---

1. Introduction

§1. The study of the scribal culture of post-medieval Iceland has progressed in leaps and bounds in recent years, propelled by an inspiring series of summer schools taught at the Arnamagnæan institutes in Reykjavík and Copenhagen (Óskarsdóttir 2009) and by doughty progress in producing online catalogs and digitizations of Iceland's complete manuscript heritage (primarily through the website Handrit.is; see also Hall and Zeevaert 2018; and Þorgeirsson 2022). That manuscript heritage is large, due to Lutheran Iceland's relatively high literacy, a market too small to support extensive commercial literary printing until the nineteenth century, and the conservatism of the Icelandic language, that allows medieval prose texts to remain readily intelligible into the present day. Particular attention has been directed to the closely related genres of medieval Icelandic romance sagas (riddarasögur) and heroic adventure stories set in pre-Conversion Scandinavia (fornaldarsögur)—which were popular from the thirteenth century into the twentieth—pre-eminently through Kalinke and Mitchell's 1985 Bibliography of Old Norse-Icelandic Romances and the online Stories for All Time catalog of fornaldarsaga manuscripts. Over the last fifteen years, a series of PhD theses and books have assembled detailed case studies of the material philology of Icelandic manuscripts containing fornaldarsögur and riddarasögur (Lansing 2011; Hufnagel 2012; Love 2013; McDonald Werronen 2016; Lavender 2020; Kapitan 2024; cf. Stankovitsová 2023) and of the people who copied those sagas over the half-millennium and more during which they circulated in manuscript (e.g. Ólafsson 2009; Magnússon 2010; Parsons 2019; Parsons 2020). Whereas some other saga genres came to be well represented in print during the eighteenth and nineteenth centuries, most romance sagas became widely available in Iceland in print only late in the nineteenth century, if at all, encouraging the continuation of manuscript culture into the twentieth century, and giving us an especially large and interesting archive of manuscripts to work with. Thus, our capacity to understand fornaldarsögur and riddarasögur both as nodes in the social relations of the people who transmitted and consumed Icelandic literary culture and as literary texts in their own right is expanding dramatically.

§2. Corresponding to these trends, recent research exhibits a renewed appreciation of the value of understanding a text’s stemma—the “family tree” by which we can map which of a text’s manuscripts served as exemplars for which subsequent copies. Once imagined primarily as a tool for reconstructing the lost archetype of a given text, the stemma is also increasingly used as a device for understanding the text’s subsequent history: stemmas can enable us to map, for example, who was copying from whom, or precisely what alterations each scribe made to their exemplar as they read and copied it (cf. Þorgeirsson 2017, 51–55; Lavender 2020, 73–131 is one virtuosic example of what can be achieved through this approach). Such research has been facilitated by the digitization of manuscripts and deployment of software originally designed for phylogenetic analysis in the biological sciences (for Iceland specifically, see Robinson 1989a, 1989b; Robinson and O’Hara 1996; Hall and Parsons 2013; Kapitan 2017; Hall and Zeevaert 2018; Stankovitsová 2023; cf. Platnick and Cameron 1977). The present publication is a case study in these approaches, applying them to Sigurgarðs saga frækna, a riddarasaga close in style to the fornaldarsögur, composed around the third quarter of the fifteenth century. Sigurgarðs saga frækna is among the medieval Icelandic sagas with the largest number of surviving manuscripts. We are aware of 61 witnesses (around twice as many as the average for a romance saga); by the count of manuscripts in Kalinke and Mitchell’s 1985 Bibliography of Old Norse–Icelandic Romances—which is now out of date but probably representative of the relative numbers—it is the seventh best attested riddarasaga, and therefore a particularly informative example of how romance sagas were transmitted (Hall and Parsons 2013, Figure 1 spreadsheet). Standing as a companion to Hall, Richardson, and Þorgeirsson’s publication of a normalized text and translation of Sigurgarðs saga frækna (2013, based on Loth 1962–65, V 39–107), the present article charts the saga’s textual transmission for the first time.

§3. Methodologically, this article is not relevant only to Iceland: Icelandic manuscripts constitute a useful testing ground for methods that can be applied to manuscript traditions from elsewhere around the world. Icelandic fornaldarsögur and romances generally survive in abundant copies, and tend to be transmitted in manuscripts that are dedicated to these genres (Hall and Parsons 2013, §1.3; Kapitan, Rowbotham and Wills 2017; cf. Kapitan 2021; Kapitan and Wills 2023), while the vast majority of Icelandic manuscripts can be consulted in just two cities (Reykjavík and Copenhagen) and are made more accessible again by the exceptionally extensive digitization of Nordic manuscript collections. Thus, Icelandic saga manuscripts constitute a rich and unusually accessible archive of a scribal culture, and a good testing ground for methodological innovation. The main methodological contribution of this paper is to demonstrate how, with the use of computer-assisted stemmatology, our knowledge of the stemmas of fornaldarsögur and riddarasögur is now approaching the critical mass needed for us to comment not only on how individual sagas were transmitted, but on how the manuscripts that contained them were anthologized: whether a given manuscript represents, for example, the wholesale copying of an exemplar or an anthology drawn from multiple sources. The ability to study anthologization in this way will enable a significant advance in our understanding of Iceland’s scribal and literary culture. The article investigates anthologization by closely examining one branch of the Sigurgarðs saga frækna stemma (manuscripts descended from Copenhagen, Den Arnamagnæanske Samling, Rask 32 4to) and integrating our new findings about that branch with a synthesis of both existing knowledge and targeted new research concerning the transmission of Ambrósíus saga og Rósamundu, Drauma-Jóns saga, Hálfdanar saga Barkarsonar, Hálfdanar saga Eysteinssonar, Nikulás saga leikara, Nítíða saga, and Sigurðar saga turnara. In doing so, we produce an early case study of what will be a step-change in our capacity to understand the literary habits of Icelandic scribes.

2. Manuscripts

§4. We aimed to filiate all manuscripts of Sigurgarðs saga frækna held in public collections, and those held privately if the opportunity happened to present itself. Our list of manuscripts was derived from Kalinke and Mitchell’s bibliography of the Icelandic romance sagas (1985), supplemented by systematic consultation of the catalog of the National Library of Iceland (Ólason 1918–37; Blöndal, Helgason, and Helgason 1947–96), electronic searches of web-based catalogs (primarily Handrit.is and Stories for All Time: The Icelandic Fornaldarsögur) and chance discoveries made by Parsons which have been catalogued in the Icelandic Immigrant Literacy Database (IILD). We also reference Kålund (1900), Stefán Einarsson (1948) and the National Library of Sweden’s online catalog Manuscripta.se. New manuscripts still come to light fairly frequently: in 2013, our count was 53 (Hall, Richardson, and Þorgeirsson 2013, 82); it now stands at 61. The 58 manuscripts which are included in our transcriptions and stemma, together with links to key online catalog records currently extant, as are shown in Table 1. Unless otherwise stated in the discussions below, information about the history of the manuscripts is sourced from these catalog entries. We also included in the stemma the 1884 popular edition, Sagan af Sigrgarði frœkna, edited and published by Einar Þorðarson in Reykjavík. This is important both as a witness to lost manuscripts—Þorðarson’s exemplar seems not to survive—and as the ancestor of extant ones.

Table 1: Manuscripts used in this study

Place	Collection	Shelfmark	Link (Handrit.is)	Link (other catalogs)	Date
Baltimore, MD	Johns Hopkins University	Ottenson 1		Stefán Einarsson	1798
Borgarnes	Héraðsskjalasafn Borgarfjarðar	MS 14 / Einkaeign 10	Handrit.is	Stories for All Time	1862–67
Copenhagen	Den Arnamagnæanske Samling	Rask 32	Handrit.is	Stories for All Time	later C18
Copenhagen	Den Arnamagnæanske Samling	AM 592a 4to	Handrit.is	Stories for All Time	C17
Copenhagen	Matthew Driscoll	Einkaeign 19	Handrit.is		1875
Copenhagen	Det Kgl. Bibliotek	NKS 1804 4to		Stories for All Time	1681
Copenhagen	Det Kgl. Bibliotek	Thott 978 fol		Kålund	late C17
Ithaca, NY	Cornell University, Fiske Icelandic Collection	Ic F75 A125 8vo			c. 1823–24
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍB 165 4to	Handrit.is	Stories for All Time	1778
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍB 426 4to	Handrit.is		1877
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍB 185 8vo	Handrit.is	Stories for All Time	1770
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍB 224 8vo	Handrit.is	Stories for All Time	1740–60
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍBR 38 8vo	Handrit.is	Stories for All Time	1828–31
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	ÍBR 44 8vo	Handrit.is		1854
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	JS 632 4to	Handrit.is	Stories for All Time	1799–1800
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	JS 411 8vo	Handrit.is	Stories for All Time	C19
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 222 fol	Handrit.is	Stories for All Time	1696
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 423 fol	Handrit.is	Stories for All Time	C18
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 354 4to	Handrit.is	Stories for All Time	C18
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 644 4to	Handrit.is	Stories for All Time	1710–50
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 998 4to	Handrit.is	Stories for All Time	C18–19
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1217 4to	Handrit.is	Stories for All Time	1817
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1305 4to	Handrit.is		1869–78
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1500 4to	Handrit.is	Stories for All Time	1880
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1637 4to		Stories for All Time	1760–1800
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1785 4to			1833
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 2316 4to		Stories for All Time	1850
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 3165 4to	Handrit.is	Stories for All Time	1870–71
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 3891 4to			late C19
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 3966 4to	Handrit.is		1869–71
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4447 4to			1868–69
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4718 4to	Handrit.is		1875
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4825 4to	Handrit.is	Stories for All Time	c. 1775–1825
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 5480 4to	Handrit.is		C20
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 5567 4to	Handrit.is		1913
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 1446 8vo	Handrit.is		1864–71
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 2484 8vo	Handrit.is		c. 1852
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 2786 8vo			1869
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 3938 8vo			1872
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4070 8vo			1862
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4847 8vo	Handrit.is		1868–74
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4867 8vo	Handrit.is		1870
Reykjavík	Landsbókasafn Íslands – Háskólabókasafn	Lbs 4977 8vo	Handrit.is		1896
Reykjavík	Stofnun Árna Magnússonar	AM 167 fol	Handrit.is		c. 1660
Reykjavík	Stofnun Árna Magnússonar	AM 556a 4to	Handrit.is		C15
Reykjavík	Stofnun Árna Magnússonar	AM 588m 4to	Handrit.is		C17
Reykjavík	Stofnun Árna Magnússonar	AM 588n 4to	Handrit.is		C17
Reykjavík	Stofnun Árna Magnússonar	AM 123 8vo	Handrit.is	Stories for All Time	c. 1600
Reykjavík	Stofnun Árna Magnússonar	GKS 1002 fol	Handrit.is		1667
Reykjavík	Stofnun Árna Magnússonar	SÁM 47	Handrit.is		1867–68
Reykjavík	Stofnun Árna Magnússonar	SÁM 131	Handrit.is		1871–90
Stockholm	Kungliga biblioteket	Islandica Papp fol 1		Stories for All Time; Manuscripta.se	1600–50
Stockholm	Kungliga biblioteket	Islandica Papp fol 66		Manuscripta.se	1690
Stockholm	Kungliga biblioteket	Islandica Papp 4to 17		Stories for All Time; Manuscripta.se	1640–71
Stockholm	Kungliga biblioteket	Islandica Papp 4to 27		Manuscripta.se	c. 1650
Stockholm	Kungliga biblioteket	Islandica Papp 8vo 6		Stories for All Time; Manuscripta.se	1674
Utah	Privately owned	Leifson 1		IILD	C19
Winnipeg, MB	University of Manitoba, Elizabeth Dafoe Library	ISDA JB3 6 8vo		IILD	C19

Omissions from our survey of which we are aware are as follows. The two publicly held manuscripts that we have omitted are New Haven, CT, Yale University, Beinecke Library, Z 113.82, dated to 1806 (unavailable due to conservation when Hall visited U.S. collections) and Reykjavík, National Library of Iceland, Lbs 5767 4to, dated to 1911 (formerly Böðvar Kvaran, Tjaldanes, MS I 2.b and in private ownership when Hall visited Icelandic collections). Kalinke and Mitchell also listed the privately held Jón Ófeigsson, Hafnarnes, Hornafjörður MS 1 (nineteenth century). It is also worth adding that two manuscripts (Rask 31 4to and Lbs 2319 4to) contain a saga entitled Sigurgarðs saga frækna and are listed by Kalinke and Mitchell as containing the saga under discussion in this article, but in fact contain the saga usually known as Sigurgarðs saga og Valbrands.

§5. Noteworthy inclusions are two North American manuscripts unknown to Kalinke and Mitchell. Winnipeg, MB, Elizabeth Dafoe Library ISDA JB3 6 8vo has been described by Parsons (Hall and Parsons 2013, §4.1). The Spanish Fork, UT manuscript was in the private ownership of the late Thor Leifson (1928–2024) of Provo, Utah. The manuscript belonged to Thor’s grandfather and namesake, Sigurður Þorleifsson (1859–1922), who left Iceland in 1884 and adopted the name Sigurdur Thor Leifson. Sigurður emigrated from Hólshús in the Westman Islands to Spanish Fork, Utah. Sigurgarðs saga frækna (ff. 1r–29r) is the first of five sagas in Sigurður’s manuscript, the others being Ajax saga frækna (ff. 29r–34r), Ála flekks saga (ff. 34v–54v), Nikulás saga leikara (ff. 55r–71r) and Friðberts saga frækna (ff. 71r– 85v). The manuscript is defective in its present state, with the final saga ending in chapter 8 in the middle of a battle scene. It is in a single hand. According to a note by Thor Leifson, the scribe was Sigurður Þorleifsson himself, who walked up Spanish Fork Canyon to borrow an exemplar from an unnamed friend, recopying it in its entirety. Photographs of the manuscript, taken by Thor Leifson, are available on the Icelandic Immigrant Literacy Database (IILD).

§6. A final manuscript in North America that should be mentioned is ISDA JB3 1 4to, which came to the Icelandic Collection of Elizabeth Dafoe Library at the University of Manitoba from the Jón Bjarnason Academy in Winnipeg, Manitoba, Canada. It is a nineteenth-century manuscript of 187 leaves that was presumably brought to Canada by an unknown Icelandic immigrant and donated to the Jón Bjarnason Academy between 1913 and 1940. It contains nine sagas in a single unknown hand: Falentíns og Ursins saga, Adonias saga, Sigurgarðs saga og Valbrands, Hálfdanar saga Eysteinssonar, Flóres saga konungs og sona hans, Sörla saga sterka, Ektors saga, Nikulás saga leikara, and Flóres saga og Leó. In the manuscript, the title of Sigurgarðs saga og Valbrands is given as “Sagan af Sigurgarði frækna og Valbrandi svikara” (“The saga of Sigurgarður the Valiant and Valbrandur the Traitor”). Examination of the manuscript confirms that it does not contain Sigurgarðs saga frækna.

3. Stemma and methods

§7. By far the most substantial labor represented by this article is manifested in Spreadsheet 1 and Figure 1. Figure 1 represents our conclusions in the familiar visual form of the stemma codicum. Spreadsheet 1 contains the data underlying this visualization, fulfilling our commitment to open-data approaches to publishing. It summarizes key information about the origin of each manuscript, including giving co-ordinates that can be used to make electronic maps, and relational information that can be used to create visualizations of our stemma. It publishes transcriptions of five sample passages spread through the saga from each of the 58 manuscripts, producing a corpus of transcriptions totaling around 45,000 words. For heuristic purposes, it reconstructs a text for every node in the stemma where we infer a lost manuscript (though we have not ventured to reconstruct the text of the archetype, feeling that this is one step further than is useful given our shaky understanding of the top of the stemma). It numbers each alternative reading to facilitate computer analysis. In its .odt and .xlsx forms, the spreadsheet is color-coded to indicate manuscript families and variants, but it is also provided in the highly future proof .csv format.

§8. The method for constructing the stemma was largely that described in Hall and Parsons (2013, on Konráðs saga keisarasonar) and Hall and Zeevaert (2018, on Njáls saga; cf. Zeevaert and others forthcoming). In constructing a stemma that takes into account (nearly) the entire manuscript tradition, with a large corpus (i.e., large survival rate), we have arrived at a stemma with three branches at the top, contrary to what might be considered a more usual two-branch stemma (on the discussion of bifurcating stemmas in Old Norse philology see Haugen 2016; see also Trovato 2017).

§9. Although it is only now reaching publication, the present research on the stemma of Sigurgarðs saga frækna was begun before the articles just cited were published: the methodological questions raised by Hall’s initial forays into establishing the first stemma of Sigurgarðs saga frækna prompted those studies, which used past research to validate innovative methods for gathering, handling, and publishing stemmatic data. In 2013, we found that two samples totaling around 317 words enabled the independent production of a stemma largely consistent with the findings of past work (whose methods were not described, but can be taken as representing what had up to then been seen as an acceptable standard in the field). Subsequent application of the same approaches to Njáls saga—whose stemma is complicated by a large number of fragmentary manuscripts and manuscripts with multiple exemplars—has largely enabled the replication of past research using samples between around 270 and 392 words. For Sigurgarðs saga frækna, we produced a stemma on the basis of five samples, encompassing the beginning and end of the text, along with three other moments chosen for their literary interest and fairly even distribution across the text. In Loth’s edition, these five passages together comprise 616 words; in our data, the average sample length was 108 words. We created an independent stemma for each of the five samples, though where individual samples pointed in the same direction but did not offer enough evidence to situate a manuscript in the stemma precisely, the evidence of multiple samples was combined to produce the highest-resolution conclusion possible. The creation of our Sigurgarðs saga frækna dataset will enable the systematic digital production of stemmas based on different numbers or subsets of samples, enabling objective assessment of how far smaller sample sizes produce different results, facilitating a deeper understanding of how large a sample researchers should expect to need to filiate Icelandic saga manuscripts reliably. (Our anecdotal impression is that extra samples are helpful for improving the resolution of the stemma when handling conservative copying, and for demonstrating the occasional case where a manuscript draws on multiple exemplars, but that single samples are generally sufficient.) In 2013 we also made use of manuscript dating in determining the stemma: if the textual evidence suggested that manuscript A was the parent of manuscript B but manuscript dating suggested that manuscript A was produced later than manuscript B, we filiated the manuscripts as siblings, children of a lost manuscript. As we discuss below in §5.3, however, manuscript dating is not always correct, and stemmatic evidence can usefully prompt checking of received wisdom on dating.

§10. The naivety of Hall’s early data-collection regarding Sigurgarðs saga frækna has some legacies in the present article. The Sigurgarðs saga frækna transcriptions are diplomatic, but we struggled to maintain consistent approaches while working with varied scripts and sometimes semi-legible manuscripts over many years of discontinuous data-collection (and even during the much more concentrated period of research represented in Hall and Parsons 2013). Thus although the evidence provided by diplomatic transcriptions is sometimes stemmatically useful (see Part 4 below), our paleographical standards are not consistent enough for future researchers to make absolutely reliable use of the transcriptions as evidence for, for example, the use of abbreviations, or spelling variation between ð and d, u and v, or i and í. By the time of Hall and Zeevaert’s work on Njáls saga (2018), we had opted for transcribing into standard modern Icelandic spelling, which makes it easy to ensure consistency, and moreover facilitates electronic analysis. We have left the Sigurgarðs saga frækna dataset, however, in its loosely diplomatic form.

§11. We also regret not selecting a sample from the short section where the earliest two surviving, fragmentary manuscripts (AM 556a 4to and AM 123 8vo) overlap. Moreover, we found that our data for the seventeenth-century siblings AM 592a 4to (which is only partly legible) and GKS 1002 fol (which happens to be highly abbreviated for the first and last of our samples in particular) was insufficient to place them confidently in the stemma. We therefore cross-checked our samples for eight early manuscripts by tabulating the 151 occasions where Loth recorded an alternative reading in those manuscripts that she referred to for her edition (AM 556a 4to, AM 588m 4to, AM 167 fol, AM 123 8vo, AM 588n 4to) and adding the corresponding readings from three more early manuscripts (Lbs 423 fol, Lbs 222 fol, and GKS 1002 fol; Spreadsheet 2). This provided a cross-section of readings throughout the saga. This sample was perhaps not ideal: Loth’s apparatus is certainly not comprehensive. Still, electronic stemmas produced using the whole dataset, only that part of the dataset where all manuscripts sampled are intact, and only those readings which we deemed likely to be significant, returned consistent results showing that the fragmentary AM 556a 4to and AM 123 8vo, along with the sibling pair AM 592a 4to and GKS 1002 fol, are most likely independent witnesses to the archetype of Sigurgarðs saga frækna.

§12. We have also opted not to produce the kind of interactive HTML stemma presented in Hall and Parsons (2013) due to our present lack of an automated process to facilitate this otherwise onerous undertaking. Our experience publishing these with Digital Medievalist in 2013 also underscores the challenges of future-proofing even simple interactive publications: in updates to Digital Medievalist’s website, these files were lost, and requests to fix them have not at the time of this publication succeeded (the stemmas remain functional, however, in the preprint version). But perusal of our spreadsheet nonetheless makes it relatively easy to check our conclusions, and the publication of our data there would at least make the production of an interactive stemma straightforward for future researchers.

4. How Sigurgarðs saga frækna corroborates what we already thought we knew

§13. The distribution of Sigurgarðs saga frækna manuscripts in time and space is consistent with what has become a well-established pattern for fornaldarsögur and riddarasögur, a picture which began to come into focus with the work of Peter Springborg (1977) and has received much empirical backing in recent case-studies of saga-transmission, most conveniently expounded by Philip Lavender through his study of Illuga saga (2020, 73–131; cf. Hall and Parsons 2013, §1.3). The data and stemma published here provide a corroborating case study of this pattern.

§14. The origins of the textual transmission of Sigurgarðs saga frækna lie in the Middle Ages, but are otherwise murky, since few saga manuscripts can be localized during this period. Indeed, Hall and Parsons (2013, §43) found that although the sampling method they used was generally successful, it did not generate enough data to reach a reliable conclusion about the top of the Konráðs saga stemma, partly because the early manuscripts of the saga were fragmentary. Some other major stemmatic work on the romance sagas has indeed declined to speculate on the top of the stemma entirely (e.g. Slay 1997 and McDonald 2014; McDonald Werronen 2016; cf. the studies listed in §5.1 below). The earliest manuscript of Sigurgarðs saga frækna is the late fifteenth-century Eggertsbók (AM 556a 4to, once part of the same manuscript as AM 556b 4to—probably, as Lethbridge has suggested, the latter part), better known as the earliest manuscript of Gísla saga Súrssonar (Lethbridge 2012b, 396). Eggertsbók lacks more than half of the beginning of the saga. The second oldest manuscript, the sixteenth-century AM 123 8vo, has a number of lacunae and quite a limited overlap with Eggertsbók. The top of our stemma must therefore be regarded as fairly arbitrary.

§15. Our finding that the surviving medieval manuscript of Sigurgarðs saga frækna is the ancestor of only one other manuscript is similar to conclusions drawn regarding other romance sagas whose stemmas have already been surveyed in detail. It is not self-evident that this would have been so: assuming that Eggertsbók is the archetype of all surviving manuscripts would be a parsimonious interpretation and so is methodologically attractive. Hast identified Eggertsbók as the archetype of all surviving Harðar saga manuscripts (1960a, 1960b), and it is thought to be the archetype of all manuscripts of the shorter version of Gísla saga (Guðjónsson 2010, 108). Our conclusion, however, is in line with Loth’s opinion (1962–65), as implied by her edition’s willingness to take readings from other manuscripts while using Eggertsbók as a base text—not least on p. 197, line 17, where she draws on text from AM 123 8vo and AM 167 fol to correct what seems likely to have been an eye-skip omission from Eggertsbók (cf. Spreadsheet 2). Strikingly, Eggertsbók (and sometimes its only child, AM 588m 4to) calls Ingigerður’s kingdom Taricia instead of the well attested place name Tartaria. Taricia can easily be explained as a misreading, showing c/t confusion, of an abbreviated form of Tartaria like Tart’ia. Admittedly, a lost copy of Eggertsbók that corrected this reading to Tartaria could stand between Eggertsbók and other Sigurgarðs saga frækna manuscripts, or copyists might independently have changed Eggertsbók’s reading. Tellingly, however, Lbs 423 fol, copied in 1733 at the behest of Bjarni Pétursson at Skarð á Skarðsströnd, contains a scattering of archaic spellings in places where Eggertsbók lacks them: v̈vine (for standard modern Icelandic óvini), showing ú- for the prefix ó-, in §1; at (for að) in §2 and §3; examples of h-loss in lute (for hluti) in §3 and liöp (for hljóp) in §4; and the Latinate accusative Tartariam in §3, a feature that tended to fall out of use as Icelandic romances were copied and so may be archaic. Taken together, such features are not of a kind which are likely to have been produced by self-consciously archaizing scribes (like Jón Þórðarson who copied Lbs 222 fol in 1696 for Magnús Jónsson í Vigur, and the scribes of ÍBR 38 8vo, written 1828–31—who wrote -r in preference to modern Icelandic -ur whether or not this was etymologically correct). These spellings are therefore likely to descend from a lost medieval exemplar.

§16. A good impression of the kind of lost manuscripts that might lie behind our surviving lines of stemmatic descent is perhaps given by AM 123 8vo (on which see Lavender 2020, 77–84), the saga’s second earliest manuscript and a major witness to the earliest version of the saga. A rare example of a sixteenth-century saga manuscript, this no-nonsense octavo volume was made to be read. Copied onto vellum originally pricked for quarto production, it is readily portable. In the century or so over which it was used before coming into the hands of Árni Magnússon, it was read almost to death, now being both dirty and fragmentary and bearing many marks of use, provoking only a laconic accession slip by Árni: “aptanaf Tiodels sógu, af jlluga gridarfostra. Sigurgardz saga. af Drauma Jons. nockur æfintir. af þorsteini bæiarmagn. recentissima membrana” (“the latter part of Tiodels saga, [some] of Illuga saga Gríðarfóstra, Sigrgarðs saga, of Drauma-Jóns saga, some exempla, and [some] of [the saga of] Þorsteinn bæjarmagn. A very recent parchment”). Given the state of AM 123 8vo, it is no surprise that other Sigurgarðs saga frækna witnesses should be lost.

§17. After about 1600, our understanding of the transmission of Sigurgarðs saga frækna becomes much more detailed, and again consistent with other evidence. The saga begins to appear in scholarly seventeenth-century copies associated with Icelandic humanism. Their production, as far as we can tell, was focused on a few centers of scholarship, particularly Iceland’s two episcopal seats: Skálholt in the south and Hólar in the north (cf. Lavender 2020, 99–107 for the case study of Illuga saga). One of the Sigurgarðs saga frækna manuscripts that can be localized for this period is the monumental two-volume GKS 1002–3 fol, produced in the late 1660s by Páll Sveinsson (1650–1703), who was based at the farm of Geldingalækur, about 25 kilometers south of Skálholt. Their commissioner was the wealthy farmer Jón Eyjólfsson of Eyvindarmúli, another 25 kilometers or so south-west of Geldingarlækur, who was the cousin of both Bishop Brynjólfur Sveinsson of Skálholt and the bishop’s wife, Margrét Halldórsdóttir. Subsequently rebound as a velvet-covered, gold-edged set, the volumes were presented to King Christian V of Denmark in 1692 by Björn Þorleifsson, the future bishop of Hólar, who belonged to the extended kinship network of the farmer at Eyvindarmúli (Parsons 2024). This exceptionally late vellum production, exhibiting the thick parchment characteristic of seventeenth-century Iceland, offers abridged and consequently innovative versions of its texts. Árni Magnússon’s catalog of its contents for the king, inserted into GKS 1002 fol as front matter and now cataloged on Handrit.is, offers a poker-faced but implicitly skeptical representation of Sigurgarðs saga frækna in an elevated Danish:

en relation om en kongeson af Ryssland, ved nafn Sigurgard, som fick til egte Ingerd, kong Herculis daatter af Tartarien, oc blef der saa siden konge. denne Roman haver aldelis intet hvor af mand kunde udleede hvad tiider dens auctor skulle hafue villet applicere den til.

an account of a prince of Russia called Sigurgard, who succeeded in marrying Ingerd, the daughter of King Hercules of Tartarien, and thus became king there. This romance contains absolutely nothing from which one may deduce what period its author might have wanted to set it in.

Enough of the damaged, unlocalized, seventeenth-century AM 592a 4to can be made out to demonstrate that it shared a lost ancestor with GKS 1002–3 fol and that the pair comprise independent witnesses to the archetype of Sigurgarðs saga frækna. The same milieu produced AM 167 fol, once part of a massive volume also comprising the present AM 123 fol, AM 163h fol, AM 163h fol and AM 164f fol, copied around 1660 by the scribe and poet Arnór Eyjólfsson (1642–95) at Flókastaðir, Rangárvallasýsla, around 50 kilometers south of Skálholt. AM 167 fol is again an independent witness to the Sigurgarðs saga frækna archetype.

§18. Associated with Iceland’s other episcopal seat, Hólar, is Papp 17 4to: in this manuscript, from the second half of the seventeenth century, Sigurgarðs saga frækna was copied by Þorlákur Sigfússon (d. 1693) of Glæsibær in Krækingahlíð, in Eyjafjörður, one fjord to the east of Skagafjörður, where the episcopal seat of Hólar lay. Some other parts of this volume were written by Brynjólfur Jónsson of Efstaland, Öxnadalur, about whom little is known but who certainly collaborated with Þorlákur Sigfússon and undertook commissions for Bishop Þorlákur Skúlason of Hólar (1597–1656) and probably Bishop Þorlákur’s son and successor Gísli Þorláksson (1631–84) (Lansing 2011, 61). If our stemma is correct, Papp 17 4to was also based on a lost exemplar, but is one of a cluster of sixteenth- to seventeenth-century Y-class manuscripts, suggesting a bustling culture of copying and recopying.

§19. This seventeenth-century humanist activity culminated in the activities of two scholars. On the one hand, Árni Magnússon (1663–1730), along with other collectors based in Continental Scandinavia, began collecting Icelandic manuscripts, removing them from circulation in Iceland while also (generally) facilitating their preservation. Accordingly, the parent-child pair of Eggertsbók and AM 588m 4to, and the sibling pair of GKS 1002 fol and AM 592a 4to, perhaps lack descendants because they were removed to Denmark in Árni’s time. On the other hand, we see at the same time Magnús Jónsson í Vigur (1637–1702), a fishing magnate based in the Westfjords, undertaking and patronizing manuscript production on a grand scale (McDonald Werronen 2020; McDonald Werronen 2016–18). To his patronage we very likely owe the Lbs 222 fol manuscript of Sigurgarðs saga frækna, copied in 1695–96 by Jón Þórðarson, a scribe who has been identified as having worked as one of Magnús í Vigur’s main copyists at that time (McDonald Werronen 2020, 43). This manuscript has left us no surviving copies but represents the lively intellectual culture of the late seventeenth-century Westfjords, which in turn fed into eighteenth-century copying in the Dalir, just to the south. More investigation would be needed, but it is tempting to fit Papp 17 4to into a story where medieval manuscripts of Sigurgarðs saga frækna were being gathered at Hólar in the seventeenth century, followed by similar activity in the Westfjords, which facilitated the production of a number of related manuscripts there over the eighteenth and nineteenth centuries.

§20. Thus, from the seventeenth century, we transition into an eighteenth-century culture producing increasingly small, inexpensive reading copies of romance sagas, and eighteenth- and nineteenth-century manuscripts account for most of those localized in Figure 2. In the west, scribes drew directly on manuscripts produced under the aegis of Magnús and people like him, and both there and elsewhere they also drew on manuscripts now lost whose ancestry goes back to medieval or early modern copies (presumably because, unlike surviving examples, they remained in Iceland, where they were worn out and discarded), perhaps mediated by men of lesser means but similar proclivities to Magnús. The most pronounced cluster of manuscripts produced in this period is of A-class manuscripts around the Dalir and Westfjords in the north-west of Iceland (cf. Lavender 2020, 107–22 on Illuga saga, which finds a similar north-western cluster). Not only is this group of manuscripts tightly grouped in space, but they are especially closely related textually. Their similarity to one another can readily be visualized through a computer-generated stemma in which branch lengths are proportional to the number of differences between manuscripts. The fact that the A-class Dalir manuscripts are all so similar to one another either tells us that these scribes were exceptionally conservative copyists, or that an exceptionally large proportion of their manuscripts survive, whereas elsewhere, intermediary copies in which textual changes built up incrementally have been lost (or both).

Much research has yet to be done before a map like Figure 2 can be interpreted in greater detail: its distributions may, for example, tell us more about where scribes were inclined to convey the autobiographical information we need to localize their manuscripts than about where manuscripts were generally produced. We have not found a clear equivalent to the extensive cluster of textually-related Nítíða saga and Illuga saga manuscripts localized respectively by McDonald (McDonald 2014; McDonald Werronen 2016, 25–59) and Lavender (2020, 107–22) to the Eastfjords, though the parent-and-child pair Lbs 4070 8vo and Lbs 4447 4to does attest to Sigurgarðs saga frækna in that region in the 1860s. On the other hand, we can see a cluster of Sigurgarðs saga frækna manuscripts in the south-west.

§21. Leifson 1, Lbs 4070 8vo, and Lbs 4447 4to are noteworthy as they are the only known manuscripts of Friðberts saga frækna, an obscure late Icelandic romance. Given that Leifson 1 and Lbs 4447 4to also both contain Ála flekks saga, one might expect a close relationship between them in spite of the geographical distance separating them, which the stemma of Sigurgarðs saga frækna confirms. The stemma, however, also draws attention to the pairing of Nikulás saga leikara and Sigurgarðs saga frækna in both Leifson 1 and Lbs 3938 8vo, the latter of which was purchased in 1970 from Einar Guðmundsson of Reyðarfjörður in the Eastfjords and may previously have spent a stint in North America. Similarities in content within this cluster of manuscripts could support the statement that Sigurður Þorleifsson copied a friend’s anthology in its entirety after leaving Iceland for Utah. Large numbers of Icelandic emigrants to North America originated from the Eastfjords, and it is entirely possible that one of them brought the parent manuscript of Leifson 1. Further research on the co-texts in this branch of the stemma would be useful.

§22. By the late nineteenth century, we can see the impact of the 1884 printed edition. The fact that an edition produced in Reykjavík has close relatives in nineteenth-century manuscripts produced around the Reykjanes peninsula and not far to the north, in Borgarfjörður, suggests that the edition arose from a local scribal milieu in which manuscripts were circulating quite intensively. Yet the three manuscripts descended from the printed text are widely dispersed, hinting at the swift and broad distribution achieved by the printed text.

5. What Sigurgarðs saga frækna can tell us that we did not know

5.1 Studying the transmission of co-texts in the Rask 32 group

§23. This section argues that, with the creation of a stemma for Sigurgarðs saga frækna, we just about have a critical mass of knowledge to start to understand not only the stemmas of individual sagas, but to compare the stemmas of sagas that co-occur in multiple manuscripts in order to investigate the processes whereby scribes produced their anthologies. This is the beginning of a step-change in our study of medieval Icelandic scribal and literary culture. In addition to the present study, fairly comprehensive stemmas have now been attempted for the following romance sagas: Dínus saga drambláta (Kristjánsson 1960), Gibbons saga (Page 1960), Konráðs saga keisarasonar (Zitzelsberger 1980, 1981, 1983; Hall and Parsons 2013), Mírmanns saga (Slay 1997), Nítíða saga (McDonald 2014, McDonald Werronen 2016), Tíodels saga (Ohlsson 2009), and Viktors saga og Blávus (Kristjánsson 1964). (Digitized versions of their stemmas (excluding Nítíða saga), provided as a .dot file.) Comparable work on fornaldarsögur is also relevant: for the present study, the stemma of Hrólfs saga kraka established by Desmond Slay is particularly important (Slay 1960a, Slay 1960b, 1970, 1981, 1994). Although Icelandic romance sagas tend strongly to appear in manuscripts containing other romance sagas (Hall and Parsons 2013, §1.3; Kapitan, Rowbotham and Wills 2017; cf. Kapitan 2021; Kapitan and Wills 2023), there are enough different sagas in this genre that even with the stemmas of eight sagas available, too few well studied sagas appear in the same manuscript for us to say very much about how they travelled together. However, by far the most frequent companion to Sigurgarðs saga frækna is Nítíða saga; the two usually co-occur in manuscripts whose Sigurgarðs saga frækna text descends from Lbs 423 fol. Moreover, manuscripts in this branch of the Sigurgarðs saga frækna stemma frequently exhibit other overlapping content, each including some (and, in the case of Lbs 998 4to, all) of Ambrósíus saga og Rósamundu, Drauma-Jóns saga, Hálfdanar saga Barkarsonar, Hálfdanar saga Eysteinssonar, Nikulás saga leikara, and Sigurðar saga turnara. Although the transmission of these sagas is less well understood, existing research on them, abetted with some targeted new investigations produced for the present article, is sufficient for us to say something about how each scribe went about compiling the texts in their respective manuscript—or, as we should perhaps conceptualize the process, how each editor went about creating their respective anthologies. Studying the literary and cultural meanings of the juxtaposition of texts in a given manuscript has in recent decades been a popular endeavor both in saga-studies and far beyond (for particularly relevant examples see Lethbridge 2012a and 2012b on Eggertsbók, Lavender 2020, 77–91 on AM 123 8vo, and Hufnagel 2016a on Lbs 2319 4to; cf. Kapitan 2021; Kapitan and Wills 2023). But, at least as far as riddarasögur and fornaldarsögur are concerned, such studies are seldom undertaken with much understanding of whether a scribe was copying their exemplar(s) more or less wholesale or selectively curating an anthology—and, if the latter situation holds, what texts they chose not to copy. Combining stemmatic research on multiple sagas helps us to fill this gap.

§24. The present article explores the possibilities for studying anthologization with a case study of one branch of the Sigurgarðs saga frækna stemma where our stemmatic knowledge is now just about sufficient to make informed comment: that is, the tight-knit group of Dalir manuscripts descended from Rask 32 4to, which we refer to as the “Rask 32 group,” mapped in Figure 4. The previous section, along with McDonald’s research on Nítíða saga (McDonald 2014; McDonald Werronen 2016, 34–44) shows that most of the manuscripts in which Nítíða saga and Sigurgarðs saga frækna co-occur belong to this group. The results of our investigations are summarized stemmatically in Figure 5, reference to which will make it markedly easier to follow the discussion below. Table 2 charts the overlapping content of Sigurgarðs saga frækna manuscripts descended from the parent of Rask 32’s Sigurgarðs saga frækna text, Lbs 423 4to, and the corresponding branch of Nítíða saga manuscripts, those descending in McDonald’s stemma from JS 166 fol. In two cases in the manuscript Lbs 3966 4to, there is no prose version of the narrative, but a rímur cycle on the same narrative has been included (see discussion below).

Table 2: Shared contents in the Rask 32 group (in order of frequency)

Classmark	JS 166 fol	Lbs 644 4to	Lbs 423 fol	Rask 32	Lbs 354 4to	Lbs 998 4to	JS 632 4to	Lbs 1137 8vo	Lbs 3966 4to	Lbs 3165 4to	SÁM 131	Lbs 5567 4to
Date	1678–79	c. 1730–40	1733	c. 1756–67	C18	1765–1805	1799–1800	c. 1819–20	1869–71	1869–71	1871–72	1913
Fragmentary?								×
Sigurgarðs saga frækna		×	×	×	×	×	×		×	×	×	×
Nikulás saga leikara		×		×		×	×		×	×	×	×
Nítíða saga	×	×		×		×	×	×	×	×
Ambrósíus saga og Rósamundu			×	×	×	×	×				×
Hálfdanar saga Eysteinssonar	×			×	×	×	×		(rímur)	×
Drauma-Jóns saga				×		×		×		×
Flóres saga og Blankiflúr			×	×		×	×
Hálfdanar saga Barkarsonar						×	×	×			×
Sigurðar saga turnara			×	×		×	×		(rímur)
Blómsturvalla saga		×		×	×
Bærings saga			×	×	×
Sturlaugs saga starfsama			×	×			×
Bósa saga			×									×
Fertrams saga og Platós			×			×
Jóns saga Upplendinga-konungs						×	×
Konráðs saga keisarasonar						×	×
Jókuls þáttur Búasonar									×	×
Rímur af Þorsteini Víkingssyni									×	×

A careful look at the stemma in Figure 5 emphasizes the obvious point that although these manuscripts contain similar texts, those texts are not necessarily textually closely related. No manuscript is a wholesale copy of another; a fact that our visualizations obscure, moreover, is that even where manuscripts contain the same texts, they do not usually contain them the same order. Nítíða saga and Sigurgarðs saga frækna co-occur in Lbs 644 4to, and we might guess for this reason that the copies of these sagas in that manuscript are textually related to the cluster of other manuscripts containing these sagas; and indeed, if McDonald’s stemma of Nítíða saga is correct, both Lbs 644 4to and Rask 32 derived their text of Nítíða saga from the same exemplar, JS 166 fol. But Lbs 644 4to and Rask 32 certainly took their texts of Sigurgarðs saga frækna from entirely different sources.

§25. It would be nice to be able to comment with this degree of precision about all the sagas listed here, or at least those that co-occur most often. We are not yet quite able to do this, and the present case study is built around our detailed understanding of the transmission of Sigurgarðs saga frækna and Nítíða saga. But we have a rough idea of the stemmas of several of the other sagas, which enables us to guess whether sagas that co-occur with Sigurgarðs saga frækna and Nikulás saga leikara were transmitted on the same lines, and to undertake targeted stemmatic research without having to establish a complete stemma for each saga in which we are interested. McDonald Werronen and Kapitan’s edition of Ambrósíus saga og Rósamunda (2018, 184) determined that of the nineteen manuscripts of that saga, all six listed here belong to the B-group (along with three others, and a further sub-group of three). Spaulding’s 1982 PhD thesis made much of the difficulty of establishing a stemma for Sigurðar saga turnara, protesting that “those trying to construct a stemma despair at the complexity of the interinfluence shown by […] groups of younger manuscripts […] No single manuscript can be definitely said to be a copy of one and only one precursor” (1982, 98). Spaulding’s comments are better understood as the cry of a graduate student in pain than an accurate assessment of the manuscript situation. Despite her reservations, Spaulding situated the four manuscripts of Sigurðar saga turnara which fall into the Rask 32 group (along with Magnús Jónsson í Tjaldanesi’s copy in Lbs 1503 4to, to which she could have added Magnús’s earlier copy in Lbs 4940 4to) as belonging to a distinct group, closely related to the archetype of the saga (1982, 99, 108): the classic pattern for Dalir riddarasaga manuscripts. A young R.I. Page expressed anguish similar to Spaulding’s as he grappled with the stemma of Drauma-Jóns saga but found that the two of our group of manuscripts which he examined, Rask 32 and Lbs 998 4to, belonged in the same branch (1957, 32). In all, we have cursorily checked the information provided by past studies of Ambrósíus saga og Rósamundu, Hálfdanar saga Eysteinssonar, Nítíða saga, and Sigurðar saga turnara, and in all cases except Hálfdanar saga Eysteinssonar, the data either supports the presumption that the sagas were transmitted on the same lines as Sigurgarðs saga frækna or, when the variation is insufficient to come to a conclusion (as is sometimes the case due the tendency of Rask 32 group manuscripts to be very precise copies of one another), it at least does not contradict it; in these cases, we have been satisfied provisionally to conclude that co-occurring sagas were indeed transmitted on the same stemmatic lines. Meanwhile, Hálfdanar saga Eysteinssonar offers a good example of the usefulness of partial stemmatic information of uncertain precision, in this case Schröder’s 1917 edition. In our Rask 32 group of Sigurgarðs saga frækna manuscripts, Hálfdanar saga Eysteinssonar co-occurs with Sigurgarðs saga frækna in JS 166 fol, JS 632 4to, Rask 32, and Lbs 354 4to, hinting that they might have copied the saga from one another. But Schröder identified JS 166 fol and JS 632 4to as manuscripts of the A-recension of the saga and Rask 32 and Lbs 354 4to as belonging to the C-recension (1917, 71–72). A glance at the manuscripts shows that Schröder was right, with not only JS 166 fol and JS 632 4to, but also Lbs 998 4to, Lbs 3966 4to, and Lbs 3165 4to belonging, at least at the opening, to the A-recension whereas Rask 32 and Lbs 354 4to contain an altogether different version. Meanwhile, consultation of JS 166 fol reveals that it has a markedly different text from Rask 32. Since Lbs 998 4to and its relatives are, where we are in a position to check, very conservative copies of their exemplars, it seems clear that they are not copied directly from JS 166 fol but from another A-recension exemplar yet to be identified.

§26. No research has hitherto been published on the stemma of Nikulás saga leikara. The saga survives in no medieval manuscripts (and so was omitted from Kalinke and Mitchell’s 1985 bibliography of the genre) but does seem to have been part of the now fragmentary fifteenth-century manuscript Stockholm, Royal Library, Perg. fol. nr 7 (Sanders 2000, 17, 21). Either way, the saga is found in over sixty post-medieval manuscripts (in two main recensions) and was twice printed in popular editions: in Winnipeg by the Heimskringlu Prentstofa (1889) and in Reykjavík by Helgi Árnason (1912). This makes it a particularly widely attested saga, an interesting example of Canadian-Icelandic literature, and an unusually late example of Icelandic readerships for printed romance sagas. The saga was edited by Wick (1996), on whose work our own depends. Given the prominence of Nikulás saga leikara in the Rask 32 group, we have used the full apparatus criticus of Wick’s 1996 critical edition—2,039 sites of variation—and the phylogenetic software Pars according to the methods described by Hall and Parsons (2013) to establish a stemma of the six manuscripts which Wick surveyed. Despite the lack of human checking, this stemma is surely highly reliable (Figure 6). We supplemented this by sampling chapter 1 of the saga in all those manuscripts which our work on Sigurgarðs saga frækna and Nítíða saga situates in the Rask 32 group, and—to ensure that the Rask 32 group is indeed a distinctive branch of the overall stemma—a further 19 more or less randomly chosen manuscripts of the main recension accessible in Icelandic collections, along with Winnipeg, Elizabeth Dafoe Library ISDA JB3–01–04. The Pars-generated stemma based on this sample shows clearly that the Rask 32 group is indeed a distinctive group and that, compared with other manuscripts, its manuscripts are unusually similar to one another, as can be seen in the tight cluster in Figure 7. This is exactly as we would expect from the much more thorough studies of Sigurgarðs saga frækna and Nítíða saga. Again, then, it is reasonable to assume that Nikulás saga leikara was indeed transmitted on the same lines, and the data shows no reason to doubt this. We have, therefore, integrated Nikulás saga leikara into Figure 5, our stemma of the Rask 32 group, accordingly.

5.2 Rask 32 and its sources

§27. Of the six sagas in Rask 32 on whose transmission we have some information, three (Sigurgarðs saga frækna, Sigurðar saga turnara, and Ambrósíus saga og Rósamundu) seem to have been copied from Lbs 423 fol, a volume which was, according to the title page, “kostgæfilega samanhentur og tildreginn af Bjarna Péturssyni að Skarði á Skarðströnd þeim til tíma og dægrastyttingar er þvílík lofsverð æfintýri girnast að lesa og heyra. Árum eftir frelsarans fæðing MDCCXXXIII” (“painstakingly collected and assembled by Bjarni Pétursson at Skarð á Skarðströnd, as a diversion for those who long to read and hear such praiseworthy examples. In the year 1733 after the Savior’s birth”). Bjarni (1681–1768) was a rich and well-educated collector and producer of manuscripts (Hufnagel 2016a, 395), successor to Magnús Jónsson í Vigur as the key literary patron of north-west Iceland in his time. As it stands, the contents of Lbs 423 fol are clearly divided into konungasögur (which comprise the first ten items in the collection) on the one hand and romances and fornaldarsögur (which comprise the remaining ten) on the other. This conception may be reflected in the title page, which positions the volume as a “fróðlegur sagnafésjóður af Norvegsríkis einvaldskonungum allmörgum […] Samt þeirra lofsverðra hertoga, greifa, jarla og annarra trúfastra þénara” (“learned treasury of sagas on very many of the monarchs of the state of Norway […] together with their praiseworthy dukes, counts, earls, and other loyal servants”). Although all the sagas involve kings and their servants, the distinction drawn on the title page might reflect the fact that the protagonists of the second half are seldom kings (though they almost invariably become kings in the end). That said, the manuscript opens with the konungasaga Heimskringla, entitled “norsku konga króníka samandregin af Snorra Sturlusyni á Íslandi og hefst með Svíþjóðskóngum, hverja hann útfærir af Schytia eða Tartaríalandi” (“the chronicle of Norwegian kings edited in Iceland by Snorri Sturluson and beginning with the kings of Sweden, whose lineage he derives from Scythia or Tartaríaland”), so Sigurgarðs saga frækna, set in Tartaría, may have seemed a contiguous part of this history; at the back of the manuscript, a printed picture of a Turk conquering Constantinople is tucked in, emphasizing an interest in Asia among readers. At any rate, the romances and fornaldarsögur in the manuscript are continuous, rather than each starting on a new page, so we can be confident that we have them in the order that Bjarni had them written. So far, no surviving manuscript sources for the romances and fornaldarsögur in Lbs 423 fol have been identified: for Sigurgarðs saga frækna, at least, Bjarni drew on a manuscript either lost or yet to be identified. We have noted above some of the archaic features in the spelling of this manuscript’s text of Sigurgarðs saga frækna (Part 4): it is likely that its exemplar was a medieval manuscript or otherwise not far removed from one.

§28. The history of Rask 32 itself has been explored by Silvia Hufnagel (2016b), who has analyzed it alongside other saga manuscripts associated with its main scribe, Ólafur Gíslason of Hvítidalur, Dalasýsla (1727–1801), pre-eminently Rask 31 and NKS 1148 fol. Ólafur produced Rask 32 in collaboration with his father Gísli Jónsson (c. 1699–1781). Hufnagel (2016b, §20) concluded that:

as it must have been difficult to carry an unbound manuscript back and forth several times without damaging or soiling it, it seems more likely that the two scribes lived under the same roof, or at least in close proximity to each other, when they took turns writing Rask 32. This would narrow down the possible time of origin to the years 1756 and 1765 when Ólafur was the assistant to his father.

For the most part, we can be confident that we have the sagas in the order that Ólafur and Gísli put them together, as they are written continuously, notwithstanding changes of scribe. The exception is ff. 155r–74v. With the end of Ragnars saga loðbrókar on f. 154v, half the verso is, uncharacteristically, left blank, and Sturlaugs saga starfsama starts a new gathering at the top of f. 155r. The manuscript proceeds continuously from Sturlaugs saga into Bærings saga. The transition from the end of Bærings saga on f. 174v to the beginning of Drauma-Jóns saga on f. 175r involves a little empty space at the bottom of f. 174v and the end of a gathering; it also coincides with a shift from Ólafur’s hand to Gísli’s. But hereafter the manuscript is again continuous.

§29. With the stemma of the Rask 32 group presented as Figure 5, we can for the first time glimpse the scholarship and discernment with which Ólafur and Gísli assembled their collection. It is not self-evident that they drew on multiple sources for their anthology: they were conservative copyists of their exemplars and it is conceivable that they aimed to replicate the contents of a source manuscript as faithfully as they replicated the text of each saga. The contents of Rask 32 are not obviously idiosyncratic and it is conceivable that they were copied wholesale from a single, seventeenth-century compendium. This was not, however, Ólafur and Gísli’s approach. They clearly wanted to develop a large collection of romances and fornaldarsögur: they did not copy any of Lbs 423 fol’s konungasögur, but they definitely copied from it Sigurgarðs saga frækna, pretty certainly Sigurðar saga turnara and Ambrósíus saga og Rósamundu, and potentially Flóres saga og Blankiflúr, Sturlaugs saga, and Bærings saga. They did not, however, copy these sagas as a block or even in the same order (with the possible exception of Flóres saga og Blankiflúr and Sigurðar saga turnara, which are a pair in both manuscripts). They could also have copied the (in)famously lewd Bósa saga (which is, moreover, a source for Sigurgarðs saga frækna, emphasizing the sagas’ generic similarity: Hall, Richardson and Þorgeirsson 2013, 85–86) or the exempla (or exemplum-like) Griseldis saga, Brita þáttur, and Lykla-Péturs saga og Magelónu fögru—and indeed Ólafur did add Griseldis saga, whether from this source or another, into NKS 1148 fol—but for Rask 32 at least they elected not to. It looks like they were focusing on producing a volume of romances, secular in tone yet staying (unlike Bósa saga) on the right side of propriety.

§30. Ólafur and Gísli also had access to JS 166 fol, which was compiled in Ísafjörður in the Westfjords, in the milieu surrounding Magnús Jónsson í Vigur. From this, if McDonald (2014, 2016) is correct, they selected Nítíða saga and potentially Hálfdanar saga Brönufóstra. They did not, however, copy its text of Hálfdanar saga Eysteinssonar, preferring a C-recension manuscript which, if it survives, has yet to be identified. Since Hálfdanar saga Eysteinssonar is the first text in Rask 32, it is possible that Ólafur and Gísli copied it before they had access to JS 166 fol and that they would have chosen the JS 166 fol version if they had had the chance: we cannot be certain that they were actively discriminating between two versions. At any rate, they also elected not to copy Trójumanna saga, Vilmundar saga viðutan, Rémundar saga keisarasonar, or Orms þáttur Stórólfssonar—though Ólafur did copy Trójumanna saga into NKS 1148 fol, possibly from JS 166 fol.

§31. All the Nítíða saga manuscripts included in the present study, including Rask 32, belong to a version whose ending says that the protagonist of Nikulás saga leikara is descended from Nítíða (McDonald Werronen 2016, 43). Ólafur and Gísli evidently wanted to complement Nítíða saga with Nikulás saga leikara, and added the latter text accordingly. JS 166 fol could not, in its present state, have provided a text of Nikulás saga leikara. Yet JS 166 fol is quite worn and was evidently in a fragile state whenever it was put into its present binding (apparently by Páll Pálsson stúdent, in the nineteenth century, since he copied out parts of the text lost during conservation), so we cannot be certain that no sagas have been lost from it. JS 166 fol also seems to be the parent manuscript to Lbs 644 4to for Nítíða saga, and the Nikulás saga leikara texts in Rask 32 and Lbs 644 4to are also closely related and could share a parent, so the possibility that JS 166 fol once included Nítíða saga and Nikulás saga leikara non-consecutively, giving them both to Rask 32 and Lbs 644, is attractive. But in JS 166 fol Nítíða saga ends with f. 190r and Hálfdanar saga Eysteinssonar starts on the verso of the same folio, so Nikulás saga leikara at least never followed Nítíða saga in JS 166 fol directly. At minimum, we can say that Ólafur and Gísli reordered their source to bring two related sagas together.

§32. We can reconstruct a similar method behind Rask 31, copied entirely by Ólafur: one source was another major compilation by Bjarni Pétursson, this time Lbs 2319 4to, which is known to have given Rask 31 Hrólfs saga kraka and Tíodels saga. Both manuscripts also contain Sigurgarðs saga og Valbrands which, unusually, they both entitle Sigurgarðs saga frækna, so Rask 31 probably took that saga from Lbs 2319 4to too. And Lbs 2319 4to potentially also supplied Rask 31 with Áns saga bogsvegis, Yngvars saga víðförla, Æfintýr af Perus meistara, Klári saga, Apollonius saga, and Haralds saga Hringsbana. But, as in Rask 32, Ólafur supplemented his Bjarni Pétursson exemplar with a major seventeenth-century compilation produced in the Westfjords, this time by Magnús Jónsson í Vigur himself, ÍBR 5–6 fol, which supplied Konráðs saga keisarasonar and potentially also Flóvents saga, Sigurðar saga þögla, and Elis saga og Rósamundu. At least one other manuscript must have supplied the manuscript’s remaining five texts. If these inferences are correct, while Ólafur reordered the texts of his exemplars, he did first copy those sagas that he wanted from Lbs 2319 4to and then the ones he wanted from ÍBR 5–6 fol, suggesting that he worked through his source manuscripts sequentially rather than having them all available at the same time.

§33. Overall, we can start to see the working methods whereby Ólafur Gíslason took the major seventeenth- to eighteenth-century editorial work of the magnates Magnús Jónsson í Vigur and Bjarni Pétursson and transferred it into a collection of less prestigious, generically more tightly focused anthologies containing fornaldarsögur and riddarasögur. Provisionally, pending further stemmatic research, we can say that Ólafur probably used one large, folio manuscript deriving from each of his predecessors as the main source for each of his quarto manuscripts, selecting significant groups of sagas from these, but also reordering them and supplementing them from other sources.

5.3 Lbs 998 4to and JS 632 4to

§34. Rask 32 supplied Sigurgarðs saga frækna, probably Ambrósíus saga og Rósamundu, and possibly Hálfdanar saga Eysteinssonar to the eighteenth-century Lbs 354 4to, about which little is as yet known. That manuscript contains many other sagas which Rask 32 could not have supplied, and the story of its compilation must await the further investigation of their transmission. But we can say quite a lot about how Rask 32 was used as an exemplar for Lbs 998 4to, and how Lbs 998 4to was used in turn as an exemplar for JS 632 4to. Rask 32 supplied a lot of sagas to Lbs 998 4to: Sigurðar saga turnara, Ambrósíus saga og Rósamundu, Drauma-Jóns saga, Sigurgarðs saga frækna, Nítíða saga and Nikulás saga leikara, as well as, potentially, Flóres saga og Blankiflúr and Fertrams saga og Platós. Lbs 998 4to also used Rask 31, certainly for Konráðs saga keisarasonar and potentially for Yngvars saga víðförla.

§35. Lbs 998 4to is an even more conservative copy of its sources than Rask 32 is; moreover, while it did not copy Rask 32 wholesale or preserve the ordering of texts there, it is clearly more closely a replica of a single anthology than Rask 32 itself was. Intriguingly, Lbs 998 4to does contain a copy of Hálfdanar saga Eysteinssonar, but did not copy this from Rask 32 (or for that matter from JS 166 fol). So there is also the possibility that Lbs 998 4to was intended largely to reproduce the editorial scope of Rask 32 but with the substitution of a copy of Hálfdanar saga viewed as superior by Lbs 998 4to’s editors. Otherwise, while Lbs 998 4to does not copy all Rask 32’s contents, it adds only sagas found in Rask 31, plus Jóns saga Upplendingakonungs and Hálfdanar saga Barkarsonar.

§36. The similarity of Lbs 998 4to to its sources makes it hard to be certain how to filiate manuscripts that are similar to both, but the substantial body of evidence in our transcriptions of Sigurgarðs saga frækna captures enough minor innovations in Lbs 998 4to that are shared by JS 632 4to, Lbs 3165 4to, and Lbs 3966 4to, and enough distinctive innovations in each of the latter three manuscripts, that we can confidently position Lbs 998 4to’s Sigurgarðs saga frækna as the child of Rask 32 and the parent of JS 632 4to, Lbs 3165 4to, and Lbs 3966 4to.

§37. The history of Lbs 998 4to has as yet been little explored. In its present form, Lbs 998 4to opens with three nineteenth-century folios, the first of which is an elaborate title page bearing the initials GES and JGJD. The former is probably Guðbrandur Einarsson (1842–1922), who was the owner of the book according to an ownership statement signed at Knarrarhöfn (or Knararhöfn) in Hvammssveit, Dalasýsla, by Magnús Jónsson (1841–1923) on 2 January 1862. The latter is likely Guðbrandur’s wife, Jóhanna Guðrún Jónsdóttir (1842–1915), who was Magnús’s sister. The hand appears to be that of one of three brothers from Fellsströnd in Dalasýsla: Guðlaugur Magnússon (1848–1917), Guðmundur Magnússon (1850–1915), or Jóhannes Magnússon (1852–1917). The brothers spent their first years at the farm of Arnarbæli but were orphaned in 1858. Guðlaugur was fostered by an aunt at Hafursstaðir, while Guðmundur and Jóhannes were raised in the household of Jóhannes Bæringsson at Breiðabólstaður. Guðlaugur and Jóhannes emigrated to Canada together in 1874, eventually settling in New Iceland, Manitoba, while Guðmundur remained at Breiðabólstaður. All three are associated with the creation of large saga anthologies, including Lbs 474 fol (Surmeli 2020) and ISDA 42 and ISDA 43 at the Icelandic Collection, Elizabeth Dafoe Library, University of Manitoba. These three folios that begin Lbs 998 4to contain a contents list and the beginning of Yngvars saga víðförla, presumably added because the beginning of the manuscript needed replacing. The original manuscript begins on folio 4r, in the main hand, a neat cursive (which occasionally shifts to Fraktur for direct speech). This hand has not previously been identified by the standard authorities but clearly belongs to Ólafur Jónsson (1722–1800), presumably writing at his farm of Arney, Dalasýsla: he is also the scribe of JS 632 4to and says so therein, as we discuss below. After the sixth line of f. 5r, partway through chapter two of Yngvars saga víðförla, the hand switches to a much larger and slanting cursive that continues to f. 9r where, two words into line 17, the hand returns to the main script. We believe, however, that these scripts belong to the same hand. The hand completes Yngvars saga víðförla and then Konráðs saga keisarasonar, Jóns saga Upplendingakonungs, Flóres saga og Blankiflúr, Sigurðar saga turnara, Hálfdanar saga Barkarsonar, and Ambrósíus saga og Rósamundu, ending the latter partway down f. 61v.

§38. So far in the manuscript, one saga starts immediately after the other ends, with no breaks in foliation. However, after Ambrósíus saga, the remainder of the folio is blank. Folio 62r is in a different hand and, to judge from the facsimile photographs at Handrit.is, begins a new gathering; this hand contributes Fertrams saga og Platós. Fertrams saga ends on f. 87v with the colophon “og endum vér svo þessa sögu, af Fertram og Plató, skrifuð á Dritvík af Ólafi Sveinssyni árið 1805” (“and thus we end this saga of Fertram and Plató, written at Dritvík by Ólafur Sveinsson in the year 1805”), partly written in Fraktur and partly in cursive, which more or less completes the folio. This scribe is agreed to be Ólafur Jónsson’s son-in-law Ólafur Sveinsson of Purkey (1762–1845), Dalasýsla, evidently writing while working at the bustling fishing station of Dritvík on Snæfellsnes. Seeing the two men collaborating on the same manuscript is no surprise: besides the elder being father-in-law to the younger, their co-productions also include ÍB 184 4to (a collection of Íslendingasögur, fornaldarsögur, and riddarasögur from the last quarter of the eighteenth century), JS 633 4to (more sagas copied by the two men, at varying times from at least 1780 to 1822), and ÍB 112 4to (religious poems anthologized by Ólafur Jónsson in 1800 and both introduced and completed by Ólafur Sveinsson). The next folio, which belongs to the same gathering, continues in Ólafur Sveinsson’s Fraktur hand with Drauma-Jóns saga. The hand shifts briefly to cursive on ff. 90v–91r, but as far as we can tell it remains Ólafur Sveinsson’s. Our small sample of Drauma-Jóns saga suggests that it was copied from Rask 32, like so many of the sagas copied by Ólafur Jónsson. Drauma-Jóns saga concludes with the end of f. 97v, with Ólafur spacing his writing to more or less fill the folio. Since 1805 was five years after the death of Ólafur Jónsson, it seems that Ólafur Sveinsson added these two sagas to the manuscript after Ólafur Jónsson completed it, and they were not part of Ólafur Jónsson’s editorial plan.

§39. From f. 98r, which may be the start of a new gathering, the main hand resumes, morphing briefly on f. 102r into a large and slanting hand, contributing Hálfdanar saga Eysteinssonar, Sigurgarðs saga frækna, Nítíða saga, and Nikulás saga leikara. The final page of Nikulás saga leikara more or less fills f. 137r. Folio 137v is blank except for a cursive scribble written up the margin in a different hand. Finally, a single bifolium is included, in the hand of the same one of the Magnússon brothers who added the material at the beginning of the manuscript. The first recto is a title page, evidently incomplete, which ought to read “Sagan af Ljósvetningunum” but lacks one of the requisite nine minims at the end of the word. The following verso and recto contain text not from Ljósvetninga saga, as one might expect, but from Yngvars saga víðförla. The final verso is blank.

§40. Having produced most of what is now Lbs 998 4to, Ólafur Jónsson copied most of it to produce JS 632 4to. On the first folio, he wrote a contents list and then added “saman tínt þeim til ánægju um tíma, sem með þess konar dægrastytting vilja nýta það[?] af einsýnum og útlifuðum, nær áttræðum karli á Purkey, við Skarðströnd byrjað 1799, endað 1800 af ÓLAFI JÓNSSYNI” (“gathered together for the enjoyment, for a spell, of those who wish to make use of it with this kind of pastime by the one-eyed and age-worn, nearly eighty-year-old man ÓLAFUR JÓNSSON on Purkey, by Skarðströnd, begun 1799, finished 1800”; f. 1v). In addition to the initial note, JS 632 4to concludes its copy of Böðvars saga Bjarka with “enduð 15. marsi 1800 af Ólafi Jónssyni 78 ára gömlum” (“finished 15 March 1800 by Ólafur Jónsson, 78 years of age”) on f. 245r. Tellingly, JS 632 4to contains all the sagas in Lbs 998 4to—albeit in an almost completely different order—apart from Yngvars saga víðförla, Fertrams saga og Platós, and Drauma-Jóns saga, the latter two of which were added to Lbs 998 4to by Ólafur Sveinsson after Ólafur Jónsson’s death. Following its Lbs 998 4to material, JS 632 4to goes on to add Sturlaugs saga starfsama, Úlfs saga Uggasonar, Böðvars þáttur bjarka, Sigurðar saga fóts, and Huga saga serka ok Skaplers konungs, from exemplars yet to be determined (Sturlaugs saga could have come from Rask 32 if Ólafur still had access to this in 1799–1800, but this is not assured).

§41. Thus Ólafur Jónsson produced Lbs 998 4to, presumably on his farm of Arney, Dalasýsla, sometime in the thirty years or so following the completion of his main source, Rask 32, as early as 1765. In 1799–1800 he then used Lbs 998 4to as the main exemplar for a new manuscript, JS 632 4to. He kept all the texts from Lbs 998 4to but seems not to have been concerned to preserve their sequence. Ólafur died in the same year as he completed JS 632 4to. The inside front cover of JS 632 4to contains the note “ég er Katrínar Þorvaldsdóttur á Hrapp. gefin 11/10.41. af Sg. Ól. Sveinssyni á Purk.” (“I was given to Katrín Þorvaldsdóttir of Hrappsey 11 October 1841 by Seignor Ólafur Sveinsson of Purkey”). This tells us that after Ólafur Jónsson’s death, not only Lbs 998 4to but also JS 632 4to came into the hands of Ólafur Sveinsson (like other manuscripts which Ólafur Sveinsson inherited, such as ÍBR 105 8vo). In 1805, Ólafur Sveinsson then copied Fertrams saga (from an as yet unknown source, but potentially Rask 32) and Drauma-Jóns saga (certainly from Rask 32, as far as our limited sample shows) into a new gathering and inserted those sagas into Lbs 998 4to.

§42. The findings in this section entail some corrections to past research. Páll Eggert Ólafsson, in his catalog of manuscripts in Iceland’s national library, dated Lbs 998 4to “á öndverðri 19. öld” (“to the beginning of the nineteenth century” 1918–37, I 415 [no. 1315]), while Handrit.is puts it at 1800–50. The manuscript is in fact mostly from between 1765 (the likely date of its exemplar’s completion) and 1799 (the date of its principal copy); presumably Páll Eggert was led astray by the 1805 date on Ólafur Sveinsson’s additions. Examining the relationship between the Lbs 998 4to and JS 632 4to texts of Konráðs saga keisarasonar, and following Páll Eggert’s dating, Hall and Parsons noted that Lbs 998 4to could in theory be the parent of JS 632 4to but added that “Lbs 998 4to [is] apparently too late to be JS 632 4to’s exemplar” (Hall and Parsons 2013, HTML stemma). With this dating problem resolved, we can confidently situate JS 632 4to as a child of Lbs 998 4to for Konráðs saga (as Zitzelsberger had previously in 1981), and this conclusion is reflected in Figure 5. In the same article, we were unsure as to whether Rask 31 was the parent or sibling of Konráðs saga in Lbs 998 4to, but the evidence amassed in this article leads us to concur with Zitzelsberger that Rask 31 is indeed the parent of Lbs 998 4to. Meanwhile, McDonald’s “possible rough stemma” of Nítíða saga for the Rask 32 group was published with the caveat that “where Lbs 644 and Rask 32 fit in relation to JS 632 and the later manuscripts is also uncertain, and it would require further detailed collation of larger text samples to unravel the intricacies of these relationships” (2014, 86). Notwithstanding these caveats, McDonald filiated the Nítíða saga manuscripts described in this article as shown in Figure 8.

We believe the data published in this article for Nikulás saga leikara, alongside the extensive data for Sigurgarðs saga frækna, are sufficient to show that our stemma (Figures 1 and 5) is far more likely.

§43. The copying of Lbs 998 4to from Rask 32 and of JS 632 4to from Lbs 998 4to emphasizes the influence of Rask 32 on Ólafur Jónsson: Ólafur Jónsson did not feel bound entirely to replicate Rask 32, but he did elect mostly to copy its contents with little change into Lbs 998 4to and from there into JS 632 4to, somewhat as Gísli Jónsson and Ólafur Gíslason seem to have made extensive use of Bjarni Pétursson’s editorial decisions in Lbs 423 fol. The changes which Ólafur Jónsson did make to the Rask 32 selection would bear closer analysis—not least his decision to replace the Rask 32 C-class text of Hálfdanar saga Eysteinssonar with an A-class text. But the overall impression is that Ólafur Jónsson was happy with his predecessors’ choices and, when he came to copy Lbs 998 4to, even happier with his own. Thus Lbs 998 4to is a good example of a manuscript which replicates the editorial decisions of a single main exemplar. These observations perhaps lend significance, however, to the fact that Ólafur Jónsson does not seem to have been concerned in what order he copied sagas into a manuscript: literary critics keen to find meaning in the juxtapositions of sagas in particular manuscripts may be more interested in those juxtapositions than were the manuscripts’ actual anthologists.

5.4 Lbs 3966 4to and Lbs 3165 4to

§44. Not too long after the copying of JS 632 4to—probably in 1813–15—Rask 31 and 32 were sold to Rasmus Rask and transported to Denmark, removing them from circulation in Iceland (Hufnagel 2016b). After Rask 31 and 32 were removed from Iceland, both Lbs 998 4to and JS 632 4to remained and continued to be copied. The stemmas of Sigurgarðs saga frækna and Nítíða saga reveal three manuscripts drawing on Lbs 998 4to besides JS 632 4to (and fuller surveys of the stemmas of other sagas in Lbs 988 4to might of course reveal others). Two of the descendants of Lbs 998 4to were produced in 1869–71 by Ólafur Þorgeirsson (1826–94) of Skáleyjar, Austur-Barðastrandarsýsla: Lbs 3165 4to (156 folios, copied for Jón Jónsson of Purkey, 1812–88) and Lbs 3966 4to (280 folios, with no evidence for an intended audience other than Ólafur Þorgeirsson himself). The folios of these volumes are the same size, meaning that Lbs 3966 4to is approaching twice the length of Lbs 3165 4to. In both cases, Ólafur selected some sagas from Lbs 998 4to. For the longer manuscript, he chose Sigurgarðs saga frækna, Nítíða saga, and Nikulás saga leikara. Alongside these he included six rímur derived from other riddarasögur and fornaldarsögur, and in four cases he gave a hint as to their exemplars, revealing that he had access to an autograph copy of rímur composed by Magnús Jónsson (1763–1840) of the farm Magnússkógar, Dalasýsla, and at least one other manuscript, whose contents included work by Magnús. In the case of two of the rímur, Ólafur could, if he had wished, have included the same story in saga form from Lbs 998 4to: Rímur af Sigurði turnara (“skrifaðar eftir tveim handritum árið 1869,” “copied from two manuscripts in 1869,” f. 30r) and Rímur af Hálfdani Eysteinssyni (“skrifaðar eftir eiginhandriti skáldsins en þó blökku og máðu,” “copied from an autograph manuscript by the poet, albeit dim and faded,” f. 244r), both by Magnús Jónsson. To these three sagas and six rímur, Ólafur added some other poems and, finally, two sagas not in Lbs 998 4to: Ála flekks saga and Knúts saga heimska, both “skrifaðar eftir einu handriti en ekki þó að öllu leyti réttu” (“copied from one manuscript, albeit not correct in all respects,” f. 280v). Overall, then, we can see that Ólafur planned Lbs 3966 4to as a manuscript dominated by fornaldarsaga and riddarasaga-type stories, but it seems that he preferred to include stories in rímur form when he had the choice, and was also selective in the prose sagas he took from Lbs 998 4to, drawing prose material from at least two manuscripts.

§45. The manuscript that Ólafur copied for Jón Jónsson, Lbs 3165 4to, is similar in scope and genre but contains somewhat different choices of material: the first hundred folios contain six fornaldarsögur and riddarasögur. Of these, five seem certainly to be from Lbs 998 4to: Sigurgarðs saga frækna, Nítíða saga, Nikulás saga leikara (all included in Lbs 3966 4to), Hálfdanar saga Eysteinssonar (included in Lbs 3966 4to in rímur form), and Drauma-Jóns saga (not included in Lbs 3966 4to). The remaining prose text is Jökuls þáttur Búasonar, presumably copied from the same source from which Ólafur copied it into Lbs 3966 4to. Ólafur followed these sagas with eight folios of short poems, and one of the Magnús Jónsson rímur that he had included in Lbs 3966 4to, Rímur af Þorsteini Víkingssyni. Thus this manuscript is not only shorter but more heavily weighted towards prose sagas: Ólafur used the prose Hálfdanar saga Eysteinssonar from Lbs 998 4to rather than Magnús Jónsson’s rímur adaptation.

§46. Overall, we might conclude—provisionally to forming a fuller understanding of the copying of Lbs 998 4to—that Ólafur had a fairly clear sense of what his favorite Lbs 998 4to sagas were, and that Jón Jónsson of Purkey shared it. But Ólafur himself preferred rímur versions of the same stories where they were available and chose to copy two prose sagas from a different manuscript rather than copy more extensively from Lbs 998 4to. Thus, we can see him inheriting some of the editorial decisions that underlay the production of Rask 32, yet purposefully developing his own anthology.

§47. We can probably glimpse a similar pattern in the other copy of Lbs 998 4to of which we are aware: the fragmentary Lbs 1137 8vo IV, the relevant fragment of which was copied around 1819–20 by one Jón Sigurðsson in Háihóll, Álftártungusókn, Mýrasýsla. Lbs 998 4to looks likely to have been the exemplar for at least Nítíða saga and Drauma-Jóns saga in Lbs 1137 8vo IV, and it might also have contributed Hálfdanar saga Barkarsonar. Yet the remaining three texts which we know to have been in this manuscript—Hrollaugs saga og Ingibjargar, Vilmundar saga væna og Hrómundar blinda, and a closing piece beginning Maður skuli og mann læra—came from elsewhere, suggesting a fresh round of editorial decisions. We cannot know whether Lbs 1137 8vo IV once contained Nikulás saga leikara, but it certainly did not follow Nítíða saga directly, suggesting that Jón was comfortable breaking the link between the sagas. Likewise, JS 632 4to gave at least Sigurgarðs saga frækna, probably Ambrósíus saga, and possibly Hálfdanar saga Barkarsonar and Nikulás saga leikara to SÁM 131, copied by Guðbrandur Sturlaugsson (1821–97) at Hvítadalur (the same farm where Rask 31 and 32, now in Denmark, had once been produced) between 1871 and 1890. But Guðbrandur did not feel the need to keep Nikulás saga leikara together with Nítíða saga, dropping the latter, and he turned elsewhere for Sigurgarðs saga og Valbrands, Ketlerus saga keisaraefnis (by Jón Oddsson Hjaltalín, 1749–1835), Úlfs saga Uggasonar, and Parmes saga loðinbjarnar (by Jón Bjarnason, 1721–85). Moreover, the latter three of these are post-medieval romances: the range of compositions available to Guðbrandur were rather different from those which had determined the editorial decisions behind Rask 32, and perhaps Guðbrandur’s interest in those more recent sagas also indicates that his tastes were changing with the times, somewhat as Ólafur Þorgeirsson preferred relatively recent rímur to sagas.

6. Conclusions

§48. In publishing the first stemma of Sigurgarðs saga frækna, this study has established one of the most thorough accounts of the manuscript transmission of an Icelandic saga; the eighth reasonably thorough stemma of a medieval Icelandic romance; and, through the use of electronic publishing and open data, probably the most extensively and transparently evidenced one. We have provided extra evidence for a now quite well-established master-narrative for the transmission of medieval Icelandic riddarasögur and fornaldarsögur in which medieval texts were copied in the seventeenth century by humanist scholars focused around Iceland’s episcopal seats. Their activity was followed by a seventeenth- to eighteenth-century translatio studii to the Westfjords, which was in turn inherited in the eighteenth to nineteenth centuries by a vigorous culture of local copying there and, even more so, slightly to the south in the Dalir. We have also been able to show localized nineteenth-century manuscript transmission in the south-west and the Eastfjords, followed by the geographically diverse distribution and copying of the 1884 printed edition of Sigurgarðs saga frækna.

§49. This study has also published the first sketch of a stemma of Nikulás saga leikara and slightly refined our understanding of the transmission of Ambrósíus saga og Rósamundu, Sigurðar saga turnara, Hálfdanar saga Eysteinssonar, Nítíða saga, and Konráðs saga keisarasonar. What these rougher but wider-ranging analyses have shown is that our understanding of the transmission of medieval Icelandic romances and fornaldarsögur is just starting to reach the level where we can understand not only the transmission of individual sagas from one manuscript to another, but also how scribes operated in their capacity as anthologists, at least in the eighteenth to nineteenth centuries. With a case study of copying in the Dalir region, we have found no reason to think that any scribe copied an exemplar in its entirety, but we have been able to show examples both of large-scale adoption of earlier scribes’ editorial choices and of much more selective combining of texts from multiple source manuscripts, as well as instances where a scribe accepted a former anthologist’s choice of story, but substituted a different recension of that story from an alternative exemplar. The same scribes’ lack of interest in replicating text-order also cautions us against assuming that the sequence of texts in manuscripts was important to their editors and readers, or that it is useful evidence for making stemmas.

§50. We have scarcely hinted here at the literary analyses that these findings might facilitate, but our publication of a stemma will allow future researchers to assess how different editors adapted Sigurgarðs saga frækna to their own tastes in ways that have hitherto been attempted in detail for sagas of this type by only a couple of commentators (McDonald Werronen 2016; Lavender 2020). For example, a glance at Spreadsheet 1 will show that the version in Borgarnes, Héraðskjalasafn Borgarfjarðar, MS 14 is among the most innovative versions of Sigurgarðs saga frækna, and its alterations to the saga can now be traced through comparison with its closest relatives among Sigurgarðs saga frækna manuscripts. Borgarnes 14 was produced from 1862 to 1867 by Jóhannes Jónsson, a farmer in the Dalir, and emerges from our data as an interesting witness to one scribe who was given to a lively and innovative engagement with his source material.

§51. We can also think with a new degree of confidence about what literary meanings might lie in the juxtapositions of sagas in manuscripts. Our case study has suggested that copyists did not much care in what order they copied sagas into a manuscript: we think, at least, that the burden of proof now lies with those who would like to see literary significance in the sequencing of texts. But copyists clearly took a critical interest in what sagas they put into a given volume. To give a quite superficial but not uninteresting example, we have shown that the three sagas Sigurgarðs saga frækna, Nítíða saga, and Nikulás saga leikara were probably brought together twice by independent anthologists in eighteenth-century Iceland (once in Lbs 644 4to and once in Rask 32); this collocation was sustained by Ólafur Jónsson at the end of the century in two manuscripts descended largely from Rask 32 (Lbs 998 4to and its child JS 632 4to), which is unsurprising because Ólafur adopted most of Rask 32’s texts; but it was also sustained in the nineteenth century in two manuscripts, copied from Lbs 998 4to for different audiences, by Ólafur Þorgeirsson in 1869–71, in a context of greater editorial selectivity. The pairing of Nítíða saga and Nikulás saga leikara surely reflects the fact that in this recension the latter is said to be the sequel to the former, showing an interest in uniting such paired sagas in anthologies; but perhaps the pairing of Nítíða saga with Sigurgarðs saga frækna reflects the fact that Nítíða saga is the most clearly proto-feminist saga in the genre, which chimes with how the protagonist of Sigurgarðs saga frækna is shown to get his comeuppance for womanizing (cf. McDonald Werronen 2016; Hall, Richardson and Þorgeirsson 2013, 94–100). That said, it is the pairing of Sigurgarðs saga frækna and Nikulás saga leikara that stood the test of time, with these two sagas remaining as a pair in SÁM 131 (1871–72) and Lbs 5567 4to (1913), and it would be worth investigating why. Many other such targeted and well-evidenced investigations await us now that electronic methods are making the extensive or even complete surveying of the manuscript transmission of whole saga-genres a practical undertaking.

Acknowledgements

Hall’s early data-collection for this project was funded by a 2008 British Academy small grant; McDonald, Fernandez-Aceves, and Simpson’s contributions were funded by a 2009 University of Leeds Teaching Fellowship award. We are grateful for both. We also thank Haukur Þorgeirsson for invaluable comments on this work, and the late Thor Leifson for photographing his grandfather’s manuscript of Sigurgarðs saga frækna and making the images available to researchers.

Appendices

1. Fundamental data

Spreadsheet 1: Sigurgarðs saga transcriptions and data

• [ODS Format]
• [XLXS Format]
• [CSV Format]

Spreadsheet 2: Sigurgarðs saga variants from Loth and other early MSS (AM 556a 4to, AM 588m 4to, AM 167 fol, AM 123 8vo, AM 588n 4to, Lbs 423 fol, Lbs 222 fol, GKS 1002 fol)

• [ODS Format]
• [CSV Format]
• [XLSX Format]
• [Raw File] (derived from spreadsheet)

Spreadsheet 3: Nikulás saga leikara variants from Wick 1996

• [ODS Format]
• [CVS Format]
• [XLSX Format]

Spreadsheet 4: Nikulas saga leikara variants from our sample of ch 1

• [ODS Format]
• [CSV Format]
• [XLSX Format]

Stemmas for all well researched romance sagas

• [DOT Format]
• [PS Format]
• [PNG Format]

2. Derived files and images

Figure 1: Stemma of Sigurgarðs saga frækna (derived from Spreadsheet 1)

• [DOT Format]
• [PS Format]
• [PNG Format]

Figures 2 and 4: Known places of production of Sigurgarðs saga frækna manuscripts (derived from Spreadsheet 1)

• [KML Format]
• All Known Places [JPG Format]
• Dalir [JPG Format]

Figure 3: Unrooted computer-generated stemma of Sigurgarðs saga frækna (derived from Spreadsheet 1)

• [Raw File]
• [PS Format]
• [PNG Format]

Figure 5: Stemma of selected sagas in the Rask 32 group

• [DOT Format]
• [PS Format]
• [PNG Format]

Figure 6: Rooted stemma of Wick’s Nikulás saga leikara recension-one manuscripts (derived from Speadsheet 3)

• [Raw File]
• [PS Format]
• [PNG Format]

Figure 7: Unrooted stemma of chapter 1 of a sample of Nikulás saga leikara recension-one manuscripts (derived from Speadsheet 4)

• [Raw File]
• [PS Format]
• [PNG Format]

Figure 8: A-branch of Nítíða saga stemma according to McDonald Werronen 2016

• [DOT Format]
• [PS Format]
• [PNG Format]

Works Cited

Blöndal, Lárus H., Grímur M. Helgason, and Ögmundur Helgason. 1947–96. Handritasafn Landsbókasafns. 4 vols. Félagsprentsmiðjun.  [Back]

Driscoll, Matthew and others. 2011–2015. Stories for All Time: The Icelandic Fornaldarsögur. [https://fasnl.net/]  [Back]

Einarsson, Stefán. 1948. “Safn Nikulásar Ottensons í Johns Hopkins Háskólabókasafninu í Baltimore, Md.” Árbók Landsbókasafns Íslands 3–4: 157–72.  [Back]

Guðjónsson, Þórður Ingi. 2010. “Editing the Three Versions of Gísla saga Súrssonar.” In Creating the Medieval Saga: Versions, Variability, and Editorial Interpretations in Old Norse Saga Literature, edited by Judy Quinn and Emily Lethbridge. University Press of Southern Denmark.  [Back]

Hall, Alaric, Steven D.P. Richardson, and Haukur Þorgeirsson. 2013. “Sigrgarðs saga frækna: A Normalised Text, Translation, and Introduction.” Scandinavian-Canadian Studies/Études Scandinaves au Canada 21: 80–155.  [Back]

Hall, Alaric, and Katelin Parsons. 2013. “Making Stemmas with Small Samples, and Digital Approaches to Publishing them: Testing the Stemma of Konráðs saga keisarasonar.” Digital Medievalist 9: https://doi.org/10.16995/dm.51 [fully functional working paper at http://www.alarichall.org.uk/working_paper_on_stemmas_from_small_samples].  [Back]

Hall, Alaric and Ludger Zeevaert. 2018. “Njáls saga Stemmas, Old and New.” In New Studies in the Manuscript Tradition of Njáls saga: The Historia mutila of Njála, edited by Emily Lethbridge and Svanhildur Óskarsdóttir. Medieval Institute Publications, Western Michigan University.  [Back]

Handrit.is. 2010–. https://handrit.is/  [Back]

Hast, Sture, ed. 1960a. Harðar saga. Editiones Arnamagnæanæ, Series A, 6. Munksgaard.  [Back]

———. 1960b. Pappershandskrifterna till Harðar saga. Bibliotheca Arnamagnæana, 23. Munksgaard.  [Back]

Haugen, Odd Einar. 2016. “The silva portentosa of stemmatology: Bifurcation in the recension of Old Norse manuscripts.” Digital Scholarship in the Humanities 31: 594–610.  [Back]

Hufnagel, Silvia V. 2012. “Sörla saga sterka”: Studies in the Transmission of a Fornaldarsaga. University of Copenhagen, Faculty of Humanities.  [Back]

———. 2016a. “Texts and Contexts: Bjarni Pétursson and His Saga Manuscript Lbs 2319 4to (1727–1729).” Scandinavian Studies 88: 393–422.  [Back]

———. 2016b. “The Library of the Genius: The Manuscript Collection of Rasmus Christian Rask.” Tabularia: Sources écrites des mondes normands. Autour des sagas: Manuscrits, transmission et écriture de l’histoire, 17 November: https://doi.org/10.4000/tabularia.2666.  [Back]

Kalinke, Marianne E. and P. M. Mitchell. 1985. Bibliography of Old Norse–Icelandic Romances, Islandica 44. Cornell University Press.  [Back]

Kapitan, Katarzyna Anna, Timothy Rowbotham and Tarrin Jon Wills. 2017. “Visualising Genre Relationships in Icelandic Manuscripts.” In Digital Humanities in the Nordic Countries, Second Conference: Conference Abstract, edited by Daniel Brodén. University of Gothenburg.  [Back]

Kapitan, Katarzyna Anna. 2017. “A Choice of Relationship-Revealing Variants for a Cladistic Analysis of Old Norse Texts: Some Methodological Considerations.” In International Symposium on Digital Humanities, 7–8 November 2016, Växjö, Sweden: Extended Papers, edited by Koraljka Golub and Marcelo Milrad, CEUR Workshop Proceedings (CEUR-WS.org): Free Open-Access Proceedings for Computer Science Workshops 2021, 52–74. Aachen: CEUR-WS Team.  [Back]

———. 2021. “Perspectives on Digital Catalogs and Textual Networks of Old Norse Literature.” Manuscript Studies: A Journal of the Schoenberg Institute for Manuscript Studies 6: 74–97.  [Back]

———. 2024. Lost but not Forgotten: The Saga of Hrómundur and its Manuscript Transmission. Taylor Institution Library.  [Back]

Kapitan, Katarzyna Anna and Tarrin Wills. 2023. “Sagas and Genre: A Case for Application of Network Analysis to Manuscripts Preserving Old Norse-Icelandic Saga Literature.” Digital Scholarship in the Humanities 38(3): 1130–1144. [Back]

Kristjánsson, Jónas, ed. 1960. Dínus saga drambláta. Riddarasögur, 1. Handritastofnun Íslands.  [Back]

———. 1964. Viktors saga ok Blávus. Riddarasögur, 2. Handritastofnun Íslands.  [Back]

Kålund, Kristian. 1900. Katalog over de oldnorsk-islandske håndskrifter i Det store kongelige bibliotek og i Universitetsbiblioteket (udenfor den Arnamagnæanske samling) samt den Arnamagnæanske samlings tilvækst 1894–99). Gyldendal.  [Back]

Lansing, Tereza. 2011. “Post-medieval Production, Dissemination and Reception of Hrólfs saga kraka.” PhD diss., University of Copenhagen.  [Back]

Lavender, Philip. 2020. Long Lives of Short Sagas: The Irrepressibility of Narrative and the Case of Illuga saga Gríðarfóstra. The Viking Collection: Studies in Northern Civilisation 25. University Press of Southern Denmark.  [Back]

Lethbridge, Emily. 2012a. “Authors and Anonymity, Texts and Their Contexts: The Case of Eggertsbók.” In Modes of Authorship in the Middle Ages, edited by Slavica Rankovič, Papers in Medieval Studies 22. Brepols.  [Back]

———. 2012b. “The Place of Þorsteins saga Víkingssonar in Eggertsbók, a Late Medieval Icelandic Saga-book.” In Uppruni og þróun fornaldarsagna Norðurlanda/The Origins and Development of the Legendary Sagas, edited by Ármann Jakobsson, Agnete Ney, and Annette Lassen. Háskólaútgáfan.  [Back]

Loth, Agnete, ed. 1962–1965. Late Medieval Icelandic Romances. Editiones Arnamagnæanæ, Series B, 20–24, 5 vols. Munksgaard.  [Back]

Love, Jeffrey Scott. 2013. The Reception of Hervarar saga ok Heiðreks from the Middle Ages to the Seventeenth Century. Münchener nordistische Studien 14. Utz.  [Back]

Magnússon, Sigurður Gylfi. 2010. Wasteland with Words: A Social History of Iceland. Reaktion.  [Back]

McDonald, Sheryl. 2014. “Two Major Groups in the Older Manuscript Tradition of Nítíða saga.” Saga-Book of the Viking Society for Northern Research 38: 75–94.  [Back]

McDonald Werronen, Sheryl. 2016. Popular Romance in Iceland: The Women, Worldviews, and Manuscript Witnesses of Nítíða saga. Crossing Boundaries: Turku Medieval and Early Modern Studies 5. Amsterdam University Press.  [Back]

———. 2016–2018. “Icelandic Scribes: Scribal Networks in 17th-century Iceland. The Patronage of Magnús Jónsson í Vigur.” https://icelandicscribesproject.com.  [Back]

———. 2020. “The Working Practices of Magnús Ketilsson: An Icelandic Scribe at the End of the Seventeenth Century.” Scandinavian Studies 92: 39–61.  [Back]

McDonald Werronen, Sheryl and Katarzyna Anna Kapitan. 2018. “An Edition of Ambrósíus saga og Rósamunda Based on BLAdd 24 969.” Opuscula 16: 179–215.  [Back]

Ólafsson, Davíð. 2009. “Wordmongers: Post-Medieval Scribal Culture and the Case of Sighvatur Grímsson.” PhD diss., University of St Andrews.  [Back]

Ólason, Páll Eggert. 1918–1937. Skrá um handritasöfn Landsbókasafnsins, 3 vols. Prentsmiðjan Gutenberg.  [Back]

Ohlsson, Tove Hovn, ed. 2009. Tiodelis saga. Stofnun Árna Magnússonar í Íslenskum Fræðum.  [Back]

Óskarsdóttir, Svanhildur. 2009. “To the Letter: Philology as a Core Component of Old Norse Studies.” Scripta Islandica 60: 7–22.  [Back]

Page, R.I. 1957. “Drauma-Jóns saga.” Nottingham Medieval Studies 1: 22–56.  [Back]

———. 1960. Gibbons saga. Editiones Arnamagnæanæ, Series B, 2. Munksgaard.  [Back]

Parsons, Katelin Marit. 2019. “Albert Jóhannesson and the Scribes of Hecla Island: Manuscript Culture and Scribal Production in an Icelandic-Canadian Settlement.” Gripla 30: 7–46.  [Back]

———. 2020. “Songs for the End of the World: The Poetry of Guðmundur Erlendsson of Fell in Sléttuhlíð.” PhD diss., University of Iceland.  [Back]

———, ed. 2023–. Icelandic Immigrant Literacy Database [https://vesturheimur.arnastofnun.is/]  [Back]

———. 2024. “Relics of the Past: Calfskin, Authenticity, and Nordic Antiquarianism.” Manuscript Studies: A Journal of the Schoenberg Institute for Manuscript Studies 9.2: 268–92.  [Back]

Platnick, Norman I. and H. Don Cameron. 1977. “Cladistic Methods in Textual, Linguistic, and Phylogenetic Analysis.” Systematic Zoology. 26(4): 380–385. https://doi.org/10.2307/2412794  [Back]

Robinson, P.M.W. 1989a. “The Collation and Textual Criticism of Icelandic Manuscripts (1): Collation.” Literary and Linguistic Computing 4: 100–5.  [Back]

———. 1989b. “The Collation and Textual Criticism of Icelandic Manuscripts (2): Textual Criticism.” Literary and Linguistic Computing 4: 174–81.  [Back]

Robinson, P.M.W. and Robert J. O’Hara. 1996. “Cladistic Analysis of an Old Norse Manuscript Tradition.” In Selected Papers from the ALLC/ACH Conference, Christ Church, Oxford, April 1992. Research in Humanities Computing 4. Clarendon Press.  [Back]

Sanders, Christopher, ed. 2000. Tales of Knights: Perg. fol. nr 7 in The Royal Library, Stockholm (AM 167 VIβ 4to, NKS 1265 IIc fol.) Manuscripta Nordica: Early Nordic Manuscripts in Digital Facsimile 1. Reitzel.  [Back]

Schröder, Franz Rolf, ed. 1917. Hálfdanar saga Eysteinssonar. Altnordische Saga-Bibliothek 15. Niemeyer.  [Back]

Slay, Desmond. 1960a. The Manuscripts of Hrólfs saga Kraka. Bibliotheca Arnamagnæana 24. Munksgaard.  [Back]

———. 1960b. Hrólfs saga kraka. Editiones Arnamagnæanæ, Series B, 1. Munksgaard.  [Back]

———. 1970. “Hitherto Unused Manuscripts of Hrólfs saga kraka.” Opuscula 4: 260–68.  [Back]

———. 1981. “More Manuscripts of Hrólfs saga kraka.” In Speculum norroenum: Norse Studies in Memory of Gabriel Turville-Petre, edited by Ursula Dronke and others. Odense University Press.  [Back]

———. 1994. “Perhaps the Last Hrólfs saga kraka Manuscript,” in Strengleikar slegnir Robert Cook, edited by Margrét Eggertsdóttir. Menningar- og Minningarsjóður Mette Magnussen.  [Back]

———, ed. 1997. Mírmanns saga. Editiones Arnamagnæanæ, Series A, 17. Reitzel.  [Back]

Spaulding, Janet Ardis. 1982. “Sigurðar saga turnara: A Literary Edition.” PhD diss., University of Michigan.  [Back]

Springborg, Peter. 1977. “Antiqvæ historiæ lepores—om renæssancen i den islandske håndskriftproduktion i 1600-tallet.” Gardar: Årsbok för Samfundet Sverige-Island i Lund-Malmö 8: 53–89.  [Back]

Surmeli, Þorsteinn Árnason. 2020. “Í búningi samtíðar sinnar.” Saga 58.1: 7–16.  [Back]

Stankovitsová, Zuzana. 2023. “Parchment, Paper and Print: The Manuscript Transmission of Króka-Refs saga.” PhD diss., University of Bergen.  [Back]

Trovato, Paolo. 2017. Everything You Always Wanted to Know about Lachmann’s method: A Non-standard Handbook of Genealogical Textual Criticism in the Age of Post-Structuralism, Cladistics, and Copy-text. Translated by Federico Pool. Libreriauniversitaria.it.  [Back]

Wick, Keren H. 1996. “An Edition and Study of Nikulás saga leikara.” PhD diss., University of Leeds.  [Back]

Zeevaert, Ludger and others. Forthcoming. “A New Stemma of Njáls saga.” DOI https://doi.org/10.17613/x9wtx-y8b37.  [Back]

Zitzelsberger, Otto J. 1980. “AM 567, 4to, XVI, 1v: An Instance of Conflation?’” Arkiv för nordisk filologi 95: 183–88.  [Back]

———. 1981. “The Filiation of the Manuscripts of Konráðs saga keisarasonar.” Amsterdamer Beiträge zur älteren Germanistik 16: 145–76.  [Back]

———. 1983. “Six MSS of Konráðs saga Unlisted in Printed Catalogues.” Amsterdamer Beiträge zur älteren Germanistik 18: 161–77  [Back]

Þorgeirsson, Haukur. 2017. “A stemmatic analysis of the Prose Edda.” Saga-Book 41: 49–70. [Back]

———. 2022. “Z-texti Laxdæla sögu–textafræðileg tilraun.” Gripla 33: 7–38. [Back]