As examples of differing genealogies of the same characters in materially related sources we may consider those of the Droplaugarsons and Helgi Ásbjarnarson as given in Droplaugarsona saga, Brand-Krossa þáttr, and Fljótsdœla saga, together with the little that is said about these people in Landnámabok. It will be seen that in all these sources there are differences both of emphasis and in the individual details.

At the beginning of Droplaugarsona saga we are given information on the ancestors of the brothers Helgi and Grímr Droplaugarson, with the main emphasis on how their paternal great-grandfather, Ketill þrymr, came to marry an earl’s daughter in a foreign land. This story is also told in Landnámabók (S 278, H 240), but here there is no mention of either the Droplaugarsons or their maternal ancestry. In the saga itself, however, there appears to be a clear intention to demonstrate a link between Droplaug and her sons and the line of Ketill þrymr:

Some of these people appear with similar genealogies in Landnámabók, but there are so many differences that there is no way that whoever put the saga together got his genealogical information directly from Landnámabók. The altogether more plausible supposition is that Droplaugarsona saga was written using traditional knowledge in the east of Iceland. If we examine the differences between two sources, it becomes clear that Landnámabók was no great mine of information when it came to the central characters of Droplaugarsona saga:

1. According to Landnámabók (S 285, H 246), Yngvildr (Hávarsdóttir, the wife of Þiðrandi and mother of Ketill þrymr in Droplaugarsona saga) was the daughter of Ævarr the Old and Þjóðhildr, daughter of Þorkell fullspakr.
2. None of Yngvildr’s paternal relations nor the in-laws of her brother Bersi as given in Droplaugarsona saga are mentioned at all in Landnámabók, with the sole exception of Egill the Red of Nes (S 292, H 253).
3. Landnámabók mentions Egill’s son Óláfr in passing, but includes neither Yngvildr’s sister-in-law Ingibjǫrg nor Egill’s father Guttormr, both of whom are mentioned in Droplaugarsona saga.
4. Of the three brothers and sisters, Þorvaldr (father of the Droplaugarsons), Hallkatla (wife of Geitir Lýtingsson), and Gróa of Eyvindará, none is mentioned in Landnámabók.
5. The Sturlubók redaction of Landnámabók (S 388) names the wife of Earl Ásbjǫrn as Álǫf, daughter of Þórðr vaggagði, rather than Sigríðr.

In Brand-Krossa þáttr there is nothing on the Droplaugarsons’ paternal relations, but in compensation we are given a considerably more detailed treatment of their mother’s side of the family, as follows:

In Droplaugarsona saga, this line is taken back only as far as the Droplaugarsons’ grandfather, who is named as Þorgrímr as opposed to just Grímr in the þáttr. The genealogy in the þáttr does not appear to be based on older written sources, and so the þáttr may be regarded as good evidence of the existence of oral accounts of the maternal ancestry of the Droplaugarsons. This is supported by the þáttr’s version of the adventure in which an ancestor of the brothers acquires a wife while traveling abroad. In Landnámabók and Droplaugarsona saga this tale is told of Ketill þrymr; in Brand-Krossa þáttr the hero is Grímr, the Droplaugarsons’ great-great-grandfather on their mother’s side (for this story, see p. 205 f below). It is notable that in the þáttr, at the beginning of chapter 2, there appears to be an assumption that the audience will already have some kind of prior knowledge of the brothers’ ancestry. The text reads: ‘Þessa rœðu segja sumir menn til ættar Droplaugarsona, þeirar er ókunnari er. En þótt sumum mǫnnum þykki hon efanlig, þá er þó gaman at heyra hana’ (‘This story is ascribed by some people to the side of the Droplaugarsons’ family that is less well known. But even though some people doubt its veracity, it is still fun to hear it’) (ÍF XI:186). This comment on the brothers’ maternal relations seems to imply that the audience is expected to be rather better informed on their paternal relations. But this knowledge can hardly have been so universally accepted that it prevented a story that we know from older written sources was associated with the better known paternal ancestors from being transferred and told of the ‘less well known’ maternal side of the family. This kind of transference would have created problems if the writer’s and audience’s knowledge of these people had been set down in black and white on the written page. If, however, their information had come largely from oral tradition it is quite understandable that a story of this kind might come to be shifted from one family to another in this way and told of different people.

Fljótsdœla saga gives yet another version of Droplaugarson genealogy:

The most interesting features of this genealogy are how different the mother’s side is from the one given in Brand-Krossa þáttr, and the fact that the tale of the wife won in foreign lands is here told of Helgi and Grímr’s parents. The father’s side, however, is similar to what we find in other sources, other than being somewhat reduced in size and scope. But even here the relationship of Þorkell fullspakr to the brothers is specified more precisely in Fljótsdœla saga than in Droplaugarsona saga, where he is said merely to be ‘frændi Gríms skyldr’ (‘a close relative of Grímr’s) (ÍF XI:176) when he accepts Hrafnkell goði’s bribe to join the hunt for Grímr. Þiðrandi too is placed more fully in Fljótsdœla saga than elsewhere (see p. 222). Fljótsdœla saga is also alone in ascribing such extraordinary longevity to Þiðrandi the Old.

Further evidence of the existence of oral genealogical lore appears in the confusion surrounding Grímr of Gil, who in Fljótsdœla saga is said to be the brother of Helgi and Grímr’s maternal grandmother. In Droplaugarsona saga, their maternal grandfather is called Þorgrímr but given the same address. In Brand-Krossa þáttr this man is named as just Grímr, but with nothing said about where he lived. It is interesting that in Landnámabók (S 388, H 342) too this story is associated with the name of Grímr, but in this case it is Grímr the eponymous settler of Grímsnes in the southwest of Iceland who marries the mother of Arneiðr, named as Álǫf. Genealogical confusion of this kind can hardly be explained other than by assuming that the people who put the texts into book form acquired their historical and genealogical information from oral sources rather than written books—unless they made the whole thing up, which seems highly improbable in view of the number of points of agreement there are between the various sources.

From the genealogies of the Droplaugarsons and their paternal relations detailed above, it seems clear that none of the four sources was based on any of the others. Each contains either more or less information on this family than the others, and where they agree in general they differ on specifics. Landnámabók makes no mention of Helgi and Grímr themselves but tells a story about their father’s family, which is told differently in the other sources and even associated with their mother’s side of the family in Brand-Krossa þáttr and Fljótsdœla saga. The fullest treatment of their father’s family is found in Droplaugarsona saga; in Fljótsdœla saga the details given are generally much sketchier and the emphasis more towards their mother and her relations, even though, to judge from Brand-Krossa þáttr (which only describes the mother’s side), this was the ‘less well known’ side of the family. Despite the contradictions, the patent similarities between the various genealogies mean that it is impossible to imagine individual ‘authors’ constructing them independently out of just a few miscellaneous ‘facts.’ The recourse often taken to ‘explain’ such inconsistencies—that the ‘author’ had misremembered material he had read or heard many years before he came to put his text together—strikes one as clutching at straws, as well as being entirely unamenable to test or proof. Though each of the texts is different (and therefore of very limited value as regards historical research into ‘real’ events), they provide clear evidence of a closely integrated tradition of story material in the background, material which, we may suppose, provided a source of both learning and entertainment to storytellers and their audiences in the east of Iceland.

Helgi Ásbjarnarson is the archenemy of the Droplaugarsons in the eastern sagas and thus it is hardly surprising that Droplaugarsona saga pays rather less attention to him and his relations than to the Droplaugarsons themselves in the genealogical information offered. For instance, it is striking that the saga makes no mention of the fact known from other sagas, that Hrafnkell Freysgoði, the hero of Hrafnkels saga, was Helgi Ásbjarnarson’s grandfather.

Helgi Ásbjarnarson’s genealogy is traced in chapter 3 of Droplaugarsona saga. The main point of the information given appears to be to link Helgi and his brother’s son, Hrafnkell goði, through marriage to Ingibjǫrg and Spak-Bersi Ǫzurarson, thereby establishing a connection with a famous and powerful dynasty known from Landnámabók, the descendants of Brynjólfr the Old and his brothers, Ævarr the Old and Herjólfr. Brynjólfr was the paternal great-grandfather of Spak-Bersi, and an audience familiar with Brynjólfr’s family and authority would have been able to deduce from this that Helgi and Hrafnkell’s goðorð (i.e. their official status as patron chieftains over client farmers) had come to them through these marriage relations. Sigurður Nordal (1940:20) considered this connection to have had a basis in historical fact, since it tied in with his belief that Hrafnkell Freysgoði could not have built up the power base described in Hrafnkels saga at a time when Brynjólfr the Old ruled the roost in the farming lowlands of Fljótsdalur; that is, the goðorð of Helgi Ásbjarnarson and Hrafnkell goði, and what it implied, provided further evidence for his view that Hrafnkels saga was a conscious work of fiction concocted by an individual ‘author.’ It is striking that Hrafnkell Freysgoði is not mentioned at all in Droplaugarsona saga; if there had been stories in general circulation about his rise to power along the lines of those found in Hrafnkels saga, we might have expected Helgi Ásbjarnarson’s authority to have come from his connection with Hrafnkell Freysgoði rather than through marriage to a descendant of Brynjólfr the Old. This silence over Hrafnkell Freysgoði might thus be viewed as lending support to Nordal’s view that Hrafnkels saga is a work of pure authorial fiction. However, the evidence is hardly unequivocal, since Brand-Krossa þáttr trumpets Helgi Ásbjarnarson’s noble ancestry by tracing his descent from Hrafnkell Hrafnsson—‘Þat viljum vér ok segja, hversu Helgi Ásbjarnarson er kominn af landnámsmǫnnum, er gǫfgastr maðr er í þessari sǫgu at vitra mann virðingu (‘We wish to say too how Helgi Ásbjarnarson, who is the noblest man in this story in the judgment of wise men, was descended from original settlers’) (ÍF XI:183)—before setting off on the details of Hrafnkell and his descendants. The þáttr’s independence is apparent from the fact that its information differs from the corresponding genealogy in Droplaugarsona saga:

According to Brand-Krossa þáttr, Helgi Ásbjarnarson is descended from the settler and chieftain Hrafnkell, who ‘fór í Hrafnkelsdal ok byggði allan dalinn sínum mǫnnum, nær tuttugu bœjum, en hann bjó sjálfr á Steinrøðarstǫðum’ (‘moved to Hrafnkelsdalur and settled the whole valley with his men, near on twenty farms, and lived himself at Steinröðarstaðir’) (ÍF XI:183). Steinröðarstaðir is not mentioned in Hrafnkels saga, where the hero is said to have been Hallfreðarson (as opposed to Hrafnsson in Brand-Krossa þáttr) and to have spent most of his life at Aðalból until he buys the little farm at Lokhilla, which later becomes Hrafnkelsstaðir. The Hrafnkell whom Helgi Ásbjarnarson shares the goðorð with is said to be Þórisson in Brand-Krossa þáttr, just as his sister is said to be Þórisdóttir in Droplaugarsona saga—but then in Brand-Krossa þáttr Helgi is not Hrafnkell’s uncle, as in Droplaugarsona saga, but his cousin, their fathers having been brothers as in Landnámabók, which adds that Hrafnkell Hrafnsson lived at Steinröðarstaðir: ‘Hans son var Ásbjǫrn, faðir Helga, ok Þórir, faðir Hrafnkels goða, fǫður Sveinbjarnar’ (‘His son was [sic., sons were] Ásbjǫrn, father of Helgi, and Þórir, father of Hrafnkell goði, father of Sveinbjǫrn’) (S 283, H 244). Sveinbjǫrn Hrafnkelsson is not mentioned in any sagas. According to Hrafnkels saga itself, Hallfreðr came and settled in Iceland with his young son Hrafnkell; Hrafnkell later married Oddbjǫrg Skjǫldólfsdóttir from Laxárdalur (who is otherwise unknown), by whom he had two sons, Þórir and Ásbjǫrn. After Hrafnkell dies in Hrafnkels saga, Þórir takes over the farm at Hrafnkelsstaðir while Ásbjǫrn gets Aðalból—in contrast to what is said in Brand-Krossa þáttr, in which Þórir lives at Steinröðarstaðir and Ásbjǫrn at Lokhellur.

For many years most scholars went along with Sigurður Nordal’s conclusion (1940:22-24) that the ‘author’ of Hrafnkels saga had taken the dream of Hrafnkell Hrafnsson in Skriðdalur from Landnámabók and adapted it point by point to create the dream of Hallfreðr in Geitdalur, and then used this as an element within a narrative of his own creation (see Jóhannesson 1950:xl-xliii), or even as a springboard for an exegesis upon the moral teachings of the medieval Church as found in learned Latin writings, as Hermann Pálsson supposed (1962a, 1971; see p. 25 f). However, in 1976 Óskar Halldórsson (1976:33) pulled the rug from under all such ideas by demonstrating that it was not possible to infer any kind of written connection between the sources, and that the arbitrary combination of both agreement and discrepancy in the dream sequences in Landnámabók and Hrafnkels saga indicated that the latter must have got at least some of its material from oral sources. Halldórsson’s research is informed by a far deeper understanding of oral tradition than we find among any of his predecessors; for instance, he takes account of all sides of the problem and displays a familiarity with the nature and fluidity of narrative in oral preservation. Nothing that has appeared since has confuted his main finding that the correspondences and differences between Landnámabók and Hrafnkels saga should be put down to the oral preservation of the story material rather than uneven use of written texts.

There is no getting away from the fact mentioned previously, that the main subject of Hrafnkels saga, the rise of a chieftain in a part of the country that the lore behind Landnámabók considered to be the fiefdom of Brynjólfr the Old, runs directly counter to the saga world of Droplaugarsona saga, which appears to associate Helgi Ásbjarnarson’s authority with his relationship through marriage to Brynjólfr and which makes no mention whatsoever of any Hrafnkell Freysgoði of Hrafnkelsdalur. On the other hand, this central theme of Hrafnkels saga ties in excellently with Brand-Krossa þáttr and Fljótsdœla saga, which follows Hrafnkels saga directly in the main manuscript. Thus it appears that ideas about Hrafnkell’s chieftaincy are restricted to written sources that are younger than Landnámabók and reveal little awareness of the settlement domain of Brynjólfr the Old and his brothers in the east of Iceland. However, the sources also suggest the existence of oral tales about Hrafnkell and his sphere of influence in Hrafnkelsdalur—and ancient audiences would hardly have been surprised to hear that this man had managed to advance himself through his own efforts and attract support to the east of the great lake of Lagarfljót when he had a grandson, Helgi Ásbjarnarson, who later turned out to be a great chieftain in the same area. The confusions in the discussion have come about through a failure to maintain a clear distinction between what is historically probable and what can happen within the internal premises of a story tradition. Since Hrafnkels saga overlaps so little with other written sagas, no firm conclusions can be reached about whether its main events are the creation of an author working with pen in hand or the product of years of reshaping at an oral stage. Either is possible, although whatever the case we have to assume that ideas of some kind were alive about Hrafnkell in the tradition that is clearly perceptible behind the saga, and so the saga of Hrafnkell must be something more than straight fiction from the bottom up.

Fljótsdœla saga picks up the narrative directly from the point at which Hrafnkels saga leaves off. In this case, therefore, there is no question as to Helgi Ásbjarnarson’s identity and family background:

None of the relatives of Helgi Ásbjarnarson listed here appear in other sagas except Ásbjǫrn, Þórir, and Hrafnkell Þórisson. Hákon is mentioned in Landnámabók, but with no details of his wife. As noted previously, in Brand-Krossa þáttr Ásbjǫrn’s wife is named as Hallbera and said to be the daughter of Hrollaugr, the son of Rǫgnvaldr earl of Møre in Norway—a line of descent presumably intended to redound to Helgi’s glory. The information on these characters in Fljótsdœla saga can thus hardly have come from other books anything like those we know today.

The sources for the genealogy of Helgi Ásbjarnarson appear to be full of internal contradictions. They frequently conflict and each has details not found in the others. There can hardly be any question of literary relations except perhaps between Brand-Krossa þáttr and Landnámabók in the account of Hrafnkell’s dream. However, the þáttr reveals signs that whoever wrote it knew more about his characters than he could have found in Landnámabók, and so the latter can hardly have acted as the sole source for the writer of the former. The same applies to the other texts; in each of them there are clear indications of writers turning to oral sources in search of their material.

The genealogical details given on Helgi Ásbjarnarson are not simply isolated and extraneous information, any more than genealogies in the sagas generally (see p. 83). They serve specific purposes in each of the sagas. In Droplaugarsona saga, Helgi and his close relation Hrafnkell are linked through ties of marriage to the descendants of Brynjólfr the Old, who according to Landnámabók was one of the leading figures in the east of Iceland in directly post-settlement times. It is not unlikely that audiences at one time would have been thoroughly familiar with Brynjólfr and fully aware of the potential significance of a family connection of this sort. In Brand-Krossa þáttr, on the other hand, the focus is on Helgi’s own noble lineage: his great-grandfather is said to have been Earl Rǫgnvaldr of Møre, and his grandfather Hrafnkell Hrafnsson, the settler and chieftain of Hrafnkelsdalur whose star appears to have been on the rise within the saga tradition of eastern Iceland. The purpose of the genealogy offered in Fljótsdœla saga is not so immediately apparent; there is perhaps what purports to be a purely informative function, or at least a pretense of comprehensiveness and verisimilitude, for here we find various names that are not known from any other source and about whom nothing further is said in the saga itself.

Far from being evidence of lapses of memory or slips of the pen as is commonly claimed in such cases, the differences in information offered by the various sources on families and genealogies are generally motivated by purposes specific to each occasion, with modifications and selection dictated by the needs of the narrative rather than any aim to provide a strictly historical record of the actual family connections among Icelanders in the 10th century. This fluidity is precisely what one would expect of learning dependent for its preservation on the unreliable medium of human memory, and the opposite of what one would expect of people in the habit of turning to written books for their information.

As it happens, even scholars steeped in the traditional approach to saga origins have been happy to admit the importance of oral sources in the case of Droplaugarsona saga. In an excellent article on the treatment and historical and cultural source value of individual episodes in the saga in light of how far they can be tested against other sources, Björn Karel Þórólfsson (1928:46) wrote: ‘Sagan er öll ausin af sama brunni, sagnafróðleik, sem gengið hefur í óbundnu máli frá kyni til kyns’ (‘The whole saga is the product of the same wellspring, historical lore, which was passed on in prose form from generation to generation’). However, the main argument that has been adduced to support the idea of oral origins in this saga has centered on its closing words, about a certain Þorvaldr ‘er sagði sǫgu þessa’ (‘who told this saga’) (ÍF XI:180); see also Ólsen (1939:14-8). Unfortunately, the source value of these words is limited, for a number of reasons: the text is corrupt at this point, making this Þorvaldr’s genealogy problematical (he is said to be the son of Ingjaldr, son of Þorvaldr, which does not tie in with what has previously been said about the son of Grímr Droplaugarson and his wife Helga being called Þorkell); and, as Björn M. Ólsen (1939:18) rightly pointed out, the words ‘told this saga’ might equally apply to dictation to a scribe, as in the famous reference to the poet Sturla who is said to have ‘sagt fyrir’ (‘dictated,’ ‘recited’) sagas. Various scholars have suggested that it is unlikely that the ‘author’ of Droplaugarsona saga composed the verses in the saga because they concur so poorly with the material in the prose; as a result the verses have been seen rather as one of his sources, though how genuine they are and how valuable as a historical source is open to question. In general, however, it must be doubted that uncontextualized fugitive verses can ever have been a major ‘source’ for saga writers—especially when this also requires us to suppose that the writers then proceeded to ‘misinterpret’ the verses in various ways.

Richard Perkins (1989) revives another old and well-worn argument in favor of oral accounts behind the sagas, viz. that place names, buildings, and other human artifacts supported the preservation of oral tales associated with them. In a general article on Droplaugarsona saga, Perkins (1986) notes that the saga says that Helgi Ásbjarnarson’s hall ‘stendr … enn í Mjóvanesi’ (‘is still standing at Mjóvanes’) (ÍF XI:155); that Grímr’s hiding place above Krossavík ‘er nú kallat at Grímsbyggðum síðan’ (‘has since then been known as Grímsbyggðir’) (176); and that, after the description of the battle in Eyvindardalur, there is ‘nú lítill grjótvarði, er þeir bǫrðusk’ (‘now a little cairn of rocks where they fought’) (162). References of this sort might point to a story tradition that survived in connection with features of the landscape. All these points might indeed be indicative of an oral tradition, but they might equally be viewed as literary devices put in by a writer with the intention of giving his saga a patina of credibility (see p. 37). For this reason, this study has attempted to approach the problem of orality behind the sagas through different kinds of arguments and methods. As it happens, these arguments and methods also appear to come down in support of Björn Karel Þórólfsson’s main contention, that of the sources used by the writer of Droplaugarsona saga by far the most important was the fund of lore and stories passed on by word of mouth from generation to generation in the areas in which the saga is set.

The battle between Þorkell Geitisson and Bjarni Brodd-Helgason in Böðvarsdalur is described in detail in Vápnfirðinga saga and mentioned in passing in Landnámabók, Ǫlkofra þáttr, and Þorsteins þáttr stangarhǫggs.

Of the fifteen who are named out of the total of thirty-three who took part in the battle in Böðvarsdalur, five are names and little more: Eyjólfr of Víðivellir on Þorkell’s side, and from Bjarni’s side Brúni of Þorbrandsstaðir, Eilífr Torfason of Torfastaðir, Skíði, Bjarni’s foster son, and Haukr Loptsson. Of these, only Eilífr achieves any note in the saga, for having ‘fallen’ in the battle (and then being cured). The names of the others appear solely as independent scraps of information among the followers of their chieftains. These men play no significant part in the plot of the saga and it is thus difficult to explain the inclusion of their names in the text other than as the result of the writer of the saga knowing stories that already included these names. It is well imaginable that these men may have been more than mere names to the original audience of the saga; background knowledge of this sort is at least probable in the case of the brothers from Búastaðir, whom the saga takes for granted that the audience will be able to identify as the sons of Glíru-Halli without needing to spell this out for them.

Moving beyond the references in Landnámabók described above, the battle of Böðvarsdalur is also mentioned in Þorsteins þáttr stangarhǫggs and Ǫlkofra þáttr, and in both texts is treated as if it were a well-known event. In Þorsteins þáttr stangarhǫggs, the first reference occurs in a scene in which the brothers Þórhallr and Þorvaldr are talking at home at Hof and one of them complains that Bjarni has failed to take action against the outlaw Þorsteinn stangarhǫgg (‘Staff-Strike’), remarking that it would have been ‘eigi verra at hafa meir vægt frændum sínum í Bǫðvarsdal ok sæti nú eigi skógarmaðrinn jafnhátt honum í Sunnudal’ (‘no worse to have shown some more mercy towards his kinsmen at Böðvarsdalur and not have the outlaw sitting there in Sunnudalur on a par with him’) (ÍF XI:72). On this reference, Jón Jóhannesson comments in a footnote that the ‘author’ is here assuming that the ‘readers’ are familiar with Vápnfirðinga saga. Jóhannesson offers no reasons to support his assumption that what is being referred to here is the saga in its written form, and in fact no such evidence exists, either here or later in the þáttr when Bjarni and Þorsteinn start to fight. During their confrontation, Bjarni takes a rest and goes to get a drink of water, and puts down his sword: ‘Þorsteinn tók upp, leit á ok mælti: “Eigi mundir þú þetta sverð hafa í Bǫðvarsdal”’ (‘Þorsteinn picked it up and looked at it and said, “You wouldn’t have had that sword in Böðvarsdalur”’) (75). Exactly what these words are supposed to imply is not clear, but a comparison with Bjarni’s performance at Böðvarsdalur also lies behind the comment of the old farmer Þórarinn when Bjarni goes to his farm and pretends he has killed his son Þorsteinn: ‘“Eigi er kynligt at því,” kvað karl, “at þungt veitti við þik í Bǫðvarsdal, er þú bart nú af syni mínum.”’ (‘“I’m not surprised,” said the old man, “that things proved tough for your opponents at Böðvarsdalur, when you’ve now gotten the better of my son”’) (77)—words evidently reflecting the high regard in which Þórarinn held his son. None of these references point to any kind of written literary relationship with Vápnfirðinga saga; it is altogether much likelier that what we have here are allusions to a generally known battle in which Bjarni of Hof fought against one of his kinsmen and came out on top.

The same applies to the reference to Böðvarsdalur in Ǫlkofra þáttr. Here Broddi Bjarnason is depicted as taunting Þorkell Geitisson with the words: ‘hitt ætla ek, ef þú leitar at, er þú munir fingrum kenna þat, er faðir minn markaði þik í Bǫðvarsdal’ (‘but anyway I reckon that if you look you’ll see on your fingers how my father left his mark on you at Böðvarsdalur’) (ÍF XI:93). This implies that Þorkell came out of the battle with an appreciable injury or disfigurement, and that it was Bjarni who caused it. Neither of these points is specified in Vápnfirðinga saga, which says only that Þorkell was wounded on the hand and had to retire from the battle, and that he was later tended at home at Krossavík. Þorsteins þáttr on the other hand clearly implies that Bjarni launched a vigorous attack on his kinsmen, and this may well be a reflection of the idea that it was Bjarni himself who wounded Þorkell. Vápnfirðinga saga does not mention this particularly, but it may well be alluded to in Þorkell’s remark before the battle that ‘frændrnir’ (‘the kinsmen,’ i.e. he and Bjarni) (ÍF XI:61) will face each other in combat.

From this it is clear that the battle in Böðvarsdalur was such a major event in the saga tradition of eastern Iceland that audiences would have been thoroughly familiar with it and known who fought in it and against whom. It was possible to name the people involved without needing to give further information on them all, perhaps because their family details were known, as in the case of the brother from Búastaðir. Later, familiarity with the parts played by Bjarni and Þorkell in the battle was used in other sagas as a spur to action or to display scorn or contempt when the leading participants are reminded of the ill-fated day in their youths when they faced each other in battle, and one of them wounded his own kinsman.

In both Landnámabók and Droplaugarsona saga, the hero of the episode is Ketill þrymr, who according to Droplaugarsona saga was the paternal great-grandfather of Helgi and Grímr Droplaugarson (see p. 192). Landnámabók (S 388, H 432) contains further information pertinent to this version of the story where it describes the viking activities of Hólmfastr and Grímr, the son and nephew of Véþormr of Jämtland on the borders of Norway and Sweden, whom Droplaugarsona saga also describes as a friend of Ketill’s. (This Grímr later moved to Iceland and became the eponymous settler of Grímsnes in the southern lowlands.) The entry says that Hólmfastr and Grímr ‘drápu í Suðreyjum Ásbjǫrn jarl skerjablesa ok tóku þar at herfangi Álǫfu konu hans ok Arneiði dóttur hans, ok hlaut Hólmfastr hana ok seldi hana í hendr fǫður sínum ok lét vera ambátt. Grímr fekk Álǫfar, dóttur Þórðar vaggagða, er jarl hafði átta’ (‘killed Earl Ásbjǫrn skerjablesi [‘Skerry-Blaze’] in the Hebrides and captured as spoils of war his wife Álǫf and his daughter Arneiðr; she fell to Hólmfastr’s share and he passed her on to his father, who made her a handmaid. Grímr married Álǫf, the daughter of Þórðr vaggagði, who had previously been married to the earl’) (S 388, H 432).

The accounts of Ketill þrymr’s marriage in the various redactions of Landnámabók are in all major respects the same as in Droplaugarsona saga, and scholars have been quick to pounce on the parallels and postulate literary relations. For instance, in a footnote in the ÍF edition of Landnámabók, Jakob Benediktsson has no hesitation in claiming that, when compiling Sturlubók, Sturla Þórðarson took the existing accounts in the (lost) Styrmisbók manuscript of Landnámabók and in Droplaugarsona saga and cobbled them together to produce a composite version: ‘In the saga Arneiðr finds the silver and reveals its existence before she marries Ketill, and Ketill offers to transport her to her kinsfolk with the money; in Sturlubók the sequence of the events is changed in order to take account of both sources’ (Benediktsson, J. 1968:295-6, translated).

The passage where Sturla Þórðarson made his changes is indicated in bold in the comparison of the Landnámabók texts. As Benediktsson points out, there appears also to be a literary relationship with the Droplaugarsona saga text (shown in italics). In Sturlubók the text reads: ‘Þá bauð Ketill at flytja hana til frænda sinna, en hon kaus þá honum at fylgja’ (‘Ketill offered to take her to her relatives, but she chose to go with him’). This sentence appears almost word for word in the saga: ‘Þá bauð Ketill henni at flytja hana til frænda sinna með þessu fé, en hon kaus at fylgja honum’ (‘Ketill offered to take her to her relatives with this money, but she chose to go with him’) (ÍF XI:139). The similarity here is so great that some kind of connection through a written text seems indisputable. However, it is very doubtful that the text in question was Droplaugarsona saga in the form in which we know it and that this single sentence can be taken as proof that Sturla had access to the saga in its written form; if this were the case, for instance, it is hard to see why he failed to bring in the Droplaugarsons at this point by continuing the line from Ketill and Arneiðr down to them in the way done so conspicuously at the start of the saga (see p. 192).

Sturla also appears not to have used Droplaugarsona saga in the matter of the name of Arneiðr’s mother. At S 388 he names her as Álǫf. (At S 278 she is not mentioned, while in the addendum in Þórðarbók in H 240 the form is Ólǫf.) In the saga, however, this woman is called Sigríðr, and had Sturla had the first two chapters of the saga in front of him he would surely have used that name at S 388 rather than Álǫf, whom he had not mentioned previously. In addition, S 388 is the only source to include the information that Grímr, the settler of Grímsnes, brought Arneiðr’s mother to Iceland with him. Sturla must either have invented this himself, or had information on these people from oral tradition, or used some text other than the Droplaugarsona saga we know; for instance, Grímr’s marriage in S 388 perhaps recalls the Grímr named in the Droplaugarsons’ maternal line in other sources (see p. 193). Whatever the case, Sturla’s use of sources indicates material differences, and therefore material independence, between the saga and the text that he probably used at S 278 and S 388.

In addition to being, unsurprisingly, considerably more detailed than the short résumé in Landnámabók, the Droplaugarsona saga account of Ketill þrymr’s marriage conflicts directly with Landnámabók on a number of other points:

1. Ketill þrymr’s settlement claim. According to Landnámabók, Ketill and his brother Graut-Atli claimed land in Fljótsdalur, on both sides of the lake, before Brynjólfr (the Old) arrived in Iceland; later Ketill goes abroad and when he returns sets up home at Arneiðarstaðir within his land claim. In Droplaugarsona saga, on the other hand, the brothers live at Húsastaðir in Skriðdalur before Ketill goes abroad; when he returns he buys land to the west of the lake and settles at Arneiðarstaðir, and Atli buys land to the east ‘er nú heitir í Atlavík’ (‘at the place now called Atlavík’) (ÍF XI:140).
2. Véþormr’s family. In Landnámabók, Véþormr is said to be the son of Vémundr the Old, the father of Hólmfastr, and the maternal uncle of Grímr. In Droplaugarsona saga he is the son of Rǫgnvaldr, son of Ketill raumr, and has three brothers, Grímr, Guttomr, and Ormarr. In both sources Grímr takes part in the raids on the Hebrides, but Véþormr goes with him only in the saga. Hólmfastr, who in Landnámabók is the one who captures Arneiðr, is not mentioned in the saga; on the other hand, the saga is alone in mentioning Guttormr and Ormarr.
3. Arneiðr’s price. Landnámabók says that Ketill purchased Arneiðr at twice the price Véþormr had set on her initially, without the actual price being specified. According to Droplaugarsona saga, Ketill got her at a discount because of his friendship with Véþormr, who says, ‘Þú skalt fá hana fyrir hálft hundrað silfrs sakar okkarrar vináttu’ (‘You shall have her for half a hundred of silver on account of our friendship’) (ÍF XI:138).

Putting the two texts together, the common core boils down to this: Ketill goes abroad and buys from Véþormr a handmaid called Arneiðr, the daughter of Earl Ásbjǫrn skerjablesi of the Hebrides; on the way home she finds a hidden hoard of silver which they take back with them to Fljótsdalur, and thereafter they live together at Arneiðarstaðir.

Both Brand-Krossa þáttr and Fljótsdœla saga also have tales to tell about an ancestor of the Droplaugarsons who wins himself a wife while abroad, though in a rather different fashion from that in Landnámabók and Droplaugarsona saga. For one, the hoard of silver is gone and no more is said of it in ancient sources. In its place we get accounts bearing the hallmarks of adventure tales, in which men rescue their future wives from the hands of trolls and giants, along with ‘gull ok silfr ok gersimar’ (‘gold and silver and jewels’) in Brand-Krossa þáttr (ÍF XI:190), or ‘stórfé’ (‘great wealth’) in Fljótsdœla saga (ÍF XI:232).

• In Brand-Krossa þáttr, a farmer called Grímr from Vík in Vopnafjörður loses his precious brindled ox with cross markings on its head into the sea and becomes so depressed that his brother Þorsteinn from Öxarfjörður suggests he go abroad to recover his spirits. The brothers go to Norway with sheepskin cloaks and arrange a deal with a man called Kárhǫfði. This man lives with a farmer called Geitir, who he tells them is ‘vel auðgan ok inn bezta í skuldum’ (‘very well off and with an excellent credit rating’) (ÍF XI:188). Grímr hands over all his goods to Kárhǫfði and goes with his brother to stay with a farmer called Þórir in Trondheim. A little while later they set off to look for Geitir. No one knows anything about him, except that an old man in an isolated valley knows of some cliffs called Geitishamrar and points them off in that direction. In a cave in the rock face they find Kárhǫfði ‘kompán sinn’ (‘their business associate’) (188), who invites them into the living room. There they see the intact skin of the ox Brandkrossi and recognize their sheepskins on the men who are sitting in the room. The proprietor of the cave is Geitir, who has a wife and a daughter called Droplaug. Geitir cheers Grímr up by offering him generous compensation for his ox, which he had lured there by magic, as well as his daughter and lots of money besides. Grímr is happy to accept and they all sleep together in a single bed throughout the winter, the brothers and Droplaug, and love blossoms between Grímr and Droplaug. When spring arrives Droplaug is sent off with a rich dowry and they sail back to Iceland. The þáttr ends by tracing the line of descent from Grímr and Droplaug to Helgi and Grímr Droplaugarson (see p. 193).

It is interesting that, with his last breath, the giant Geitir puts a curse on the sword: ‘Verði þeim sízt gagn at, er mest liggr við’ (‘May it be of least help to them [the relatives of Þorvaldr] when the need is greatest’) (ÍF XI:229). In Droplaugarsona saga (ÍF XI:157-8, 164) it comes out that Helgi Droplaugarson does not have his usual sword with him when he is killed, having left it at Eyvindará for sharpening and borrowed another in its place. In a footnote Jón Jóhannesson comments that the writer of Fljótsdœla saga assumes that Helgi Droplaugarson’s sword is the same one that Þorvaldr got from the giant. It is not really permissible to link saga texts in the way Jóhannesson does, but it is tempting to turn the whole matter on its head and ask whether the Droplaugarsona saga account of Helgi leaving his sword behind might have provided the inspiration for stories about its being cursed. It should be noted that later in Fljótsdœla saga (235) it is made clear that Þorvaldr’s sword has magical properties, for Droplaug asks him to leave it behind at home before he sets out on his final journey—since it would be of no help to him anyway. The saga breaks off before the point at which Helgi Droplaugarson is killed and so it is not possible to say whether there was any reference to Geitir’s curse at this point in the Fljótsdœla saga account.

Though different in most respects, Fljótsdœla saga and Brand-Krossa þáttr are clearly linked by the names of Droplaug and Geitir the giant. The two Droplaugs seem to have similar personalities. In the þáttr she is described as ‘væn kona ok stórmannlig, umsýslumikil ok drenglunduð ok ómálug. Hon var stórlát ok staðlynd, ef í móti henni var gǫrt, fálát ok fengsǫ m ok staðfǫst vinum sínum, en mjǫ k harðúðig óvinum sínum’ (‘a fine-looking and impressive woman, active and noble-minded and reserved. She was strong-willed and unswerving if crossed, composed and unstinting and steadfast toward her friends, but very grim toward her enemies’) (ÍF XI:190). This chimes well with her portrait in Fljótsdœla saga, where the earl warns Þorvaldr about his daughter’s strong character and she is described thus: ‘við aðra menn var hun heldr skapstór, en þess í milli fálát ok steigurlát, en þó var hun afbragð annarra kvenna, bæði at yfirlitum ok atgjörvi’ (‘toward other men [than Þorvaldr] she was rather self-willed, and at other times reserved and haughty, but for all that she stood out from other women both for appearance and attainment’) (ÍF XI:232). Later she is described as ‘ríkilát’ (‘headstrong,’ ‘assertive’) (233). Though the only word common to both descriptions is ‘fálát’ (‘reserved’), it is clear that in many respects they are describing similar personal characteristics, which the audiences were probably meant to associate with Droplaug’s supernatural parentage in the þáttr and her time spent in captivity with the giant in the saga.

The account in Fljótsdœla saga shares with Droplaugarsona saga the fact that the female protagonist is an earl’s daughter (from the Hebrides here, rather than from Shetland) and the appearance of the names Arneiðr and Grímr, in both cases in connection with the ancestors of the Droplaugarsons. However, there are no direct verbal correspondences to indicate that any one of the sagas with a version of this tale was written using any other as its source—with the exception of the Sturlubók redaction of Landnámabók and Droplaugarsona saga, by way of some unknown written source. It therefore seems much likelier that people knew oral stories about some ancestor of the brothers winning a wife in adventurous fashion in a foreign land, but the details had become muddled and varied according to who was telling the story, until no one could say for sure whether the woman in question was Arneiðr, an earl’s daughter taken as spoils of war in the Hebrides, or Droplaug, the daughter of the rock-dweller Geitir in Norway, or Droplaug, the daughter of the earl of Shetland and his wife Arneiðr rescued from a malevolent giant called Geitir. What we have here, then, is an example of a single story appearing in various guises in the preserved texts, a story that changes and develops according to the knowledge and stylistic preferences of the individual saga writers who clothed their oral learning in written form. This kind of variation can hardly be explained other than as the result of a common oral tradition behind all the texts—which by no means excludes the possibility that written texts may have acted as the inspiration for new oral tales, as may be the case with the story in Fljótsdœla saga of the curse laid upon what was later to become Helgi Droplaugarson’s sword.

Both Droplaugarsona saga and Fljótsdœla saga give accounts of how the first wife of Helgi Ásbjarnarson drowns when the ice gives way underneath her on the lake of Lagarfljót:

The sagas agree that Helgi’s wife was on her way home from a visit to her parents (to her mother Ingibjǫrg in Droplaugarsona saga, to her father Bersi in Fljótsdœla saga) when she drowns in Þrælavík on the lake at Lagarfljót together with her oxen and two slaves. About the rest of the events leading up to and following on from this tragedy there are significant differences, as follows:

From this comparison it becomes apparent just how little the two texts agree on specific details, details that ought really to remain the same if this were a case of direct use of written sources, e.g. the name of the wife, the number of children, the purpose and length of the visit, who witnessed the accident, and the means employed to help Helgi get over his grief. One thing that is clear, however, is that Fljótsdœla saga provides a large amount of additional material. Here we find an explanation for the deep personal hatred between Helgi Ásbjarnarson and Helgi Droplaugarson, viz. that the former had married the girl the latter had set his heart on (even if he was only 12 years old at the time). The saga also extends Helgi’s history at Oddsstaðir by having him stay on there and have a child by his housekeeper in between wives; this then provides an opportunity for demonstrating Þórdís’s nobility of spirit when she accepts the child and sends its mother off well provided for. Both these incidents shed light on the characters’ circumstances and spell out what might have been unstated background knowledge in the less detailed account in Droplaugarsona saga: that the two Helgis fell out over a matter of the heart, and that Þórdís todda was a woman of great spirit and strength of mind who came into her own the moment she got free of her bullying brother at Hof, cf. Fljótsdœla saga’s description of her prior to her marriage to Helgi: ‘Hun var skapstór ok skörungr mikill, skafinn drengr ok líklig til góðs forgangs, en þó var hun lítils virð heima’ (‘She was a woman of determination and spirit, a woman of her word and likely to prove a good leader, though she was little valued at home’) (ÍF XI:239). This picture of Þórdís’s miserable treatment in her family home is reflected later in the saga when Helgi Ásbjarnarson urges her not to surrender Gunnarr Þiðrandabani to the mercies of her brother: ‘Máttu ok muna, hversu mikils þú vart virð, meðan þú vart heima. Vartu þá í einum sloppi ok gekkt þar fyrir búi. Sá ek þig ekki betr haldna en eina ambátt, áðr en ek tók við þér’ (‘You might also remember how little respect you were shown while you lived at home. All you had was a single shift to wear as you went about the running of the household. I saw you being kept no better than a slave girl before I took you on’) (282).

The expanded narrative in Fljótsdœla saga ties in with what appears to be its general policy of providing fuller details of the characters and events than we find in the comparable texts, suggesting that the writer was unable to assume the same degree of background knowledge of his audience as the writers of these other texts (see p. 182). But all things considered, it seems likelier that the writer of Fljótsdœla saga got his information on how Helgi Ásbjarnarson’s wife drowned in Þrælavík from hearing stories about it rather than from reading about it in Droplaugarsona saga.

Fljótsdœla saga and Droplaugarsona saga again treat comparable material in their accounts of Helgi Droplaugarson’s killing of Þorgrímr torðyfill (‘Dung-Beetle’), a resident laborer of Þórir of Mýnes, as a result of slanders spread by Þorgrímr about his mother Droplaug. But the circumstances of the killing, and the events leading up to it and following on from it, are reported very differently in the two sources:

These two passages share more material than was the case in the drowning of Helgi Ásbjarnarson’s wife:

1. The slander of one of Þórir of Mýnes’s domestic retainers, Þorgrímr torðyfill, that Droplaug has been unfaithful to her husband, reaches the ears of Droplaug and her sons.
2. Helgi and Grímr have a reputation as keen ptarmigan hunters.
3. Helgi kills Þorgrímr with a spear as Þorgrímr tries to make his getaway from the haymaking on horseback.
4. Helgi and Grímr visit their kinswoman Gróa at Eyvindarstaðir on their way.
5. Reparation is made for Þorgrímr by:
   a. in Droplaugarsona saga, Þorkell Geitisson paying money to Helgi Ásbjarnarson (Þorgrímr being his freedman);
   b. in Fljótsdœla saga, Gróa paying money to Þórir of Mýnes.

The names of the corresponding characters in each saga are the same, only Droplaugarsona saga is more precise in that:

1. It supplies the name Þorfinnr for the man who tells Droplaug about Þorgrímr’s slander; Þorfinnr has no other function in the saga.
2. It names Ásmundr as the farmhand who is working on the hay with Þorgrímr.

Ásmundr does not appear elsewhere in the saga and in fact plays a much less important role than the unnamed farmworker in Fljótsdœla saga, who we accompany throughout the day on a range of outdoor chores—‘en Þorgrímr sat á baki einart meðan’ (‘with Þorgrímr sitting there on the horse’s back the whole while’) (ÍF XI:247)—before returning to the house and telling Þórir what has happened. Other features in Fljótsdœla saga with no parallels elsewhere are the slave Svartr mentioned in Þorgrímr’s accusations and the long, comic episode about Þorgrímr’s brother Nollarr. In this coda to the story, however, Fljótsdœla saga works in with notable skill the idea that at the heart of Helgi Droplaugarson’s problems lie unresolved disappointments in love after suffering, aged only twelve, the loss of the two girls who were dearest to him: first the daughter of his foster father Bersi, who is married off to Helgi Ásbjarnarson; and second Helga Þorbjarnardóttir of Skeggjastaðir, who refuses to accompany him to Eyvindará because she suspects that Nollarr will misrepresent her in front of Bersi. After this experience, the saga says, Helgi Droplaugarson never looked at a woman again.

At first sight it might appear that the Þorgrímr torðyfill episode in Fljótsdœla saga is the work of a conscious author, created by adding to and filling in on the picture given in Droplaugarsona saga. Thus we might imagine the author improvising on the conversation around the fire at Mýnes and spicing it up by making Þorgrímr’s remarks even more unambiguous, and providing a more explicit description of Þorgrímr’s innate nastiness, which in Droplaugarsona saga comes across entirely through his actions: now he is also said to be ‘lítill maðr vexti ok kvikligr, orðmargr ok illorðr, heimskr ok illgjarn, ok ef hann heyrði nökkurn mann vel látinn, þæstist hann í móti ok mátti þat eigi heyra, ok varð hann þeim öllum nökkura flýtu at fá’ (‘small in stature and with quick movements, garrulous and disparaging, stupid and malicious, and if he heard good things being said about anyone he would puff himself up and refuse to hear, and would always have to come back with some put-down’) (ÍF XI:240).

However, there are certain points that argue specifically against such an interpretation:

1. Droplaugarsona saga supplies the names of Þorfinnr and Ásmundr, while the corresponding characters in Fljótsdœla saga remain nameless.
2. The ages of the brothers are given differently: thirteen and twelve in Droplaugarsona saga; twelve and ten in Fljótsdœla saga (which, however, attributes to Helgi significant experience in affairs of the heart at the time of this episode).
3. The time scales are different: in Droplaugarsona saga the whole affair takes place within a comparatively short period; in Fljótsdœla saga the brothers allow things to drift on toward winter before they decide to act.
4. Following the killing of Þorgrímr, the two sagas give different details of the people involved in the settlement, and in Fljótsdœla saga there is the added burlesque episode about Þorgrímr’s otherwise unknown brother, Nollarr.

Other perhaps than Helgi’s reply regarding the hunting, there are no correspondences of diction in the Droplaugarsona saga and Fljótsdœla saga accounts of the killing of Þorgrímr torðyfill that might suggest any kind of literary relationship. So the conclusion remains unaltered: that these texts are obviously related, but this relationship most probably goes back to an oral tradition in which people told stories about the two young ptarmigan hunters from Arneiðarstaðir who attacked the freedman Þorgrímr torðyfill of Mýnes and killed him in revenge for derogatory remarks he had made about their mother. Beyond this, there was nothing in any of the sources to specify how this killing fitted into the overall scheme of the lives of the Droplaugarsons as reconstructed so carefully in the saga we now know and which bears their name. It is precisely the whole—the overall saga—that does not achieve fixed form until set down on the page; up to this point we can only assume the existence of individual, unconnected, and fluid stories about the heroes people chose to tell their stories about.

About Gunnarr Þiðrandabani and the events leading up to and following on from the killing that kept his name alive we have a specific Gunnars saga Þiðrandabana. The same people and events are also mentioned in Fljótsdœla saga and Laxdœla saga, and Gunnarr is also named as a well-known character in Droplaugarsona saga. In broad outline, the common elements of the story begin with Ásbjǫrn vegghamarr moving from the south of Iceland to the east and taking temporary lodgings with the Kóreksson brothers. He sets up farm on his own but falls into debt and flees from his creditors, taking refuge with Ketill þrymr at Njarðvík. Þiðrandi Geitisson happens to accompany the Kórekssons when they go to summon Ásbjǫrn to court, but things go badly wrong and first Ketill is killed and then Þiðrandi, cut down by a spear thrown by a Norwegian merchant called Gunnarr to avenge his host Ketill. Þiðrandi’s powerful relatives seek vengeance and track Gunnarr down to the home of a man called Sveinki or Sveinungr in Borgarfjörður, but he manages to conceal Gunnarr’s presence by a series of tricks. Gunnarr is passed on to the chieftain Helgi Ásbjarnarson who shelters him into the summer and then sends him on to Guðrún Ósvífrsdóttir in the west of Iceland, who gets him out of the country. The main points of this story are shared by all the sources, but their accounts differ so much in other respects that it is worth looking into the connections between them in greater detail. The first place to turn is the improvident builder of drystone walls, Ásbjǫrn vegghamarr himself.

Ásbjǫrn vegghamarr. Ásbjǫrn vegghamarr appears in Gunnars saga Þiðrandabana and Fljótsdœla saga:

There are thus an unusual number of verbal correspondences in the two accounts of Ásbjǫrn vegghamarr, even if they are largely formulaic, and so it is hardly possible to rule out some kind of literary relationship entirely. Nevertheless, the points on which the sagas disagree are far greater in number and seem more compelling in nature:

From the comparison above, the conclusion seems inescapable that the connection between these texts can hardly come down to the direct use of written sources—at least, not in the form in which we know them. It appears that both writers knew about this unappealing southerner called Ásbjǫrn from an oral story tradition and were familiar with the details of how he set in motion the course of events that led to the killing of one of the best loved men in the east of Iceland, Þiðrandi Geitisson. In such a tradition, it is perfectly conceivable that the formulas mikill ok sterkr, taka á skeið heim at bæ, and bakask við eld constituted handy tools in the storytellers’ repertoire.

There is also the possibility that Gunnars saga Þiðrandabana is an original source that served as the inspiration for other sagas in the east of Iceland (see p. 137) and that these sagas were then used by the writer of Fljótsdœla saga. Counting against this, however, is the familiarity with which the writer of Gunnars saga Þiðrandabana refers to people and events. As mentioned earlier in connection with Þorkell Geitisson (p. 162), Þiðrandi is never actually introduced in the saga; instead, it is assumed that the audience realizes that he is a member of the Krossavík clan and thus Þorkell’s brother. This has been criticized as an apparent failure on the part of the writer (see Jóhannesson 1950:xci) but is in fact a typical feature of works written for an audience that is already familiar with the subject matter. In addition, in at least three places the saga makes direct reference to background knowledge of this kind:

1. ‘Nú er þat sumra manna sǫgn, at í þessari ferð hafi verit Helgi Droplaugarson með Þorkatli, frænda sínum, en eigi vitum vér, hvárt satt er’ (‘According to some people, Helgi Droplaugarson was on this expedition with his kinsman Þorkell, but we do not know whether this is true’) (ÍF XI:204; see p. 180).
2. ‘Þar bjó Helgi þá’ (‘Helgi lived there [sc. at Mjóvanes] at the time’) (ÍF XI:207).
3. ‘Ok eigi miklu síðar tókusk til skipti þeira Helga Ásbjarnarsonar ok Gríms Droplaugarsonar, at Helgi var veginn’ (‘And not much later things between Helgi Ásbjarnarson and Grímr Droplaugarson took the turn that Helgi was killed’) (ÍF XI:209).

Jón Jóhannesson (1950:lxxxviii) used these passages to argue that the ‘author’ of Gunnars saga Þiðrandabana knew either an original version of the saga of the Droplaugarsons or an ‘ævi’ of them (a ‘life,’ i.e. a written biographical synopsis), since, to translate his words: ‘The demands made of the readers at these points are greater than can be allowed for solely by reference to oral sources.’ This is very doubtful: the demands made can hardly be viewed as enormous and there seems no real reason not rather to assume a purely orally based background. Jóhannesson’s argumentation is particularly complex and requires the acceptance of a large number of unknowns. To start with, he has to posit an original version of Droplaugarsona saga rather different from the one we have now. Then, to explain the large number of differences between this—a saga we cannot be sure ever existed in written form—and Gunnars saga Þiðrandabana, he is forced to hypothesize that the ‘author’ of the latter based his work on the former solely through memory. The appeals made to memory become even greater when we bear in mind that the text was written for a particular readership and needed to take their knowledge into account too. So, to take Jóhannesson’s premises a step further, even if the writer had read this particular (hypothetical) text, which could hardly have existed in many copies, it stretches credulity beyond breaking point to suppose that the entire intended audience for his newly written saga would also have known this same text through ‘background reading’—a text that the ‘author’ himself had forgotten large parts of.

To further complicate the matter, Droplaugarsona saga itself assumes a familiarity with Gunnarr the Norwegian on the part of its audience in a passage where it describes Grímr’s journey abroad after he is declared guilty of killing Helgi Ásbjarnarson. Grímr has found a ship and arrived at Sogn in Norway: ‘Þá mælti Þorkell stýrimaðr við Grím: “Mat spari ek eigi við þik, en traust hefi ek ekki til at halda þik fyrir Gunnari Austmanni ok engum þeim, er þik vilja feigan”’ (‘Then Þorkell the helmsman said to Grímr, “I don’t begrudge you food, but I don’t trust myself to be able to keep you safe from Gunnarr the Norwegian or any of those who want you dead”’) (ÍF XI:177-8). This is the first and last we hear of Gunnarr the Norwegian in the saga and the audience is patently supposed to know who it is that is being referred to, just as if the reference were to the king himself. It seems very likely, as Jóhannesson points out in a footnote, that this Gunnarr the Norwegian is the same man as Gunnarr Þiðrandabani. But in order to understand the threat he poses to Grímr in Norway the audience needs to know the story that is told in Gunnars saga Þiðrandabana. Here literary relations are out of the question: on the one hand, such an approach would require us to believe that the ‘author’ of Droplaugarsona saga is referring at this point to the written saga of Gunnarr; on the other, we have the widely accepted view, the view of Jóhannesson himself, that the ‘author’ of Gunnars saga used Droplaugarsona saga as a source. This leads us into a vicious circle from which there is no obvious way out. A more plausible explanation for the familiarity with which Droplaugarsona saga refers to Gunnarr the Norwegian, and also the allusions in Gunnars saga Þiðrandabana to the subject matter of Droplaugarsona saga, is that both works are making reference to an orally based body of knowledge of this material, knowledge that was available to both the writers and their audiences. We can therefore say with some confidence that each of Droplaugarsona saga, Gunnars saga Þiðrandabana, and Fljótsdœla saga probably drew on material from oral tradition, and this in turn reduces the likelihood that the oral story matter in Fljótsdœla saga comes mainly by way of written sagas—though it by no means rules out the possibility that people’s appreciation of oral narratives may have been enhanced by the knowledge that the stories they heard contained material that they also knew existed elsewhere in written form in books, even if they may never have seen or read these books themselves.

The killing of Þiðrandi. Þiðrandi Geitisson is killed in the same sources as tell about Ásbjǫrn vegghamarr, that is Gunnars saga Þiðrandabana and Fljótsdœla saga. As already described (p. 194), he is introduced differently in these two texts, as the foster son either of Ketill of Njarðvík in Gunnars saga (which does not mention that Ketill is the brother of Þiðrandi’s mother), or of Hróarr Tungugoði in Fljótsdœla saga (which provides comprehensive details of all Þiðrandi’s genealogy and circumstances). As also noted, the events leading up to Þiðrandi’s participation in the Kórekssons’ journey to Njarðvík to serve the summons on Ásbjǫrn are also described differently in the two sources. In light of this, it is worth comparing how the sources deal with the killing of Þiðrandi itself.

In summary, these accounts agree on the following points:

1. The names of Ketill, Gunnarr, and Þiðrandi;
2. Both Ketill and Þiðrandi are killed;
3. An unnamed woman urges Gunnarr to kill Þiðrandi.

On everything else each account goes its own way:

Once again, the differences between the texts are so great that the most obvious explanation is that each is drawing on material from an unfixed and amorphous oral tradition in the background. The central motif of a guest who shoots the person who least warrants it from a party coming to serve a summons has similarities to the scene in Hœnsa-Þóris saga where a Norwegian fires his bow and hits the boy Helgi Arngrímsson on Hœnsa-Þórir’s ill-fated journey to serve a summons on Blund-Ketill in Örnólfsdalur (see p. 325 ff). Despite there being nothing in the diction to recommend such a connection (other than that the name Ketill appears in all the texts), people have even suggested this may be a case of literary relations: see, for instance, Jóhannesson 1950:lxxxviii. But the circumstances in the eastern sagas are quite different from those in the Örnólfsdalur incident; despite what Gunnarr says afterwards in Fljótsdœla saga, the killing of Þiðrandi can hardly be viewed as in any way accidental or involuntary. In both Fljótsdœla saga and Gunnars saga Þiðrandabana Gunnarr makes direct inquiries about Þiðrandi, either by having him pointed out to him or by asking whose death would constitute the greatest loss to the attackers, and in both cases the bowman is fully aware who it is that he is aiming at. On top of this, in Fljótsdœla saga Þiðrandi is already so badly injured from Ketill’s sword, with an open wound in the back that exposes his lungs, that he would hardly have lived much longer even if Gunnarr had not shot him through the chest with his arrow. It also seems highly improbable that the audience is intended to be taken in by Gunnarr’s expressions of regret addressed to the housemaid when he holds her to blame for having pointed Þiðrandi out to him—since this is precisely what he asked her to do! The underlying idea behind the killing of Þiðrandi is thus quite unlike the incident in Hœnsa-Þóris saga, where the fact that it is an innocent boy who is hit by the Norwegian’s arrow is a matter of pure chance, as in the myth of Hǫðr and Baldr: see p. 325; see also Magerøy 1991. Explanations for the parallels in these texts’ accounts of the expeditions to summons Blund-Ketill/ Ketill þrymr, in which the person who is most innocent ends up being killed, are thus better sought among ideas about formalized narrative themes and migratory motifs than through appeal to the direct use of written sources.

A further dramatic instance of the differences between the sources, and therefore why it is unlikely that one is based on the other, concerns the Kórekssons themselves. Firstly, they are given different names in Gunnars saga Þiðrandabana and in Fljótsdœla saga; and secondly, the sources disagree on their fate after Þiðrandi is killed—in the former, they simply disappear from the story; in the latter we follow Þorkell and Gunnsteinn as, grievously wounded, they struggle home from Njarðvík. In this account, they get as far as Kiðjahvammur but can go no farther; Þorkell takes off his clothes, whereupon his guts pour out ‘ok lét Þorkell þar líf sitt’ (‘and there Þorkell died’) (ÍF XI:266). Gunnsteinn has become so stiff that he cannot move on but is saved by his father having an ominous dream at home at Kóreksstaðir and sending out a shepherd to look for his sons; the shepherd finds Gunnsteinn, gets him into the saddle and leads him home, where Kórekr tends his wounds.

The search for Gunnarr at the home of Sveinki/Sveinungr. As may be imagined, the Norwegian merchant Gunnarr does not get away scot-free after killing the man who Fljótsdœla saga calls ‘hinn fjórði maðr … bezt menntr á öllu Íslandi’ (‘one of the four most accomplished men in all of Iceland’) (ÍF XI:220)—the other three being Kjartan Óláfsson (from Laxdœla saga), Hǫskuldr Þorgeirsson Ljósvetningagoði (Ljósvetninga saga), and Ingólfr the Handsome Þorsteinsson (Vatnsdœla saga), ‘ok er svó mikit af sagt ásjónu þessara manna, at margar konur fengu eigi haldit skapi sínu, ef litu fegrð þeirra’ (‘and this much is said of the looks of these men, that many women were unable to control themselves if they looked upon their beauty’) (221). The first person to take up the case is Þorkell Geitisson, but his part in it varies greatly according to which of the sources one looks at.

It is worth noting that Sveinungr is in any case not entirely the creation of the writer of Fljótsdœla saga and that some kind of knowledge of him must have been passed on by tradition, since Landnámabók (S 287, H 248) records that a man called Þórir lína settled and lived in Breiðavík, the next bay to the south of Borgarfjörður, and had two sons, Sveinungr and Gunnsteinn. These are presumably the brothers that the writer of Fljótsdœla saga places in Borgarfjörður and clearly has more information about than he could have got from Landnámabók. There also seems to have been some common conception of the character of the farmer of Bakki, judging from the way he is described in the various sources:

There is nothing in these descriptions that invites explanation through a literary relationship between the written texts as we have them. On the other hand, the material allows us to suppose that there were stories going around about a big, strong, and independently minded man who had once lived at Bakki in Borgarfjörður. Some people might have told stories about the time when he bested the big shots of the district by concealing the fugitive Gunnarr Þiðrandabani, either from Þorkell Geitisson or Helgi Droplaugarson, before sending him on to Helgi Ásbjarnarson for protection. Others might have known other stories about this man helping Helgi Droplaugarson against Helgi Ásbjarnarson. Certainly, these two incidents would go together badly within a single written saga but might easily have coexisted in an oral storytelling tradition where the common factor was that this Sveinki/Sveinungr of Bakki was enough his own man to stand up against all the most important chieftains in the east of Iceland—in a way not necessarily consistent with their own stories as they eventually came to be written in books.

Helgi Ásbjarnarson shelters Gunnarr and sends him on to Guðrún Ósvífrsdóttir—and the question of ‘saga Njarðvíkinga’. Bjarni Brodd-Helgason’s unsuccessful mission to his sister at Mjóvanes to seek out Gunnarr Þiðrandabani is described in detail in Gunnars saga Þiðrandabana and Fljótsdœla saga. The two accounts were compared earlier (pp. 152 f) and found to exhibit no more signs of direct literary relations than the passages analyzed in this chapter. But at this point in the story of Gunnarr’s tribulations the picture is complicated by the addition of a further source, Laxdœla saga (ÍF V:202-4). Here, just as the wedding celebrations of Guðrún Ósvífrsdóttir and Þorkell Eyjólfsson are getting under way, it is said that Gunnarr Þiðrandabani had ‘verit sendr Guðrúnu til trausts ok halds; hon hafði ok við honum tekit, ok var leynt nafni hans. Gunnarr hafði sekr orðit um víg Þiðranda Geitissonar ór Krossavík, sem segir í sǫgu Njarðvíkinga’ (‘been sent to Guðrún for security and safekeeping and she had accepted him, and his name was kept secret. Gunnarr had been declared guilty of the killing of Þiðrandi Geitisson of Krossavík, as related in “saga Njarðvíkinga”’) (202).

Much scholarly ink has been spilt over this passing mention of an otherwise unknown ‘saga Njarðvíkinga’ (‘saga of the house of Njarðvík’). Many have taken the view that this is simply a reference to Gunnars saga Þiðrandabana and that therefore the ‘author’ of Laxdœla saga used Gunnars saga as a source. This idea was first put forward in an edition of Laxdœla saga published in Copenhagen in 1826 (xviii, with ‘Þáttur af Gunnari Þiðranda-bana,’ i.e. Gunnars saga Þiðrandabana, printed in an appendix) and repeated in the introduction to Kålund’s Copenhagen edition (1889-91:lx). Andreas Heusler (1930:215 [1969:326-7]) regarded the connection in the same light, arguing that, since the account in Laxdœla saga was twice the length of the corresponding passage in Gunnars saga Þiðrandabana, Gunnars saga was the source and the Laxdœla saga account represented an artistic reworking of the material found there. The main dissenting voice was that of Finnur Jónsson. Jónsson (1923:440) saw no reason to view Gunnar saga as the source of the account in Laxdœla saga simply because the latter was longer, the more so since the Laxdœla account contained at least one significant point of substance not found in Gunnars saga. To Jónsson, the account in Gunnars saga Þiðrandabana read rather like a synopsis of the material from Laxdœla saga. In a chapter dedicated to Gunnars saga Þiðrandabana—which, despite its Norwegian hero, he classified as an Íslendingaþáttr on the grounds that it is set in Iceland and deals with Icelandic chieftains and their families—Jónsson (1923:544) made the point that the ‘saga Njarðvíkinga’ mentioned in Laxdœla saga cannot well be Gunnars saga Þiðrandabana since this would be a singularly inappropriate title for a short saga (or þáttr as he would have it) with a Norwegian as its central character. He did not, however, reject the idea that the story of Gunnarr might have formed a part of such a saga.

In his book on the origins of the sagas, Knut Liestøl (1929:41-2) analyzed several examples of parallel accounts of the same incidents in different sagas, including the example from Gunnars saga Þiðrandabana and Laxdœla saga. Liestøl paid particular attention to the difference in length between the two accounts and to what he felt was a striking difference in the incidence of direct speech—virtually absent in the Gunnars saga Þiðrandabana account but comprising between a quarter and a third of the corresponding passage in Laxdœla saga (nine instances). This he compared with what he viewed as equally striking correspondences in subject matter: both accounts of the scene where Þorkell Eyjólfsson recognizes Gunnarr take place in the evening as people are getting washed, and Gunnarr is wearing a hat and gives a false name. The difference, as Liestøl saw it, lay in the use of literary devices on the one hand and the simpler narrative techniques of oral performance on the other. A little later in his study, Liestøl (1929:47-8) points out that the features that Laxdœla saga has over and above the other sagas have the effect of elevating Guðrún’s part, both in word and deed, in accordance with the general tendency throughout the saga. However, there were so many differences between the texts, e.g. in the use of direct speech and in the part played by Snorri goði, that they could hardly have been grounded in the same tradition, nor was it possible to view Laxdœla saga as an expansion or amplification of Gunnars saga Þiðrandabana. Thus, so far as Liestøl was concerned, there was no question of a literary relationship between the two works. These views were vigorously contested by Johannes van Ham (1932:129-32), who felt that the verbal correspondences in the scene where Gunnarr first appears carrying the water for washing were so great that there could be no doubt about the literary dependence between the two accounts. The changes made by the author of Laxdœla saga he considered to be completely natural and thus concluded that there was no reason not to think that ‘saga Njarðvíkinga’ and Gunnars saga Þiðrandabana were one and the same thing (1932:132).

In his introduction to the Íslenzk fornrit edition, Einar Ól. Sveinsson (1934) discussed the sources of Laxdœla saga, including ‘saga Njarðvíkinga.’ Sveinsson (xxxviii-xxxix) was also in no doubt that this referred to Gunnars saga Þiðrandabana (which he calls Gunnars þáttr Þiðrandabana), countering the objection that a ‘þáttr’ should be termed a ‘saga’ in Laxdœla saga on the grounds that this kind of variation in nomenclature was perfectly normal and of no significance. He considered the diction of the two accounts to be so similar and the differences in subject matter so slight that the changes could be put down to ‘authorial taste’ aimed at the greater glory of the heroine (xxxix); for example, in Laxdœla saga it is Guðrún herself who supplies Gunnarr with a ship, as opposed to him having to rely on the help of Snorri goði to get him out of the country as in Gunnars saga Þiðrandabana. According to Sveinsson, the most conspicuous differences lay in the actual presentation, the episode in Laxdœla saga being twice as long and more dramatic and impressive than the one in the ‘þáttr,’ which he characterized as little more than a ‘well worded summary’ (1934:xxxix).

The name Þórir Englandsfari also appears in the second Halldórs þáttr Snorrasonar, which is generally reckoned to be older than Heimskringla. Jóhannesson takes this as evidence of a further literary relationship, with Gunnars saga Þiðrandabana again as the giver, despite the fact that Halldórs þáttr is set in the middle years of the 11th century, some while later than the events described in Gunnars saga. Jóhannesson (1950:lxxxvii) put this anomaly down to the author of the þáttr overlooking the fact that Þórir had died in the encounter at Njarðvík. On the basis of this supposed connection, Jóhannesson moved the date of Gunnars saga back as early as the first quarter of the 13th century. Jóhannesson sees oral influences in the saga in Þorkell’s threefold search for Gunnarr at Sveinki’s farm and in the killing of the calf when Þorkell forces Eyjólfr Ketilsson to divulge the whereabouts of Gunnarr’s hiding place, an incident which he took to be a mark of clumsiness on the part of the author and probably a legendary motif of foreign provenance (1950:lxxxix).

Jóhannesson’s idea that the supposedly poor knowledge of local place names and topography indicated that the saga was written in the west of Iceland has not been subject to thorough reassessment, nor his view that the ‘author’ based his work on stories from the east and that the finished saga was later taken east and acted as the catalyst for the development of saga writing there (1950:xci). But despite this apparent broad scholarly consensus on the literary relations among these sources, it now seems apposite to take the case up afresh and re-examine the three extant accounts of how Guðrún Ósvífrsdóttir held a protecting hand over Þorkell Geitisson’s convicted enemy and got him out of the country.

Bold text indicates verbal correspondences in all three texts; italics indicate correspondences between Gunnars saga and Laxdœla saga; and underline between Laxdœla saga and Fljótsdœla saga.

Certain central points are common to all three texts:

1. The names of the main characters are the same in all cases.
2. Gunnarr Þiðrandabani comes to Guðrún Ósvífrsdóttir.
3. Þorkell Eyjólfsson opposes Guðrún’s protecting Gunnarr.
4. There is a threat of a breakdown in relations between Þorkell and Guðrún.
5. Guðrún gets Gunnarr out of the country.

But there are also several clear differences in how the sagas present the events and the texts are related in a number of different ways. In the comparisons above and below, italics indicate points where there are verbal similarities shared by two texts but not by all three.

Gunnars saga Þiðrandabana	Laxdœla saga	Fljótsdœla saga
1. Þórdís sends Gunnarr to Guðrún.	1. The person who sends Gunnarr to Guðrún is not specified.	1. Helgi Ásbjarnarson has Gunnarr escorted to Guðrún.
2. Gunnarr arrives during the wedding of Guðrún and Þorkell.	2. Gunnarr arrives before the wedding of Guðrún and Þorkell.	2. Guðrún and Þorkell are living at Helgafell and have a grown-up son, Gellir.
3. Gunnarr and Þorkell speak while the guests are washing their hands; Gunnarr holds vatni (‘water’).	3. Gunnarr and Þorkell speak while the guests are washing their hands, i.e. go to vatns (‘water’).	3. Gunnarr and Þorkell speak by the fire.
4. Gunnarr gives an unspecified false name.	4. Gunnarr gives an unspecified false name.	4. Gunnarr gives his name as Gestr.
5. Þorkell threatens to leave the feast if Gunnarr is not sent away.	5. Þorkell gives an order for Gunnarr to be arrested.	5. Þorkell aims a blow at Gunnarr.
6. —	6. Þorkell has promised (heitit) Þorkell Geitisson that he will kill Gunnarr.	6. Þorkell has promised (heitit) Þorkell Geitisson that he will kill Gunnarr.
7. Guðrún is indifferent to whether Þorkell leaves or not.	7. Guðrún gets her own men to protect Gunnarr.	7. Guðrún tells Þorkell to leave Gunnarr in peace and threatens to divorce him.
8. The combined forces of Guðrún and Snorri goði are enough to stop Þorkell from carrying out this threat.	8. Snorri goði intervenes and tells Þorkell to put more value on the trust of him and Guðrún than of Þorkell Geitisson.	8. —
9. —	9. Gunnarr is led away from the feast.	9. —
10. With the support of Snorri goði, Guðrún gets Gunnarr out of the country.	10. Guðrún gets Þorkell to provide Gunnarr with a ship to get abroad.	10. Guðrún provides Gunnarr with the ship belonging to her and Þorkell’s son, Gellir.
11. Gunnarr sends Guðrún fine gifts and invites Sveinki to come to him.	11. Gunnarr lives on his riches in Norway and is out of the saga.	11. Gunnarr is the son of a nobleman and returns the ship with fine gifts.