100

Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

The Dojoji Tale

and
Ancient Bronze Metallurgical Traditions

REINIER H. HESSELINK1
The University of Northern Iowa, USA

Abstract: A comparison of the motifs in the different versions of

the famous Japanese folktale about the Dōjōji Temple leads the
author to other tales which share some of the same motifs. In this
way, he isolates two motif clusters, one on the theme of the mixed
marriage between humans and snakes, and another one centered on
the traditions of bell casting. In the Dōjōji Tale these motif clusters
seem to have been joined by Buddhist authors eager to employ local
pagan tales to write new stories with a Buddhist message. Next, the
author translates two Chinese tales in which the same motif clusters
appear. Although these tales show unmistakably the influence of
popular Taoism, the similarities with the Dōjōji Tale are too striking
to be coincidental. In his conclusion, the author proposes that the
tales with these motifs were brought from China to Japan before the
introduction of Buddhism (552).

A Folktale from Japan

Two Buddhist priests, one old and one young, go on a pilgrimage to

Kumano. On their way, in Muro-no-kōri, they spend the night at the house
of a widow. She takes a fancy to the young priest, and visits him at night.
The priest refuses her, but to get rid of the woman he promises to come
back to her after his pilgrimage is over.

One their way back, however, the two priests avoid the house of the widow.
When the woman finds this out, she becomes enraged and suddenly dies.
Shortly after her death, a giant snake is seen leaving her bedroom. The
priests, still on their way home, hear about the animal and understand the
danger they are in.

They take refuge in the nearby Dōjōji temple, which closes its gates. The
young priest is hidden under the temple bell, which the priests of the temple
have let down. The snake, however, crosses right over the closed gate of the
temple compound. It glides around the hall once or twice and then finds the
door to the bell tower, which it crushes with one hundred slaps of its tail.

1 The author is a Professor of Japanese History and Civilization at the University of Northern Iowa.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 101

It then wraps itself around the bell and pounds the cannon with its tail.
Tears of blood drip from the snake’s eyes, and its tongue licks its lips. After
two or three hours, the animal leaves in the same direction from where it
has come. The priests notice the bell is burning hot from the snake’s
poisonous breath. No one can go near it. The priests cool the bell down with
and when they are finally able to move the bell, they see that all that is left
of the young priest is a few ashes.

Later, the senior priest has a dream in which the young priest appears as a
giant snake, who tells him how the woman has caught up with him and
that only the “Longevity” chapter of the lotus sutra can set them both free.
The priest copies the chapter and offers it to the two snakes during a temple
festival. After that the young priest and the woman again visit the elder
priest in his dream to thank him for their rebirth in heaven.

It is generally assumed that this famous tale told about the Dōjōji temple of
Wakayama prefecture in slightly differing versions in the Hokke Genki (1040?),2
Konjaku Monogatari (1120?-1240?),3 the Genkō Shakusho (1322),4 the Dōjōji Engi
Emaki (1450?),5 the Noh play Dōjōji (1510?),6 and other sources,7 has as its subject
the terrible wrath of a woman who is disappointed in love.8 As such it is believed to
be a Buddhist tale and its interpretation hinges on the danger of rejecting a
woman’s love. The sexist bias implicit here (a woman so rejected will turn into a
dragon!) can be reduced to the Buddhist quandary over whether women can be
saved at all. This is a dilemma typical of the late Heian and medieval periods in
Japanese history. Here I will assume this interpretation to be a relatively late
version of the tale, and one which tells us more about the interpreters than about
the story itself. I want to present an alternative interpretation, one that deals with a
different level underneath the Buddhist garb, which is misleading us into thinking
that Buddhist priests are sexist. As we shall see, in reality they are saving women.
Before proceeding. I must admit to making a few other assumptions as to the nature
of the material we are dealing with and the methods of analysis permitted. I assume
that the tale to be researched in this chapter is a folktale that is considerably older
than its oldest known version, that of the Hokke genki (1040). Further, I ascribe to

2 Yoshiko Kurata Dykstra, Miraculous Tales of the Lotus Sutra from Ancient Japan. Hirakata:
Kansai University of Foreign Studies, 1983, p. 23.
3 Robert H. Brower, The Konzyaku Monogatarisyū: An Historical and Critical Introduction with

Annotated Translations of Seventy-Eight Tales. Unpublished PhD thesis University of Michigan,

1952, “Chapter VI: The Problems of Authorship and Date”, pp. 274-342.
4 Maria B. Ury, Genkō Shakusho. Japan’s First Comprehensive History of Buddhism: A Partial

Translation with Introduction and Notes. PhD dissertation. Berkeley: University of California, 1970,
pp. 336-9.
5 Virginia Skord Waters, “Sex, Lies, and the Illustrated Scroll: The Dōjōji Engi Emaki” in MN, vol. 52,

1 (1997): 60.
6 Tokue Motomasa, Muromachi geinōshi ronkō. Tokyo: Sanbii Shoten, 1984, p. 402.
7 Matsumoto Takanobu in his “Zōtei Muromachi jidai monogatarirui genzonhon kanmei mokuroku”

in Otogi zōshi no sekai. Tokyo: Sanseido, 1982, mentions the Kengaku zōshi and the Hidakagawa
sōshi (inaccessible to me).
8 For the standard interpretation, see: Nihon koten bungaku daijiten. Tokyo: Iwanami shoten, 1984,

p. 414.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

102 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

the idea of folklorists that in folktales it is not the order in which things happen
which is of importance, but what motifs appear in the tale. In other words: it is the
motifs of the tale which form its constant features, whereas the manner in which
these motifs are combined changes with time and place, producing a “different”
story.9 This approach differs considerably from the usual literary analysis of tales,
which consider them as entities to be taken as wholes, and analyze their context
both inside and outside the tale as such.10 I do not dispute the value of the latter
approach, but I think there are new historical insights to be gained from the
methods of the folklorists, especially when the motifs are analyzed within a certain
culture area, such as Europe or East Asia. I will, therefore, first try and isolate the
important motifs from the different known versions of this tale, and then I will
proceed to relate these motifs to other tales, which share some but not all of the
motifs of the Dōjōji tales. In this manner I will come up with a motif-cluster which
will point to a possible alternative interpretation. This, in turn, will lead us to ask
the question where the motif-cluster came from, and I will show how the cluster
came to Japan, from where, and in what context.

The Dōjōji tales of the Hokke Genki and the Konjaku Monogatari are sufficiently
similar to be treated as the same tale.11 It is when these tales are compared to the
ones found in the Genkō Shakusho and the Dōjōji Engi Emaki12 written down more
than two hundred years later, that we see some important differences emerge. The
account in the Genkō Shakusho, while firmly in the Japanese setsuwa tradition of
written tales, may be considered the bridge connecting the tales and the Dōjōji Engi
Emaki, a painted scroll with images and text. These later tales, it should be pointed
out, have become much more specific. Although it is still unclear when exactly the
miracle of the transformation happened, the Genkō Shakusho is the first version to
identify the young priest by name as Anchin and to locate him in the Kuramadera in
Yamashiro.13 The Dōjōji Engi Emaki takes over the name, but not his place of
origin, which has now become “Michinoku” or northeastern Japan. In the Genkō
Shakusho, again, the woman is still anonymous, but in the Dōjōji Engi Emaki, she
finally has acquired a name, i.e. Kiyohime (Pure Princess!) and the tale is dated to
the year 928, when Kiyohime is said to have lived in Manago village, in the district
of Muro. In the older tales, everything is vague. The story happened “once upon a
time”, the protagonists have no names, and the geographical locations are simply:
Kumano-Muro-Dōjōji.

9 The monument to this approach is Stith Thompson, Motif-Index of Folkliterature, a Classification

of Narrative Elements in Folktales, Ballads, Myths, Fables, Mediaeval Romances, Exempla,
Fabliaux, Jest-Books and Local Legends. Bloomington, Ind.: Indiana University Press, 1955.
10 Susan B. Klein, “When the Moon Strikes the Bell: Desire and Enlightenment in the Noh Play

Dōjōji” Journal of Japanese Studies, vol. 17, 2 (1991): 291-322.

11 Dykstra (1983), pp. 145-46; Brower (1952), vol. 2, pp. 415-20. Also translated in Marian Ury, Tales

of Times Now Past. Berkeley/Los Angeles/London: University of California Press, 1979, pp. 93-6.
12 Skord (1997): 75-84.
13 The temple is near the Kifune shrine, another place where a snake-god is worshipped. In the Akai

Hall of the Kuramadera, Benzaiten is honored as a female serpent water deity, see: Cornelius
Ouwehand, “Some Notes on the God Susanoo” in MN, vol. 14, nos. 3-4 (1958-9): pp. 396-8.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 103

The Kumano shrines are located in the ancient district known as Muro-no-kōri, so
there are really but two place names which can be located on a map: Kumano, deep
in the eastern mountains of the Ki peninsula (now: Wakayama prefecture), and the
Dōjōji temple situated not far from the mouth of the Hidaka river on the western
coast of the same peninsula. It is between these two sacred places that the action
takes place in the Hokke Genki and Konjaku versions. If the later versions of the
tale are more specific, we may be sure that most of these details are later
elaborations, and that, for our purposes here, they can be neglected. The emaki of
the temples were scrolls shown and explained to the public on fixed days of the year
by a knowledgeable monk who would embellish the story told in the pictures.14
Viewed in this context, it is not surprising that the emaki should be more elaborate
than the setsuwa tradition, and the surviving picture scroll was probably not the
first to exist.

More remarkable, however, is what the Dōjōji Engi Emaki has lost, when we
compare it to the earlier versions. In the early setsuwa tales, there were two priests,
whereas in the Dōjōji Engi emaki there is only one.15 For the Buddhist explanation
of the tale this feature is unimportant; it is therefore all the more fortunate that the
earlier Buddhist versions have preserved it, because by the fourteenth century it
seems to have been discarded already. We will see later that the older priest may
very well have been the center of a different but related tale in an earlier
combination of the same motifs.

A third difference between the two versions concerns the change of the woman into
a snake. In the Konjaku version, it is said that the woman dies of anger, and that her
grieving servants see a snake leaving her bedroom shortly after.16 In the Dōjōji Engi
Emaki version, however, we are treated to a full metamorphosis: first Kiyohime
starts spewing fire, then she is overcome by dizziness and grows a dragon head, and
when she finally enters the Hidaka river, she loses all her human features to become
a giant snake. It is possible that, for this metamorphosis, the later tellers of the tale
drew upon an old tradition connected with the Dōjōji temple which stated that an
image enshrined in the temple had been recovered from the water by a female
diver.17 This indicates a connection with the well-known Japanese tales of women
who marry snakes living in ponds, and change into snakes themselves after entering
the pond. These latter tales, it is interesting to note, are still current folktales in

14 Barbara Ruch, “Medieval Jongleurs and the Making of a National Literature” in John Whitney Hall
and Takeshi Toyoda [eds.], Japan in the Muromachi Age. Berkeley/Los Angeles: University of
California Press, 1977, pp. 279-309.
15 Komine Kazuaki, Konjaku monogatarishū no keisei to kōzō (Tokyo: Kasama Shoin, 1985), p. 137.

Already in the Genkō Shakusho version, the second priest only makes an appearance in the very
beginning of the tale. Skord has pointed out that even in the text of the picture scroll there remains
one instance where a reference to the other priest is made [Skord (1997): 77].
16 As Komine (1985, p. 138) has pointed out, the details of the grieving servants and the

metamorphosis in the bedroom are lacking in the Hokke Genki version, on which the Konjaku
version seems to be based. Therefore, we may consider these to be later additions by the Konjaku
compilers.
17 Skord (1997), p. 61.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

104 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

Wakayama prefecture.18 Moreover, this difference between the two texts indicates,
first, that these two versions of the stories have independent histories, and, second,
that the transformation into a snake (or dragon) is the crucial element connecting
the earlier part to the later part of the story. It should be clear that this link was the
work of Buddhist monks trying to fit the morals of their world view to local pagan
tales.19

If we now distill the motifs all versions have in common, we would come up with
the following:

1. A holy man is tempted by a woman.

2. The woman takes the shape of a serpent
3. The holy man hides under a temple bell
4. The serpent heats up the bell until it is red hot.
5. When the bell has cooled off, there is nothing left of the man.

For the moment I will concentrate on these five motifs. Motifs 1 and 2 still combine
in Japanese folklore, as we have already seen, as the marriage between a human
and a snake. Tales like these are not only still told in Japan,20 they are also among
the oldest tales of which we have a record, appearing in the Kojiki (712)21 as well as
in the Hitachi Fudoki (around 725).22

The Hitachi Fudoki tale involves a woman giving birth to a snake, whereas in the
Kojiki the snake changes into a handsome man who comes to spend the night with a
girl, but disappears mysteriously in the morning. His true identity is found out by
sticking a needle threaded with hemp into his clothes. In many of the later folklore
versions the iron needle causes the death of the snake. In other tales, it is said that
iron dropped in a snake pond will create a storm.23 In the Kojiki version, the girl
follows the thread and reaches Mt. Miwa, the deity of which is a snake. In the
Kojiki’s Chinese style companion volume, the Nihongi (720), Sukaru no Chiisakobe
goes out to catch the deity of Mt. Miwa: “So he ascended the Hill of Mimoro and

18 Seki Keigo, “Hebi muko iri dan no bunpu” in Minzokugaku kenkyū, vol. 6, 4 (1940): 1-40.
Translated as “The Spool of Thread: A Subtype of the Japanese Serpent-Bridegroom Tale” in R.M.
Dorson [ed.], Studies in Japanese Folklore Bloomington, Ind.: Indiana University Press, 1963, pp.
267-88.
19 For a description of this phenomenon, see W. Michael Kelsey, “Salvation of the Snake, the Snake of

Salvation: Buddhist-Shinto Conflict and Resolution” in Japanese Journal of Religious Studies, vol. 8,
1-2 (1981): 83-113.
20 F.J. Daniels, “Snake and Dragon Lore of Japan.” In Folklore, vol. 71 (1960): 163-4. Fanny Hagin

Mayer, Ancient Tales in Modern Japan: An Anthology of Japanese Folk Tales. Bloomington:
Indiana University Press, 1984, pp. 28-30.
21 For an English translation, see Donald L. Philippi: Kojiki. Tokyo: University of Tokyo Press, 1968,

pp. 203-4.
22 Akimoto Kichirō [ed.], Fudoki in: Nihon Koten Bungaku Taikei 2. Tokyo: Iwanami Shoten, 1958,

pp. 78-80.
23 Dorson (1963), p. 280.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 105

caught a great serpent, which he showed to the emperor, who had not practiced
religious abstinence. Its thunder rolled and its eyeballs flashed.”24

This same connection of a snake with the thunder god also appears among the tales
of the Nihon Ryōiki (822), the earliest Buddhist tale collection of Japan, where the
thunder god is also said to be a small child, or in Japanese chiisako.25 Sukaru no
Chiisakobe reappears in these tales and brings the thunder god to the palace at the
orders of the emperor. He is called Thunder-catcher Sukaru. 26 The collection
contains other tales of women marrying with or being sexually assaulted by snakes
as well.27 But it is again in the Konjaku monogatari that we find the clearest linking
of all these motifs in the story “How a Dragon King was Kidnapped by a Tengu”.28
In this intriguing tale, however, they seem to be hopelessly mixed up: a dragon
lives in a pond, from which he emerges in the shape of a little snake to bathe in the
sun. A tengu (or mountain demon) in the shape of a hawk swoops down suddenly
and brings the dragon to its nest on Mt. Hira. Once there, the tengu is unable to
devour the snake, while the dragon cannot move for lack of water. This stalemate
lasts until the demon brings a monk, whom he has kidnapped on Mt. Hiei, to his
lair. The monk has a jar with a little water left in it and this serves to revive the
dragon, who then changes into the shape of a little boy and takes the monk on his
back. At that moment, the sky darkens, thunder rolls, and lightning flashes. Before
the monk knows what is happening, he is back on Mt. Hiei.

Interesting here (and suggestive about the connections with archaic pagan
traditions) is the presence of a monk from Mt. Hiei, which itself is a Tendai
Buddhist temple complex superimposed on a spot of ancient mountain worship.
For our purposes here, the existence of a combination of these motifs without the
human-snake marriage is important. In the tales about such marriages, it is usually
a woman who marries a snake-man, whereas in the Dōjōji tales, it is a man who is
harassed by a snake-woman. This feature is in itself unusual, and may be explained
by the efforts of the Buddhist monks who authored these tales to connect different
sets of motifs into one single coherent Buddhist tale. However this may be, the key
is the mixed marriage, and to this we can now connect the following chain of
related motifs: marriage of human and snake → snake and thunder → thunder
and water → water and ponds → ponds and metal → metal and snake.

24 Nihongi, Yūryakki, XIV, 18. NKBT 67, p. 472. Translated by William G. Aston, Nihongi. Tokyo:
Tuttle Reprint, 1971, vol. 1, p. 347. If the emperor “had not practised religious abstinence” and so
caused the snake to act up, it must be deduced that Sukaru no Chiisakobe, who was handling the
snake, must have been a holy man of extreme “religious purity.” He may have been the chief of a
powerful guild of rainmaking magicians, the chiisakobe. For the connection between snakes and the
ability to make rain see also below.
25 Nihon Ryōiki, I, 3. NKBT 70, pp. 71-2. For a summary of these tales in modern folklore see: Hiroko

Ikeda, “A Type and Motif Index of Japanese Folk-Literature” in Folklore Fellows Communications,
no. 209 (Helsinki 1971): 106-7.
26 See also: M.W. de Visser, “The Snake in Japanese Superstition” in Mitteilungen des Seminars für

Orientalischen Sprachen an der Königlichen Friedrich-Wilhelms Universität zu Berlin, vol. XIV

(1911): 275.
27 Nihon Ryōiki, II, 8. NKBT 70, pp. 201-3, 293-6.
28 Konjaku monogatarishū 20, 11.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

106 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

The other three motifs of the five I distilled above belong, I think, to separate cluster
and are only combined with the first two in the Dōjōji tales. This combination of the
two sets of motifs is, in fact, the essence of the tales connected with the Dōjōji
temple.

The Noh play Dōjōji

The Noh play Dōjōji is the youngest of the tales analyzed here.29 It dates in its
present form from the early sixteenth century, and is said to be an adaptation of an
earlier play called Kanemaki, or “wrapping around the bell.” From the title of the
earlier play, which survives in the countryside,30 but not as a play performed by
professional Noh actors, we can surmise that the central element of the story is the
heating of the bell. The play, then, represents a third version of the tale dealing
mainly with the relationship of the woman to the bell. To explain this relationship,
the play falls back on the same Buddhist interpretation used in the tales discussed
above in an explanatory interlude by the waki or supporting actor of Noh plays,
who represents the abbot of the Dōjōji temple.

This explanation differs but little from the Dōjōji Engi Emaki version: a yamabushi,
or mountain priest, makes his yearly pilgrimage to Kumano. As usual, he stops at
the house of the steward of Manago, where he is in the habit of bringing little
presents for the daughter of the house. The steward has once told his daughter as a
joke that someday she will marry the visitor. This time she thinks she has waited
long enough and asks the priest how long she still has to wait before he will make
her his wife. But the priest does not respond and makes off in the middle of the
night. The girl follows him, and turns into a giant snake when she finds she cannot
cross the Hidaka river.

It is important, however, that the action of the play itself is a follow-up on this story.
It stresses the continuing connection of the Dōjōji temple bell with the snake
woman. The play takes place at the dedication ceremony of a new bell, and strict
orders have been issued not to allow any women near it. As Tokue Motomasa has
pointed out, the taboo on women approaching the site where bell casters work is a
common one.31 This taboo must have originated in the time when the only females
allowed near the casting site were those who were sacrificed (married) to the dragon
of the furnace. The motif of the sacrifice of a virgin to the forge or furnace is a
common among metalworkers all over the old world.32

29 Yokomichi Mario, Omote Akira [eds.], Yōkyokushū, vol. 2 NKBT 41. Tokyo Iwanami shoten, 1963,
pp. 129-42. Extract in Herman Bohner, Die einzelnen Nô MGOA Supplementband XXII. Tokyo:
Deutsche Gesellschaft für Natur und Völkerkunde Ostasiens. Wiesbaden: Otto Harrassowitz, 1956,
pp. 255-9. A complete translation can be found in René Sieffert, Nô et Kyôgen. Printemps,Été. Paris:
Publications Orientalistes de France, 1979, pp. 390-403.
30 According to Donald Keene [ed.], 20 Plays of the No Theatre. New York/London: Columbia

University Press, 1970, p. 238.

31 Tokue (1984), p. 418.
32 Mircea Eliade, The Forge and the Crucible. Chicago/London: University of Chicago Press, 1978,

pp. 65-70. For China, see: Lionello Lanciotti, “Sword Casting and Related Legends in China” in East
and West, vol. 6 (1955): 109.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 107

The dragon, then, and its everyday manifestation the snake, is in these tales on the
one hand the symbol of the powerful fire blast the casters have to create (and
contain!), and on the other hand it symbolizes the liquid metal itself which is slowly
poured into the mold. Tales claiming that the casting of large bells can only be
successful through the sacrifice of a girl are attested in China33 and Korea.34 The last
part of the bell to be cast is the cannon at its very top from which it will be hung.
Fittingly, the part of the bell is still called “the dragon’s head” (ryūzu),35 and one of
the most suggestive images of the Dōjōji Engi Emaki shows the dragon’s head at the
top of the bell.

As expected, a woman does turn up and manages to gain access to the temple
grounds where she begins her magic dance. She reaches the spot just underneath
the newly hoisted bell and jumps up into it, causing it to fall down from the belfry
again. When the temple servants hear the noise and arrive on the spot, they find
that the bell is burning hot. The climax of the play is the second raising of the bell.
This is done with the chorus invoking all the dragons of the Buddhist universe to
help contain the terrible transformation which is occurring underneath the rising

33 E.T.C. Werner, Myths and Legends of China. New York: Arno Press Reprint, 1976, pp. 394-8.
34 The Emille bell in Kyongju is still said to ring with the voice of the girl sacrificed at its casting. See:
Takahashi Moritaka, “Jinshōkō” in Chōsen Gakuhō, no. 3 (1951), p. 83-100.
35 Percival Price, “Japanese Bells” in Calvin L. French [ed.], Studies in Japanese Culture II Center for

Japanese Studies, Occasional Papers no. 11. Ann Arbor, MI: University of Michigan, 1969, p. 50.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

108 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

bell. The woman has changed into a fire-spitting dragon once again but the monks’
dragon prayer is too powerful for her and she disappears into the Hidaka river.36

This detail of the prayer is significant because it is different from the standard
Buddhist redaction which ascribed the solution of the whole dilemma of the snake
incarnation to the power of the Lotus sutra. Again, we see how at different times
different “solutions” were applied to the anomalies of the pagan tales.37 More clearly
than in the versions of the Dōjōji tales we have analyzed above, the Noh play has
preserved the connection of the motifs of the tale with metalwork (i.e. bronze
casting) and its traditions, and it therefore represents another distinct version of the
tale, the one in which, in my opinion, the motifs are combined in the most archaic
fashion.

There is another Noh play dealing with the motif of the bell, i.e. Miidera. The
temple bell of the Miidera in Shiga prefecture is called Mii-san, one of three snake
sisters who live in Lake Biwa.38 It is said in the play that even the dragon-woman
would be able to reach Buddhahood if she can ring the bell.39 Even though the Noh
play is more properly a Buddhist play than Dōjōji, the pagan element of the dragon
in the lake in still there. Furthermore, a folk tradition connected with this bell
associates it with an evil woman, managing to come near it in spite a temple
ordinance forbidding women to enter, whose touch once spoiled its bright luster.40
The resemblance with the “evil woman” in the Dōjōji tales is, of course, not
coincidental and must be considered yet another instance of a vestige in folklore of
the ancient metallurgy tradition of human sacrifice. For our second set of motifs we
may now note the following chain of associations: woman and dragon → dragon
and fire → fire and casting → casting and bell → bell and its inside → inside and
transformation.

Two “Chinese” Folktales

Wolfram Eberhard, whose work on the local cultures of China has done so much to
illuminate the composite nature of classical Chinese civilization, has, without
mentioning the Dōjōji tales, pointed to some very intriguing parallels in south-east
China from the culture area of the ancient Yue.41 The tales for which he gave
references turn out amazingly similar to the cluster of motifs we found in Japan.
The first tale is called “The Dark-Green Snake Emissary.”

36 The dragons invoked are: “the green dragon of the east, the white dragon of the west, the yellow
dragon of the center, and all the countless dragon kings of the three thousand worlds.” NKBT 41, pp.
141-2.
37 Used here in the sense established by William R. LaFleur, The Karma of Words. Buddhism and the

Literary Arts in medieval Japan. Berkeley/Lon Angeles/ London: University of California Press,
1983, pp. 33-45.
38 Cf. the story of O-Mii-san told by Fukumoto Mie in Kasai Noriko [ed.], Ōmi no mukashibanashi,

vol. 6 in Inada Kōji, Nihon no mukashibanashi. Tokyo: NHK, 1973, pp. 28-30.
39 NKBT 41, p. 489.
40 See the story “Benkei and the Bell”, in William Elliot Griffis, Fairy Tales of Old Japan. London:

George C. Harrap, 1908, pp. 43-4.

41 Wolfram Eberhard, The Local Cultures of South and East China. Leiden: Brill, 1968, p. 418.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 109

The San-Jie temple is also called the Temple of the Dark-Green Snake
because once there was a small snake living in a hole [under the altar] of
the temple. It had a line on its back and a red belly. People brought it food
and worshipped it as a god. If you held it in your hands it would be very
gentle and playful. If you made a vow and did not honor it, the snake
would come and demand satisfaction, even if your home was several
hundred li away. That is the reason why it was called the divine emissary.

San Jie’s family name was Xu and he came from Ping-nan county. One
day, when he was gathering firewood, he found himself a coat as light as a
leaf and without any seams on top or bottom. With the coat there was a
belt, which had the character for “returning” on it and could cause the wind
to blow, the rain to fall, and predict the future. People would gather by its
magic.

During the reign of Hong-zhi,42 the local government sent an official to the
temple and he covered the snake with a big bell. Then he had firewood
stacked up around the bell, which he lit and burned from morning till
nightfall. [When the official later checked underneath], there was nothing
there. Probably, San Jie had put on his magic coat, [vanished], but stayed
alive. The day before this happened, I talked with my two friends Yuan-
yuan and Zhao Ping Si about this snake. I told him it must be a metal43
spirit . . . [follows a literary discussion].

One morning, the three of us went to the temple and poured wine into the
hole of the snake under the altar, and lit the alcohol. Then we obtained a
sword of a pure green color. That same night, my two friends both had the
same dream in which they saw a young man in green clothes putting a
piece of jade on my belt, after which the green snake man disappeared
forever.44

In this story we find the following chain of associated motifs. First, there is the dark-
green snake, revered in its own temple, but which is also referred to as the divine
being San Jie, whose family name is Xu. Obviously, San Jie is a snake-man, i.e. a
being able to take both the shape of a snake and that of a man. In his human shape,
San Jie finds “a magic coat without seams” when collecting firewood in the
mountains. This coat is “as light as a leaf” and is, on the one hand, another proof of
his snake nature, because as we all know snakes discard their “seamless coats” every
year. On the other hand, the magic coat refers to the eternal quest of the
metalworkers who need to be invulnerable to fire. There are, indeed, other
indications in this tale that San Jie must originally have been a metalworker
because in the end the temple-snake itself is changed into a deep-green (that is:

42 A vague era during the Ming dynasty, lasting from 1488 to 1506.
43 Lit. “gold”.
44 From Wang Sen汪森 (1653-1726), Yue-xi cong-zai 粤西叢載, Taipei: Guang wen shuju, 1969. vol.

14, 9b-10a.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

110 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

bronze) sword. Wang Sen, the 17th century author of the text translated here, quotes
the Chi Ya as his source for this tale, which obviously dates from the Ming period. It
is unclear how much older the story of the Green Snake was at the time it was
written down. Some elements (the magic coat, the snake under the heated bell) are
clearly of a much earlier date. These are the motifs also occurring in the following
tale:

Once, Feng Ke-li from Gui county went into the Northern Mountains to look
for herbs to make incense. There he came upon eight immortals who were
playing chess together. They gave him a magic coat without any seams.
Upon coming home, he noticed that in his absence several generations had
passed by. He reported this to the authorities and an official was sent down
to his county to investigate. This official had Feng and his servant put
underneath a large bronze bell, after which firewood was lit all around it.
When the inside of the bell was inspected [after it had cooled off], Feng was
still sitting straight up, but his servant had been reduced to ashes.
Therefore, he was regarded as a true immortal, and a document was
drawn up to that effect to be presented to the authorities. Later an edict
was issued naming Feng “Heavenly Traveller Accomplished Taoist of the
Third Level”. When he reached the Cang-wu river on his way home, [after
receiving the certificate] he vanished into thin air. Nowadays, many people
of that neighborhood worship him as a god. (Quoted from the Gazetteer of
Xun-zhou Prefecture).45

Going into the mountains to look for herbs to make incense is a clear reference to
the quest for immortality. Feng Ke-li indeed meets with the immortals who give
him a magic coat without any seams. As we have seen, a snake skin in a coat
without seams, so Feng Keh-li is a snake man in the same traditions as San Jie.
When he returns home, several generations have passed: again a common element
in stories about immortality, also known in Japan, for example in the story of
Urashima Tarō. The Taoist quest for immortality is closely related, furthermore, to
earlier shamanistic traditions, and from shaman to smith/bronzecaster is but a
small step. “Like the shamans”, writes Eliade, “the smiths were reputed to be
‘masters of fire’. And so in certain cultures, the smith is considered the equal if not
superior to the shaman. ‘Smiths and shamans come from the same nest,’ says a
Yakut proverb.”46 Feng Ke-li, the shaman, has the necessary powers to deal with the
ordeal of sitting under a burning hot bell, but his servant lacking these powers
perishes. The reader will recall that in the older Dōjōji tales of the Hokke Genki and
Konjaku monogatari there were two priests pursued by the snake/dragon. In the
early Buddhist recombinations of the old motifs, then, the older priest still figured
but he survived the ordeal – more rationally! – outside of the bell. The younger
priest (the servant in the Feng Ke-li version), however, perished as of old.

45 Wang Sen, op. cit. vol. 11, 14b. Also: Wolfram Eberhard, “Volksmärchen aus Südost-China” in
Folklore Fellows Communications, no. 128 (Helsinki, 1941) pp. 181-2.
46 Eliade, op. cit., p. 81.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 111

Just as in Japan where the traditional stories of the bronze casters were adopted by
the Buddhist monks and given a new interpretation, we find them told in Taoist
terms in these Chinese folktales. In a way this is fortunate because it shows how in
spite of their extensive similarities (holy man under bell / bell heated by external
fire / testing powers) all of these written tales have totally independent histories,
and that if they have a common origin, as is assumed here, this must go back to
remote antiquity. In other words, it is highly unlikely that the similarities should be
explained as borrowings from the time after Buddhism was brought to Japan. The
many different Japanese versions of the tale, with each their own idiosyncrasies,
point on the contrary to the fact that the stories all go back to a living oral tradition
around the Ki peninsula until at least the late Middle Ages, and in some cases until
today. Elements from this tradition were picked up from time to time by Buddhist
monks, who worked hard at giving them a Buddhist interpretation.

It may be, therefore, that the dengaku, or rural play, Kanemaki itself was a direct
result of this oral tradition, which was transformed by Kanze Kojirō Nobumitsu
(1435-1516) into the Noh play Dōjōji in the beginning of the 16th century. This would
explain why the combination of the motifs in this play is the most archaic, even
though the play itself the youngest of all the different Dōjōji traditions we have
considered. That the Japanese countryside should have preserved the most ancient
version is, of course, not surprising.

Conclusion

The implications of the existence of these Chinese tales may reach very far. Even
though there are still many lacunae in our knowledge of early Japanese and Chinese
bronze traditions as well as concerning the distribution of Chinese folktales, it
seems clear that here we have a piece of evidence for the theory already proposed by
Torii Ryūzō in the early part of the 20th century that the Yayoi period may have
been started by the Yue people from south-east China, who brought their
knowledge of seafaring, wet rice agriculture, and bronze casting to Japan.47 The Yue
are known to have been among the oldest seafarers along the Chinese coast, and
Torii made particular mention of Yue bells which he called the only available
prototypes for the later Japanese dōtaku, the bronze bells of central Japan for
which the Yayoi period is famous. Tanigawa Ken’ichi has tried to unravel the
mystery of all the different bands of metalworkers coming to Japan from this area.48
Eberhard, moreover, gives numerous examples of human sacrifices to a snake god
who lives in rivers and ponds in the area of the Yue.49 Willem van Gulik has taken
up Torii’s suggestions again in his study of Japanese tattoos, and documented the
extensive use of snake and dragon symbolism, notably in connection with the fire-

47 Torii Ryūzō, “Populations préhistoriques de la Mandschourie méridionale” in Journal of the

College of Science, vol. 36, 8 (Tokyo, 1915): 39-40. And Yūshi izen no Nihon Tokyo: Isobe Kanyōdō,
1918, pp. 302-6.
48 Tanigawa Ken’ichi, Seidō no kami no ashiato. Ch. 4: “Kōnan to no tsunagari to dōtaku” in:

Tanigawa Ken’ichi chosakushū, vol. 5. Tokyo: San’ichi Shobō, 1985, pp. 241-65.
49 Eberhard, “The Cult of the Snake” in Local Cultures, pp. 381-90.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

112 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

fighters of the Edo period.50 More recently, Charles Holcombe has made the same
argument in his study of the emergence of the East Asian culture sphere.51 I agree
with these authors and think that the cluster of motifs found in the Dōjōji and
related tales were brought to Japan by groups of Yue during the Yayoi period. It is
possible that particular groups bringing these tales as well as the technology of
bronze casting settled on the coasts of the Ki peninsula and those of Shikoku on the
other side of the Ki Suidō, before moving further inland. Finally, it is intriguing to
note that the site of the Dōjōji temple is right in the middle of one of the few areas in
Japan were more than ten dōtaku have been found that are counted among the
oldest known.52 This would seem to indicate that, at the very least, bronze bells
were venerated in this general area long before the advent of Buddhism and the
establishment of the temple (AD 708) famous for its bell.53 Even more intriguing is
the fact that, in 1762, on the grounds of the Dōjōji temple itself, a dōtaku was found
buried on the spot where a three-tiered pagoda was being repaired.54

What is more, the place name of Kumano (or: Manago in Muro-no-kōri) is a

constant feature in the Japanese tales we have analyzed, just like the name of the
Dōjōji temple is. Kumano, on the eastern shore of the Kii peninsula, is the place
where, in all the Japanese versions, the priests are traveling and where the
transformation between a human and a snake actually takes place, even though in
the Dōjōji Engi Emaki version that transformation is only complete once the
woman enters the Hidaka river on the western side of the peninsula. I have already
stated the probable reason for this latter feature, i.e. the addition of another ancient
element to the tale told on this scroll, that of the snake wife in the water. The real
question here is thus why the location of Kumano is so consistently associated with
the Dōjōji temple in these tales.

The two temple complexes are a considerable distance away from each other, on
foot at least a week of arduous walking crossing several mountain ridges.
Admittedly, part of the way would take one along the banks of the Hidaka river.
Still, taking our tales literally, it is hard to see how or why, if one felt threatened by a
dragon in Muro-no-kōri, one would flee all the way to find refuge in the Dōjōji
temple on the other side of the peninsula. Therefore, I think it is safer to take this
connection of Dōjōji with Kumano as another linkage, albeit a rather ancient one
(because it occurs in all the Japanese tales), made by the Buddhist retellers of the
ancient motifs found in the folktales of the area. If we assume, for example, that the

50 Willem R. van Gulik, Irezumi, the Pattern of Dermatography in Japan (Leiden: Rijksmuseum
voor Volkenkunde, 1982.
51 Charles Holcombe, The Genesis of East Asia. Honolulu: University of Hawaii Press, 2001, p. 20.
52 H. Mabuchi, Y. Hirao, and M. Nishida, “Lead Isotope Approach to the Understanding of Early

Japanese Bronze Culture” in Archaeometry, vol. 27, 2 (1985): 143, fig. 8.

53 Even in the name of the temple itself, there seems to be a vague connection with metallurgy. Dōjō

is written with characters that can also be read as Michinari, and thus Dōjōji means ‘Michinari’s
temple’. The Michinari referred to is the temple’s founder, said to have been Tachibana no Michinari,
a semi-legendary figure, whose only other appearance in Japanese folklore is in the Noh play Kokaji
(“Sword smith”) as an imperial messenger sent to order a sword from the smith Munechika. See: C.K.
Parker and S. Morisawa, “Kokaji” in MN, vol. 3, 2 (1940): 259-69.
54 Umehara Sueji, Dōtaku no kenkyū: Shiryōhen. Tokyo: Ōokayama shoten, 1927, p. 275.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions 113

original layer of what later became the Kumano faith was also based on
metallurgical traditions,55 just as the stories connected with Dōjōji were, the linking
of the two locations in the Buddhist tales becomes more logical.
In the case of Kumano, bands of metal prospectors and bronze casters seem to have
become mountain priests or yamabushi.56 In the Noh version of the Dōjōji tale, it
will be remembered, the protagonist is expressly identified as a yamabushi from the
northeast making his yearly pilgrimage to Kumano.57 Among the repertoire of Noh
and dengaku or shrine plays performed by yamabushi today, furthermore, Dōjōji,
Kanemaki, as well as Miidera still figure prominently. 58 The latter temple,
moreover, functioned until the fourteenth century as the link of Kumano Shugen
with the Tendai school,59 and we have seen how it figures in the folklore about the
bronze casters. The yamabushi have preserved many traits that directly link them
to metalwork, and more especially with bell casting. 60 The most prominent
Buddhist deity of the yamabushi is Fudō Myōō, a deity represented as sitting
immobile within a halo of fire, and feats such as fire-walking still feature among
yamabushi ceremonies. This alone reminds one of the Chinese shaman Feng Ke-li.

To the standard equipment of the yamabushi, moreover, belong several metal

objects, for example a staff with rings of metal attached to its upper part. Some
yamabushi carry a staff which is entirely made out of iron. Others, like Benkei, are
said to wear geta made of iron.61 Until very recently, women were not allowed to
climb the holy mountains of the yamabushi, and it has often been said that
Shugendō is nothing but “a Buddhistically modified, ancient and native form of
Nature-worship.”62 Therefore, we may find in the fact that the transformation into
the snake/dragon occurs at Kumano, while the redemption occurs at Dōjōji, a faint
echo of the time when Dōjōji was already functioning as a Buddhist center, while
Kumano was still pagan.

A form of human sacrifice, indeed, continued far into historical times on the
Kumano side of the Kii peninsula. We only have to think of the custom of Fudaraku
tokai, “crossing the sea to reach Mt. Potalaka,” to realize the difficulties the
Buddhist clergy must have faced here.63 According to this custom, Buddhist “saints”

55 This is Sawamura Tsuneo’s thesis in: Kumano no nazo to densetsu. Tokyo: Kōsakusha, 1981.
56 Wakao Itsuo, “Kumano shūgen to kinkō” in his Oni densetsu no kenkyū – Kinkōshi no shiten kara.
Tokyo: Yamato Shobō, 1981, pp. 103 ff.
57 NKBT 41, p. 139.
58 Thomas Immoos, “Das Tanzritual der Yamabushi” in MGOA, vol. 50 (1968): 9. Takatani Shigeo,

“Yamabushi odorikō” in Kinki minzoku, vol. 47 (1969): p. 1833; Inoura Yoshinobu, A History of the
Japanese Theatre I. Noh and Kyogen. Tokyo: Kokusai Bunka Shinkokai, 1971, Ch. 10: Shugen Noh,
p. 81.
59 Miyake Hitoshi, “Rethinking Japanese Folk Religion: A Study of Kumano Shugen” in Peter F.

Kornicki and I.J. McMullen [eds.], Religion in Japan. Arrows to Heaven and Earth. Cambridge:
Cambridge University Press, 1996, p. 123.
60 Wakao Itsuo, “Imoji no kami wo hōzuru Kumano shugen” in Oni densetsu no kenkyū, pp. 96 ff.
61 One of the most famous tourist attractions of Kyoto, the Kiyomizu temple, features a pair of huge

iron geta, said to have belonged to the legendary yamabushi Benkei.

62 U.A. Casal, “The Yamabushi” in MGOA, vol. 46 (1965): 15.
63 See: Nei Kiyoshi, Shugendō to kirishitan. Tokyo: Tōkyōdō shuppan, 1988, pp. 275-316. Also by the

same author, “Chūsei no Fudaraku tokai” in Indogaku bukkyōgaku kenkyū, vol. 34, 2 (1986): 209-

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899

114 Reinier H. Hesselink: The Dōjōji Tale and Ancient Bronze Metallurgical Traditions

and their followers sacrificed themselves by entering a hermetically closed wooden

box mounted on a small vessel which was abandoned at sea in the kuroshio current,
never to return to Japan.64 This custom is attested in historical sources since 868,65
with most of the suicides starting at Hama no Miya on the beach in front of Mt.
Nachi. The last of these “voluntary sacrifices” occurred in 1722.66
Although it can be argued that this was a Buddhist custom and had nothing to do
with the Kumano faith or with metallurgy, I believe on the contrary that this is
another example of the process of assimilation and absorption of local traditions by
the Buddhist clergy as I have demonstrated above occurred at Dōjōji. Buddhism
incorporated and, in some ways, sanitized and civilized many of Japan’s native holy
places. Mountains yielding metal, furthermore, figure prominently among such
holy places, and of Mt. Nachi it is said that it yielded copper until the Tokugawa
period.67 The human sacrifices that seem, originally, to have been common in
connection with metalwork were anathema to the Buddhists. Stories based on the
practice were, however, plentiful on the Kii peninsula, and not easily forgotten by
the local people. They, therefore, had to be changed into forms that stressed
eventual redemption in paradise.

14; and Hashikawa Tadashi, “Waga kuni ni okeru Fudaraku shinkō” in Miyaie Jun [ed.], Kumano
shinkō. Tokyo: Yuzankaku, 1990, pp. 99-102.
64 See: D. Max Moerman. Localizing Paradise: Kumano Pilgrimage and the Religious Landscape of

Premodern Japan. Cambridge, MA/London: Harvard University Asia Center, 2005, p. 117, and its
review in Journal of Asian Studies, vol. 67, 1 (2008): 326.
65 The first saint attested to have sacrificed himself in this manner was, intriguingly, called

Keiryū慶龍, or “happy dragon,” information from the Kumano nendaiki『熊野年代記』quoted in Nei

(1988), p. 276.
66 104 people sacrificed themselves upon twenty different occasions in this manner between 868 and

1722. I am not suggesting, of course, that, except maybe for the earliest ones, their conscious reasons
had anything to do with metallurgy. The fact that 77 of these sacrifices, divided over 10 occasions,
occurred between 1499 and 1594 suggests, on the contrary, that the misery of sengoku times must
have played a large role in their decision to go and look for paradise. See: Nei (1988): 290.
67 Sawamura (1981), p. 152.

Comparative Mythology December 2016, Volume 2, Issue 1 ISSN 2409-9899