A British Myth of Origins?

by John Carey
A British Myth of Origins?
John Carey
History of Religions
Start Page: 
End Page: 
Select license: 
Select License


Most readers of the Middle Welsh tale Math uab Mathonwy, also known as the Fourth Branch of the Mabinogi, would probably agree with Gwyn and Thomas Jones that its "confusion and complexity . . . are extreme."' The tale as a whole has, perhaps in consequence, received comparatively little scholarly attention: Celticists have pre- ferred to concentrate upon whatever individual episodes and characters seemed most amenable to explication.

A monumental exception to this tendency is W. J. Gruffydd's Math vab Mathonwy: An Inquiry into the Origins and Development of the Fourth Branch of the Mabinogi, published in 1928. Gruffydd's strategy was to "disentangle . . . the various traditions which, in different ways and to different degrees, have acted and re-acted upon one another to produce the mabinogi as we possess it."2 With tireless ingenuity and enormous learning, he traced a hypothetical prehistory of the tale involving the merging and mutual contamination of at least eight tale

A version of this article was presented at the ninth Harvard Celtic Colloquium, May 12, 1989. 1 am grateful for the comments of those who attended, and also for the reactions and suggestions of Professors John Koch, Proinsias Mac Cana, and Gregory

Nagy. Gwyn Jones and Thomas Jones, The Mabinogion (London: J. M. Dent, 1949),

p. xvii.

2 W. J. Gruffydd, Math vab Mathonwy: An Inquiry into the Origins and Development of the Fourth Branch of the Mabinogi (Cardiff: University of Wales Press, 1928), p. 350.

el991 by The University of Chsago. rights reserved. 0018-271Oi92i3101-WSOl 00

types, accomplished in the course of a seven-stage evolution beginning in pre-Christian Ireland and concluding in medieval Wales.

Subsequent studies, differing to varying degrees from Gruffydd's in their conclusions, have approached the text with other goals and methodologies. Saunders Lewis, in one of a series of articles on the Four Branches, argued that the military tactics described in Math point to Norman influence and a twelfth-century date.3 In one of his last essays, the late Georges Dumezil sought to show that the tale's principal characters constitute a pantheon with a trifunctional struc- t~re.~Most recently, Roberta Valente has proposed a synchronic reading of a different kind, based on the complex dynamics of the characters' sex roles.' Other interpretations of the story have been conditioned by theories of the overarching unity of the Branches as a whole: John Bollard advocates a literary appreciation of the Mabinogi as an "interlaced" examination of the ethos of relati~nshi~s;~

while Brinley Rees has put forward a comprehensive structural analysis based on the theories of ~umezil.'

In the remarks which follow I shall argue, against Gruffydd, that Muth as it survives is a largely faithful reflection of a single early narrative. I differ from Bollard and Valente in believing that the tale's primary significance is inherited, not synchronic, and see it as impor- tant evidence for the beliefs of the pre-Christian Britons.

The theoretical foundation of Gruffydd's approach was the view that narrative elements which are puzzling in juxtaposition must be dispa- rate in origh8 This attitude has been the mainstay of innumerable

3 Saunders Lewis, "Math fab Mathonwy," in Y Traerhodydd(Hydref,1969), pp. 185-

92. For a detailed critique of Lewis's position, see T. M. Charles-Edwards, "The Date of the Four Branches of the Mabinogi," Transacttons of the Honourable Societ. of Cymmrodorion (1970), pp. 263-98.

4 Ceorges Dumezil, "La quatrieme branche du Mabinogi et la theologie des trois fonctions," in Rencontre de Religions, ed. Proinsias Mac Cana and Michel Meslin (Paris: Belles Lettres, 1986). pp. 25-38.

5 Roberta Valente, "Gwydion and Aranrhod: Crossing the Borders of Gender in

Marh," Bullerin of the Board of Celtic Studies 35 (1988): 1 9.

6 John Bollard, "The Structure of the Four Branches of the Mabinogi," Transacfions of the Honourable Society of Cymmrodorian (1975), pp. 250-76. 7 Brinley Rees, Ceinciau'r Mabinogi (Bangor: Sili-wen, 1975). 8 For a summary of Gruffydd's working methods. see the "general assumptions"listed

on pp. 114-15. His attitude is shared even by those who find its consequences unappeal- ing; thus Bollard states that "the study of origins has the disadvantage of fragmenting the text into more or less well-defined parts rather than seeking to find its unifying elements and its inherent meaning"(Bo1lard. p. 252). As 1 hope to show, this is not necessarily the case.

students of traditional storytelling and has of course led to the solution of many difficult questions. It has a weakness, however, in that it excludes the possibility of a juxtaposition intended to be puzzling: the implicit assumption, in fact, is that no tale is paradoxical in its original state. Such an assumption is difficult to justify: it is surely a truism that paradox has been of fundamental importance to every culture of which we have detailed knowledge. Before dismissing an anomaly as evidence of corruption or confusion, we must ask whether it may not be signifi- cant as it stands; we should also look for possible parallels, in hopes of obtaining some further idea of its age and range.

Alwyn and Brinlqy Rees have discussed the significance of anomalies in Celtic conception tales, arguing persuasively that they are "symbols of the transcendental meaning of birth, of what birth is from the point of view of the unseen world. . . . The child may derive its biological inheritance from its earthly parents, but it is also the incarnation of a supernatural es~ence."~

This transcendence is expressed as mystery: a child may be "born without being begotten. In other cases the mystery lies, not in the absence of a cause, but in an embarrassing multiplication of contradictory causes to produce an effect which by the dictates of common sense can have only one cause."1° A particularly striking example of such multiple causation is provided by the Irish tale Compert Con Culainn;" below I present a list of its main incidents, set beside a similar outline of the first part of Math:

Cornpert Con Culainn Math
  1. The king Conchobar leaves his a) The king Math has a virgin foot-
    court to pursue a flock of otherworldly holder Goewin.
  2. His unmarried daughter Deichtine b) He leaves his court to fight for a is his charioteer. herd of swine from the Otherworld.

9 Alwyn Rees and Brinley Rees, Celtic Heritage (London: Thames & Hudson, 1961),

226. 10 Ibid., p. 237.
G. van Hamel, ed., Compert Con Culainn and Other Stories (Dublin: Dublin Institute for Advanced Studies, 1933), pp. 1-8. Referring to this story in particular, the Rees brothers observe that "it would be wrong to try to make sense of these traditions by tearing them to fragments on the assumption that they are a confusion of unrelated motifs. Their very ambiguity is of the essence of mythology" (Rees .and Rees, 235). Compare Alwyn D. Rees, "Modern Evaluations of Celtic Narrative Tradition," in Proceedings of the Second International Congress of Celtic Studies (Cardiff: University of Wales Press, 1961), pp. 31-61; the present study is much indebted to this lecture's methodological insights.
  1. The Ulaid spend the night in an c) Goewin is raped by Math's ne- Otherworld house, where Deichtine phew Gilfaethwy, abetted by Gil- assists their hostess in giving birth to faethwy's brother Gwydion. a son.
  2. Two foals are born at the same d) Gilfaethwy and Gwydion are time as the child, to whom they are turned into animals, and produce given. Deichtine fosters him. a fawn, a piglet, and a wolf-cub.
  3. The child dies. Thirsty after keen- e) Gwydion suggests that his sister ing him, Deichtine inadvertently drinks Aranrhod become Math's new foot- a "small creature" (mil mbecc). holder. Although she declares her-

self to be a virgin, she gives birth to a son.

  1. Lug appears to her in a dream, f) She also drops a "certain small tells her that the child she fostered thing" (ryw bethan), which Gwy- was his, that she is now pregnant by dion rears to be a boy. him, and that the foals are to be given to their son.
  2. Puzzled by her pregnancy, the g) The dialogue between Gwydion Ulaid conjecture that Conchobar is and Aranrhod implies that the boy responsible. is their incestuously begotten son.
  3. Betrothed to Sualtaim, Deichtine h) Aranrhod does all she can to induces a miscarriage on her wedding prevent the boy from obtaining an night. adult identity.

While there are obviously striking and significant differences here in both sequence and detail, I find the correspondences significant: in both tales we find a quest for Otherworld animals, a king with an attendant virgin, the incidental birth of animals identified with human youths,'2 the close association of a mysterious tiny creature with the pregnancy of a woman thought to be a virgin, indeterminate paternity giving rise to imputations of incest, and the mother's attempt to deny her son existence either before his birth (Deichtine) or afterward (Aranrhod). More generally, both stories deal with a supernatural conception which leads to the birth of a hero, despite many obstacles and-particularly important-several explicit ruptures in the causal

12 The nature of this identification is, of course, different in the two legends. The foals are, as it were, congenital doubles of the Otherworld child; while the offspring of Gwydion and Gilfaethwy are transformed into young men by Math.

sequence. It is probably no coincidence that the name of the father in one tale (Lug)is a cognate of the name of the son in the other � leu).'^

The simple fact of these resemblances constitutes a formidable, perhaps insuperable, obstacle for Gruffydd's analysis.14 If Math has indeed resulted from the fortuitous agglomeration of heterogeneous themes, then we are obliged to suppose that in the case of Compert Con Culainn either closely analogous materials came together inde- pendently to produce a similar story or an entirely different concatena- tion of chances led to a comparable outcome.I5 In the absence of other considerations, neither hypothesis appears plausible.

If we accept it as a likelihood that the opening section of Math reflects a tale of the "impossible" events attendant on the birth of a supernaturally begotten hero, a unifying theme becomes apparent in the Fourth Branch as a whole. For Lleu's death is as "impossible" as his birth: he can be killed neither inside nor outside, neither on horse nor on foot; and the narrative implies that these restrictions are rein- forced by others-he may be slain neither naked nor clothed, neither on land nor on water, and only with a spear made at a time when work is forbidden. Gruffydd has shown that this series of prohibitions can be parallelled elsewhere in European legend: the idea is always that of an apparent impossibility, resolved by mediating the terms of a dilemma.16 The Rees brothers have linked this paradoxicality with the concept of death itself, a change as anomalous as (and precisely symmetrical with) that of birth: their analysis of Celtic death tales strikingly complements the assessment of conception tales summarized above.''

13 This link is still closer if we understand Compert Con Culainn to state that Lug has begotten himselfon Deichtine. The wording is ambiguous: "He told her that she was pregnant by him . . . that the boy she had fostered was his, and that it was he [Lug or the child?] who had come into her womb, and that he was Lug mac Ethnenn" (van Hamel, ed., p. 5).

14 It is noteworthy that Gruffydd, who draws in his study on a wide range of material from all parts of Gaelic tradition, makes little use of Compert Con Culainn; for a rather vague allusion to another version of the tale, cited as if it were a straightforward incest account, see Gruffydd, p. 134, n. 36.

15 Many of the resemblances linking these tales are, of course, characteristic of super- natural conception tales generally (thus Rees and Rees, pp. 223 ff.); the series of possible correspondences to which I am calling attention may or may not be evidence of a more specific connection here.

16 Gruffydd, pp. 301-2. Stories of the "difficult death" tend to involve the protagonist's betrayal by a woman to whom he confides his secret: this element is of course present in Math as well, and it is reasonable to inquire whether the primary theme here is simply that of a hero rendered vulnerable by a woman. Without seeking to minimize this motif's significance, 1 would maintain that in Math it should be subordinated to the overall structure of the tale: fundamental to this structure, it seems to me, are the counterpoised mysteries of birth and death.

17 Rees and Rees, pp. 333 -35; of Diarmait mac Cerbaill they observe that "in a world of 'eitherlor' or 'nothing but' he could not have died at all." The transcendent abnormality of death is reflected in Lleu's statement that "Unless God kill me, my killing is

Bracketed by enigmas, Lleu's life is further determined by the "desti- nies" (tynghedau)imposed on him by his hostile'mother: she seeks to deny him a name, weapons, and a wife. The obtaining of each of the three is, of course, a crucial rite de passage in the journey to manhood:'' we may observe for instance that individual tales are devoted to each of these episodes in the career of C6 chulainn.19 For Lleu, each transition seems at first to have been rendered an impossibility: only the decep- tions and enchantments of Gwydion can bring him to maturity.

The Fourth Branch as a whole, therefore, may be read as a tale which presents life's natural stages as preternatural anomalies, negoti- ated only by magic and enigma. The broader significance of this will emerge more clearly when we have considered other aspects of the story.

It may be helpful at this point to make another extended comparison. Here I shall use material drawn from farther afield, however: the Greek myths concerning Prometheus.

Prometheus Math
  1. Prometheus deceitfully divides the a) Gwydion tricks Pryderi into ac- carcass at a sacrifice, so that Zeus cepting illusory horses, hounds, and selects bones and fat disguised to look ornaments. like cuts of meat.20
  2. Prometheus steals fire from Olym- b) Gwydion steals the swine brought from the ~therworld.~~
  3. Zeus and the other gods create the c) Math and Gwydion create the

beautiful but deceitful ~andora.~~ beautiful but unfaithful Blodeuedd.

not easy" (Ifor Williams, ed., Pedeir Keinc y Mabinogi [Cardiff: University of Wales Press, 19511, p. 86; hereafter PKM). 18 Compare Bollard, p. 268: "The swearing of the destinies on him is an attempt by Aranrhod to deny existence to the boy."

19 The gaining of a name and of weapons conclude the series of his boyhood exploits (macgnimrada):thus C. O'Rahilly, ed., Tbin 86.Ctiailnge: Recension I(Dub1in: Dublin Institute for Advanced Studies, 1976), lines 540-824. His winning of a wife is the subject of Tochmarc Emire (van Hamel, ed., pp. 16-68).

20 Hesiod Theogony 535-60.

21 Hesiod Works and Days 47-59, Theogony 561-69.

22 Gwydion's theft of the swine is important as the cause of the war; Gregory Nagy stresses the thematic connection between the "strife" occasioned by the actions of Prometheus and the Zpt~which brings about the Trojan War (The Best of the Achaeans [Baltimore: Johns Hopkins University Press, 19791, pp. 218-20).

23 Hesiod Works and Days 60-105, Theogony 570-89.
    1. Prometheus is chained to a pillar, d) Lleu is transformed into an his liver torn by an eagle.24 eagle after being wounded with a
    2. spear, and roosts in the top of a tree.
  1. Prometheus is released when Chi- e) After Lleu is released, Goronwy ron agrees to die in his stead.25 submits to the same death.

The choice of items here is obviously fairly selective, and any attempt to reconstruct a common prototype from this evidence would be rash in the extreme.26 The point of the comparison is that incidents similar to episodes drawn from all parts of Math can be found elsewhere, as components of a single myth and indeed in the same sequence. Again, we find evidence that a perplexing and complicated narrative structure does not necessarily require an atomistic analysis.

How is the Prometheus analogy to be harmonized with the Cu Chulainn analogy? I would suggest that in Math we have the reflex of a myth which combined two themes: that of the life cycle as a heroic progress past supernatural dilemmas, and that of the culture hero who wins for mankind benefits hoarded by the gods. In an account of humanity's beginnings, the culture hero's exploits and the conquest of birth and death could be elements in a single drama, functioning respectively on the social and the individual planes.

We may now consider one of Math's most puzzling passages: "And at that time Math son of Mathonwy would not be alive, except while his feet were in the lap of a virgin, unless it were the disturbance of war which prevented him."27

As Gruffydd and others have noted,28 the tale itself provides no elucidation of this mysterious statement; that there existed in medieval Wales a functionary known as the king's "foot-holder" (tr~edawc)~~

24 Hesiod Theogony 521-34. A further intriguing parallel here is that of Odin: in one

legend he obtains knowledge for mankind by hanging wounded from a tree (HbvambZ,in

Edda: Die Lieder des Codex Regius, ed. G. Neckel [Heidelberg: Carl Winter, 19831,

1:40), in another by assuming the form of an eagle (Bragaroedur, in Die prosaische

Edda, ed. E. Wilken [Paderborn: F. Schoningh, 18771, p. 100).

25 Apollodorus 2.5.4.

26 The most striking inconsistency is, of course, among the characters: some of the deeds of Prometheus may be likened to those of Gwydion, others to those of Lleu.

27 PKM, p. 67: "Ac yn yr oes honno Math uab Mathonwy ny bydei uyw, namyn tra

uei y deudroet ymlyc croth morwyn, onyt kynwryf ryuel a'y Ilesteirei."

28 Gruffydd (n. 2 above), p. 50.

29 Ibid., p. 54, n. 6.

does not explain why in Math the king's feet must be held by a virgin, or why his life should depend on this relationship. DumCzil suggested that the motif was in some way connected with the ideology of kingship, but had little to say concerning the details peculiar to the Welsh account;30 his suggestion may, however, point the way to a fuller explanation.

The Celtic king-and-goddess myth is amply attested in Irish, British, and Continental sources, and has been discussed at considerable length by several scholar^.^' In outline, the king's successful exercise of sovereignty was envisaged as his marriage to the goddess of the land: the goddess chose him as her mate, offered him a vessel of drink, and then slept with him. The kingdom's prosperity and fertility reflected this ritual union.

The opening of Math presents a starkly different picture. Although not portrayed as an unrightful king, Math is unmarried. He is closely associated with a woman, but his life depends on her virginity. Instead of being married to her, he rests his feet upon her.

It seems to me likeliest that this discrepancy with Celtic royal norms is a matter of deliberate contrast, not simple inconsistency. I suggest that Math's relationship with Goewin symbolizes a time in which mankind's connection with the earth was not yet one of laborious partnership: instead of cultivating the land (mating with the goddess) men simply stood on it (rested their feet on her). As long as this situation endured, there was no death.32

The correlation of agriculture, sexuality, and death is of course attested in other traditions. Perhaps the most familiar parallel is the biblical account of the Fall, where God's curse upon Adam and Eve is threefold: painful childbirth and sexual subordination for women

30 Georges Dumkzil, The Destiny of a King, trans. Alf Hiltebeitel (Chicago: University of Chicago Press, 1973), p. 126: "The reflections we have just made on the relations between virgins and the rex explain this strange condition, at least in principle ifnot in form" (my emphasis).

31 See, e.g., R. A. Breatnach, "The Lady and the King, a Theme of Irish Literature," Studies 42 (1953): 321-36; Proinsias Mac Cana, "Aspects of the Theme of King and Goddess in Irish ~iterature," ~tudesceltiques 7 (1955-56): 76-1 14,350-413; 8 (1958-59): 59-65; Rachel Bromwich, "Celtic Dynastic Themes and the Breton Lays," ~tudes celtiques 9 (1960-61): 439-74; Kim McCone, "Firinne agus torthulacht," LPachtai Cholm Cille 11 (1980): 136-73. Some Gaulish accounts are transmitted by Diodorus Siculus 5.24; Athenaeus 13.576; Justin Philippic Histories 43.3.

32 That the turning point of Math is the fall into mortality as a result of the "invention of sex" may explain various gender-related peculiarities of the narrative: the transforma- tion of two males into a series of mating couples; Aranrhod's apparent uncertainty as to whether or not she is a virgin; and the creation of a new kind of woman (Blodeuedd). Math's marriage to Goewin after her rape, by contrast, initiates the normal relationship of king and goddess.

(Gen. 3:16), the toil of growing food in a hostile environment for men (Gen. 3:17-18), and mortality for the human race as a whole (Gen. 3:19).~~

Very similar ideas are found in Hesiod's account of the Golden Age, immediately following one of his versions of the legend of ~rometheus:~~

If you wish, I will sum up for you another story . . . : how gods and mortal men have a single origin [bp60sv yeyhao~].For first the deathless dwellers on Olympus made a golden race of human beings, who lived while Kronos ruled in heaven. They lived like gods, possessing a carefree spirit, far removed from sufferings and sorrow. There was no wretched old age: with unchanging limbs they revelled in feasts, safe from all evils. They died as if overtaken by sleep. The grain-giving earth yielded produce of its own accord [ac~opa~q],

abundant and unfading [nollbv 7s ~ai

Eicp0ovov]. Glad and at ease, they inhabited their lands with many good things, rich in sheep, dear to the blessed gods.35

The inhabitants of this pre-agricultural paradise are not immortal but are at any rate unaging; I would call particular attention to the emphasis on their close relations with, and resemblance to, the gods. We are not informed that they did not marry;36 but it should be remembered that the reign of Kronos, which Hesiod synchronizes with the Golden Age, lasted only until his displacement by a heroic son (Zeus) whose birth Kronos took elaborate precautions to prevent.37 Gruffydd has pointed out many telling parallels between Math and the innumerable versions of the story of "The King and His Prophesied Death"; we should consider the possibility that these resemblances may be due to the Fourth Branch's derivation, not from a garbled folktale,

33 Similar ideas can of course be found in the figurative diction of many literatures. In

Ireland we may note the opening of Tochmarc Becfola, where a lone woman proposition-

ing a king says that she is journeying "to seek seed-wheat . . .1 have good arable land

[dagirhir] but lack seedwhich is suitable for it" (Maire Bhreathnach, "A New Edition of

Tochmarc Becfiola," Eriu 35 [1984]: 59-91, quote on 72, 77). Compare in Brittany the

remarkable agricultural epithalamion in De sancro Iudicaelo rege historia, where the

royal consort is spoken of as terra bona bene arara (R. Fawtier, "Ingomar, historien

breton," in Melanges d'histoire du moyen cige offerrs b M. Ferdinand Lor [Paris:

Edouard Champion, 19251, p. 191); this text and its significance were called to my

attention by John Koch.

34 The end of the Golden Age and the quarrel between Zeus and Prometheus are

directly correlated by Nagy (n. 22 above), pp. 214-21.

35 Hesiod Works and Days 106-20.

36 This assertion is, however, found in Plato Politicus 272A, where the statement that

men did not have wives or children in the age of Kronos immediately precedes a passage clearly based on Hesiod: "They had unfading [&(pCJovouq] fruits from trees of every kind, not produced by agriculture, but yielded by the earth of its own accord [afiropa~qq]." Earlier we are told that "being begotten by one another" [TO pbv kt bhhjhov. . . ysvvwp~vov]was unknown to Kronos's subjects (ibid., 271A).

37 Hesiod Theogony 453-506.

but from the myth of a god who like Kronos "learned from Earth and starry Heaven that it was his fate to be overthrown by his own

Having proposed that the opening of Math reflects a myth describing man's fall from a condition in which he was not yet, in Eliade's phrase, "a mortal, sexed, and cultural being,"39 I would like to turn to a consideration of the punishment of Gwydion and Gilfaethwy, which mediates between this opening and the remainder of the tale. Math enchants his nephews so that they become, in successive years, mating pairs of deer, wild swine, and wolves; they exchange sex with each transformation, and produce three offspring whom he changes into handsome youths.

I have suggested elsewhere that a sequence of metamorphoses may function in Celtic narrative as "a device to connect the present with its origins, whether the beginnings of history or the transtemporal eternity of the ~therworld";~~

this is an appropriate occasion for considering some of the relevant evidence in more detail.

The earliest Irish document in which such a sequence is attested is probably the obscure text in which Mongan describes to Colum Cille a splendid kingdom which once flourished in territory now covered by the waters of Lough Foyle; he then describes his successive existences as a deer, a salmon, a seal, a wolf, and a man.41 This notion of a flood followed by a series of metamorphoses is again implicit in Immram Brain, where Manannan, in a paradisal "present" apparently equivalent to the antediluvian past,42 predicts that Mongan will pass through the shapes of a dragon (or hero), a wolf, a stag, a salmon, a seal, a swan, and a king.43

38 Ibid., 463-64. A similar destiny hung over Zeus himself. Kronos swallowed his children by Rhea; Zeus swallowed Metis lest she bear him a son (ibid., 886-900; Apollodorus 1.3.6). The two myths are linked still more closely by the tradition that it was Metis who assisted Zeus in causing Kronos to vomit up his other offspring (Apollodorus 1.2.1). According to Aeschylus, one reason for Prometheus's torture was his refusal to elucidate the prophecy of Zeus's downfall (Prometheus Bound 907-27); we may perhaps compare Math's punishment of Gwydion for his involvement in the rape of Goewin.

39 Mircea Eliade, Myth and Reality, trans. Willard R. Trask (New York: Harper & Row, 1963), p. 6. 40 John Carey, "Time, Space, and the Otherworld," Proceedings of rhe Harvard Celtic Colloquium 7 (1987): 10. 41 Kuno Meyer, ed., "The Colloquy of Colum Cille and the Youth at Carn Eolairg," Zeitschrifi fur celtische Philologie 2 (1 899): 3 13-1 7.

4* I discuss this scene in "Time, Space, and the Otherworld."

43 Seamus Mac Mathuna, ed. and trans., Immram Brain (Tiibingen: Niemeyer, 1985), pp. 42, 55.

In the Mongan tales, the inundation which precedes the metamor- phoses is specific to a single locality; similar legends are found through- out the Celtic countries.44 A tale involving the biblical deluge is ScPl Tuain, whose long-lived protagonist is part of the first settlement of Ireland after the Flood, and survives until the conversion period by regenerating himself at intervals as a stag, a boar, a bird of prey, and a salmon: these transformations roughly correspond to the subsequent settlements of Ireland by the followers of Nemed, the Fir Bolg, the Tuath Dk, and the sons of ~i1.~~

ScPl Tuain provides, it seems to me, a convenient tertium quid for comparing Math and the Hesiodic material: the shape-shifting sequence is found in both the Welsh and Irish tales, but Tuan also lives through the settlement of Ireland by four peoples, comparable to the races (silver, bronze, "demigods," and the "iron" folk of the present) who follow the "golden race" in the Works and Days.

A similar episode in Greek mythology is of particular interest because it serves the same narrative function as some of the other Irish ex- amples. In both Tochmarc ~taine and De Chophur in Da Muccida, the serial metamorphoses of supernatural beings lead to the births of avatars who trigger pivotal "historic" events: in the former case ~tain, whose beauty is ultimately responsible for the death of Conaire Mar in Togail Bruidne Da Derga; in the latter case the bulls which provide a pretext for the massive conflict of the ~hin.~~

Athenaeus quotes from the lost epic Cypria a tradition that Helen, cause of the Trojan War, had a similar origin: "Helen, a wonder to mortal men, whom Nemesis of the lovely hair once bore to Zeus, who lay with her by force. For she fled, not wishing to be joined in love to her father Zeus, the son of Kronos. . . . She fled through land and through dark barren water, and Zeus pursued, for his heart longed to possess her. . . . At one time in the shape of a fish she raced through the ample sea; at another (she went) upon the clodded earth. She changed perpetually into whatever fearsome creatures the land sustains, in order to escape him."47

44 1 have-examined some of this material in "Origin and Development of the Cesair Legend," Eigse 22 (1987): 37-48. Prometheus is associated with a deluge also, in that his son Deucalion is the ancestor of postdiluvian r?ankind (e.g., Apollodorus 1.7.2).

45 John Carey, "Sctl Tuain Meic Chairill," Eriu 35 (1984): 101-2.

46"Tochmarc Etaine," ed. and trans. Osborn Bergin and R. I. Best, fiiu 12 (1938): 137-96; De Chophur in 06 Muccida, ed. and trans. Ulrike Roider, Innsbriicker Beirrage zur Sprachwissenschaft, vol. 28 (Innsbruck: lnstitut fiir Sprachwissenschaft der Uni- versitat Innsbruck, 1979).

47 Deipnosophistae 8.334B; that the chase leading to Helen's conception is directly connected with her destiny as cause of the war is suggested by the fact that her mother here is Nemesis, goddess of avenging fate. It is suggestive that Metis also flees Zeus's advances by changing "into many shapes" (~iq nohhaq iGQa5, Apollodorus, 1.3.6); for her connection with transition between epochs, see n. 38 above.

The theme of incest is also of interest here. A myth in which a metamorphosis sequence is explicitly associated with the creation of gender occurs in another ancient source, the ByhadCranyaka Upanisad. The "he" of the account is the First Being, brooding in primordial solitude:

He was not happy thus. . . . He desired a companion [dvitiyarn].He became as great as a woman and a man linked in an embrace [stripurndmsau sampari~- vaktau]. He caused his own self to split in two: thus they became husband and wife. . . . He united with her: hence men were born.

Then she considered: "How is it that he now unites with me, having generated me from himself [md dtmana eva janayitvd]? Come, let me be concealed [tiro Sdni]." She became a cow, and he a bull: he united with her, hence cattle were born. One became a mare, the other a stallion; one became a she-ass, the other a he-ass: he united with her, hence single-hoofed animals were born. One became a she-goat, the other a he-goat; one became a ewe, the other a ram: he united with her, hence goats and sheep were born. Thus he generated [asyjata]every couple [mithunam] whatever, down to ants.48

Here the incest motif has, I think, its original significance: the first female is both wife and daughter of the original entity. It is also noteworthy that in this case the series of transformations involves procreation at every stage, a feature which the Indic account shares with Math, but with none of the other comparanda cited.

The considerations set forth above suggest that in Math we have the remnant, much changed but in many respects remarkably intact, of a myth describing the end of a paradisal state of being, and the emer- gence of the conditions of morality. The first rape sets in motion the cycle of generation, leading to the birth of a hero whose life blazes the path followed by subsequent humanity. The hero is nurtured and protected by the wizard-trickster who presided over the passing of the primordial age.

48 Brhaddranyaka Upanisad 1.4.3-4; 1 have used the edition of mile Senart (Paris: Belles Lettres, 1967). The relevance of this myth and of that in the Cypria has been pointed out by Joseph Campbell, The Hero with a Thousand Faces (New York: Pantheon, 1949), pp. 278-79; cf. Emily B. Lyle, "The Twa Magicians as Conception Story,"Scottish Studies 23 (1979): 79-82. 1 am grateful to Laurie Patton for pointing out to me that the Brdhmanas contain a similar tale: PrajBpati's daughter becomes a doe, and he assumes the form of a stag in order to mate with her; thus Aitareya Brdhmana 13.9, trans. A. B. Keith in Rigveda Brdhmanas: The Aitareya and Kau~itaki Brdhmanas of the Rig Veda (reprint, Delhi, 198l), pp. 185-86; other references in A. A. Macdonell, Vedic Mythology (reprint, Delhi, 1981), p. 1 19.

This analysis of course leaves many questions unanswered: further study of the tale is clearly needed,49 and many of its mysteries will doubtless continue to frustrate our curiosity. I am for instance uncer- tain of the significance of the comparanda which I have adduced from Greece and India: are they cognate survivals or merely typological parallels? I should emphasize that my interpretation of the Fourth Branch does not depend on any particular answer to this question: the primary purpose of the comparisons is simply to demonstrate that story patterns like those in Math can be found elsewhere as features of coherent narratives. The correspondences are so numerous, however, and the underlying symbologies so similar, that I think it likely that Math reflects Indo-European doctrines of cosmogony.

If Math does reflect a myth of origins, how is it that it exists at all? It is essentially unique in the Celtic literatures. A native account of the beginning of things would inevitably have challenged the tenets of ~hristianit~:"how would such an account, even in significantly altered form, have survived the conversion and found its way into writing? The only suggestion which I can offer involves a solution obsciri per obscurius: I believe that the clue is to be found in the apocryphal Taliesin poetry.

This material presents many difficulties of translation and interpre- tation. Until very recently, no reliable editions of the poems existed; we now have several excellent studies by Marged Haycock, but much of her research is yet to be published.51 A few general observations may, however, be made with relative confidence. The poet has assumed the persona of the historical sixth-century bard Taliesin; but his composi- tions are concerned with arcane lore rather than with royal eulogy. His knowledge encompasses the mysteries of the physical universe, the future, and the legendary past; he speaks frequently of having been

49 One promising line of inquiry has been outlined by Georges DumCzil, "Remarques sur le dieu scandinave Heimdallr," ~tudesceltiques 8 (1959): 283, n. 1: he associates the Mahdbhdrata's account of Bhisma's birth from the Ganges with aquatic elements in the birth tales of Lleu and (in some folktales) Lug. This suggestion should be reexamined in light of the role of floods in the legends of Tuan and Mongan, as well as the aquatic rebirth of Taliesin (discussed below); but such an investigation cannot be undertaken here.

50 Compare the apposite remarks of Proinsias Mac Cana, Celtic Mythology (New York, 1983), pp. 136-37. 51 Marged Haycock has published editions of poems in the Book of Taliesin in her articles "'Preiddeu Annwn' and the Figure of Taliesin," Studia Celtica 181 19 (1983-84): 52-78, "Dylan ail Ton," Ysgrifau Beirniadol 13 (1985): 26-38, and "'Some Talk of Alexander and Some of Hercules': Three Early Medieval Poems from the Book of Taliesin," Cambridge Medieval Celtic Studies 13 (1987): 7-38; cf. further her studies "Llyfr Taliesin," National Library of Wales Journal 25 (1988): 357-86, and "Metrical Models for the Poems in the Book of Taliesin," in Early Welsh Poetry: Studies in the Book of Aneirin, ed. Brynley F. Roberts (Aberystwyth: National Library of Wales, 1988), pp. 155-77. 1 have not seen her unpublished dissertation "Llyfr Taliesin: Astu- diaethau ar rai agweddau" (Ph.D. thesis, University of Wales, 1982). present at celebrated events in former ages and of being able to remember previous existences in other shapes. The attitude to Chris- tianity in the poems is variable: there are many pious formulae and some Latin phrases, but also scornful references to the ignorance of the clergy. This anticlerical poetic tradition, preoccupied with the mastery of native esoterica, seems a favorable milieu for the transmission of any mythical material which might have survived into the early medieval period; it may therefore be significant that the pseudo-Taliesinic corpus contains several references to the characters of the Fourth Branch. Thus one poem speaks of the "magic wand of Mathonwy" (hutlath vathon~~),~~ and a another alludes to the skill of Lleu and ~w~dion,'~ third laments the death of Lleu's brother ~~lan;'~ the poem Kadeir Kerrituen mentions Gwydion's creation of Blodeuedd, and the theft of swine from Dyfed." The most important allusions are found in Kat godeu, where the poet describes his metamorphoses, speaks of having been present with Gwydion at "the battle of the trees,"56 and.claims to have been created by Math and Gwydion "before the world" (kyn byt) in a manner similar to that used to create Blodeuedd in ~ath.'~ Elsewhere in the poem he speaks as if he were Lleu himself, saying that he was attacked by Goronwy, and perhaps alluding to the twin birth of himself and ~~lan.'~ Pseudo-Taliesin's concern with metamorphosis and creation is echoed in the tale Hanes Taliesin, surviving only in Modern Welsh versions but based on a narrative to which the poems contain clear allusion^.^^ A shape-shifting pursuit which leads to a supernatural 52 J. Gwenogvryn Evans, ed., The Book of Taliesin (Llanbedrog: privately printed, 1910), p. 28, line 26; hereafter BT. 53 BT 1 (Red Book of Hergest, col. 1054, line 16). 54 BT67.9-17; cf. Haycock, "Dylan ail Ton." 55 BT 36. 56 Compare in the poem "Golychafi gulwyd arglwyd pop echen" the lines "Bum yg kat godeu gan lleu a gwydyon / wy a rithwys gwyd eluyd ac elestron" (ibid., 33.23-25). In Angar kyvyndawt the phrase aches gwyd gwydyon "eloquence of the trees of Gwydion (?)" (ibid., 22.3) may refer to the same battle. Further references in Welsh poetry are listed by Rachel Bromwich, Trioedd Ynys Prydein (Cardiff: University of Wales Press, 1978), pp. 277,400-402,420-21,448-49. 57 This may be seen as another parallel with Greek tradition: not only do the gods create Pandora as a snare for men, but Prometheus himself is credited with having formed the first human from mud (e.g., Apollodorus 1.7.1; Pausanias 10.4.3). 58 In the lines Neu bum yn yscor / gun dylan eil mor, "I was in an enclosure with Dylan Eil Mor" (BT26.9-lo), I tentatively take yscor, "enclosure, pen, rampart, strong- hold," to refer to the womb of Aranrhod. For the text of Kat godeu, see BT 23.9-27.12; Patrick Ford has furnished a translation in The Mabinogi (Berkeley: University of California Press, 1977), pp. 183-87. 59 Thus BT 23.1-3: "A hen conceived me, with red claws and . . . comb: for nine nights I rested as a lad in her womb." It may be relevant that the Hanes, like Math, is set primarily in Gwynedd. conception is an essential part of this story, as has frequently been noted;60 a less explicit but highly significant feature of the legend is the association between the metamorphoses and a deluge, recalling the adventures of Monghn and Tuhn. Several traces of this inundation survive in the story. a) Ceridwen, the witch who prepares the caldron which is the source of Taliesin's knowledge, is the wife of Tegid Voel, whose territory is identified as having been "in Y Bala in Penllyn, and the lake is called Llyn Tegid to the present day."61 In other words, the tale opens in a region subsequently covered by a lake. b) When the caldron has produced the three drops in which all knowledge is distilled, it explodes and pours forth a flood of poison; this poison kills the horses of Gwyddno Garanhir, a king whose realm was engulfed by the sea.62 We may reasonably see here traces of an earlier legend, which directly attributed the loss of Gwyddno's realm to the bursting of the caldron. c) Ceridwen pursues Gwion Bach, the boy who received the inspira- tion: after they have passed through several shapes, she swallows him and gives birth to him nine months later. The reborn Gwion is Taliesin. She puts him in a sack and throws him into the sea. He is fished up by Gwyddno's son Elphin: that he should be taken from the sea by the survivor of a flood again strongly suggests that his previous existence was an antediluvian one. The Taliesin legend, then, has strong thematic links with the complex of ideas which I have posited as the background for the Fourth Branch: its association with this background goes well beyond the tantalizing allusions in the poetry. Under the circumstances, it does not appear unreasonable to suggest that the poems likewise represent a tradition deeply concerned with the lore of origins, and the myth of the fall into mortality; it may have been this tradition on which Math's author drew. Harvard University 60 Thus Campbell (n. 48 above), pp. 197-98; and Lyle (n. 48 above), p. 80: both associate the episode with the passages cited above from the Cypria and the ByhadGranyaka Upani~ad. 61 Patrick K. Ford, ed., "A Fragment of the Hanes Taliesin by Llywelyn Sidn," fiudes celriques 14 (1975): 454, lines 1-2, p. 457. Other legends explicitly attribute Lake Bala to a deluge; thus John Rhys, Celtic Folklore Welsh and Manx (Oxford: Oxford University Press, 1901), pp. 376-78,408-10. 62 Ford, ed., "A Fragment," 454.26-30; for the legend of Gwyddno, see Rachel Bromwich, "Cantre'r Gwaelod and Ker Is," in 7'he Early Culture of North- West Europe, ed. Cyril Fox and Bruce Dickins (Cambridge: Cambridge University Press, 1950), pp. 215-41.

  • Recommend Us