We have now seen enough data to draw some firm conclusions. Chapters 5 and 7 can each stand alone as studies of particular phenomena—in one case the globally distributed dragon idea, and in the other, globally distributed beliefs about the rainbow. The purpose of this chapter is to align the two sets of observations to show that many of the traits that are attributed to dragons in widely separated parts of the Earth are also attributed to rainbows, and that the number of such correspondences clearly eliminates chance as an explanation for the agreements.

In some cases, these traits have a transparent physical explanation, as where both dragons and rainbows are considered to be controllers of the weather, while in others they do not, as where both are considered to be offended by menstruation. However, where agreement exists it is evidence of common origin, whether the trait has a transparent physical explanation or not. Moreover, many of these traits are shared not only in cultures in which the dragon and the rainbow are all but indistinguishable (as in aboriginal Australia, parts of Southeast Asia and South America, and much of Africa), but also in cultures (as those of Europe, India, China or aboriginal North America) in which the dragon has no generally accepted connection with the rainbow. In many ways this is reminiscent of Tylor’s long-abandoned, but certainly not useless theory of ‘survivals’—that is, cultural practices which no longer have a conceptual underpinning, but once did. Like many of the examples Tylor cited in defense of his proposal, dragons show their heritage through the presence of traits that can only plausibly be derived from earlier beliefs about the rainbow, regardless of the present state of the cultures, and in this sense, they may be seen as survivals from a time when dragons were universally conceived as the Rainbow Serpent (Blust 2023 to appear). Finally, because of their inherent interest, the ethnology of the rainbow in Chapter 7 also included traits that have no known connection to the dragon, but for obvious reasons these are ignored here.

This book begins with two primary divisions for a reason. Most readers will come from cultures in which the connection between the dragon and the rainbow is far from obvious. The dragon is a creature of folklore and mythology—to those of European cultural ancestry it is an abhorrent beast that preys on the human community, especially young women, and so must be killed, while to those of East Asian cultural ancestry it is a beneficent creature that brings the fertilizing rains, although sometimes only under compulsion, and that commanded sufficient respect to become a symbol of the emperor of China. The rainbow, on the other hand, is real rather than mythological, a thing of beauty suitable for a place in poetry or a symbol of divine promise rather than a monster to be avoided or destroyed. From this perspective few things could be more different. But once we incorporate attitudes about the rainbow from a wide sample of the world’s peoples, we see a very different picture.

To some readers it may be apparent that I have found it difficult to keep Parts I and II separate. While it is possible to discuss the dragon without reference to the rainbow in Europe, the Near East, North America, and to a large extent in the Pacific, and Central and East Asia, it is much more difficult to do this in mainland and island Southeast Asia, New Guinea, Central and South America, Africa, and particularly Australia, where the two blend into one so often that a distinction between them seems pointless. That others have had a similar problem can be seen in e.g. Benedict (1975:274), where a single entry is given for ‘dragon’ and ‘rainbow’.

Table 2 (on facing page) summarizes agreements between globally distributed traits of dragons, and similar traits attributed to rainbows. In the interest of systematic presentation I adhere to the outline in Chapters 5 and 7, using numbers to indicate the areas where the trait is documented, as follows: 1. Europe, 2. the ancient Near East, 3. South Asia, 4. Central and East Asia, 5. North America and Mexico, 6. Central and South America, 7. Mainland Southeast Asia, 8. Insular Southeast Asia, 9. New Guinea and satellite islands, 10. Australia, 11. The Pacific islands, starting with the Solomons, 12. Africa.

What does Table 2 tell us? Most fundamentally, it tells us that many traits that have been widely reported for dragons are also reported for rainbows, and there is no plausible way to explain these multiple agreements as a product of chance. The simplest explanation for the correspondences in Table 2 is that at some point in the history of the dragon idea what is unambiguously considered a dragon today must have been a Rainbow Serpent, and hence a rainbow. The second thing that Table 2 tells us is that about half of the traits of dragons that appear to be whimsical inventions by myth-makers are transparently derived from attempts to explain the causality and nature of the rainbow. Although they have already been discussed separately in Chapters 5 and 7, traits 1–26 will now be considered together to show that many traits of the dragon that have puzzled scholars for over a century must be due to its origin as the Rainbow Serpent.

1. Controller of rainfall. As already shown, the idea that dragons have the power to give or withhold rain is found in at least the folk traditions of Europe, the ancient Near East, India, East Asia, and Central and South America, with one questionable case from North America. If this trait were arbitrary, why would it be so widespread? And if the dragon was inspired by living creatures, mammalian memories from the Age of dinosaurs, line dancing, breaking into burial mounds, or other clever but unenlightening proposals, why would it include the ability to control rainfall?

The answer to this question is found in Table 2: rainbows appear only at that tipping point when the rain is about to stop, but hasn’t yet done so. They can thus be visualized as either bringing the rain because they cannot appear until it falls, or as preventing the rain because they cannot appear until the sun breaks through the clouds. For this reason the rainbow is widely seen as a giant snake that either drinks water from a terrestrial source and sprays it out to make the rain, or that drinks the rain from the sky and makes it stop. The ambivalent attitude toward the rainbow in relation to rainfall, even among closely-related peoples, is thus easily understood, and at the same time provides a ready explanation for the often-mentioned ambivalence of the dragon as a force of good or evil.

2. Guardian of water sources. The second thing shown in Table 2 is that the dragon is regarded as a guardian of springs, waterholes, rivers, lakes and the like virtually everywhere that it is found, except in the arid regions of the Middle East (since data on this area is limited, this may change if further information becomes available). Here, the idea of a real water reptile may initially seem to carry more weight than it does as a controller of rainfall. However, the best candidate for such a dragon model is the crocodile, or water python, neither of which is present in Europe, most of North America, including Mexico, the Chaco of South America, or the Australian continent apart from the far north. Moreover, dragons guard wells in Europe and China, and these are not places where one would expect to find a crocodile or water python.

Again, the connection of dragons and rainbows here is readily understandable. Preliterate humans had to explain where rainbows went when they were not in the sky. Since rainbows had to drink from a terrestrial source to produce rain, it is hardly surprising that in their terrestrial form as giant snakes or dragons they would be guardians of these water sources.

3. Found in waterfalls. As seen in Table 2, dragons have been reported as residing in waterfalls in North and South America (multiple locations in the latter case), insular Southeast Asia and Africa. In fact, as noted earlier, every major waterfall for which it has been possible to collect folkloric information has its resident dragon. So far as is known, this unexpected association is limited to waterfalls with a significant water volume. Slender falls with little splash presumably have no dragon, but cataracts that produce spray which rises high into the air are, in effect, creating reverse rainfall, and hence the conditions for producing rainbows when the sun shines through the spray. This is clear at Niagara, where one of several cataracts by which the Niagara River plummets to the water below is called ‘Rainbow Falls’, since the rising mist is filled with fleeting rainbows that come and go at the whim of the sun. It is not unexpected, then, that the Seneca Indians whose territory included the falls, believed that a horned water serpent inhabited Cayuga Creek, which flows into the Niagara River. It is, moreover, noteworthy that in this case the local version of a dragon resided above the falls, rather than in a cave at the base, since this is where the rainbows appear—high above the cascade that generates them.

The British social anthropologist A.R. Radcliffe-Brown (1930–1931:343) noted that the Rainbow Serpent of Australia is particularly associated with waterfalls in the New England tableland because of the prevalence of rainbows seen there. In this case the connection of the local version of a dragon with a waterfall is almost superfluous, since the Rainbow Serpent is already seen as a rainbow. But in cases like the dragon of Niagara the association of the horned water serpent with a waterfall must reflect an earlier belief that this mythical creature was a Rainbow Serpent that, for whatever reason, lost its connection with the rainbow.

4. Lives in caves. Although it is not a fixed feature of the dragon idea, the use of caves as dragon dens is fairly common. As seen in Table 2, this feature appears everywhere that unambiguous dragons (i.e., dragons that have no explicit connection with rainbows) are found, except (so far) in India. Caves may, of course, be either wet or dry, and the connection with dragons does not distinguish between them, but the basis for this belief can again be traced to the association of dragons with waterfalls, since natural hydraulic erosion of the cliff base produces caves at many major waterfalls, and rainbows may appear in front of them.

5. Is capable of flight. Why dragons should he capable of flight when they are basically giant snakes is unexplained by any theory that has been proposed to date. But as descendants of the Rainbow Serpent their association with the sky needs no explanation. Rainbows naturally do not fly, but they appear suddenly in the sky and then as quickly disappear, and to the minds of preliterate humans this must have been because they travelled between their watery homes on earth and the realm of the clouds. This is expressed in some traditional societies, as the Murle of the Sudan, who say that the rainbow is a large dragon-like snake which sleeps in a cave when not flying in the sky. We can therefore assume that all rainbows ‘fly’.

Dragons, on the other hand, seem to vary in this trait. Although there is some evidence that the dragons of each of the above geographical regions can fly, there are differences in how robustly attested this trait it. As seen earlier, in Mesoamerica there is debate about whether the plumed serpent is capable of flight, since it has feathers but no wings. In Europe dragons are sometimes portrayed in the air, but far more commonly on the ground (despite their wings). In the ancient Near East the evidence for flight is limited, but persuasive. In India the nāga is normally portrayed as residing in underwater splendor within bejeweled palaces, but there are distinct types of nāgas, and some of these are said to fly. In China and the largely derivative dragon tradition in Japan the dragon is generally seen flying among the clouds, making it the most airborne of all dragons, despite the absence of wings. The one area where flying dragons have not been found is North America, as here the horned water serpent is normally confined to rivers and lakes, and no longer has a clear association with the rainbow. Once the association with their natural source was lost, the need to see dragons as capable of flight could have vanished, although this did not happen everywhere.

6–9. Combines hot/cold, fire/water. As noted many times already, what makes a dragon distinct from all other creatures, both real and imaginary, is its bodily form, which fuses the trunk of a cold-blooded reptile with various embellishments from mammals or birds. In addition to the scales of a reptile this includes horns, hair or feathers, depending on the region in which the dragon is found.

There is an inherent contradiction in this combination of traits, since it could not happen in the real world. However, once again we are reminded that seemingly arbitrary traits of dragons that appear in widely-separated areas must have some natural basis or they could not have a global distribution. This basis is, of course, the nature of the rainbow, which is universally seen as a fusion of hot and cold, or fire and water, since neither the sun alone nor the rain alone can produce it, but when these combine the rainbow appears. Preliterate people may not have had the powerful tools of science that have become available in recent centuries, but they made accurate observations about the natural world, and this is one that was made everywhere.

10. Is opposed to thunder/lightning or the sun. Despite the lack of any obvious reason why dragons should be opposed to thunder/lightning or the sun, Table 2 shows that this belief is found almost everywhere that dragons exist, often taking the specific form of the dragon being slain by a thunderbolt hurled at it by a celestial deity. Again, other theories of the dragon are silent on this trait, as they are on most of the preceding ones. But, given its distribution over several continents, it cannot be ignored.

If dragons were modelled on real creatures, why would they be opposed to thunder and lightning, or in some cases even to the sun? Moreover, if any other theory claims to explain the universality of the dragon idea, how could this well-established feature of its nature be accounted for? Again, once we recognize that the dragon idea has evolved at different rates in different parts of the world from the Rainbow Serpent, and that the latter is almost certainly the first human attempt to understand the nature of the rainbow, everything falls into place. Rainbows simply cannot occur during thunderstorms, since these drown out sunlight; nor can they occur during sunny days with no rain. All over North America the conflict of the horned water serpent with the Thunderbird figures prominently in mythology and folklore, and here the reason has been lost because of the nearly universal separation of the dragon idea from the idea of the Rainbow Serpent, although in some parts of Mexico the connection is still made. In Bantu-speaking central Africa, however, the conflict between the lightning and the rainbow, both of which are usually represented simply as weather phenomena (although the rainbow’s alter ego is the water serpent), is completely transparent, as described in detail by de Heusch (1982).

11. Is androgynous. Virtually no one who has ventured a theory about the origin of dragons seems to be aware that dragons are regarded as androgynous in widely separated cultures. In Europe this is understandable, since the dual gender of the dragon appears only in the tradition of medieval alchemy, although in that context it is robustly attested. But the same trait surfaces again in the Taoist metaphysics of ancient China and in modern Chinese folk tradition, as well as with the plumed serpent of Mesoamerica. No biological model can explain this feature, and in most theories it is simply not mentioned.

Once again, however, we are led back to the rainbow to explain an imputed trait of dragons that otherwise seems completely mysterious. Rainbows naturally occur double (physical treatises on the rainbow speak of n-order rainbows, where n can be up to ten, although most of these are imperceptible to human vision). Given the pan-human tendency in preliterate societies to animate nearly every unusual feature of the natural environment, it is not surprising that people have struck upon the same interpretation independently in widely-separated parts of the Earth, namely that one of the bows is male and the other female. If rainbows are simultaneously male and female, and if dragons evolved from the Rainbow Serpent (which is androgynous in much of Aboriginal Australia), the recurrent theme of double gender for the dragon is seen to have a natural basis.

12. Is colorful/red. There is no reason why dragons should be colorful, or in some reports, red. Yet this trait is reported from Europe and the ancient Near East (where some sources report the dragon as ‘red’), and in China, and North America (where the dragon is described as ‘colorful’, ‘multi-colored’, or ‘colored like the rainbow’). That dragons are sometimes described as ‘multicolored’, or even as showing the colors of the rainbow, should not come as a surprise, if they are derived from the Rainbow Serpent.

The rainbow is, by its nature, ‘colorful’ or ‘multi-colored’, but in Table 2 I have indicated only those areas where the primary color of the rainbow is reported to be red. In most of these cases there is a specific reference to the redness of the rainbow being due to the blood of humans that has risen skyward. While the latter reference is not known to apply to dragons, the red color of dragons in some reports, and the occasional heavy emphasis on red as the basic color of the rainbow provides an additional tie linking independent dragons in places like Europe or North America, with the Rainbow Serpents from which they evolved.

13. Guards a treasure/has a jewel in head. European dragons are particularly known for guarding hoard of gold or other treasure, but they are not unique in this respect. The Indian nāga, or at least some types of nāgas, guard collections of precious jewels, the Chinese dragon is often depicted as pursuing a round object that has been interpreted as a pearl, and both European dragons and North American horned water serpents are sometimes described as carrying a jewel in their foreheads.

At the same time, cultural traditions in at least Europe, Central America, insular Southeast Asia, and Africa maintain that there is a treasure at the end of the rainbow. Although the nature of this treasure varies somewhat for the dragon, the treasure at the end of the rainbow is almost always gold. As noted earlier, there may be a natural basis for this association, since gold particles are commonly found in river sands, and are visible because of their lustre. Since it is widely believed that the rainbow ends in a water source from which the Rainbow Serpent drinks, the association of a treasure, and in particular of gold, both with the end of the rainbow, and with the dragon that guards terrestrial water sources, is arguably one and the same thing.

14. Preys on young women. It does not need to be emphasized that the dragon in Europe is seen as a menace to young women, as this theme is common in the history of Western art. In China the same trait is expressed more subtly, as in the passage from the Chinese classics that describes a naked woman used to manipulate a dragon into producing rain. In North America the horned water serpent also preys on young women, and can cause demonic pregnancies.

It is remarkable that the rainbow has a similar menacing role in relation to young women. While in most parts of the world this is clearly the rainbow qua Rainbow Serpent, in North America the horned water serpent and rainbow are generally seen as distinct entities, yet both menace young women. A psychoanalytic interpretation has been proposed for this trait of the Rainbow Serpent (Tuschman 2008), but what matters here is that it is found in both dragons and rainbows.

15. Is offended by menstruation. In some ways, traits 14 and 15 can be combined, although they are clearly separate in some parts of the world. The theme of menstruation as an irritant in relation to both the dragon and the rainbow is very robust in North and South America, and Australia, with some reports also from New Guinea, and Africa, but it is unattested elsewhere. In both the Americas and Australia a girl in her menarche must be especially careful to avoid drinking from, or especially bathing in springs that are guarded by a serpent. In North America this is the horned water serpent, with no connection to the rainbow. In both South America and Australia, it is the Rainbow Serpent, and hence the rainbow. Once again, then, whether or not a natural basis for the behavior can be found, the dragon and the rainbow are virtually indistinguishable.

16. Is connected with hoofed mammals. Dragons are chimeras that combine a basic reptilian body form with additional traits derived from mammals, or less commonly, birds (specifically the Central American quetzal). While the general fusion of a cold-blooded and warm-blooded animal is readily intelligible since the rainbow needs both fire and water to exist, the specific association of dragons with hoofed mammals, and in particular with horses, is puzzling.

Surprising as this connection may seem, it is considerably more surprising to see that rainbows have a similar connection, even in areas where they are sharply distinguished from dragons, as in Europe, India or China. And when we read that the Uduk of the Sudan talk about a “hungry, meat-eating rainbow” that lives in a cave, and has ears and a mouth like a camel (recall that in one of the Chinese classics the dragon is said to have the head of a camel) the line between rainbow and dragon is erased completely.

17. Has fiery breath. Dragons breathe fire. In Europe it is the fire of blast furnaces, while in China it is different from ordinary fire. Why such a trait should be attributed to a mythical animal that is modeled on real reptiles has never been explained, and much the same shortcoming is found in virtually all other theories of the dragon. Did the ‘dragons of Eden’ (Sagan 1977) breathe fire? Was this trait simply imagined by people who conceived of dragons while line-dancing?

In Mexico, Indonesia and Africa the end of a rainbow is said to start fires where it touches earth. The basis for this belief may be that the rainbow ‘burns up’ the rain and makes it stop (as seen earlier, this belief occurs among Bantu speakers, where the rainbow is seen as antagonistic to the desired rainfall, and is thus associated with the dry season). Since the end of the rainbow corresponds to the head of the Rainbow Serpent, which must drink from the Earth to cause the rain, the idea that fire proceeds from the end of the rainbow provides a ready transition into the dragon breathing fire.

18. Has fetid or toxic breath. In North America the horned water serpent normally does not breathe fire. However, it is often said to have toxic breath, and a similar trait has been noted in Europe, China and Africa. Why the dragon should be maligned in this way is not immediately clear, but the wide distribution of the belief suggests that it has a natural basis. Again, we find the same trait in the rainbow, even where it is not associated with a Rainbow Serpent, as in northern California or northern Mexico. Because rainbows produce no odor, we can only speculate that this association might be connected instead with lighting strikes, since organic materials that have been struck by lightning may have unusual smells. Once again, however, it must be stressed that even without evidence of a natural basis for this trait, its connection with both dragons and rainbows further strengthens the argument that dragons began as Rainbow Serpents.

19. Causes earthquakes. In Europe, North America and Mesoamerica dragons are said to sometimes produce earthquakes. Dragons are often portrayed as large, heavy animals, so it might be imagined that by lumbering over the Earth they would create vibrations strong enough to trigger an earthquake. However, the same trait is attributed to rainbows in mainland and insular Southeast Asia (including the long-isolated Andaman Islands), and in at least some parts of Africa. Since the rainbow is a celestial phenomenon and the earthquake a terrestrial one, it is hard to see why the appearance of a rainbow should trigger an earthquake. The only connection that readily comes to mind is that the end of the rainbow is thought in various places to be spiritually highly-charged, and since it is thought to cause other disturbances, it might also cause earthquakes.

20. Causes whirlwinds/storms. Both dragons and rainbows are said to cause whirlwinds, waterspouts and the like. The first of these associations has been reported in Europe, China, Japan, and western North America. Given the mythology of the dragon as a large and somewhat violent creature that can rise skyward from its watery haunts on earth at the beginning of the rainy season, the belief that its flight might precipitate whirlwinds or storms is perhaps not surprising. However, it is hard to say this for rainbows, which share the same feature. The reasons that are given for rainbows starting whirlwinds or waterspouts, however, are quite varied. In the Balinese myth of the origin of the rainbow seen earlier, it is the severed penis of the hermaphroditic goddess Uma that fell to earth which causes whirlwinds, among the Suau of southeast New Guinea it is simply the spiritually-charged end of the rainbow that has the same effect, and among the Maori of New Zealand, it is the union of a male and female rainbow that triggers this natural disturbance.

21. Causes floods. There is no need to enter into detail in connection with this trait, since it is very similar to traits 19 and 20. Again, it is not clear why dragons should trigger floods, although this is reported almost everywhere they are found, and the same trait is associated with the rainbow or Rainbow Serpent in South America, mainland Southeast Asia, New Guinea and Australia.

22. Is a sign of war. I have taken some liberties here. So far as is presently known the dragon is a sign of war only in Europe where, as noted previously, the army of the later Roman Empire used a draco (dragon) as the standard of a cohort (about 500 men), and later carved dragon heads figured prominently on the prows of Viking ships when they struck out on expeditions of rapine and pillage. Ingersoll (1928:129) mentions that some Indo-European speaking peoples in the Near East, such as the Persians, Scythians and Parthians also had dragon standards that they carried into war, and that this was also true of the Semitic-speaking Assyrians. However, given the geographical contiguity of these groups, the Assyrians may have acquired this trait by contact.

While this association with dragons is not known in other parts of the world, the rainbow has been recorded as a sign of war everywhere except the ancient Near East, North America, Australia and Africa. Whatever the basis for associating the rainbow with war (red = blood = bloodshed?), the fact that this association is widely shared by the dragon and the rainbow must be counted as further evidence for their ultimate unity.

23. Causes sickness, disease or trouble. Dragons are not dangerous just because of their aggressive behavior, but they may also cause sickness through their toxic breath, or for other reasons. Among indigenous people in southern Canada and Paraguay land serpents or water serpents could sicken people, in some cases just by being seen. The same trait is associated with the rainbow in Europe, North and South America, mainland and insular Southeast Asia, New Guinea, the Pacific islands, and various parts of Africa.

24. May have human traits. Since dragons are usually opposed to humans, it is surprising to see them regarded as having human traits, or as interbreeding with humans virtually everywhere they are found. To forge one more link with the Rainbow Serpent, the same trait is described for the rainbow in China, British Columbia, southern Mexico, Guatemala and in at least the Northern Territory of Australia.

25. Can be personified. Closely related to the preceding is the tendency to personify dragons and rainbows by giving them individual names in myths or tales. In Europe this is a feature of the mythology of classical Greece and the Germanic epics, in the ancient Near East it is done in the Babylonian creation epic, and in India it appears in the jātakas—stories of the many incarnations of the Buddha. However, it has never been reported for the Chinese dragon or the North American horned water serpent. In the latter case the story of the dragon of Niagara is instructive, as the keeper of the thunderbolts (Hi’no), who kills the horned water serpent, is named, and hence personified, while the serpent is only referred to as a “monster”, even when it has human traits.

26. Encircles the world. This last connection is provisional, but nonetheless worth noting. As already pointed out, in ancient Egypt, Norse mythology, and European alchemy a dragon forming a circle to bite its own tail figures as a prominent symbol, the ouroboros. The earliest known manifestation of this idea is from Egypt, where the dragon appears “as Sito, a serpent-god who encircles the world, either with many coils, or with its tail in its mouth, or walking—with legs provided!” (Clark 1959:239–240).

As seen in Chapter 9, circular rainbows or rainbow-like atmospheric phenomena exist, and could have been seen by people at any time in the past provided they were at a sufficient height looking downward, and it is perhaps this type of observation that gave rise to the ouroboros.

This chapter lists 26 traits of the dragon (22 if we combine 6–9 and 14–15), all but one of which (its use as a sign of war) are widely distributed in the world’s cultures, and which align closely with traits of the Rainbow Serpent. The probability that this agreement is fortuitous is too small to take seriously. Rather, the simplest conclusion is that the dragon idea arose from earlier conceptions of the rainbow as the Rainbow Serpent.

Although descriptions of the Rainbow Serpent in standard compendiums of world mythology and folklore treat it as a uniquely Australian phenomenon, it should be evident from the preceding documentation that this is not true. Essentially the same idea must have once been the common property of all humanity. In culturally conservative areas, such as Aboriginal Australia, parts of South America and sub-Saharan Africa, the view of the rainbow as a gigantic serpent has remained largely intact, while in culturally more innovative areas, where agriculture, urbanization and literacy began, the Rainbow Serpent has divided conceptually into the dragon and the rainbow.

The preceding point cannot be emphasized too strongly. It is known that species and languages change over time, so why wouldn’t cultures? If the dragon began as an attempt to explain the causality and nature of the rainbow, its earliest form would have been that of the Rainbow Serpent—not that of the ancient Near Eastern dragons, as some writers, who confuse documentary history with culture history, have claimed. For much the same reason, the useful paper by Taçon, Wilson, and Chippendale (1996), which implies that the Rainbow Serpent in Australia was first conceived some 6,000 years ago in Arnhem Land because that is where it makes its first known appearance in rock art, must be viewed through a critical lens.

The evidence considered here suggests instead that the Rainbow Serpent entered Australia with the first human settlers some 60,000 years ago, and preceded the dragon everywhere on the planet by tens of thousands of years. In culturally conservative areas it changed only in details over countless generations, while in culturally innovative areas such as Europe, the ancient Near East, India and China, it experienced major transformations. In short, like any biological species, the dragon has had an evolutionary history, but one mediated by culture, not by genes. In considering the antiquity of the dragon idea in human history, it is important to remember that the brain is produced by nature, but the mind is produced by culture. The archaeologist Paul Mellars (1996:366) states the problem nicely in relation to inferences about the possible use of symbols by Neanderthal populations:

There is an obvious danger of making simplistic equations between ‘simplicity of behavior’ and ‘simplicity of mind’ which somehow short-circuits scientific analysis, and effectively assumes what one should be attempting to find out. Obviously, the less complex patterns of technological or economic behavior of modern communities such as Australian Aborigines, Bushmen, or sixteenth century Europeans in no sense imply that they had an inherently inferior intelligence or simpler cognitive capacities than those of modern industrial societies.

While the point Mellars makes is concerned with inferences about the cognitive capacities of different species of the same genus, its general import is that wherever evidence of biological development is fundamentally similar we have no right to assume that mental abilities were radically different in the past, no matter how distant that past is. Reasoning from the uniformitarian principle that the present is our best guide to the past tells us that as long as our ancestors had brains equivalent to ours, their thinking processes probably would have been similar, and the mystery of the rainbow would likely have triggered the same reaction it still does in what remains of the animistic world. This means that the Rainbow Serpent could well have a history that began with the appearance of modern humans, making it the oldest symbol in the mental life of our species, and discourse about it Man’s oldest story.

The universality of the dragon idea raises at least one other important question, namely, was it unique to us? Based on skeletal evidence, most current views are that modern humans appeared some 150,000–200,000 years ago (Fagan 2010:83 ff.) making it unlikely, or even highly unlikely, that their thinking processes would have been fundamentally different from those of hunter-gatherers in the recent past. In the absence of definitive evidence, then, it is reasonable to assume that the Rainbow Serpent complex was already taking shape with the appearance of anatomically modern humans. However, when we look back beyond about 30,000 years ago, we find not one human species, but at least two: Homo sapiens, and Homo neanderthalensis (with the still poorly understood Denisovan population of central Asia possibly representing a third).

Papagianni and Morse (2013:80 ff.) bring the Neanderthal story up to date by revising a number of earlier misconceptions about differences between these close relatives and our own genetic line (which diverged from the Neanderthals some 500,000 years ago). Among the misconceptions that they revise is the idea that Neanderthals lacked the cognitive capabilities and forward-planning abilities of modern humans because the archaeological record shows little use of symbolic systems such as art. However, a remarkable animal kill site (La Cotte de St Brelade) on the island of Jersey in the Channel Islands between Britain and France, suggests a very different story. This site, dated to before 130,000 years ago, shows unmistakable evidence of a method of hunting that was still practiced by Plains Indians in North America until the horse was accidentally reintroduced to them by Spanish explorers in the sixteenth century. This was a method whereby herd animals were tempted to follow a human leader (in the Native American case one dressed in a buffalo robe) toward a cliff. Other humans, concealed behind converging mounds on either side of the herd then began to shout, shake animal skins, and otherwise try to panic the animals in the rear to move forward, forcing the entire herd toward the brink. At the base of the cliff the animals that still lived after the fall were dispatched by hunters who awaited them.

Needless to say, this type of hunting requires considerable forward-planning, coordination among individuals with different roles, and—unmistakably—language. If nothing else were known about the cognitive capabilities of Neanderthals, the evidence for such animal drives would strongly suggest that their brains worked in ways not very different in most respects from those of Homo sapiens. And if this were the case, we must ask ourselves, how would they have responded when seeing a rainbow? Almost certainly they would have asked the same questions as early Homo sapiens: ‘What is it?’, ‘How did it get there?’, ‘Where does it stay when it isn’t in the sky?’ The only factor that might have caused a difference would have been a lack of familiarity with snakes, which could not survive in Ice Age Europe, but may well have been familiar to them during the warmer interglacials.

In summary, the Rainbow Serpent concept, which gave rise to the dragon idea after the advent of agriculture, urbanization and literacy, could have arisen independently in both of the two major hominid lines that co-existed in Europe and the Middle East until the past 30,000 years. Although we may never be able to test this supposition, it is at least plausible, provided that the physical environment was similar enough for both species to envision the rainbow as a gigantic spirit snake.