Dragons are portrayed in widely separated cultures as simultaneously male
and female. While this trait may appear arbitrary, its global distribution implies that it is
motivated by observations available to all humans. Many seemingly arbitrary traits of
dragons are shared by conceptions of the rainbow, which is widely portrayed as having
both male (primary) and female (secondary) arcs. This correlation of features in mythical
and natural counterparts shows that the nineteenth century doctrine of naturalism, which
saw various elements of myth and folklore as reflecting features of the natural world, was
valid, despite excesses that eventually caused its abandonment.
Robert Blust received his doctorate in linguistics from the University of Hawai’i in 1974.
He subsequently worked at the Australian National University for two years and at the
University of Leiden in the Netherlands for nearly eight years before returnng to Hawai’i,
where he has been a professor of linguistics since 1984. He has conducted linguistic
fieldwork on 100 Austronesian languages in Borneo, Papua New Guinea and Taiwan, and
has authored over 250 publications, including the first single-authored book on the entire
Austronesian language family (2009/2013. The Austronesian languages. Canberra:
Pacific Linguistics, 824pp.; 2nd rev. ed. 2013. Berlin: de Gruyter-Mouton. 851pp.), as
well as the ongoing Austronesian Comparative Dictionary (www.trussel2.com/acd), by
now uncontroversially the largest research project ever undertaken on the Austronesian
language family, containing nearly one-fifth of the world’s languages. He was elected a
Fellow of the Linguistic Society of America in 2016, and in 2017 he received a Lifetime
Achievement in Research award from the University of Hawai’i. His interest in why
there is a global belief in dragons goes back over two decades, and resulted in an earlier
publication in Anthropos (Blust 2000).
Publications since 2015
2018 101 problems and solutions in historical linguistics: a workbook. 492pp.
Edinburgh: Edinburgh University Press (With the editorial assistance of Hsiu-
2015 The case markers of Proto-Austronesian. Oceanic Linguistics 54: 443-498.
2015 Longhouses and nomadism: is there a connection? Borneo Research Bulletin
2015 Southeast Asian islands and Oceania: Austronesian linguistic history. In
Peter Bellwood, ed., The global prehistory of human migration:276-283.
Chichester: Wiley Blackwell.
2015 In Memoriam, George William Grace, 1921-2015. Oceanic Linguistics 54:589-
2016 The Liangdao skeleton and the dangers of overinterpretation. Journal of Chinese
2016 Kelabit-Lun Dayeh phonology, with special reference to the voiced aspirates.
Oceanic Linguistics 55:246-277.
2016 Avoidance of dissimilar labial onsets: the case of Subanon. Oceanic Linguistics
55:620-633 (with Elizabeth Nielsen).
2017 The Lowland Kenyah posterior implosives: a typological reversal. Language and
2017 Historical linguistics and archaeology: an uneasy alliance. In Philip J. Piper,
Hirofumi Matsumura and David Bulbeck, eds., New perspectives in Southeast
Asian and Pacific prehistory (Terra Australis 45):275-291. Canberra:
Australian National University.
2017 The linguistic history of Austronesian-speaking communities in Island Southeast
Asia. In Peter Bellwood, First islanders: Prehistory and human migration in
island Southeast Asia:190-197. Oxford: Wiley Blackwell.
2017 Regular metathesis in Batanic (northern Philippines)? Oceanic Linguistics
2017 Odd conditions: context-sensitive sound changes in unexpected contexts.
Journal of Historical Linguistics 7.3:322-370.
2017 The challenge of semantic reconstruction: Proto-Malayo-Polynesian *suku
‘lineage; quarter?’. Oceanic Linguistics:247-256.
2017 The pitfalls of negative evidence: Nuclear Austronesian, Ergative Austronesian
and their progeny. Language and Linguistics 18.4:579-623 (with Victoria Yen-
2017 Review of Greg Mellow. 2014. A dictionary of Owa, a language of the
Solomon Islands. OL 56.2.508-512.
2018 Ancient DNA and its contribution to understanding the human history of the
Pacific islands. Archaeology in Oceania 53.3:205-219 [invited comment on pp.
2018 In Memoriam: Richard Bernard McGinn, Jr., 1939-2018. OL 57:359-362.
2018 Historical linguistics in the raw: my life as a diachronic fieldworker. In Hannah
Sarvasy and Diana Forker, eds. Word Hunters. Field linguists on fieldwork:29-
42. Amsterdam and Philadelphia: John Benjamins.
2018 The ‘mystery aspirates’ in Philippine languages. OL 57:221-247.
2018 The challenge of semantic reconstruction 2: Proto-Malayo-Polynesian
*kamaliR ‘men’s house’. OL 57:335-358.
2018 Two birds with one stone: the aerodynamic voicing constraint and the languages
of Borneo. Journal of the Southeast Asian Linguistics Society 11.2:1-18.
2019 Why dragons are bisexual: a defense of Naturalism. Anthropos 114:169-180.
2019 The Austronesian homeland and dispersal. Annual Review of Linguistics 5:417-
2019 Review of Ritsuko Kikusawa and Lawrence A. Reid, eds. (2018), Let’s talk about
trees: genetic relationships of languages and their phylogenetic representation.
Senri Ethnological Studies 98. Osaka: National Museum of Ethnology. JSEAS
to appear The resurrection of Proto-Philippines. Ms., 107 pp. To appear in Oceanic
Linguistics 58.2 (Dec. 2019).
to appear More odd conditions? Voiced obstruents as triggers and suppressors in
Miri, Sarawak. To appear in Phonology.
to appear Some remarks on etymological opacity in Austronesian languages. To
appear in a festschrift. Ms., 18pp.
to appear Linguistic approaches to Austronesian culture history. To appear in a
volume edited by K.A. Adelaar and Antoinette Schapper. Ms., 25pp.
4. cross-cultural comparison
Table 1: The geographical distribution of draconic traits (1 = Europe, 2 = the ancient Near
East, 3 = India, 4 = the Far East, 5 = Mesoamerica, 6 = North America; + = trait reported)
(1) (2) (3) (4) (5) (6)
g) + + + + +
h) + + +
i) + + +
j) + +
k) + + +
l) + + + + +
m) + + + +
o) + + +
p) + + +
q) + + +
r) + +
s) + + +
t) + +
w) + + +
x) + +
z) + +
aa) + +
a = giver/withholder of rain
b = guardian of springs or other bodies of water
c = capable of flight
d = appears when rain and sun are closely interspersed
e = can change shape or size, or suddenly disappear
f = has scales
g = has horns
h = has hair (mane, whiskers, etc.)
i = has feathers
j = is equine or hippophidian
k = is bisexual/androgynous
l = is opposed to thunder/lightning, or the sun
m = is colorful/red
n = is offended by a menstruating woman
o = terrifies young women; can impregnate them with demonic child
p = has fiery breath
q = has fetid or poisonous breath
r = causes tornadoes
s = causes floods
t = causes earthquakes
u = is an omen of catastrophe
v = is an emblem of war
w = guards a treasure
x = is connected with longevity/immortality
y = is connected with fertility
z = encircles the world
aa = lives in waterfalls
WHY DRAGONS ARE BISEXUAL: A DEFENSE OF NATURALISM
University of Hawai’i
1. The abandonment of scientific ideas. The abandonment of scientific ideas provides
one of the most revealing windows on the sociology of science. Many theoretical
orientations that received a ready hearing in nineteenth century scientific circles were
subsequently found to be flawed in some way. In social and cultural anthropology the
reaction of the scientific community to the discovery of imperfection in theory generally
was unconditional rejection of the original idea. Only rarely has there been a willingness
to reconsider the potential value of core concepts that were initially associated with
untenable corollaries. Nineteenth century Evolutionism is a case in point. As originally
formulated the central concepts of cultural evolution were inextricably bound up with
ethnocentric notions of cultural superiority, and in the early twentieth century the entire
enterprise was abandoned and even considered a source of intellectual embarrassment.
Shorn of these trimmings, however, and situated within a different matrix of assumptions,
the study of cultural evolution made an important comeback in the middle to late
twentieth century in the work of such anthropologists as Leslie White, Robert Service and
In most cases where scientific ideas have been abandoned, however, there has been no
return to the status ante quo in some other guise. Rather, rejection by the community of
scholars has resulted in a sense of ineradicable taint. One of these nineteenth century
ideas that went from widespread acceptance not just to rejection, but to treatment as a
notion that was fundamentally and irrevocably wrong is Naturalism.
2. Naturalism. The term “Naturalism” has several referents. Its most common usage in
contemporary scholarship is in philosophy, where it refers to “the view that the spatio-
temporal universe of entities studied in the physical sciences is all there is”, and is
consequently opposed to Theism (Craig and Moreland 2000:xi). However, in nineteenth
century anthropology, and more particularly folklore, the term ‘Naturalism’ referred to
the position that myths and folkloric motifs have their origin in attempts to represent the
natural world in anthropomorphic or theriomorphic shapes. My point of departure is a
brief remark by the anthropologist Claude Lévi-Strauss in his 1962 book, The Savage
Mind (p. 95):
“…even when raised to that human level which alone can make them intelligible, man’s
relations to his natural environment remain objects of thought: man never perceives them
passively; having reduced them to concepts he compounds them in order to arrive at a
system which is never determined in advance: the same system can always be
systematized in various ways. The mistake of Mannhardt and the Naturalist school was
to think that natural phenomena are what myths seek to explain, when they are rather the
medium through which myths try to explain facts which are themselves not of a natural
but a logical order.”
Who was Lévi-Strauss targeting in this critique? He mentions Wilhelm Mannhardt, a
nineteenth century German folklorist who authored a 1,007 page tome entitled Wald und
Feldkulte. The most famous representative of this school in the English-speaking world,
however, undoubtedly was Friedrich Max Müller, a German national who spent his
academic career at All Soul’s College, Oxford, and who is perhaps best remembered for
establishing the field of comparative religion, and for his theories of solarism --- that is,
the purported solar origin of many anthropomorphic characters in myths around the
world. To members of this school myths often involve only slightly disguised
personifications of recurrent events in the heavens, and myth was thus a way of
representing these events (which were often important in agricultural societies) in a
memorable form. As Müller put it his 1891 book, Physical Religion, “It was in these
very phenomena of nature that ancient man perceived for the first time something that
startled him out of his animal torpor, and that made him ask, What is it? What does it all
mean? Whence does it all come? --- that forced him to look behind the drama of nature
for actors and agents … whom in his language he called superhuman, and, in the end,
Müller combined his interest in mythology with an interest in Indo-European
comparative linguistics, and insisted that many myths have arisen from “a disease of
language,” meaning that the original sense of a term has been forgotten, leading to folk
etymologies that are not only fanciful, but even myth-producing. He illustrates this with
the myth of the barnacle goose “reported by sailors and travelers who had seen birds
hatched from shellfish.” To explain this belief Müller found a twelfth century Irish
version of the tale, noting that geese were called Hiberniculae, a name eventually
shortened to Berniculae, which easily becomes Bernacula, and is confused with
‘barnacles.’ As with many schools of thought the solarism and etymologizing of Max
Müller went too far, and despite his fame at the time he became the subject of scholarly
ridicule even in his own day. Today the ideas that he represented are generally viewed as
scholarly, but quaint. However, in rejecting Naturalism as a theory of myth or folklore
there is reason to believe that the baby may have been thrown out with the bath.
The thrust of Lévi-Strauss’s criticism appears to be that widespread mythological motifs
frequently appeal to recurrent processes of Nature, but the intent of the myth-maker is not
to represent or explain natural processes. Rather, the intent of the myth-maker is to
express abstract ideas regarding Man’s relationship to Man and to the natural order by
exploiting the concrete symbolism readily made available by the phenomena of the
natural world. This may be true in some cases, but does it mean that myth and folklore
are never motivated by an anthropomorphic or theriomorphic portrayal of the forces of
Nature? The aim of my argument is to present evidence contrary to Lévi-Strausss that in
at least some cases what folkloric beliefs are trying to represent symbolically, and even to
explain, is nothing more or less than the world of natural phenomena --- particularly
phenomena that can be viewed as etiologically challenging to prescientific minds.
3. Myth and folklore. Myth and folklore are sometimes distinguished for particular
purposes, but there is to my knowledge no basis for distinguishing the types of themes or
motifs that appear in myths from those that appear in what is commonly called ‘folklore.’
Folkore and mythology are frequently treated together in general discussions (as Funk
and Wagnall’s Standard Dictionary of Folklore, Mythology and Legend), and the tacit
view of most folklorists appears to be that similar recurrent themes are found both in the
less structured and more anecdotal medium of folk belief, and in the more purely
narrative domain of myth. For the purpose at hand, then, the motif inventory of folklore
and mythology will be treated as equivalent.
4. Dragons, familiar and otherwise. In dealing with any scientific problem, to attain a
satisfactory explanation of a phenomenon it is first necessary to have an adequate
description of the thing to be explained. Our notions of dragons are often colored by the
way they are represented in our own culture. How do we think of these unseen, yet
somehow familiar creatures? They breathe fire, they guard treasures, and somewhat
oddly they have wings, however vestigial, on bodies they obviously could not lift.
The comparative study of dragons shows both cross-cultural variations and recurrent
themes in the characterization of these mysterious creations of the human mind. While
European dragons have wings (hence implying flight), for example, they have no stated
connection with the weather. Chinese dragons, on the other hand, with their long,
sinuous bodies and powers of flight, are intimately associated with rainfall and the
control of weather in general. European dragons have come to have pagan associations
within the Christian tradition, and so are regarded in many contexts as negative (although
the essential moral ambivalence of dragons is apparent in the genuinely affectionate
treatment they often receive in children’s literature). By contrast, Chinese dragons have
overwhelmingly positive associations as bringers of rainfall, and ultimately as symbols of
imperial authority. In comparative perspective dragons must be seen as a family of
related creatures, or rather creations, which share overlapping characteristics, rather than
a single invariant type. The ubiquitous ‘horned serpent’ of aboriginal North America, like
both European and Chinese dragons, guards springs or other terrestrial water sources, and
has horns, but unlike them it apparently does not fly. On the other hand, the plumed
serpent of Mesoamerica flies and guards springs, but has no horns, while the rainbow
serpent of aboriginal Australia has traits that vary from region to region, but include
horns and an association with menstruation, or more particularly menarche.
Blust (2000:520) summarizes the geographical distribution of traits associated with
dragons in six major regions of the globe: 1. Europe, 2. the Near East, 3. India, 4. the Far
East, 5. Mesoamerica, and 6. North America. To facilitate discussion this information is
repeated with minor changes in Table 1:
(TABLE 1 HERE)
Several of these traits are discussed in Blust (2000), but my purpose in this paper is to
focus on just one, namely k, the statement that dragons have traditionally been regarded
in Europe, the Far East and Mesoamerica as being simultaneously male and female, as
this characteristic is far less commonly mentioned than such traits as the ability to fly, the
possession of horns, whiskers and fiery breath, or the habit of guarding treasures. Given
the seeming arbitrariness of this trait we naturally will want to understand why it exists,
but before seeking explanations our first task is to document the reality of the claims.
5. The sexual ambivalence of dragons, far and wide. To begin in Europe, few classic
descriptions mention the sexuality of the dragon. Indeed, since they are often portrayed
in myth and art as menacing young women in ways that are vaguely erotic, most dragons
would appear to be unambiguously male. The classic ‘damsel in distress’ who is rescued
by a knight in shining armor is commonly shown with a dragon in the background that
must be skewered by the lance of Christianity for the tale to end well.
However, there is a second context in which dragons appear in the European tradition,
and that is medieval alchemy. The psychologist Carl Jung (1967, 1968) has written, in
particular, about dragon symbolism in the alchemical process of transforming base metals
into (philosophical) gold. Regarding Mercurius, the symbol of the philosophical union of
opposites, he says (1967:217-18):
The two substances of Mercurius are thought of as dissimilar, sometimes opposed; as the
dragon he is “winged and wingless.” A parable says “On the mountain lies an ever-
waking dragon called Pantophthalmos, for he is covered with eyes on both sides of his
body, before and behind, and he sleeps with some open and some closed.” There is the
“common and the philosophic Mercurius”; he consists of the dry and earthy, the moist
and viscous. Two of his elements are passive, earth and water, and two active, air and
fire. He is both good and evil........Because of his united double nature Mercurius is
described as hermaphroditic. Sometimes his body is said to be masculine and his soul
feminine, sometimes the reverse.
Hogarth and Cleary (1979:130) illustrate the philosophical concept of the conjunction of
opposites in much the same way, with a painting of a semi-anthropomorphic dragon that
has sprouted two human heads from its tail, one male, the other female, and is further
topped by a Janus-like king-queen figure joined in one body. It is notable that in
European dragon traditions this ambivalent sexuality surfaces only in the context of the
alchemical quest for gold (precious knowledge), where it can be seen as a special
application of the more general observation that dragons embody a fusion of opposed
traits (reptilian body with feathers or hair, a creature of the sky and at the same time of
earthly springs or waterfalls, massive bulk with sometimes tiny wings, etc.). Given these
traits it is not difficult to see how the dragon could serve well as a concrete representation
of the philosophical union of opposites. But this still leaves us with an unanswered
question: just because dragons have traits that appear to be chimerical, why must they
also be conceived as bisexual?
The foregoing question would be puzzling enough if this seemingly fanciful feature was
restricted to a single, historically unitary cultural tradition (as we can perhaps assume for
Europe). However, it becomes much more pressing (and intriguing) when it is also found
in cultural traditions that we can reasonably assume to have been historically independent
of European alchemy.
Commenting on the Chinese concept of the dragon, which shows every indication of
having developed in isolation from the similar idea in Europe, the same writers note that
in Taoist metaphysics the universe is governed by the interaction of opposed but united
male (Yang), and female (Yin) principles, and a propos of the dragon they comment
Whether dragons derived from the Yang principle or from both was the subject of much
learned debate in which the number of scales on a dragon was of great significance ....
Some experts insisted that the sacles of a true dragon numbered exactly eighty-one,
equalling nine times nine. According to Chinese philosophy, the number nine is Yang ....
Other experts, arguing that dragons were not purely Yang but a combination of the
qualities of Yang and Yin, put the number of scales at 117, made up of 81 imbued with
Yang and 36 (six times six) with Yin. 1
The details about the sexuality of dragons in European alchemy and Taoist metaphysics
are, of course, different, but it would be hard to deny that they share a common element
1 Hogarth and Cleary do not cite a primary source, but Victoria Yen-hsin Chen has drawn
my attention to some short passages in the Zhuangzi that discuss the nature of the dragon
and point in the same direction. The most relevant of these passages is the following:
孔孔 孔 '孔 孔 孔 孔孔孔 孔 孔！
Confucious say I PART today finally see dragon
孔孔 孔孔 孔 孔孔 孔孔 孔 孔孔 孔 孔孔 孔 孔孔 孔 孔.
dragon gather form entity spread form essay ride clouds and form from yin
Translation: 'Confucious said: Now I finally see a dragon. Dragons, they can gather into
shape and disappear; they ride on clouds and are formed by the yin and the yang (....).
in portraying the dragon as at once male and female, and this convergence of beliefs in
two widely-separated folkloric traditions is bound to be startling when first encountered.
It is well to take a moment to consider why this would be our normal reaction. Dragons
are, afterall, creatures of the imagination, and in contemplating the imagination we are
apt to conceive of this aspect of human thought as unfettered, and its products as
consequently not motivated by anything in the real world of sense impressions. So what
would compel human beings who had no close contact with one another for millennia to
produce supposedly arbitrary folkloric results that turn out to be remarkably convergent?
It is these kinds of widely distributed, but seemingly arbitrary culture traits that have fed
into radical diffusionist theories of all kinds, including theories of the dragon, in earlier
times (Smith 1919). For those with a different mental disposition the same observations
have led to a kind of philosophical fatalism, perhaps best represented in the oft-quoted
remark of Jorge Luis Borges (1967:7) “We are as ignorant of the meaning of the dragon
as we are of the meaning of the universe.” 2 However, from the standpoint of general
scientific method, such convergence should be taken as an indication that the traits in
question are not as arbitrary as they might initially seem. Rather, if they have arisen
2 The full quote in the original edition reads: “Ignoramos el sentido del dragón, como
ignoramos el sentido del universo, pero algo hay en su imagen que concuerda con la
imaginación de los hombres. y así el dragón en distintas latitudes y edades.” The English
translation of Hurley (Borges 2005:xii) renders this as “We do not know what the dragon
means, just as we do not know the meaning of the universe, but there is something in the
image of the dragon that is congenial to man’s imagination, and thus the dragon arises in
many latitudes and ages.”
independently in two different regions of the world, the probability is high that they are
motivated by some shared feature of human perception or psychology (or both).
Or could this agreement in the conception of dragons as bisexual both in Europe and in
China simply be a fluke --- a random product of chance? Given only two witnesses this
possibility cannot be completely ruled out, despite the low probability that it is true. But
again, basic scientific method gives us the guideposts we need to answer this question:
the surest way to eliminate chance as an explanation for agreements in cross-cultural
comparison is to multiply the number of witnesses that support a given inference. What
that means in this case is that we search for a third historically independent cultural
tradition in which the local version of a dragon is also conceived as [+male], [+female].
There are, in fact, at least two other known cultural traditions in which this requirement is
met. The first is the Maya of Yucatan and Guatemala.
In a generalist account of dragon beliefs, Huxley (1979:9) compares Varuna, the Hindu
god of water and the celestial sea, with Itzam Na, the celestial iguana of the Maya, “Itzam
meaning iguana and Na, house or woman --- whose name also has to do with milk, dew,
wax, resin and sap. Itzam Na is bisexual, the male principle being in the sky ‘in the midst
of the waves’ while his consort is the unfaithful Earth, goddess of weaving and painting,
whose moon-lover yearly emasculates her spouse.” Huxley gives no source for this
statement, but he presumably drew, at least in part, on Spinden (1957), and Thompson
(1970), who themselves rely heavily on early Spanish sources, and the comtemporary
ethnography of Mayan-speaking peoples. The latter writer (1970:212) observes that
Itzam Na means “Iguana House.” Itzam is defined in the Vienna dictionary as “lagartos
like iguanas of land and water.” Lagarto can mean anything from lizard to crocodile ....
Thirty years ago I .... tentatively identified Itzam Na with the celestial monsters, so
common in Maya art, which are part crocodile, lizard, or snake and may even have deer
features (antlers or cleft hooves). At that time the Vienna dictionary was undiscovered.
Now that we have its entry defining Itzam as “iguana,” the case for that identification is
Elsewhere in the same publication he notes (1970:21) that
Among the Lacandon, Itzam Noh Ku, “Itzam the Great God,” is god of hail, lord of Lake
Pelha in which he dwells, and according to a recent source, lord of crocodiles ....Itzam is
a Pokom deity .... and among the Kekchi of the Alta Verapaz and southern British
Honduras who claim him as both male and female, he is a world directional mountain
The evidence, then, points to a reptilian creature that lives in the sky, or that perhaps
alternates between sky and earth, where he may dwell in terrestrial waters. Interestingly,
Thompson further points out that in one of his various aspects (that of Itzam Na T’ul) he
“could withhold good rains”, a trait attribed to dragons in China and various other parts
of the world (Table 1, point a)).
One could, of course, quibble with the definition of ‘dragon’ across distinct cultural
traditions. While no one has any doubt about the classification of European and Chinese
dragons as members of the same mythological category, despite differences of detail in
both physical appearance and behavior, purists may wish to exclude dragon-like creatures
in other parts of the world from the same category of imaginary beasts. However, as
more information is collected it becomes clear that, despite differences of detail, there is
also a substantial body of seemingly arbitrary physical and behavioral traits that links
dragon-like beasts as members of a common type, distinct from all others (cf. Table 1).
Moreover, writers from Smith (1919) to Huxley (1979) have not hesitated in including at
least the Plumed Serpent of Mesoamerica, and the Horned Water Serpent of North
America as members of the dragon category. On the face of it, then, we have now
established that European, Chinese and Mesoamerican dragons all are conceived in at
least some contexts, as being simultaneously male and female.
Before venturing into deeper waters there is a fourth candidate for a dragon that is
described as bisexual. The rainbow serpent of Australia was first introduced to a wide
reading public through the pioneering ethnography of the British social anthropologist
A.R. Radcliffe-Brown (1926, 1930). Since then it has become known as one of the
defining features of many, if not most Australian aboriginal cultures. In a cameo piece
Mercatante (1988:546) describes it as
a gigantic snake whose body arches across the sky as the rainbow ... Known as Taipan
among the Wikmunkan people, he is associated with the gift of blood to humankind,
controlling the circulation of blood, as well as the menstrual cycle of women ... Medicine
men and rain makers invoke Taipan by using quartz crystals and sea shells in their rituals.
Called Julunggul among the people of eastern Arnhem Land, the Rainbow Snake is
believed to swallow young boys andl later vomit them up. This is symbolic of their
rebirth, or the transition from youth to manhood. Known as Kunmanggur in a myth told
by the Murinbata of the Northern Territory to W. Stanner, the Rainbow Snake is either
bisexual or a woman. Sometimes he is described as a male but is portrayed with female
By this point a pattern has clearly emerged: dragon-like creatures that dwell (at least part
of the time) in the sky, and are associated with the rains, are conceived in widely
separated parts of the world, and in vastly different cultural traditions as simultaneously
male and female. It goes without saying that this many independent indications of a
seemingly arbitrary trait in a creature that is thought by many to be an invention of pure
imagination leaves little choice but to assume that there is a natural basis for the belief
that dragons are bisexual. But if dragons do not exist in the real world, what basis could
such a belief possibly have?
6. Beyond dragons --- and into the real world. Before proceeding it will be helpful to
remind the reader how I am using the term ‘Naturalism’, since it has radically different
meanings in philosophy, as against folkore or cultural anthropology. My use of the term
is closely similar to that of the Sanskritist, classical philologist and folklorist Friedrich
Max Müller over a century ago. Max Müller was convinced that many myths that feature
human or animal figures were actually inspired by direct observation of natural events in
the heavens (sun, stars, constellations, winds, etc.), rather than being the brainchild of
some highly creative thinker detached from the real world.
As noted already, the type of interpretation that Max Müller defended, namely that myths
with anthropomorphic or theriomorphic figures, were descriptions of natural events in the
heavens that were important in the cultures of the mythmakers, has been criticized by
later writers. In particular, the anthropologist Claude Lévi-Strauss took issue with the
claim that the intent of the mythmaker is to portray or explain natural processes, and in
general it is fair to say that Max Müller is thought of today as a scholar who represented a
particular school of thought that is no longer supported by a majority of scholars.
What has often happened when the ideas of earlier scholars are rejected as having gone
too far in the pursuit of a particular approach to data, is the phenomenon of ‘throwing out
the baby with the bathwater’: scholarly excesses are used as an excuse for denying the
usefulness of anything that was claimed within a given school. However, we know that
this kind of Hegelian dialectic in the discourse of science has sometimes led to scholars
discarding valid and potentially useful ideas as part of the larger theoretical package that
was considered either erroneous or uninteresting. The understanding of mana --- the
impersonal supernatural force associated with chiefs and other high-status individuals in
Polynesia and other parts of the Pacific --- is a case in point. Scholars since Codrington
(1891) have described the functioning of the mana concept in contemporary societies, but
without any interest in how the concept itself might have originated. However, when the
full range of meanings of this term is collected, it becomes clear that in a number of
widely-separated Austronesian languages mana refers not only to the impersonal force
associated with individuals of high hereditary rank, but also to thunder and violent storm
winds (Blust 2007). In considering the possible meaning of this term in the past, and its
likely paths of change, it is safe on various grounds to rule out the possibility that it
originally referred only to an impersonal supernatural force possessed by people that was
then transferred to dramatic and sometimes frightening natural events. Nor is it likely
that the term in its original function referred both to powerful natural forces and to the
power of a high-status individual. Rather, the most plausible semantic evolution of the
term mana is that it began with exclusive reference to intimidating or awe-inspiring
forces of Nature that were then attributed to human personalities as a concrete
manifestation of their sociocultural power.
With this much in mind, it will be useful to return to the fundamental question that this
paper is intended to answer, namely ‘why are dragons portrayed in widely separated
cultural traditions around the world as being simultaneously male and female?’ A good
place to start is perhaps where we left off near the end of the previous section ---- with
the rainbow serpent of Australia.
Although aboriginal Australia is not included in Table 1, it is clear that the rainbow
serpent must be regarded as a type of dragon since, like other members of the dragon
category, it is a guardian of springs or other bodies of water (b), is capable of flight (c),
appears where sun and rain are closely interspersed (d), is colorful (m), terrifies young
women (o), resides in waterfalls (aa), and to our present point, is portrayed as combining
male and female features in a single body (k). Most important of all with regard to
finding an answer in Nature for why dragons are conceived as bisexual, the rainbow
serpent is identical with the rainbow itself.
It is important to digress briefly to justify this claim, since no description of the rainbow
serpent has ever, to my knowledge, called it a dragon, and the treatment of the topics
‘dragon’ and ‘rainbow serpent’ in standard encyclopedias of folklore and mythology such
as Leach and Fried (1984) do not cross-reference the terms, or show any inclination to
regard them as referring to the same mythic entity.
In examining the ethnology of the rainbow it takes little time to discover that the
conception of this phenomenon in the Judaeo-Christian tradition is not at all typical of the
world’s preliterate cultures. In these the rainbow is commonly portrayed as a bridge
between heaven and earth, as a weapon or other accoutrement of a divinity, or most
typically, as an enormous spirit serpent that drinks water from a river or spring and sprays
it out to cause the rain, or that drinks up the rain and causes it to cease. Rather than being
an object of beauty or a sign of promise, in most traditional cultures the rainbow is feared
(along with sunshowers), leading to a globally-distributed taboo against pointing at it
with the index finger, lest the finger be permanently bent into the shape of the rainbow,
become infected, severed, etc. (Blust, 1999, n.d.). Above all, despite its natural origin,
the rainbow is traditionally seen as a supernatural presence to be respected, and in some
ways avoided, since it is dangerous to all except powerful shamans.
To understand this conception better a brief thought-experiment might be helpful.
Imagine yourself living in the Palaeolithic, with no literacy or knowledge of scientific
principles, but with a quick and inquisitive mind. If you are a man, for much of the time
you probably are out hunting, and if you are a woman, you probably are out gathering
useful plants. The sky darkens, and with a dramatic prologue of thunder and lightning
rain begins to pummel the earth. You take cover to wait it out, and in due course the
clouds part, and the sun shines through, across the lessened rainfall that continues to fall.
At this moment an enormous colorful arc appears out of nowhere, reaching from horizon
to horizon while fire (sun) and water (rain) compete for control of the sky. What is it?
Where was it before it suddenly appeared high above the earth, and where will it go when
it disappears? These are all natural questions for any intelligent being to ask, and there
can be little doubt that Palaeolithic humans asked them of themselves and of one another.
As already noted with reference to conceptions of the rainbow in tribal societies, the
answer that was usually given is this: the rainbow is an enormous serpent snake
(suggested by its elongated, colorful body). When it appears in the sky it is either
causing the rain to fall by drinking from a terrestrial water source and spewing it out as
rain, or it is causing the rain to cease by drinking it up (rainbows cannot appear in heavy
storms or clear skies, but only when sun and rain are in competition for control of the
sky). Where does it go when it suddenly disappears, or whenever it is not present in the
firmament? That is when it dwells in springs or waterholes, acting as a guardian of this
precious resource. It can hardly escape notice that this is exactly what dragons do, and
the resemblance between rainbows and dragons does not stop there.
Describing the surprising similarity of the rainbow serpent myth across aboriginal
Australia, Radcliffe-Brown (1930:343) pointed out that “the rainbow-serpent lives in
deep permanent lagoons and waterholes. In the New England tableland it is particularly
associated with waterfalls, possibly because at such places rainbows may frequently be
seen.” Each of these well-documented features of the rainbow serpent is matched closely
by accounts of more familiar dragons that have no known connection with the rainbow.
The rainbow serpent lives in waterholes when it is not in the sky ---- this provides an
answer to the basic question ‘Where is it where it is not visible in the heavens?’. But
dragons also guard springs in widely-separated parts of the world, and every major
waterfall for which adequate data has been collected, has its resident dragon.
In a well-known Greek myth King Cadmus, founder of the city of Thebes, sent his
companions to fetch water from the Spring of Ares, but when they arrived they found that
the spring was guarded by a fierce dragon which killed most of them before the king
himself returned to seek revenge against the monster. According to Huxley (1979:5),
quoting an earlier source, not much over a century ago rural Macedonians in the Balkans
spoke of horned serpents that guarded the “dragon springs” (wells) of their countryside.
Very similar accounts of ‘dragon-rearing wells’ are reported in China by de Visser
(1913:63), who states that “Nobody dared draw water from this well, because if one did
so, strange things happened, and the person who had ventured to thus arouse the dragon’s
anger fell ill.” Similarly, the Horned Water Serpent of North America and the Plumed
Serpent of Mesoamerica and parts of the American Southwest are well-known guardians
of springs. To cite only one of many possible accounts, Spicer (1980:64) states that
among the Yaqui of northwest Mexico respected spiritual powers were connected with
certain springs, where “snakes with rainbows on their foreheads lived and swam in the
water.” It is clear, then, that European and Chinese dragons, and the Horned Water
Serpent and Plumed Serpent of the Americas, which are commonly treated as types of
dragons, are depicted in various accounts as dangerous guardians of springs.
The association of rainbows with waterfalls is due to basic physics: water crashing
against the rocks or river below sends spray high into the air, and sunlight shining
through it refracts light in the same way that is seen when sunlight shines through
raindrops. We can therefore expect any major waterfall (one that sends significant spray
high into the air) to be a generator of rainbows on sunny days. Given this observation, it
is of no small interest that every major waterfall for which data has been collected has its
resident dragon. The Horned Water Serpent of Niagara, as conceived in the belief system
of the local Seneca Indians, has been thoroughly documented by several writers,
beginning with Lewis Henry Morgan (1851). Frazer (1922.2:156) observed that “The
Oyampi Indians of French Guiana imagine that each waterfall has a guardian in the shape
of a monstrous snake, who lies hidden under the eddy of the cascade, but has sometimes
been seen to lift up its huge head.” Shortly after this passage, with reference to a very
similar belief in southern Africa, he adds “in Basutoland the rivers Ketane and
Maletsunyane tumble, with a roar of waters and a cloud of irridescent spray, into vast
chasms hundreds of feet deep. The Basuto fear to approach the foot of these huge falls,
for they think that a spirit in the form of a gigantic snake haunts the seething cauldron
which receives the falling water.” Finally, a concerted effort to discover whether the falls
at Iguassu, on the Brazilian-Argentinian border, have a resident dragon, turned up a
positive result in the legend of Taroba and Naipi. According to Antonio Andres-Lopez,
who collected the Spanish text, the local Guarani Indians tell a story of how the Iguassu
falls were formed. From the earliest times the Iguassu river was inhabited by an
enormous and monstrous serpent which demanded that a beautiful young woman be
sacrificed annually to appease him. One year a young priest named Taroba had the duty
of delivering the victim, named Naipi, to the monster, but before doing so he fell in love
with her, and so refused to carry out his duty. Mboi, the monster (said to mean ‘viper’ in
the local form of Guarani) was driven into a fury, whipping his body about and dividing
the course of the river to form the cataracts seen today. It is also clear, then, that
wherever waterfalls generate rainbows, dragons appear.3
Without entering into further details, then, it should be clear that the rainbow serpent of
aboriginal Australia and the more familiar dragons of Europe and the Far East are
members of the same category, and that both have arisen from preliterate attempts to
understand weather phenomena of the natural world.
7. Why rainbows are bisexual. We began this inquiry with a simple question ‘Why, in
widely-separated cultural traditions, are dragons conceived as bisexual?’. Like many
basic questions, this one has turned out to be complex, but the first step has now been
taken, namely to establish the cognitive equivalence in tribal societies of dragons with
rainbows. The mere fact that dragons and rainbows are radically separate concepts in the
Judeo-Christian tradition should not blind us to the clear connections they have in many
tribal societies up to the present time. Rather, the concept of the rainbow serpent, which
3 Given its similarity to the classical Greek story of Theseus and the Minotaur, it is
possible that the legend of Taroba and Naipi shows postcontact influence from the
dominant Spanish culture. However, this cannot explain why the tale is designed to
explain the origin of the Iguassu Falls --- very much like the Seneca story of the origin of
the Horseshoe Falls at Niagara --- and most particularly, why its central character is a
gigantic serpent that lives in the spray of the cataract.
is still plainly visible in traditional Australian aboriginal belief systems, and only lightly
disguised in others, has been transformed in societies with a longer tradition of literacy
by a conceptual separation of the rainbow (with its generally positive associations) from
the serpent (with its generally negative ones). Within the European tradition this
separation appears to be complete, but in China there is an interesting split between
courtly and folk traditions: in the courtly tradition, where a five-toed dragon became the
symbol of the emperor, dragons and rainbows have little clear connection, but in the folk
tradition the rainbow serpent motif remains vividly present. According to a personal
communication from Professor Lo Chin-tang, formerly of the Department of East Asian
Languages at the University of Hawai’i, the common people of Lanzhou in Gansu
province see the rainbow as an immense dragon that drinks water from the sea and sprays
it out as rain. Remarkably, then, the folk tradition in at least some parts of China more
closely resembles that of culturally simple tribal peoples in other parts of the world than
it does the rich and elaborate courtly tradition in its own culture. 4
But we haven’t finished our search. If the idea of dragons evolved from preliterate
conceptions of the rainbow, how did the notion of sexuality arise with dragons, since
rainbows are inanimate, and obviously have no sexual identity? Again, we need to
remind ourselves that human perceptions of nature are mediated by culture, and that
4 Victoria Yen-hsin Chen has further drawn my attention to the Mengxi Bitan, a
collection of personal anecdotes written by the Song dynasty scholar Shen Gua (1031-
1095) between 1086 and 1093, in one of which he notes that “People used to spread the
saying that the rainbow can enter the river and drink from it,” confirming the veracity of
this claim by stating that he himself has seen the end of a rainbow reaching into a stream.
cultural interpretations involve projections from the human realm onto that of the natural
world. However, before looking at cultural interpretations we must return to the basics of
the rainbow itself.
Any account of the optics that produce rainbows will explain that rainbows naturally
occur double, the primary arc being lower and brighter, and the secondary arc higher,
fainter, and with the color pattern of the spectrum reversed. To a physicist the two arcs of
the rainbow are primary and secondary, but to tribal peoples accustomed to animating
nature with properties that govern their own social organization and lives, it would be
surprising if they did not characterize this feature of the rainbow in a more personalized
way. This kind of information is not often collected or reported, even in the most detailed
ethographies, but bits and pieces have been noted in enough widely-separated societies to
make it clear that the two arcs are often called the ‘male’ and ‘female’ arcs in cultures
spanning the planet.
Among Austronesian-speaking peoples Malays call the rainbow pelangi (although in the
closely-related Minangkabau it is ular minum = ‘snake drinking’). However, a double
rainbow is called pelangi sekelamin, where se- is a prefixal form of ‘one’ and kelamin is
‘married couple’. To traditional Malays, then, the double rainbow was a married pair,
male and female (Skeat 1900: 15, fn. 2). Among Palauan speakers in western Micronesia
the clear arc is said to be female, and the indistinct one male (Sandra Chung, p.c.). On
the other side of the Pacific Ocean the Totonac of Mexico describe the rainbow as
simultaneously male and female (Ichon 1969:137), the Chibchan-speaking Cuna of
Panama call the brighter arc male and the fainter one female (Nordenskiöld 1938:394),
and the Quechuan-speaking Inga of Colombia say that a double rainbow is both male and
female (S.H. Levinsohn, p.c.). Across the Atlantic Ocean among the Hausa of Nigeria,
according to Tremearne 1968:340) “Gajjimare [rainbow] is in shape something like a
snake, but is hermaphrodite, or at least double gendered, the male part being red, the
female blue.” Elsewhere in Africa de Heusch (1982:37) notes that among a whole array
of Bantu-speaking peoples”the rainbow effectively embodies a contradiction: at once
male and female, it unites fire and water, high and low.” Among the Venda this
contradiction is expressed by the opposition of masculine water to feminine fire.
Finally, although general English-language treatments of the rainbow in China are hard to
find, it turns out that writers of the Chinese classics repeatedly over a period of many
centuries refer to the two arcs of the rainbow as ‘male’ and ‘female’. I am very much
indebted to Victoria Yen-hsin Chen for bringing these sources to my attention, and for
helping me understand some of the details. Briefly, they can be summarized as follows:
In what is perhaps the earliest known source, the Erya 爾爾, described by Karlgren
(1931:46) as “a collection of direct glosses to concrete passages in ancient texts” and by
others as a dictionary, glossary, thesaurus or encyclopedia, it is stated that the rainbow has
both male (hung 爾) and female (ni 爾) arcs. The dating of this source is problematic, as
glosses evidently were added at different times over a period of several centuries, but
Karlgren states that most of them are securely dated to the 3rd century BC.
In the Book of Han 爾爾, authored by Ban Gu, who lived from 32-92 AD there is a
description of the rainbow which notes its two arcs, commenting that the one with salient
color is the male arc (called hung 爾), and the one with light color is the female arc (called
ni 爾). Next, in what is perhaps the most important work of literary scholarship during the
Han dynasty, the Shuowen Jiezi 爾爾爾爾, compiled by the scholar Xu Shen in 100 AD and
belatedly presented by his son X Chōng to the emperor An of Han in 121 AD, contains ǔ
the observation that “Ni, is a bent rainbow, with red, green, or white color. It is the yin 爾.
” Furthermore, toward the end of the Eastern Han dynasty the Yueling zhangju 爾爾爾爾,
authored by Tsai Yung (133-192 AD) comments that the rainbow is a ‘type of worm’, the
primary arc being male and the secondary arc female. Considerably later, during the
Song dynasty (960-1279 AD) the Guangyun 爾爾, authored by Pang-nian Chen and others,
states that the secondary rainbow is the female rainbow.
Attention to the Chinese classics over a period of more than a millennium, then,
confirms that not only was the dragon conceived as either yin 爾 or yang 爾, or as both, but
that the rainbow was also represented in its natural form as displaying two arcs, a primary
(brighter) one that was called the ‘male’ (hung 爾) arc, and a secondary (fainter) one that
was called the ‘female’ (ni 爾) arc.
The assignment of genders to celestial objects should not come as a surprise, since
European languages commonly assign opposite grammatical genders to the sun and
moon, and among many tribal peoples the sun and moon are conceived in various ways
as being one sex or the other, but never the same one (Lévi-Strauss 1976). However,
most cultures stop with the sun and moon, leaving the stars, rainbows and other celestial
phenomena unclassified. A notable exception is the Panare Indians of Venezuelan
Guiana, where Dumont (1979:25) recorded the following attributions of sexuality to
celestial bodies (with plus and minus values marked for “male” and “female”: sun (+ --),
moon (-- +), stars (-- --), Milky Way (-- --), rainbow (+ +). In other words, while the sun
and moon are regarded as male and female respectively, the stars and Milky Way are
regarded as being neither male nor female, and the rainbow is considered bisexual.
8. Conclusions. The question of why dragons are conceived in widely separated cultural
traditions as being simultaneously male and female is initially a baffling one. However,
once a thorough study of the ethnology of the rainbow is undertaken, it becomes clear
that the idea of the dragon must have evolved from the once pan-human conception of the
rainbow as an enormous snake that drinks water from terrestrial sources and spews it out
as rain, or that drinks up the rain and causes it to stop. The rainbow serpent continued its
existence with only minor changes over much of aboriginal Australia, and in a somewhat
modified form among Bantu-speaking peoples in central Africa, but traces of its former
existence are found over a much broader band of humanity.
Despite whatever shortcomings the ‘physical religion’ or Naturalism of Friedrich Max
Müller may have had over a century ago, the problem addressed here and its solution
show without any doubt that some globally-distributed features of culture that at first
appear totally arbitrary, are in fact motivated by careful observation of the natural world.
So-called ‘primitive man’ may have lacked the scientific tools to understand the physical
mechanisms behind the natural phenomena he observed, but his observations were
accurate, and his explanations understandable within the animistic world in which he
lived. Given this perspective the famous quote by Borges that “We are as ignorant of the
meaning of the dragon as we are of the meaning of the universe” must be seen as unduly
pessimistic. While the meaning of the universe may lie forever beyond human
comprehension, the meaning of the dragon has yielded to careful scientific inquiry, and in
the process has shed light on an important chapter in the history of human thought.
1999 The fox's wedding. Anthropos 94: 487-499.
2000 The origin of dragons. Anthropos 95: 519-536.
2007 Proto-Oceanic *mana revisited. Oceanic Linguistics 46: 404-423.
n.d. Pointing, rainbows, and the archaeology of mind. Ms., 42pp.
Borges, Jorge Luis
1967 El libro de los seres imaginarios, with the assistance of Margarita Guerrero
(original 1957). Buenos Aires: Emecé.
2005 The book of imaginary beings. Trans. by Andrew Hurley. New York: Viking.
1972 The Melanesians: studies in their anthropology and folk lore (original 1891). New
York: Dover Publications.
Craig, William Lane, and J.P. Moreland (eds.)
2000 Naturalism: a critical analysis. London and New York: Routledge.
de Heusch, Luc
1982 The drunken king, or, the origin of the state. Trans. and annotated by Roy Willis
(original 1972). Bloomington: Indiana University Press.
de Visser, Marinus Willem
1913 The dragon in China and Japan. Amsterdam: J. Müller.
1979 Not in ourselves, but in our stars. In D.L. Browman and R.A. Schwarz (eds.)
Spirits, shamans and stars. Perspectives from South America, pp. 241-254. The
Frazer, James George
1922 The golden bough: a study in magic and religion (12 vols.). London: Macmillan
Hogarth, Peter J., and Val Clery
1979 Dragons. New York: Viking.
1979 The dragon: nature of spirit, spirit of nature. New York: Collier Books.
1969 La religion des Totonaques de la Sierra. Paris: Centre National de la Recherche
1967 Alchemical studies. (Bollingen Series 20). Trans. by R.F.C. Hull. Princeton,
New Jersey: Princeton University Press.
1968 Psychology and alchemy. (Bollingen Series 20, 2nd ed.; original 1953). Trans. by
R.F.C. Hull. New York: Pantheon Books
1931 The Early History of the Chou Li and Tso Chuan Texts. Bulletin of the Museum
of Far Eastern Antiquities 3: 1–59.
Leach, Maria, and Jerome Fried (eds.).
1984 Funk and Wagnall’s standard dictionary of folklore, mythology, and legend (2
vols.). San Francisco: HarperCollins.
1972 The savage mind. Chicago: The University of Chicago Press.
1976 The sex of the sun and moon, In Claude Lévi-Strauss, Structural anthropology,
vol. II, pp. 211-221. New York: Basic Books, Inc.
1875 Wald und Feldkulte. Berlin: Gebrüder Borntraeger.
Max Müller, Friedrich
1891 Physical religion. The Gifford Lectures delivered before the University of
Glasgow in 1890. London: Longmans, Green and Co.
1988 The Facts on File encylopedia of world mythology and legend. New York: Facts
Morgan, Lewis Henry
1954 League of the Ho-de’-no-sau-nee, or Iroquois (original 1851). New Haven:
Human Relations Area Files.
1938 A historical and ethnological survey of the Cuna Indians of Panama. Göteborg:
1926 The Rainbow-Serpent myth of Australia. Journal of the Royal Anthropological
Institute of Great Britain and Ireland 56: 19-25.
1930 The Rainbow-Serpent myth in South-East Australia. Oceania 1: 342-347
Skeat, Walter William
1900 Malay magic, being an introduction to the folklore and \popular religion of the
Malay peninsula. London: Macmillan and Co.
Smith, Grafton Elliot
1919 The evolution of the dragon. Manchester: Manchester University Press/London
and New York: Longmans, Green & Co.
Spicer, Edward H.
1980 The Yaquis: a cultural history. Tucson: The University of Arizona Press.
Spinden, Herbert J.
1957 Maya art and civilization. Indian Hills, Colorado: Falcon’s Wing Press.
Thompson, J. Eric S.
1970. Maya history and religion. Norman: University of Oklahoma Press.
1968 The Ban of the Bori. Demons and demon-dancing in west and north Africa
(original 1914). London: Frank Cass and Co.