Old Norse Mythology—Comparative Perspectives

  Hermann, Pernille, Stephen A. Mitchell, and Jens Peter Schjødt, eds., with Amber J. Rose. 2017. Old Norse Mythology—Comparative Perspectives. Milman Parry Collection of Oral Literature 3. Cambridge, MA: Milman Parry Collection of Oral Literature. http://nrs.harvard.edu/urn-3:hul.ebook:CHS_HermannP_etal_eds.Old_Norse_Mythology.2017.

Baldr and Iraj: Murdered and Avenged

Emily Lyle, University of Edinburgh

Abstract: Comparing the Old Norse myth about Baldr with the Persian Iraj story, this essay deals with methodological considerations about comparativism and structural models as heuristic tools for reconstructing ancient traditions. The essay points to common aspects of the narratives focusing on familial relationships among the gods, which are analyzed by using a two-eras model (involving old gods and young gods) focusing on murder, revenge and avengers.


This article treats the death of Baldr as related by Snorri and Saxo and the death of Iraj as told in the Shahnameh (The Book of Kings) by Ferdowsi of ca. 1000 CE. The Baldr story is mainly recounted in Snorri’s Gylfaginning 49 (pp. 48–49) and Saxo’s Gesta Danorum Book 3 (1.69–79) and has been the subject of an illuminating study by John Lindow, Murder and Vengeance among the Gods: Baldr in Scandinavian Mythology (1997). The Iraj story occurs near the beginning of the Persian epic history, in the reign of the sixth king, Feraydun (Davis 2007: 33–62). The common narrative features which form a basis for comparison are:

  • A murder of a half-brother by two other brothers in conjunction who make complementary contributions to the act.
  • The inhibition of immediate revenge by the presence of a kinship tie.
  • A revenge killing by an avenger who was not yet conceived at the time of the murder.

The two stories also share the mythic context of a family of gods.

The hypothesis underlying a study of this kind is that it is possible to make fruitful comparisons between Old Norse and Iranian mythic narratives when an appropriate methodology is brought to bear. It was apparently the absence of an effective methodology that led to the shunning for a time of broad comparative study at the Indo-European level by many scholars in Old Norse studies as well as in other areas where it might reasonably have been undertaken. I have attempted to open a way to reenter the field in my book Ten Gods: A New Approach to Defining the Mythological Structures of the Indo-Europeans (Lyle 2012a). [1] I argue here that greater coherence can be expected of a posited early form of myth than of the splintered end results of millennia of transmission in changing environments, and that the coherence is related to the connection between myth and the organization of society. On the other hand, I also demonstrate that the end results are capable of carrying information derived from a remote past and that comparison can help to bring it to light.

It seems obvious that this approach could be of value in comparative mythology as well as in comparative linguistics. Within the Indo-European field, the question can be posed: “How can we know the nature of the mythologies that come to us through written records without hypothesizing the nature of the common ground from which they were derived?” Does anyone seriously think that the gods entered the equation within the historical period? It is only if this position is being defended that it makes sense to ignore the wider context that would potentially enable us to supplement the information from the relatively recent times from which direct records are available.

When Martin West surveyed myth at the Indo-European level, he deliberately excluded the structural component exemplified by the functional theory of Georges Dumézil that can be seen as one possible modeling device (West 2007: 4). The Dumézilian concept of a functional triad of the sacred, physical force, and fertility (Littleton 1982) has to be applied with discrimination since this triad is likely to be only a part of a larger structure that incorporates other triads besides the one stressed by Dumézil (Lyle 2006, 2012b). Modeling need not be confined to the well-known functional sequence, however, and this method used in a more general way is likely to prove a valuable conceptual tool as we extend the bounds of our enquiry.

The heuristic value of modeling

A model can be a useful aid to perception, as recently argued by Jens Peter Schjødt (2013a: 37–39, 2013b: 9–10). It can also serve as an intermediary, when both A and B have perceptible similarities to the model C, while a claim of a direct similarity between A and B would be unconvincing.

A model is a means for sorting out information from background noise. We necessarily come to any body of material with built-in expectations of what is relevant. Researchers may simply assume that what is relevant is already known and, when the assumption is shared by scholars in a field, it may take on the force of authority and shut down lines of exploration. A model is overt and does away with the possibility of undeclared presuppositions, in so far as these relate to the model. A model is also what has been called an “external memory device” (Donald 1991: 274, 296–97; Lyle 2012a: 10). It has an objective existence and can be explored both by the originator and by other enquirers.

A model, however, is not reality. A model may be especially valuable when it approaches reality but its usefulness is not confined to this. That is to say that a scholar is free to operate with a model without any commitment to its truth value. A model is a heuristic device.

A model and a metaphor have much in common. In introducing the value of using models in the mythological context, Jens Peter Schjødt has employed the metaphor of a jigsaw that we are trying to put together when we find ourselves in possession of a jumble of pieces. He writes:

Now, in order to do anything with these pieces we have to have an idea of what the picture could have been like. Part of this idea will come from the pieces themselves, but it is a necessary prerequisite that we have an idea of the structure of the picture. The blue will probably belong to the sky, the green to the grass and the trees, there may be pieces with parts of persons that must belong together with other parts of persons, and so on. Because of that idea we do not have to assemble every piece with every other piece: the possibilities can be delimited to those that seem to be part of the same motif. (Schjødt 2013a: 37)

A valuable model to apply in the Indo-European context is the division by eras between the old and the young gods which can be seen as corresponding to a culturally defined division at the human level between the living and the dead. The two-era scheme has a simple outline that can be filled out in a similar way to the “blue sky, above” and “green grass, below” scheme, but with the components “old gods/the dead, above” and “young gods/the living, below”.

The Indo-European two-Eras model and the Old Norse and Iranian murder-and-Revenge protagonists

The human capacity for using models is not just of value for modern scholarship, but would have been employed in the past to build up the imaginative structures containing the gods. The basis of a two-era system of gods would have been the human experience of living in one generation (era 2) but having a memory of a limited number of earlier generations (era 1). This is the temporal field that Jan Assmann has referred to as that of communicative memory and that I have called the memory capsule (Assmann 1992: 56; Lyle 2012c; see also Erll 2011: 29). The evidence seems good that a conceptual structure with four generations was emphasized among the Indo-Europeans in prehistory. This is the Common Celtic grouping, which corresponds to the Irish derbfine, as well as the Greek angkhisteia and the Indian sapiṇḍa (Charles-Edwards 1993: 55, 187, 213–14, 471–72; Foxhall 1995: 134; Dumont 1980; Lyle 2012a: 40, 2013). In the Indian case, it is explicit that only the last generation contains the living. The sacrificer in this generation makes offerings to his ancestors in the three previous generations. Michael Witzel in his recent study of myth has independently seen the relevance of the four-generation block at both the human and the divine levels (Witzel 2012: 64, 425–27, 576n24).

In the Old Norse context, there was certainly awareness of schemes with a limited number of ancestors, although the number of past generations counted in this way varied and could be as many as five (Laidoner 2015: 52–53, 104; Brink 2012: 103–6). To capture something of the flavor of the proposed Proto-Indo-European four-generation series in this context, we can turn to the case of Auðr hin djúpúðga (Auðr the Deep-Minded) at the time of Iceland’s settlement (Laidoner 2015: 155–60; Jesch 1991: 80–83, 194–95). Auðr’s great-grandson Þórðr gellir, when he was about to enter into manhood, went to the place where she was commemorated, so establishing a connection with the first-settler foremother to whom he was linked through his father and grandfather (Landnámabók 97; Laidoner 2015: 159). This specific case can be shown as in Table 1.

Eras Generations
Prior generations Auðr hin djúpúðga
  Þorsteinn rauða
  Óláfr feilan
Youngest generation Þórðr gellir

Table 1. Auðr hin djúpúðga and representatives of three following generations.

The scheme in Table 1 can be generalized to one in which the predecessors are all defined as deceased and are referred to as ancestors (Table 2).

Eras Generations
Ancestors foremother
The living male ego

Table 2. An apical ancestress and three following generations, divided into eras of the living and the dead.

It is a human model like this that I argue gave rise to a two-era scheme of the gods, as shown in Table 3.

Eras Generations
Old gods Generation of primal being
  Generation 2
  Generation 3
Young gods Generation 4

Table 3. The posited eras and generations among the Indo-European gods in prehistory.

Turning now to the Iranian Shahnameh, where the gods are reflected in epic, we find that it is structured in terms of succession, and that the two eras among the gods can be posited to relate to the first six reigns.

Eras Regnal Series Years of Reign Kings
Old gods 1 30 Kayumars
  2 40 Hushang
  3 30 Tahmures
Young gods 4 700 Jamshid
  5 1000 Zahhak (usurper)
  6 500 Feraydun

Table 4. The first six kings in the regnal series of the Shahnameh.

The first three kings (equated with the old gods) are quite briefly treated and have relatively short reigns assigned to them while the next three are treated more fully and have impossibly long reigns in human terms (Robinson 2001: 11–18, 153). As young gods, in terms of the model, these three belong to a single generation and, in the narrative, despite the great extent of time, this is also true of the kings, for Zahhak and Feraydun marry the two sisters of Jamshid. Salm, Tur, and Iraj, the sons of Feraydun, are best understood in terms of the model as belonging to this generation also, with the authority of Feraydun over them being that of king, not father. Jamshid (whose name means King Jam) corresponds linguistically to the Indian Yama.

Eras Generations
Old gods Kayumars
Young gods Jamshid, Zahhak, Feraydun, Salm, Tur, Iraj

Table 5. The succession sequence related to four generations and two divine eras.

As regards the eras in the Old Norse context, the major young gods are of the same generation and so conform to the model. Members of a following generation are also known, and of course their position should be considered, but initially it seems that we can take the major gods as a guide; doing so requires no modification in the number of generations in the case of the young gods. On the other hand, there is a missing generation in the era of the old gods since the three old gods born from the primal being are presented as of the same generation (Gylfaginning 5–7). They are shown in Table 6 along with the three young gods who feature in the murder story.

Eras Generations
Old gods Bestla
  Óðinn, Vé, Vili
Young gods Baldr, Loki, Hǫðr

Table 6. The Old Norse murder triad shown in relation to the two eras.

Clearly succession is dominant in the Iranian case and totally absent in the Old Norse one. Since succession can be traced elsewhere in the Indo-European area (Lyle 2012a), we can understand the absence in Old Norse as being related to the exceptionally dominant place of Óðinn who just is king in the stories we have, without having any need to replace a predecessor or to fear the coming of a successor. The whole mythic pattern associated with royalty is liable to have been affected by this, and we can expect to find truncated, skewed, or lost narratives. One apparently truncated story occurs in Ynglinga saga in the episode where Vé and Vílir/Vili sleep with Óðinn’s wife Frigg (Ynglinga saga 3). Lindow comments:

According to Ynglinga saga, ch. 3, when Odin is away, Frigg is possessed by his two brothers Vili and Vé and is then returned to Odin like some kind of street-walker upon his return […] Basing his insult on this episode, which he goes on to mention, Loki tells her in Lokasenna 26 that she has ever been vergiorn (eager for a man). (Perhaps, however, she was simply filling the role of consort of the chief god[s]). (Lindow 1997: 50)

This multiple union of Frigg and three gods does not result in offspring whereas comparable mythological stories of a female lying with three males concern the begetting of the young king (Bek-Pedersen 2006: 332; Lyle 2012a: 59–68, 77). [
4] Although this completion of the narrative is not included in the corpus, the most likely young-king figure, Þórr, shorn of royal attributes, is given a compensatory cosmic birth as son of a primal being, the Earth.

Feud as a binding motif in the Scandinavian context

The interesting point that emerges from the present study is that there is a third strand besides birth and succession through which the sequence of generation-periods can be expressed—that of feud. The generation-periods are tied together through the negative reciprocity of murder and revenge. When succession is preceded by killing, feud can be seen as the shadow side of the regnal sequence. In Saxo, both Balderus and Høtherus have claims to the kingship, and Høtherus becomes undisputed king after he kills Balderus.

Generally, human feuding is carried out between family groups. A significant difference when we look at feud among the gods is that the population is extremely limited; therefore, revenge has to be taken within the confines of a single family. However, the implication of the feud motif is that there are two branches within the family of the gods that are opposed to each other. After the murder of Baldr/Balderus, his father Óðinn/Othinus takes steps to bring about the killing of the murderer (Hǫðr/Høtherus) by a killer named Váli or Bous. This aligns the gods or god-descendants in two opposed camps over three generations or quasi-generations. I say quasi-generations since Óðinn/Othinus is the biological father of Váli or Bous as well as of Baldr/Balderus and Hǫðr/ Høtherus but, as the latter two are adult before the birth of the avenger, they can be regarded as being of a different social generation (see Tables 7 and 8).

The short solutions to the riddles are: 1. Vilinn, and 2. Óðinn. The explanations of the solutions tell the story. Vilinn, slain by the enemy, was compensated for through the suffering of a woman [Vrindr]. A son was born to the Kinsman [Óðinn] who, at the age of ninety, engendered Þórr in compensation for Vilinn.

  Allied Opposed
Father Kinsman aged 90  
Victim and murderer Vilinn enemy
Avenger Þórr  

Table 7. The oppositions in the Rök stone runic inscription.

  Allied Opposed
Father Óðinn/Othinus  
Victim and murderers Baldr/Balderus Loki or Gevarus, Hoðr/Høtherus
Avenger Váli or Bous  

Table 8. The oppositions in Snorri and Saxo.

Feud is also a binding motif in the Iranian context, but the Shahnameh narratives will be introduced before the oppositions are presented.

The deaths of Iraj and Jamshid

As will be seen, discussion of the death of Iraj should probably not be considered in isolation from that of the death of Jamshid, but the murder-and-revenge story of Iraj is treated first.

I have divided the summary into two parts to bring out the distinction between the separate motifs of distribution and murder, which will be discussed below.

The distribution

Feraydun tests his sons by turning himself into a dragon and observing their reactions. The oldest, Salm, flees; the second, Tur, is impetuously brave; and the youngest, Iraj, chooses “the middle way between earth and fire” and is brave without rashness. After this Feraydun divides the world into three parts and makes Salm King of the Western Lands and Tur King of Turan and China. Finally, he makes Iraj Lord of Persia and gives him his crown (Shahnameh pp. 33-36).

The murder

Salm is dissatisfied and sends a message to Tur saying that the throne of Persia should have come to Salm as the oldest, or to Tur, if Salm is set aside. He says that they have been wronged by their father’s decision to favor the youngest. When Tur receives the message, he leaps up in rage and says that they should not hesitate to act. Iraj wishes for reconciliation and arrives with a few companions for a meeting with his brothers. That night Salm and Tur decide to murder him. The next day, when they are alone together, Iraj tells his brothers that he does not value the kingship and is prepared to abdicate. Tur is angered, snatches up the golden throne he is sitting on, and strikes him on the head. Iraj asks him “How can you long to spill your brother’s blood and torment our father’s heart with such a crime?” and begs him to desist, but Tur draws his dagger and kills him. He cuts off his head and sends it to Feraydun. Feraydun curses the two unjust brothers and, addressing himself to God as lord of justice, says, “I ask only that I be given a little time, my lord, until I see a child from Iraj’s seed, who will bind on his belt for vengeance.” Feraydun is happy to find that Mah-Afarid in Iraj’s harem is pregnant, “hoping that her child would be a means to vengeance for his son’s death.” Mah-Afarid, however, gives birth to a daughter called Canopus and, when she grows up, Feraydun marries her to his nephew Pashang. Canopus gives birth to a fine boy who is called Manuchehr. The two unjust brothers pretend to seek Feraydun’s favor and try to lure Manuchehr into visiting them but Feraydun is not deceived and says they will only see Manuchehr at the head of an army seeking vengeance “for his grandfather’s death.” Feraydun adds: “Previously, we did not seek vengeance for [Iraj] for we did not think the times were propitious; it would not have been suitable for me to fight against my two sons.” Manuchehr kills Tur in battle, cuts off his head, and sends it to Feraydun. In a later battle, Salm flees for the safety of a castle but Manuchehr pursues and kills him, and sends his head to Feraydun. Manuchehr is enthroned and crowned, but Feraydun spends the rest of his life grieving for his sons (Shahnameh pp. 36-62).

So far as I am aware, the Baldr and Iraj stories have not previously been explored as mythic parallels. The murder-and-revenge theme in relation to Salm, Tur, and Iraj seems to have been masked by the theme of the distribution of territories, which was studied extensively by Georges Dumézil (1973: 9–20, esp. 13–14, 1971: 251–64, esp. 257–58, 1968: 586–88). It can be argued that the distribution is attached primarily to the three characters named while the murder-and-revenge story is a secondary development that has been displaced onto them from the figures who appear in the king list: Jamshid as victim and Zahhak and Feraydun as perpetrators. The revenge motif is found only in the Iraj story but the resolution is the same in both cases. Manuchehr is the successor of Fereydun in the king list and is the avenger of Iraj. The oppositions as expressed in the Iraj story are presented in Table 9.

  Allied Opposed
Father Feraydun  
Victim and murderers Iraj Salm, Tur
Avenger/Successor Manuchehr  

Table 9. The Iraj story murder-and-revenge protagonists.

We have two ways to explore the story. We can see it simply as told of Salm, Tur, and Iraj, as actually present in the epic and as shown in Table 9, or we can see it, more speculatively but probably more fundamentally, as displaced from the Jamshid sequence. There is a narrative weakness in the Iraj story which points to the probability of adaptation. The brothers kill Iraj because they covet his kingship and yet Salm and Tur continue to reign in their own realms in the West and East, with neither taking over the vacant throne of Iraj. The motivation of grasping royal power, on the other hand, is strongly present in the main king series when Zahhak usurps the kingdom and then has the former king, Jamshid, put to death, and when Feraydun takes over the kingdom and has Zahhak bound. If we take it that the murder was primarily that of Jamshid we can sketch out the process of revision in the following way.

In the initial, hypothesized, story Jamshid becomes king but his rule is contested by Feraydun who, with the help of the magician Zahhak, kills him and succeeds to the throne, as shown in Table 10. It is supposed that it was found inappropriate to present the heroic Feraydun as the murderer of his predecessor and that the story was converted to one where Zahhak alone was responsible for the death of Jamshid and became a usurper who was later removed from the throne by Feraydun. This would have made the theme of long-delayed revenge redundant here, but the story material was too good to throw away and it was conflated with the story of distribution which was already present as an independent entity.

  Allied Opposed
Father Tahmures  
Victim and murderers Jamshid Zahhak, Feraydun
Avenger/Successor Manuchehr  

Table 10. The reconstructed Shahnameh murder-and-revenge protagonists.

There is no trace remaining of Feraydun’s guilt in the Shahnameh but his hypothesized murder of Jamshid does appear to relate to the murder motif involving royal pairs in other Indo-European contexts, notably the murder of Remus by Romulus. The Indo-European royal pair may or may not be treated as twins, and they are certainly to be distinguished from the divine twins who appear in the Roman context as Castor and Pollux (Ward 1968: 6–7). There is an ambiguity about who has the right to be king, which comes out in various ways in different Indo-European stories (Lyle 1990: 105–18), and it is worth dwelling on this since the parallel element in Saxo has not hitherto been pointed out in this connection. In the Roman case, the twins and their supporters separate and watch for the appearance of augural birds in order to determine which of them has the right to be king (Livy 1, 6, 4–7, 3). Remus sees birds first, but there are only six of them and, when Romulus sees birds later, there are twelve. Clearly, a case could be made for either of the contestants on the grounds of priority or greater number. Similarly, in Saxo (3, 3,1–2, 1.73–74) Høtherus and Balderus both have claims to the kingship of Denmark and are “hungry for sovereignty”. When Høtherus arrived, “The Danish people came out to meet him and appointed him their king” but later, when Høtherus was absent, Balderus “gained the speedy acquiescence of the Danes to all his demands for royal honour”. Saxo stresses the ambivalence, commenting: “Such was the wavering determination of our forbears”. Although both the claimants arguably had good cases, the matter was only settled by the brute act of killing one of the contenders, both here and in the Roman example.

The fettering of Loki and Zahhak

The death of Baldr/Balderus was the result of a throw or thrust by Hǫðr/Høtherus, but was there a single murderer or a pair? John Lindow, although deciding to leave Loki’s role intact in his discussion of the texts, refers to this as a long-standing question in the scholarship (1997: 68):

Since both Baldrs draumar and V ǫ lusp á mention only Hǫðr as the slayer of Baldr and leave Loki’s connection to the deed unclear, there has naturally arisen a huge debate as to the “original” presence or absence of Loki in the myth.

We can usefully refer to the last revision of Dumézil’s book on Loki, where he argues for the two-murderer scheme (1986: 102–22), and also to a recent comment on the subject by Joseph Harris, who finds the one-murderer scheme the more likely at an early stage in the story’s life (2010: 98). It is clearly still a live issue. If the Iranian parallel is accepted, it supports the view that the two-murderer scheme is primary.

After Baldr’s death, Hǫðr is killed but Loki is fettered (Gylfaginning 50–51). These various fates could be regarded as an execution and very lengthy prison sentence. Hǫðr’s “crime” is simply the killing of Baldr but Loki’s fettering is ascribed to more than a single cause (Lindow 1997: 163), and it could be seen as the containment of a more generalized evil, of which Baldr’s murder was but one manifestation.

Treating Loki a good many years ago, Anna-Birgitta Rooth refers to “a myth of the fettered monster, a cosmological myth of the Enemy of the World who is to be released at the End of the World”. She adds: “The models for the Old Norse myth are found both in biblical and Christian Mediaeval tradition as well as in the Classical myth of Prometheus with which the Scandinavian myth is closely related” (Rooth 1961: 185). Elsewhere, she notes similarly that the “fettered giant or devil thus is found both in Classical and Christian tradition,” but at this point she adds: “It is however an interesting fact that also in Mohammedan tradition we encounter the motif of the bound Ḍaḥḥāk and his release at the end of the world” (Rooth 1961: 84–85). This is Zahhak of the Shahnameh.

Rooth’s reference is to page 327 of A. J. Carnoy’s contribution, Iranian Mythology, in the Mythologies of the World series (1917). Here Carnoy refers to the end of the world “when Ḍaḥḥāk (Azhi Dahāka), fettered by Farīdūn on Mount Damāvand, will be released by the powers of evil, who will rally for the last struggle against good”. A few pages earlier, Carnoy had described the exploit of Feraydun’s capture of the dragon-king Zahhak and continues: “He conveyed the captive to Mount Damāvand, where he fettered him in a narrow gorge and studded him with heavy nails, leaving him to hang, bound by his hands, to a crag, so that his anguish might endure” (Carnoy 1917: 323; cf. Davis 2007: 27). Carnoy then sums up the reign of Feraydun, which he has dealt with in terms of the Shahnameh, and includes references to the Pahlavi Dinkard and Bundahis passages from the ninth century CE (Dinkard VIII, xiii, st. 9; Bundahis xxxi, st. 9–11).

Tradition knows little of Farīdūn outside of his healing power and his victory over the dragon. Nevertheless the Dīnkart mentions the division of his kingdom between his sons Salm, Tūr, and Īraj; and the Būndahish explains that the two former killed the latter, as well as his posterity, with the exception of a daughter who was concealed by Farīdūn and who bore the hero Manushcithra, or Minūcihr, the successor of Farīdūn. The legends concerning these princes thus date back to a fairly ancient period, although it is doubtful whether they had the amplitude and the character which they assume in Firdausī’s epic. These stories are not mythical, but merely epic. (Carnoy 1917: 323)

The last statement may give us pause. If there is one thing that has emerged as a consensus on myth in the century since Carnoy was writing, it is that epics may contain myths and that stories about human heroes may reflect the actions of the gods. Jaan Puhvel, appositely, applies this insight to the Shahnameh in a chapter on “Epic Iran” (Puhvel 1987: 117–25, esp. 118–20).

As regards the position of Zahhak, approaching the matter through the Iraj murder and its postulated displacement, I arrived at the view that Zahhak was the counsellor who joined with Fereydun in the murder of Jamshid, and so found him to correspond to Loki in relation to the murder without any reference to their parallel fates. If this view is accepted, the tie between Loki and Zahhak, already made by Rooth in 1961, will be very much strengthened: Loki/Zahhak aids the killer of Baldr/Jamshid and is fettered to a rock.

The specially begotten Avenger as first king

In both the Baldr/Balderus and Iraj cases, the avenger is not yet conceived when the killing takes place and, in both, the father of the murdered man takes a hand in the process of revenge. In the first case, the victim’s father discovers that the avenger has to be born of a specific woman/giantess and begets him with her. In the second case, he finds the woman who is pregnant by the murdered man and cares for her in the hope that she will bear a son who will carry out the revenge. When she bears a daughter, he cares for the daughter until she conceives and bears the avenger.

The reign of Manuchehr does not fall in the part of the Shahnameh that has been equated here with the eras of the gods but immediately follows it. It seems that the story of revenge carries over from the eras of the gods into the era of humans and covers the transition to the human institution of royal succession with Manuchehr in the position of first human king. Bous (in Saxo) dies of his wounds the day after he avenges the murder, while the Norse Váli simply does not become king; there is thus no succession in this revenge story in the Old Norse context.

In the Old Norse materials as well as in the Indo-European ones more generally, the gods do not go on proliferating indefinitely, and the interest turns to the human kings who are their descendants. It seems that the avenger in the generation after that of the gods fundamentally belongs in the human category, and that the murder-and-revenge motif negotiates this crucial transition.

Fjǫlnir, the son of Freyr and Gerðr, was said to be the first king at Uppsala and members of the Svea line of kings there could be called the descendants of Freyr (Sundqvist 2000: 39, 152–55). This raises the question of how a mortal could inherit from a god. If we take the Iranian parallel as guide, it seems that the first king could be the avenger of the death of a god—but we can ask how that revenge could be accomplished. The killing of the murderer is straightforward if it is placed in the historical context as it is in Saxo; a man is killed. On the other hand, Hǫðr is a god in Snorri but is put to death; a god is killed. What does it mean to kill a god, who is an immortal? Clearly the death of Baldr (and of Yama) is cosmologically significant and is a tremendous event. This seems to be the exceptional case when a god really dies. The killing of the murderer is not highly charged in this way. Possibly we should invoke the idea of compensation found in the Rök carving. I have found no statement to this effect, but perhaps the story ran that the last god in a genealogical reckoning yielded the throne to the first human in compensation for the murder.


When stories are attached to the gods, it can be expected that they would have been in some sort of structured relationship. This is not to say that stories, once the bonds of the original sequences were loosened, did not take off on their own and become elaborate, largely independent units that generate their own interest. Compared narratives from different branches of Indo-European tradition can sometimes display a revealing lack of connection that highlights the extent of diachronic change.

What can we learn from the present comparison that might inform the study of Old Norse mythology? It reinforces the view that two males were responsible for the death of the Baldr figure, one as counsellor and one as actor. It reinforces the idea that the avenger has to be specially conceived and introduces the idea that this special conception is followed by the birth of a son who would be king as well as avenger. That point can lead to a reinterpretation within the Old Norse material that would allow the convergence of the birth of an avenger (from a giantess) and the birth of a founder king (from a giantess).

In the stories considered, comparison becomes especially rewarding when we allow for two possible major shifts: the rise to dominance of one of the old gods (Oðinn) in the Old Norse context, and the purification from guilt of the young king-god (Feraydun). The Old Norse and Iranian strands of narrative offer supplements to each other and it seems feasible, by taking them together, to advance our understanding of a key sequence of Indo-European myth.

In the Old Norse context, it is the feud motif, with murder and revenge, that ties the generations together. The theme of feud has allowed the survival of a sequence (formerly one of succession), otherwise lost in the Old Norse context through the increase in the power of Óðinn relative to the other gods.

The two-era model enables us to focus on potentially relevant connections. As indicated by Schjødt in the quotation above, an advantage of the use of modeling is that it directs attention to a specific, limited area where comparison is likely to be valuable. I argue that the default position should be that connections are present between Old Norse stories and cult on the one hand and Indo-European myth on the other. It remains for these connections to be widely and deeply explored in the critical climate that has been established through the detailed study of sources in recent decades of scholarship.

Works cited

Primary sources

Baldrs draumar: see Poetic Edda


Bundahis. In Pahlavi Texts 1: The Bundahis, Bahman Yast, and Shâyast Lâ-Shâyast. Transl. E. W. West. Sacred Books of the East, 5. Oxford: 1880. Pp. 3–151.


Dinkard. In Pahlavi Texts 4: Contents of the Nasks. Transl. E. W. West. Sacred Books of the East, 37. Oxford: 1892. Pp. 3–398, 406–17.

Eyvindr skáldapillir Finnsson: see Háleygjatal

Gesta Danorum

Saxo Grammaticus. Gesta Danorum: Danmarkshistorien. Ed. Karsten Friis-Jensen and transl. [into Danish] Peter Zeeberg. Copenhagen: 2005.


Saxo Grammaticus. The History of the Danes: Books 1–9. Ed. Hilda Ellis Davidson and transl. Peter Fisher. Cambridge: 1996.

Gylfaginning: see Prose Edda


Eyvindr skáldapillir Finnsson. “Háleygjatal.” Ed. and transl. Russell Poole. In vol. 1 of Skaldic Poetry of the Scandinavian Middle Ages: Poetry of the Kings’ Sagas 1: From Mythical Times to c. 1035. Ed. Diana Whaley. Turnhout: 2012. Pp. 195–213.


Íslenzk fornrit, 26–28. Ed. Bjarni Aðalbjarnarson. Reykjavík: 1941.


Snorri Sturluson. Heimskringla. Transl. Alison Finlay and Anthony Faulkes. London: 2011.


Íslenzk fornrit, 1. Ed. Jakob Benediktsson. Reykjavík: 1986. Pp. 29-397.


The Book of Settlements: Landnámabók. Transl. Hermann Pálsson and Paul Edwards. Winnipeg: 1972.


Livy. History of Rome: Books 1–2. Transl. B. O. Foster. Loeb Classical Library, 114. London & Cambridge, MA: 1919.

Lokasenna: see Poetic Edda

Poetic Edda

Eddukvæði I–II. Ed. Jónas Kristjánsson and Vésteinn Ólason. Reykjavik: 2014.


The Poetic Edda. Transl. Carolyne Larrington. Oxford World’s Classics. Oxford: 2008.

Prose Edda

Snorri Sturluson. Edda Snorra Sturlusonar: Udgivet efter håndskrifterne. Ed. Finnur Jónsson. Copenhagen: 1931.


Snorri Sturluson. Edda. Transl. Anthony Faulkes. London: 1987.


Ferdowsi, Abolqasem. Shahnameh. 12 vols. Ed. Jalal Khāleghi Motlag. Bibliotheca Persica. New York: 1988–2009.


Ferdowsi, Abolqasem. Shahnameh: The Persian Book of Kings. Transl. Dick Davis. London: 2007.

Skaldskaparmal: see Prose Edda

Skírnismál: see Poetic Edda

Ynglinga Saga: see Heimskringla

Secondary Sources

Abram, Christopher. 2011. Myths of the Pagan North: The Gods of the Norsemen. London & New York.

Assmann, Jan. 1992. Das kulturelle Gedächtnis: Schrift, Erinnerung und politische Identität in frühen Hochkulturen. Munich.

Axel Kristinsson. 2012. “Indo-European Expansion Cycles.” Journal of Indo-European Studies 40: 365–433.

Bek-Pedersen, Karen. 2006. “Interpretations of Ynglingasaga and the Mabinogi: Some Norse-Celtic Correspondences.” In Old Norse Religion in Long-Term Perspectives: Origins, Changes and Interactions. Ed. Anders Andrén, Kristina Jennbert, and Catharina Raudvere. Vägar till Midgård, 8. Lund. Pp. 331–35.

Brink, Stefan. 2012. “Law and Legal Customs in Viking Age Scandinavia.” In The Scandinavians from the Vendel Period to the Tenth Century: An Ethnographic Perspective. Ed. Judith Jesch. Woodbridge. Pp. 87–117.

Carnoy, Albert J. 1917. Iranian Mythology. The Mythology of All Races, 6. Boston.

Charles-Edwards, T. M. 1993. Early Irish and Welsh Kinship. Oxford.

Clunies Ross, Margaret. 1994. Prolonged Echoes: Old Norse Myths in Medieval Northern Society. 2 vols. Odense.

Davis 2007. See Shahnameh.

Donald, Merlin. 1991. Origins of the Modern Mind: Three Stages in the Evolution of Culture and Cognition. Cambridge, MA.

Dumézil, Georges. 1968. Mythe et Épopée 1. Paris.

———. 1971. Mythe et Épopée 2. Paris.

———. 1973. The Destiny of a King. Transl. Alf Hiltebeitel. Chicago & London.

———. 1986. Loki. Rev. ed. Paris.

Dumont, Louis. 1980. “La dette vis-à-vis des créanciers et la catégorie de sapiṇḍa.” Puruṣārtha: sciences sociales en Asie du sud 4: 15–37.

Erll, Astrid. 2011. Memory in Culture. Transl. Sarah B. Young. Basingstoke.

Foxhall, Lin. 1995. “Monumental Ambitions: The Significance of Posterity in Greece.” In Time, Tradition and Society in Greek Archaeology. Ed. Nigel Spencer. London. Pp. 132–49.

Harris, Joseph. 2010. “Varin’s Philosophy and the Rök Stone’s Mythology of Death.” In New Perspectives on Myth: Proceedings of the Second Annual Conference of the International Association for Comparative Mythology, Ravenstein the Netherlands, 19–21 August, 2008. Ed. Wim M. J. van Binsbergen and Eric Venbrux. Haarlem. Pp. 91–105.

Jesch, Judith. 1991. Women in the Viking Age. Woodbridge.

Laidoner, Triin. 2015. Ancestors, their Worship and the Elite in Viking Age and Early Medieval Scandinavia. PhD diss., University of Aberdeen.

Lindow, John. 1997. Murder and Vengeance among the Gods: Baldr in Scandinavian Mythology. Folklore Fellows Communications, 262. Helsinki.

Littleton, C. Scott. 1982. The New Comparative Mythology: An Anthropological Assessment of the Work of Georges Dumézil. 3rd ed. Berkeley.

Lyle, Emily. 1990. Archaic Cosmos: Polarity, Space and Time. Edinburgh.

———. 2006. “The Importance of the Prehistory of Indo-European Structures for Indo-European Studies.” Journal of Indo-European Studies 34: 99–110.

———. 2012a. Ten Gods: A New Approach to Defining the Mythological Structures of the Indo-Europeans. Newcastle-upon-Tyne.

———. 2012b. “Entering the Chimeraland of Indo-European Reconstruction.” Retrospective Methods Network Newsletter December: 6–10.

———. 2012c. “Stepping Stones through Time.” Oral Tradition 27 (1): 161–70.

———. 2013. “Defining the Religion that Lay behind the Self-Colonization of Europe.” In Critical Reflections on Indigenous Religions. Ed. James L. Cox. Farnham & Burlington, VT. Pp. 93–101.

Mallory, J. P., and D. Q. Adams. 2006. The Oxford Introduction to Proto-Indo-European and the Proto-Indo-European World. New York.

Poole, Russell. 2007. “Myth and Ritual in the Háleygjatal of Eyvindr skáldaspillir.” In Learning and Understanding in the Old Norse World: Essays in Honour of Margaret Clunies Ross. Ed. Judy Quinn, Kate Heslop and Tarrin Wills. Turnhout. Pp. 153–76.

Prior, Daniel. 2012. “Integral or Incidental? Indo-European Mythic Frag­ments in Inner Asia.” Paper presented at The Steppes: Crucible of Eurasia, Miami University Art Museum, Oxford, Ohio, Nov. 30–Dec. 1. http://muamgrassroutes.wordpress.com/symposium2012/ Accessed on 04/04/2016.

Puhvel, Jaan. 1987. Comparative Mythology. Baltimore.

Robinson, B. W. 2001. The Persian Book of Kings: An Epitome of the Shahnama of Firdawsi. Richmond.

Rooth, Anna Birgitta. 1961. Loki in Scandinavian Mythology. Skrifter utgivna av Kungl. Humanistiska Vetenskapssamfundet i Lund, 61. Lund.

Schjødt, Jens Peter. 2013a. “Reflections on Some Problems in Dealing with Indigenous Religions of the Past: The Case of Pre-Christian Scandinavian Religion.” In Critical Reflections on Indigenous Religions. Ed. James L Cox. Farnham & Burlington, VT. Pp. 29–47.

———. 2013b. “The Notions of Model, Discourse, and Semantic Center as Tools for the (Re)Construction of Old Norse Religion.” Retrospective Methods Network Newsletter May: 6–15.

Steinsland Gro. 2012. “Rulers as Offspring of Gods and Giantesses: On the Mythology of Pagan Norse Rulership.” In The Viking World. Ed. Stefan Brink in collaboration with Neil Price. London. Pp. 227–30.

Sundqvist, Olof. 2000. Freyr’s Offspring: Rulers and Religion in Ancient Svea Society. Uppsala.

Ward, Donald J. 1968. The Divine Twins: An Indo-European Myth in Germanic Tradition. Berkeley and Los Angeles.

West, M. L. 2007. Indo-European Poetry and Myth. Oxford.

Witzel, E. J. Michael. 2012. The Origins of the World’s Mythologies. New York.

Yarshater, Ehsan. 1986–. Encyclopædia Iranica. New York. http://www.iranicaonline.org. Accessed on 05/30/2016.


[ back ] 1. In relation to the ten gods of the title, I may add that Christopher Abram names as major divine figures the ten Old Norse gods that I propose for this set, plus Hǫðr and Heimdallr (Abram 2011: 61). I see Hǫðr as a hypostasis of Þórr and not as a separate figure (Lyle 2012a: 82). For a suggested comparative context for Heimdallr (whom I do not treat), see Prior 2012.

[ back ] 2. For a recently proposed location of the Indo-European homeland in the area of present-day Romania and Ukraine, see Axel Kristinsson 2012.

[ back ] 3. This is by way of contrast with the earlier male pseudo-procreation (Clunies Ross 1994: 1.158, 185–86).

[ back ] 4. The most striking case adduced in these studies is the begetting of the cosmologically significant Irish royal figure, Lugaid of the Red Stripes, by the three fathers whom he resembles in the three parts of his body divided by the red lines round his neck and waist (Lyle 2012a: 62–64).

[ back ] 5. I follow Harris (2010) but with some modifications, giving ‘Riddle’ and ‘Solutions’ instead of ‘Question’ and ‘Answers’, and replacing ‘the descendants of Ing-Vald’ by ‘the Inguldings’, ‘for’ (in Riddle 2) by ‘in compensation for’ and ‘Thor’ by ‘Þórr’.

[ back ] 6. Dumézil comments: “Il y a, dans l’Edda en prose, le conseil de Loki et l’acte de Hoðr, il y a conjointement, chez Saxo, a) le conseil de Gevarus qui donne le moyen de tuer Balderus, b) l’acte formel d’Høtherus tuant Balderus” (Dumézil 1986: 108–9) (There are in the Prose Edda, the advice of Loki and the act of Hǫðr, there are taken together, in Saxo, a) the advice of Gevarus which provides the means of killing Balderus, b) the formal act of killing Balderus.)

[ back ] 7. I use the translation of the Shahnameh by Dick Davis (2007). Full discussion of the main protagonists may be found under Aždahā, Ferēdūn, Iraj, and Jamšid in Yarshater (3: 181–205, 9: 531–33, 13: 200–2, 14: 501–28).

[ back ] 8. I have discussed elsewhere the possible importance of the mother of the avenger understood as the primal goddess (Lyle 2012a: 85–86).

[ back ] 9. For discussion of the context, see Poole 2007: 154–55, 161–66 and Abram 2011: 127–34.