Reading Animal-Human Relations: Sámr and Gunnarr in Njáls saga

Introduction

Í því sér hundrinn, at þar eru menn fyrir, ok hleypr á hann Þorkel upp ok grípr í nárann; Ǫnundr ór Trǫllaskógi hjó með ǫxi í hǫfuð hundinum, svá at allt kom í heilann; hundrinn kvað við hátt, svá at þat þótti með ódœmum, ok féll hann dauðr niðr. (Brennu-Njáls saga 1954, 185–6)

(In that moment, the dog sees that there are men in front of him, and he leaps up on Þorkell and catches hold of his groin. Ǫnundr from Trǫllaskógr struck with his axe into the head of the dog, so that the blade went into the brain; the dog cried out loudly, so that it seemed to them unprecedented, and he fell down dead.)2

In one of the most memorable scenes of Njáls saga, Gizurr hvíti and Geirr goði ride out to Hlíðarendi to attack Gunnarr at his home farm. To approach Gunnarr, they know that they must first dispatch his dog, Sámr, and do so by coercing a neighboring farmer to lure the dog away from the house while they wait to sink an axe into Sámr’s head. Before he dies, Sámr seizes an enemy’s crotch (or stomach) in his jaws and lets out a cry that wakes Gunnarr, apparently warning him of the impending attack (Brennu-Njáls saga 1954, 185). Gunnarr’s subsequent comments have often been interpreted as Gunnarr lamenting the death of a favored pet, and suggesting that he, too, soon will die, because Sámr’s cry has warned him of the intruders:

Gunnarr vaknaði í skálanum ok mælti: “Sárt ertú leikinn Sámr fóstri, ok búð svá sé til ætlat, at skammt skyli okkar í meðal.” (Brennu-Njáls saga 1954, 186)

(Gunnarr woke in the hall and said: “Painfully are you played with Sámr foster-kin, and it may be intended that a short time should be between us-two.)

However, this article argues that such an interpretation does not take into account the full range of meanings conveyed in the two Sámr episodes in the saga. Both Sámr’s cry and Gunnarr’s response, while part of a relationship based on dependence, are also placed in a context of loyalty and a mutually beneficial relationship. That Sámr and Gunnarr share their moment of death has consequences for the way we read these figures, particularly Sámr, in the saga.

Interactions between animals and humans in the sagas are complex, and multi-layered: there was not one way of expressing such relationships across Old Norse textual sources, and each relationship, with its particular context and inter-textual connections, must be considered on its own terms. This article will unpick the various levels of meaning contained within the exchange highlighted above, and the Sámr-Gunnarr relationship depicted throughout Njáls saga, arguing that Sámr’s death and Gunnarr’s fatalistic comment have more in common than a simple matter of warning, and that the figure of Sámr is both dog and social companion to Gunnarr. This relationship will be discussed alongside legal conceptualizations of the canine-human relationship as indicated by the lawbook Grágás, and depictions of relationships between humans and humans, and between humans and animals elsewhere in the Íslendingasögur—specifically, depictions and usage of fóstri and fictive kinship bonds. This article will first introduce the theoretical work conducted around animals in the Old Norse-Icelandic sagas, before considering the various aspects of the depiction of Sámr and his relationship with Gunnarr in Njáls saga, and finally concluding with a consideration of how close analysis of these episodes suggests that we need to alter how we read animals and animal-human relationships in the sagas, arguing for the multiplicity of meanings associated with the concept of the “animal” in Old Norse-Icelandic texts.

Animals in Old Norse-Icelandic Texts

Domestic animals were important figures in both the material and narrative settlement of Iceland, and in the way in which the place of the Icelandic home developed in both physical structure and legal concepts (Evans Tang 2022). This importance is interwoven into the stories that Icelanders told about their past, including texts such as Landnámabók and the Íslendingasögur, as well as compilations of laws such as Grágás. Discussions of animals in Old Norse-Icelandic literature have only recently begun emerging in English scholarship (Evans Tang 2022; Bourns 2018), and prior to this, literary studies saw only two publications explicitly focused on animals and animal-human relations in these texts—one in German and one in Norwegian, and both with their specific focus. Simon Teuscher (1990) provides a discussion of animals and men in the Íslendingasögur, but uses the sagas as evidence for society, with little linguistic or literary analysis, while Lena Rohrbach’s comprehensive study Der tierische Blick: Mensch-Tier-Relationen in der Sagaliteratur (2009) uses literary analysis to access a wide range of animal-human relations in Old Norse-Icelandic texts. However, Rohrbach primarily considers the uses of animals as narrative features of the sagas rather than analyzing the relationships represented between animals and humans in these texts in the context of the animal-human social networks of farming landscapes.

While Rohrbach suggests that the interactions between domestic animals and humans depicted in the Íslendingasögur are primarily used to reinforce masculine human behavior in comparison to an inferior animal figure, not all animal-human relations conform to this model of symbolic expression (Rohrbach 2009, 294). Thinking about domestic animals specifically, these figures are not used solely by saga writers to mirror the attributes or characteristics of humans in these texts, and confining our view of the animal as imitative, metaphorical, or symbolic limits our ability to interpret these animals as meaningful agents in their own right. While the depictions between animals and humans often show common features (for example, boisterous attractive young men are often paired with lively, well-presented but untried horses), the animal and human partners in the Íslendingasögur should be considered to be placed on a more level ontological footing than in the conclusions proposed by previous studies (Teuscher 1990; Rohrbach 2009).

Torfi Tulinius has suggested that the compilation of the Íslendingasögur may have been triggered by social anxiety expressed by Icelandic elites in the thirteenth century, especially toward circulating narratives of the origins of the Icelanders and increasing redefinition of social roles within medieval Icelandic society (2003, 527, 536). As such, these sagas can be seen as formative narratives in a program to make sense of the medieval Icelandic world and the society and ancestors who constituted that world. Animal-human interactions were a vital and daily occurrence for these people and this society, and it is therefore unsurprising that the Íslendingasögur likewise show clear awareness of the importance of, and risks around, close animal-human relationships, specifically in association with changing economic and environmental conditions (Hartman, Ogilvie, and Hennig 2017, 134; Evans 2016; Evans Tang 2022; Ingimundarson 1995; 1992).3 A refocused analysis of the domestic animal-human interactions in the Íslendingasögur is required, that recognizes the representation of certain animals in these texts as active players in networks of social exchange and kinship. The relationships with these animals echo human social organization, and they are often attributed what we consider “human” characteristics.

Animals in the Íslendingasögur can be close companions, worthy or beloved members of the household, or loathed enemies. These ordinary, non-magical, non-human figures can act for themselves, drive the action of saga episodes, and take part in human legal and social networks, understanding the responsibilities required of them as agents in these systems. Indicators of human personhood are expressed, while the figure often retains their expression of “real-world” animal behavior. Such representations of certain animals echo the ambiguities found in the Grágás laws, in which dogs and other animals occupy a legal space that cannot be called “human,” but does not appear as entirely “animal” (Evans Tang 2022, 128–33). The language of homosociality is extended to them, through phrases such as fóstri (male foster-kin), and these extensions suggest that the compilers of these sagas believed this to be a plausible way in which to phrase these relationships. These animals are not human, nor are they simply like men—rather, the categories of human and animal seem in themselves not quite right for describing the depictions of certain animals in the sagas.

Sámr the dog is such a figure.

Sámr’s Place

In Sámr’s two appearances in Njáls saga, the space or changing spaces occupied by the dog are explicitly highlighted. In the first, Sámr is placed at the side of Óláfr, his human master, before he moves himself to Gunnarr’s feet at Óláfr’s command; in the second, Sámr is positioned on the roof of Gunnarr’s house (a reversal of his position at Gunnarr’s feet) before being lured away from the house to be killed. The spatial positioning of these interactions is important. In the first, although Óláfr clearly recognizes the abilities of Sámr, their relationship is one based on a hierarchy of position that places Óláfr above Sámr, and Sámr is initially placed below Gunnarr as he moves to Gunnarr’s feet. In the second, Sámr is literally above Gunnarr, an integral part of the outlaw’s home, and a figure that must be dispatched before the attack against Gunnarr can take place.

Gunnarr’s enemies want to attack Gunnarr at his home. This is not unusual for the sagas, as groups of men often descend on farmsteads or shielings for the purpose of responding in a feud, offering a dramatic depiction of the isolated defender(s) surrounded by enemies. Although this motif often centers around attacks at shielings—for example, the killings of Bolli and Helgi in Laxdœla saga—the environs of the home are often the places in which animal-human interactions are depicted in the Íslendingasögur (Laxdœla saga 1934, 165–6, 191; Evans Tang 2022).4 The emphasis here on the main farm places the outlawed figure of Gunnarr in the center of the social sphere from which he is supposed to have been excluded.

Sámr is explicitly linked with, and even integrated into Gunnarr’s home. Gunnarr’s enemies know this, and they know they will need to lure Sámr away from the home-place before they are able to successfully attack Gunnarr. The build-up to the attack relies on the careful distinction made between certain spaces, into which figures may or may not enter. The boundary between the place of the home and the place outside of it is controlled by Sámr, as it is Sámr’s ability to discern friend from foe (discussed below) that necessitates Þorkell’s solitary approach to the house:

Traðir váru fyrir ofan garðinn at Hlíðarenda, ok námu þeir þar staðar með flokkinn. Þorkell bóndi gekk heim, ok lá rakkinn á húsum uppi, ok teygir hann hundinn braut með sér í geilar nǫkkurar. (Brennu-Njáls saga 1954, 185–6)

(There were animal-pens at the top of the enclosure at Hlíðarendi, and the band of men stopped there at that place. Farmer Þorkell went toward the home, and the dog lay up on the house, and he entices the dog away with him into a certain lane.)

While “traðir” is sometimes translated as “beaten sunk road” (Dasent 1861, 241), here, I propose it should indicate the location of animalpens—not an unlikely presence within a home enclosure of an Icelandic farm. This translation is in line with Cook (2001, 125), and such a spatial description in this passage foregrounds the progression from animal-space to the house. Both house and animal-pens are surrounded by the enclosure, and Þorkell must move through this boundary, past the animal-pens, and toward the home to reach Sámr. Sámr himself is explicitly placed on the roof of the house and must be drawn downward and away from that place to be killed.5 Sámr’s presence on the roof is significant, given that the attackers later peel away the roof of the house in their attack against Gunnarr (chap. 77). The luring away of Sámr from the house then, is stage one in a multi-step process of deconstructing Gunnarr’s home.

The position of Sámr literally above Gunnarr in this later chapter stands in stark contrast to his earlier role of “honorable gift” implied by his introduction in chapter 69:

“Hann er mikill ok eigi verri til fylgðar en rǫskr maðr. Þat fylgir ok, at hann hefir manns vit; hann mun ok geyja at hverjum manni, þeim er hann veit, at óvinr þinn er, en aldri at vinum þínum; sér hann ok á hverjum manni, hvárt honum er til þín vel eða illa; hann mun ok lífit á leggja at vera þér trúr. Þessi hundr heitir Sámr.” Síðan mælti hann við hundinn: “Nú skaltú Gunnari fylgja ok vera honum slíkr sem þú mátt.” Hundrinn gekk þegar at Gunnari ok lagðisk niðr fyrir fœtr honum. (Brennu-Njáls saga 1954, 173)

(“He is great of size and not worse at support than a brave man. Indeed, he has a man’s knowledge; he will also bark at each man who he knows is not your friend, but never at your friends; he sees in each man whether by him is wished to you well or ill, and he will lay down his life in order to be true to you. This dog is called Sámr.” Afterward he said to the dog: “Now you must accompany Gunnarr and be to him such as you are able.” The dog goes at once to Gunnarr and lays himself down at his feet.)

However, despite Sámr’s apparent nature here as a passive figure in the social bonding of two men, there are clear indications that the dog has something more to offer than simply reinforcing the bond between Gunnarr and Óláfr.

In this description, we see the writer ascribing a variety of behaviors and attributes to Sámr, echoing those of dogs found in medieval encyclopedic and bestiary texts (for example, the thirteenth-century De Proprietatibus Rerum), especially in the focus on cleverness, loyalty, and willingness to lay down their lives for their masters. However, the bestiary tradition depicts the human-dog relationship as a masterservant relationship, and while this seems to be how Ólafr relates to Sámr (by ordering him to follow Gunnarr), and how Sámr, at this point, accepts the relationship by taking a low position at Gunnarr’s feet, this relationship (as we will see) is leveled to a more horizontal understanding by Gunnarr’s comments and actions later in the saga. It may be suggested that the depiction of Sámr and Gunnarr revolts against the vilification of dogs as servile creatures in the writings of early medieval thinkers (see Salisbury 2011, 104).

The text states that Sámr is no worse as a companion than a rǫskr maðr (brave man) and possesses manns vit (intelligence or understanding of a maðr). The text does not say that Sámr is like a brave man, but rather, that he is no worse at providing support than one—here, the distinction is not explicitly between dog and man, but implicitly between a brave man and an un-brave man, and places Sámr on a comparable social level with a brave man. In addition to its meaning of “following,” “support,” or “party,” “fylgð” can also indicate “guidance” (Cleasby and Vigfússon 1874, 179), and this passage therefore might imply that Sámr is no worse at providing guidance than a brave man. Such a depiction is continued in the explicit reference to his possession of “manns vit” (Brennu-Njáls saga 1954, 173), and these attributes and capabilities place Sámr on the border of, if not within the sphere of maðr-ness. In “Regardless of Sex” (1993), Carol Clover argues that power is distributed to those who can act as a social man, and it might be wondered whether a similar continuum of social identity can be applied across the animal-human boundary. Biologically, Sámr is a dog, but through this description, he is assigned attributes of a social human, and specifically a male warrior companion. Sámr also understands human speech, as he obeys Óláfr’s command that he is now to “Gunnari fylgja” (Brennu-Njáls saga 1954, 173) [follow Gunnarr] and be a worthy companion to him, acting with all the special abilities Óláfr has outlined: “Vera honum slíkr sem þú mátt” (Brennu-Njáls saga 1954, 173) [Be to him such as you are able]. This command can be read in two ways—Óláfr may refer here to Sámr’s limitations as a dog, telling him to “be to him such as you are able as a dog,” or his words may reinforce the outstanding, maðr-like capabilities and status of Sámr: “Be to him such as you are able as a possessor of these abilities,” reminding the dog that he has these abilities and should use them in this partnership with Gunnarr.

William Sayers has drawn attention to the nature of Sámr as a gift, and as an explicitly Irish gift. The gifting and ownership of large dogs were prominent features of early Medieval Irish culture, especially among the higher echelons of society, with both legal and literary texts emphasizing their symbolic and practical importance (Sayers 1997, 44–5). Such a vitally useful Irish dog is found elsewhere in Old Norse tradition in the figure of Vígi, the dog acquired by Ólafr Tryggvason from an Irish farmer in Ólafs saga Tryggvasonar (1941, 269). In both cases, the dogs are gifted by Irish persons, are attributed (unusual?) levels of intelligence, demonstrate the characteristics of fierce warriors, and remain loyal to their owners unto death. However, there are a few key differences in their descriptions, and the circumstances of their transference from one owner (or partner) to another. The dog acquired by King Ólafr is explicitly a hjarðhund (herding-dog), whereas Sámr appears to be a companion dog—a guard dog, yes, but specifically guarding Gunnarr and wherever he may be (Ólafs saga Tryggvasonar 1941, 269). In addition, while Vígi’s cleverness in sorting and herding the cattle seems a wonder to the king and his men, it may be suggested this disbelief stems from an unfamiliarity with herding animals and the capabilities of herding dogs. The dog’s cleverness then is depicted by Snorri as rooted in his animal nature, rather than extending into manns vit, as emphatically attributed to Sámr.6 It is perhaps through his relationship with the king that Vígi (only named in the saga after his transference to the king) displays a greater number of social human features in becoming effectively a member of the king’s retinue, as opposed to a herding animal. Vígi, like Sámr, is only given one opportunity in Snorri’s text to prove his worth in his new partnership: in this, he is explicitly shown following the king (as Sámr follows Gunnarr) and apprehending his enemy (Ólafs saga Tryggvasonar 1941, 325).7

While it might be said then that Sámr’s attributes are part of a tradition regarding the depiction of Irish dogs, rather than dogs in general, I believe it is just as important to note that Sámr’s intelligence and loyalty, and in particular his association with the home and his crying out, are attributes shared by a number of animals in the saga corpus, not just dogs, Irish or otherwise.8 Perhaps here we see a merging of traditions, on the one hand, a respect and admiration for the intelligence, fierceness, and loyalty of dogs of Celtic origin and, on the other hand, a native conceptualization of animals as capable of forming affective relationships with humans and acquiring social human-ness that extends beyond the canine.

The depiction of Sámr as expressing more than ordinary canine loyalty in this scene (and the subsequent death scene discussed above) is emphasized elsewhere in Óláfr’s description. Óláfr’s statement that “hann mun ok lífit á leggja at vera þér trúr” (Brennu-Njáls saga 1954, 173) [he will also lay down his life to be true to you] suggests that Sámr possesses a human sense of social obligation, mutual benefit, and selfsacrifice that goes beyond that of a normal guard-dog. It is not Sámr’s defensive capabilities that are expressed here, but his ability to be trúr (true, faithful, trustworthy) (Cleasby and Vigfússon 1874, 643). This is not a concept rooted in any behavioral trait, but rather a concept implying faith and belief, and linked with ethical and spiritual awareness (Cleasby and Vigfússon 1874, 643). This description of Sámr’s traits and response to Ólafr’s command simultaneously depicts Sámr as both animal and more than animal. Alternatively, it questions what it is we mean when we refer to “animal” in these texts—in the world of the saga, this hundr is ascribed ethical and intelligent decisionmaking abilities and is enclosed within human social bonds through his description as a follower, his later presence in the home-place, and Gunnarr’s use of the term fóstri to refer to him (discussed below).

A Legal Relationship

This relationship from Njáls saga, however, is one that we can approach from both a literary and a legal perspective, as the wording around the relationship between dogs and human persons in Grágas explicitly suggests this relationship was perceived as a mutual exchange of benefits.

The medieval Icelandic compilation of laws known as Grágás has a whole section of laws devoted to the regulation of dogs and their interactions with humans.9 While domestic animals of all kinds are generally listed in Grágas as the responsibility of both their owner and any man who encounters them, dogs seem to occupy a more nuanced position in the legal categories of medieval Icelandic legal traditions expressed in the Grágás texts (Dennis, Foote, and Perkins 2000, 167–8, 174–5; Evans Tang 2022, 132–3). On encountering a dog, a man would have had a choice, according to Grágas, on whether he wished to take it into his company and form a relationship with it:

Ef hundr kømr ifor með manne oc biðr hann mat gefa honum eða syslir vm hann er þeir coma til húss. Þa abyrgiz hann hund þótt aNaR eigi. eN eigi ef hann sciptir ser ecki af. (Finsen 1852b, 188)

(If a dog goes along with a man and he [the man] asks for food to be given to him or works for him when they come to a house, then he is responsible for the dog even if another owns it; but not if he does not concern himself with the dog.)

This quotation shows a remarkable situation in terms of animal-human relations. Nowhere else in Grágás can a man legally choose whether to look after an animal, nor is the relationship between man and animal laid out so explicitly as an exchange of action. The dog, as the subject of the verb and the instigator of the encounter, chooses to go along with a man, and the man chooses to work for him in the social networks of this farming society. Neither figure is passive in this relationship, as it is action that creates the legal bond, and both agents have a choice in how they approach each other. The role of food in cementing bonds between man and dog is found emphasized in other early medieval laws, such as those of Alfred and the Norwegian Gulaþing (Attenborough 1922, 75; Larson 1935, 134), although unlike in these earlier laws, the depiction of feeding and caring for dogs in Grágás is not explicitly concerned with the responsibility of a man for a violent animal but rather a recognition and legal prescription of the relationship in general. There were also no degrees of responsibility when it came to human-canine relationships: a person was either with a dog or not.

The partnership of dog and man is also indicated by the legal categorization of a dog as having “eigi hælgi” (Finsen 1852b, 187) [no legal immunity] unless that dog is correctly leashed by a human figure.10 While other animals lose their legal immunity for committing crimes such as killing (bulls) or trespassing on another man’s land (pigs), dogs are unique in this law text for having no inherent legal immunity. If a dog attacks a man or an animal while unleashed, a scale of punishments is laid out for the man who has responsibility for the dog, ranging from a three-mark fine to full outlawry (Dennis, Foote, and Perkins 2000, 201–2). In contrast, figures who approach a leashed dog are themselves responsible for any harm that comes to them, rather than the dog or its human partner (Dennis, Foote, and Perkins 2000, 201; Finsen 1852b, 187). Furthermore, the line is blurred between human and canine action, as the bite of a dog resulting in death is treated as a “víg söc” (Finsen 1852b, 187) [manslaughter charge], the same wording that is used for a man killing another man (Finsen 1852a, 147). The law does not specify whether, in this case, the dog or the human responsible for the dog is to be punished for such a killing—if the dog is considered responsible, then the canine is placed on an ontological level with human action; if the man is responsible, then the dog is considered as an extension of the human, and the actions of the one were the actions of the other. This mirroring can be seen in our Njáls saga episode, as Sámr scores the first wound in the fight against Gunnarr’s enemies, biting a man in the stomach (Brennu-Njáls saga 1954, 185), just as Gunnarr’s first act is to stab Þorgrímr “á hann miðjan” (Brennu-Njáls saga 1954, 187) [in his middle].

Legally then, dogs in Grágás are both figures to be cared for and dangerous creatures, akin to outlaws in their potential lack of legal security. If a dog is not cared for by a man and properly leashed, it is not part of legal society. In Njáls saga, we see that Sámr is very much part of Gunnarr’s home and the social networks of the farmstead, having been cared for by Gunnarr and having a place on his house; however, Sámr is unleashed in his final scene, suggesting that Sámr’s status echoes that of his human partner: outside of society and outside of the law. Both Gunnarr and Sámr are outlawed.

The outlawed figure in medieval Icelandic law is the figure who has forfeited their legal immunity through an action for which no legal response has been given, or that was considered so heinous as to be irredeemable. Such figures were set in clear contrast to the homeplace of the medieval Icelandic farm, being described as óheimilt (Króka-Refs saga 1959, 123) [un-homed, without domicile] and as skógarmenn (Cleasby and Vigfússon 1874, 556) [men of the wood]. With such descriptions and designations, outlawed men occupy a distinctive narrative space in the Íslendingasögur (Poilvez 2012). In Njáls saga, Gunnarr is an outlaw who refuses to leave his home, sharing this space with his unleashed dog.11 The inappropriate nature of the continued presence on the farm of both the outlawed man and unleashed dog is manifested in physical form by the dissection of the house. The home is dissected and destroyed by their enemies in stages, beginning with the killing of Sámr, developing through the rolling of the roof from the house, and culminating with the death of the householder, Gunnarr. The home-farm in medieval Iceland is a legal category and status marker, key for a man’s ability to participate in society. Through the destruction of the dog, house, and householder, the man and dog outside the law are removed from the legal society of the home, allowing social order (in theory) to be re-established in the text.

Fictive Kinship and the Death-Vow

The legal status of Gunnarr and Sámr is not the only link between them in the saga. Sámr’s moment of death and Gunnarr’s response reveal a more socially conceived bond between the two. Gunnarr’s statement on hearing Sámr’s death-cry is a moment emphasized by the saga-author. Gunnarr is explicitly within his house when he wakes, but still able to hear Sámr’s voice. He responds with what appears to be a sentimental and fatalistic comment on the loss of a beloved pet and a summation of what the attack will subsequently mean for him. Sámr’s cry in this case seems to act as a warning to his human partner. However, such a reading ignores the inter-textual associations implied by Gunnarr’s words:

Gunnarr vaknaði í skálanum ok mælti: “Sárt ertú leikinn Sámr fóstri, ok búð svá sé til ætlat, at skammt skyli okkar í meðal.” (Brennu-Njáls saga 1954, 186)

(Gunnarr woke in the hall and said: “Painfully are you played with Sámr foster-kin, and it may be intended [to come to pass] that a short time shall be between us-two.”)

The term fóstri used here is often translated in a way that implies a dependent relationship in which the human is superior to the animal, suggesting a paternalistic reading of the relationship between Sámr and Gunnarr, if not one based on dominance and ownership: for example, “pet” (Coles 1882), “fosterling” (Dasent 1861), “foster-son” (Gunnell 1997), and “foster-child” (Cook 2001). However, in the Old West Norse used in medieval Iceland, the masculine noun fóstri does not refer only to foster-sons, but also to foster-brothers and fosterfathers (Cleasby and Vigfússon 1874, 168). This term is used in only one other place in the Íslendingasögur to refer to an animal (Freyfaxi in Hrafnkels saga).12

Fosterage was a key component in the social networks of medieval Iceland, and the extended household, including blood relations, fosterkin, and hired workers were vital in providing defense of the home-place, as well as legal support and economic benefit (Christiansen 2002, 39). It was a system that required work, as just like blood-relationships, fictive kinship bonds were not automatic bonds of affiliation, and relied on mutual exchange for mutual benefit (Christiansen 2002, 47–8). Legally, however, a foster relationship could be guarded with the same responsibilities as those of blood, and legal rights of vengeance as listed in Grágás are equivalent for blood and foster-relations (Dennis, Foote, and Perkins 1980, 154ff.; Parkes 2004, 603). This suggests that it was not only considered appropriate to avenge injuries done to one’s foster-relations, but that people would have actively done so, necessitating the recording of regulations for such acts.

However, foster-relationships were not depicted as wholly positive in the sagas, and the interweaving of natal and foster kinship often causes issues in the society of the sagas (Parkes 2004, 604). It has been suggested that practices such as allegiance fosterage, designed to cement loyalties between families from different social groups, may have been viewed with ambivalence as an artificial way of bonding figures who would not naturally be connected (Bremmer 1976; Parkes 2004, 607). Such anxieties seem also extended to close relationships between animals and humans in the Íslendingasögur, as such relationships in these sagas are often viewed suspiciously or sought to be destroyed by those outside of the partnership.13 It is particularly worth noting that the compiler of Njáls saga seems specifically interested in the nature of different fosterage relationships, and (often) their failures, or disruptive influences on society (for example, Hallgerðr Höskuldsdóttir’s foster-father Þjóstólfr, and Njáll’s fosterage of Höskuldr Þrainsson). It may be that we can see the use of “fóstri minn” here highlighting the suitability or even usefulness of comparing certain animal-human relationships with human-human fictive kinship relationships.

Applying the fictive kinship model to animal-human relationships is a useful way of conceptualizing some animal-human relationships in the sagas, as the Íslendingasögur present a range of ways in which fóstri is used, and the most common of these refer to figures caring for children outside of their immediate biological kinship circle (Hansen 2008, 73, 76). This parenting model is not specifically echoed in the legal definitions of fosterage that we find in Grágás, but such usage may be legitimately extended to include animal members of a social group, such as dogs or horses, as individuals outside of the normal bounds of a family for whom the householder accepts legal responsibility and provides food and shelter (Dennis, Foote, and Perkins 2000, 46). These relationships would have required duties of care and protection, such as those Parkes suggests were necessary for the maintenance of human-human kinship bonds (Bremmer 1976; Parkes 2004, 607). A less anthropocentric reading of the Íslendingasögur opens up the possibilities for certain animal-human relationships to be included within these social structures. The meaning of “pet” for fóstri, offered by Cleasby and Vigfússon, is inadequate (1874, 168). Compared to the wide range of uses for this word in foster-kin relationships, it is shortsighted to accept this word as meaning one thing about humans, and another when referring to animals. A translation of fóstri as “pet,” or “fosterling” as used by some translators, assumes a childlike relationship, in which the human cares for the animal, when, as we know, fóstri can mean both child-parent, parent-child, and sibling fictive relationships dependent on context.

Fóstri is a polysemous word, containing many social meanings, tied in with dependence but also mutual exchange. The exact meaning of the word is often discerned by context; in the case of Sámr, this context is determined by the attitude of the translator toward animals and their possible roles in relationships with humans. On first reading, it can seem as though there is no other indicator to suggest the meaning of this term used in this passage. However, when Gunnarr’s use of the term fóstri and his subsequent statement on his impending death are considered together, there is a clear implication of the nature of the relationship between the two figures, which suggests that rather than seeing Sámr’s death-cry as a warning of impending attack, Gunnarr hears this as an indicator of his impending death in a different way: not as a result of the encroaching enemies, but as a result of Sámr’s death itself.

Gunnarr’s response to Sámr’s death, his comment on Sámr’s mistreatment, the reference to foster-kinship, and the acknowledgment of his impending death—these features suggest the relationship between Gunnarr and Sámr is here, at the point of death, depicted as akin to that expressed by “sworn-brothers” in Old Norse-Icelandic sources. These bonds of fictive brotherhood are formed through male friendships, sometimes articulated at decisive moments in the saga narratives (for example, Grettis saga 1936, 14, 85), and most of all based upon the ideal of mutual support and allegiance (Vohra 2018, 122). Such pairings are sometimes referred to in the narrative sources as a collective unit (Vohra 2018, 121–2), and this expression of a socially useful bond might be compared to the joining together of dog and man in Grágás (discussed above). The formation of fosterbrother bonds are sometimes accompanied in the sagas by the phrase “eitt skal yfir oss ganga” (Grettis saga 1936, 14, 85) [one fate shall go over us], which provides an inter-textual reference for Gunnarr’s suggestion that his own death will follow Sámr’s. This expression of attachment, duty, and commitment to Sámr on the part of Gunnarr, partnered with Óláfr’s correct assertion that Sámr would lay down his life to be true to Gunnarr, echoes the attachment and obligation expressed between human foster-brothers in the Íslendingasögur. It may even be suggested (albeit tentatively) that the fate of Gunnarr is in part influenced by a choice, subconscious or otherwise, to die in this encounter, acknowledging the obligation of the sworn-brother to share the fate of his partner.

Reading the Animal: Listening to the Animal

How we read animal-human relations in these texts relies on the mind-set we take to our reading. How far are we prepared to go, in our understanding of the relationship between Gunnarr and Sámr? Animals in the Íslendingasögur are complex figures. They don’t fit neatly into one animal category, or even into one idea of “dog-ness” or “horse-ness,” but occupy a space in which they exhibit personal traits, traits of personhood, of social maðr-ness. Sámr is clearly a dog, but a dog not quite in the way that we expect a dog to be. He speaks in a human way, but not in a human voice or with human words; and yet he is included within social bonds normally reserved for humans. He is a member of Gunnarr’s household, and part of his home-place, and dies, not to save Gunnarr or warn Gunnarr, but because an attack against one partner in this relationship is an attack against both. Sámr is not a symbol of Gunnarr, but an integral part of understanding Gunnar’s death-scene. This article proposes that the role of other animal-human partnerships in the Íslendingasögur needs to be closely examined in the same way as performed here for Sámr and Gunnarr, to fully understand the animal-human interactions in these Old Norse-Icelandic texts.

The capability for human-animal communication in Njáls saga has already been hinted at by Sámr’s understanding of Óláfr’s command to follow Gunnarr; and this is extended by the vocalization of Sámr’s cry using the verb “kveða,” which clearly places it in the sphere of human communication:

Hundrinn kvað við hátt, svá at þat þótti með ódœmum, ok féll hann dauðr niðr. (Brennu-Njáls saga 1954, 185–6)

(The dog cried out loudly, so that it seemed to them unprecedented, and he fell down dead.)

Old Norse contains verbs for different animal sounds, including geyja (to bark), used by Ólafr to describe Sámr’s barking at enemies in his initial gift to Gunnarr, yet the compiler of the saga here chooses to use the verb “kveða” that is most often used for human speech.14 While the official Christian view on the boundary between animals and humans was often predicated on the use (or not) of speech, this strict distinction was not consistently upheld, and some medieval writers were prepared to explore the potentialities of the animal voice and the subsequent implications for reasonable behavior (Langdon 2018, 2).15 Sámr’s cry is a communication understood by Gunnarr.

The depiction of animals undergoes two levels of translation to be received by the reader of an English translation of an Icelandic saga. Animal-human relationships are translated first from life, from everyday experiences and oral stories, into written narrative, and subsequently from Old Norse into modern English. Translations that fail to consider the capabilities of the animals depicted, or the full possibilities of these animal-human relationships contained in their descriptions, alter the way in which these relationships are read within this imaginative world. Reading animal-human relations in the sagas is a matter of looking past the translations, past the pre-conceived notions of what an animal—for example, a dog—should or should not be able to do, or to be, or to say, and also a matter of considering the depictions of figures and their relationship with others on their own terms.