumstances favourable to the perpetuation of a story devoid of incident in itself and redounding in no way to the credit of the chief actor in it. It would not be surprising to find that even in Greenland Bjarni’s adventure was not long remembered. The disappearance of the tale from Iceland is _a fortiori_ immensely more probable. The interest of narrator and audience alike were in that country exceptionally domestic. It is the rarest possible exception to hear in Icelandic sagas of the exploits of anyone who had permanently left the country, and whose life never again threw him in contact with Icelanders. Bjarni, from the time he set sail from Eyrarbakki, was, short of a miracle, ‘out of the story’, as the Icelandic narrators would have put it. That the popular account of the voyages of Karlsefni and his predecessors should contain no mention of Bjarni is in accordance with every probability. The alternative appears to me to violate everything that experience teaches us of the development of tradition here and elsewhere. A person, possibly it is said fictitious, at best wholly devoid of interest for Icelandic audiences, is credited with an extremely featureless voyage, from which he derives no sort of kudos, the effect of which is—if anything—to some extent to impair the glory of the Icelander Karlsefni. Such inaccuracy as characterizes tradition has, it may be said with the utmost confidence, the effect of merging the exploits of the less well known with those of the more popular hero: the creation of a fictitious hero in addition to the real one is, I submit, the reverse of the normal process.

Thus, the legends which grew up about Charlemagne endowed that hero with the achievements of earlier Frankish kings and chieftains, and in particular absorbed and confused with Charlemagne his ancestor, Charles Martel. The national traditions of centuries were annexed and grouped round Charlemagne and his circle. On a smaller scale, much the same sort of process can occasionally be traced in saga literature. For instance, the earlier versions of the Landnámabók mention a certain Helgi Thorbrandson, who sailed with Eric to Greenland, and was accordingly less known in Iceland than his brothers, who figure largely in the Eyrbyggja Saga. This saga, therefore, ignores Helgi, and does not mention him among the sons of Thorbrand of Alptafjord. Similarly later editions of Landnáma substitute for Helgi’s name that of his brother Snorri, who went out later to Greenland, and was better known in Iceland. The less-known figure disappears and his history becomes absorbed in that of the better-known character. Such is the normal and natural working of tradition.

Prof. Gustav Storm, in his _Studier over Vinlandsreiserne_, makes a great point of the fact that though Bjarni’s voyage is represented as taking place about A.D. 986 nothing was done in the nature of further exploration for a period of about sixteen years. I fail to see the force of this argument. It was not till about a century had elapsed from the time when Gunnbjörn, son of Ulf Kráka, sighted an unknown coast to the west of Iceland, that Eric the Red, having made his adopted country too hot to hold him, followed in his track to Greenland. The battered and storm-tossed remnant who successfully accomplished the emigration to Eric’s new colony had little motive, in Bjarni’s bald description of unattractive coasts sighted from shipboard, to induce them to tempt Providence again. Leif, Eric’s son and the explorer of the future, was born in Iceland after the death of his grandfather, and was in all probability still a child. He is the only son of Eric mentioned in Landnámabók, which is concerned with the Icelandic pedigrees only.

On coming of age, and accomplishing the remarkable voyage from Greenland to Norway, having next carried out the difficult task of converting his countrymen to Christianity, it was time for him to look about for fresh worlds to conquer. The old story was recalled, the ship was manned, and the first real discovery and exploration of the new countries was effected, an exploit for which, in the Flatey Book as elsewhere, Leif receives the entire credit, just as his father, and not Gunnbjörn, is everywhere described as the ‘discoverer’ of Greenland.

_Leif’s Voyage._

Next it is said that whereas, in the Flatey version, Leif’s discovery is represented as the result of an expedition deliberately equipped to investigate Bjarni’s reports, it is uniformly described in every other account as an accidental episode of his return voyage from the court of Olaf Tryggvason in Norway. Here again it must be remembered that Leif was by this time a Greenlander, as to the exact details of whose exploits Iceland was likely to be imperfectly informed and but little interested. The main facts of his career might be known: that he was a son of Eric the Red, that he sailed to Norway and introduced Christianity to Greenland, that he rescued a crew of shipwrecked persons—more especially if, as related in the Flatey Book, one of these was the Icelandic heroine Gudrid—that he discovered somehow and at some time Wineland the Good, and thereby gave rise to Karlsefni’s subsequent expedition. More exact knowledge was not necessary as a prelude to the story of the adventures of the Icelandic hero Karlsefni; in fact, in so far as there is likely to have been any conscious interference with the truth, it may be observed that the less Leif’s voyage was dwelt on the greater would be the credit attaching to the later explorer, in whom alone Icelanders were likely to be generally interested. Such a state of things was eminently calculated to produce the fusion by tradition of two voyages into one, which was likely to be more generally known for two obvious reasons. In the first place, Leif’s voyage to Norway and his return with Olaf Tryggvason’s mission to Greenland was an important fact in the history of that proselytizing king. In the second, it was of interest to the priests who became the historians both of Iceland and Norway. As I have urged already, merger rather than expansion is the normal trend of tradition. The ‘man in the street’ at the present day might well be acquainted, for example, with an incident in the career of Captain Cook, without being able accurately to assign it to the correct voyage of the navigator, or indeed without being certain as to the exact number of the voyages for which he was distinguished. It is far more likely, in my opinion, that such a merger took place in Leif’s story as usually summarized in Iceland than that an imaginary and distinct voyage should have been invented and described with much circumstance and detail.

But, it is said, the Flatey Book’s account stands alone, while that of Hauk, short as it is, is corroborated elsewhere, by a body of independent evidence. On examination, however, this body of evidence shrinks to the dimensions of a single passage, repeated in one context with unimportant verbal variations in a number of different manuscripts.

The oldest extant version of this passage, that occurring in the Friis codex of the Book of the Kings of Norway, will be found included in the Appendix to our translation (p. 74). Another example, from the great Olaf Tryggvason Saga, may be usefully given here, for purposes of comparison:

‘That same spring when Olaf the King sent Gizur and Hjalti to
Iceland, as has already been written, he also sent Leif Ericson to
Greenland, to preach Christianity there. The King got him a priest
and other holy men, to baptize the people there and teach them the
right faith. Leif went that summer to Greenland. He took at sea a
ship’s crew, who were then in misfortune, and lay on a completely
broken wreck of a ship, and on that voyage he found Wineland the
Good, and came at the end of that summer to Greenland, and went
home to Brattahlid to his father Eric. Men called him afterwards
Leif the Lucky. But Eric his father said that the account was
balanced, since Leif had preserved and given life to the men of
the ship’s crew, and had brought the hypocrite to Greenland, so he
called the priest.’

A similar passage in the Heimskringla may also be compared.