Priority 1 — Are Supernatural Serpents Actually Exceptional?

Why this matters

This is the foundational unresolved question.

Throughout the investigation we accumulated examples of serpent beings from China, Mesopotamia, India, the Levant, Mesoamerica, Australia, Europe, and elsewhere. But abundance of examples is not the same thing as statistical exceptionalism.

Perhaps serpents really are unusually productive supernatural animals.

Or perhaps birds, felines, canids, crocodilians, bears, horses, and other animals are equally common in mythology—and modern comparative scholarship simply happens to group serpent beings under the unusually memorable label “dragon.”

If the second possibility is correct, a substantial portion of the original mystery disappears.

This is now realistically testable. The Standard Cross-Cultural Sample contains 186 societies deliberately selected partly to reduce cultural relatedness, with each society anchored to a particular time and place. (eHRAF World Cultures) D-PLACE adds cultural, linguistic, geographic, and environmental variables for more than 2,000 societies and permits variables to be examined against linguistic phylogenies. (D-PLACE)

Key questions

The investigation should ask whether supernatural serpents occur more frequently than supernatural birds, felines, canids, bears, crocodilians, or other animal classes; whether serpent beings possess a distinctive cluster of water, weather, underworld, guardianship, fertility, sovereignty, or adversarial functions; whether those associations survive controls for language family, geographic proximity, historical contact, ecology, and documentation intensity; and whether the apparent pattern remains after indigenous categories replace the generic English labels dragon and serpent.

The most consequential sub-question is:

If we had never invented the comparative category “dragon,” would the underlying statistical pattern still be visible?

Evidence to seek

Start with eHRAF World Cultures and the Standard Cross-Cultural Sample and D-PLACE. D-PLACE is especially valuable because its cultural observations are associated with location, focal date, language, and source material, making ancestry, geography, and environment separable variables rather than background impressions. (D-PLACE)

A second test bed is Pulotu, the Database of Austronesian Religions. It covers 137 Austronesian cultures and was explicitly designed for evolutionary analysis of supernatural beliefs; importantly, robust language phylogenies allow common ancestry to be modeled rather than accidentally counting related traditions as independent inventions. (Pulotu)

The project should create a preregistered coding scheme covering morphology, habitat, religious function, narrative role, moral valence, social setting, and degree of hybridization—and apply the same coding scheme to control animals.

Possible outcomes

A strong serpent overrepresentation surviving all controls would materially strengthen the idea that there is a genuine human-serpent phenomenon requiring explanation. Cognitive, ecological, and deep-historical hypotheses would become considerably more important.

A weak effect disappearing after ancestry/contact controls would suggest that inheritance and diffusion account for much of the apparent recurrence.

An effect disappearing only when other supernatural animals are introduced would be more disruptive: the “worldwide serpent mystery” itself would largely dissolve.

And a particularly interesting result would be functional exceptionalism without numerical exceptionalism—serpents are no more common than other animals but repeatedly occupy an unusual set of roles. That would redirect the investigation from why serpents recur to why serpents are repeatedly assigned particular meanings.

Expected informational value: Extremely high.

Priority 2 — The Deep-History Test: Can Any Serpent Tradition Be Reconstructed Before Recorded Civilization?

Why this matters

The investigation left one enormous possibility unresolved.

Perhaps supernatural serpents arose repeatedly.

But perhaps some portion of the pattern is far older than civilization itself.

This branch should directly test deep inheritance rather than merely discussing it.

Key questions

How far back can serpent-related mythic structures actually be reconstructed using comparative linguistics, population history, archaeology, and phylogenetic methods? Do rare trait bundles follow ancient population relationships? Can Indo-European, Austronesian, Australian, African, or Native American traditions preserve reconstructible serpent complexes substantially older than their earliest texts?

Most importantly, does similarity increase with shared ancestry once ecology and later contact are controlled?

Robert Blust's Rainbow Serpent hypothesis deserves a dedicated adversarial test here. Rather than asking whether examples can be found that resemble his reconstruction, ask what observations should exist if he is right and whether those predictions actually hold.

Evidence to seek

This branch should integrate reconstructed linguistic phylogenies, ancient population-genetic migration histories, archaeological serpent imagery, dated ethnographic traditions, comparative mythology, and phylogenetic cultural analysis.

Pulotu is unusually useful because its Austronesian sample was explicitly constructed to permit reconstruction of ancestral cultural states and tests of correlated trait evolution using established language phylogenies. (Pulotu)

The research should also return to Calvert Watkins for the strongest established positive control: a case where comparative linguistic evidence genuinely does support inherited serpent-slayer material. That provides a benchmark against which much deeper reconstructions can be judged.

Possible outcomes

Discovery of rare, arbitrary serpent trait bundles tracking ancient population ancestry would substantially increase the plausibility of deep inheritance.

Finding only generic characteristics—large, dangerous, watery, underground, celestial—would favor convergence.

If Blust's proposed distinctive Rainbow Serpent traits track ancestry in ways ecology and contact cannot explain, his hypothesis would deserve a major confidence upgrade.

If the supposed package fragments once cultural relatedness is controlled, the deep worldwide-serpent hypothesis would weaken considerably.

Expected informational value: Very high.

Priority 3 — Why Serpents? The Cognitive and Biological Experiment

Why this matters

Our present model contains a major causal gap:

real snakes → cognitive salience → supernatural serpents → dragons

Evidence exists for the first relationship. The remaining arrows have largely been inferred.

Instead of continuing to interpret myths psychologically, this branch would ask whether serpent-based supernatural beings possess experimentally measurable advantages in human cognition and cultural transmission.

Key questions

Are supernatural serpents noticed faster than comparable supernatural animals? Are they remembered better? Do people spontaneously exaggerate snakes into monsters? Are serpent hybrids transmitted more accurately through chains of retelling? Do people independently add similar features—size, wings, horns, water, venom, fire, intelligence—to serpent concepts?

Most provocatively:

If participants who know nothing about the dragon hypothesis repeatedly transform ordinary snakes into similar supernatural beings during cultural-transmission experiments, have we experimentally reproduced part of the process that mythology may have undergone?

Evidence to seek

The investigation should review experimental psychology of snake detection and fear learning, primate studies, cognitive science of religion, Pascal Boyer's work on minimally counterintuitive concepts, cultural-transmission experiments, memory research, threat detection, and developmental psychology.

Then design hypothetical or actual experiments comparing matched invented beings:

snake + bird snake + feline bird + feline crocodile + bird horse + bird neutral animal hybrids.

Participants could encounter and retransmit stories across multiple generations of retelling.

Possible outcomes

A strong serpent-specific transmission advantage would materially strengthen cultural-attractor explanations.

A general predator advantage would revive a weaker version of predator psychology without supporting the discarded snake-cat-raptor dragon template.

A general hybrid-monster advantage would suggest there is little specifically serpentine about the process.

And no serpent advantage at all would significantly weaken cognition as an explanation for the cross-cultural pattern.

Expected informational value: High.

Priority 4 — The Serpent–Water Problem

Why this matters

One association survived almost every stage of the investigation without ever becoming secure enough to call a finding:

serpent ↔ water

Vṛtra blocks waters. Nāgas inhabit or govern them. Chinese dragons govern rain. Lotan and Leviathan occupy marine imagery. Rainbow Serpents connect water and sky. Other traditions place serpent beings at springs, rivers, lakes, or rainfall.

But the relationship is frustratingly elastic.

The serpent can be water, live in water, bring water, withhold water, control water, guard water, or emerge from water.

That could represent something profound.

Or it could be exactly the sort of pattern comparative mythologists manufacture by grouping unlike relationships together.

Key questions

Are supernatural serpents associated with water more frequently than other supernatural animals? Which relationship is actually common: habitat, rainfall, obstruction, guardianship, fertility, flood, or sovereignty?

Does serpent-water symbolism correlate with rainfall variability, river dependence, drought risk, irrigation, monsoon climates, or actual aquatic snakes?

And does the relationship remain after cultural ancestry is controlled?

Evidence to seek

This branch should combine ethnographic coding with historical hydrology, rainfall variability, river systems, irrigation dependence, snake biogeography, subsistence patterns, and political organization.

D-PLACE is particularly promising because cultural and environmental information can be examined together and mapped onto linguistic classifications. (D-PLACE) Political-complexity variables are also available, allowing the old dragon/water/state hypothesis to be tested rather than assumed. (D-PLACE)

Egypt should remain a deliberate negative control.

Possible outcomes

A strong association surviving animal, ecological, and ancestry controls would identify the first quantitatively defensible cross-cultural serpent complex.

If rainfall variability rather than water itself predicts the mythology, the environmental model becomes much more specific.

If ancestry explains the effect, inheritance replaces ecological convergence.

If the effect disappears when “water” is decomposed into precise variables, one of the investigation's most persistent apparent patterns would turn out to be a classificatory illusion.

Expected informational value: High.

Priority 5 — How Dragons Are Made: Historical Transformation of Serpent Beings

Why this matters

One of the investigation's most interesting second-order conclusions was that the dragon may often be the endpoint rather than the origin.

That proposition can be tested historically.

Instead of comparing civilizations horizontally, this investigation would follow several traditions vertically across centuries and ask exactly how a serpent becomes something later observers recognize as a dragon.

Key questions

When do limbs, horns, wings, feathers, claws, fire, treasure, weather powers, intelligence, cosmic opposition, kingship, and moral character enter particular traditions?

Do traits accumulate gradually?

Do major transformations coincide with religious change, political centralization, literary borrowing, artistic innovation, or intercultural contact?

And are apparently continuous dragons actually sequences of different beings later joined retrospectively?

Evidence to seek

Use several high-resolution case studies:

China: Hongshan → later Neolithic imagery → Shang lóng → imperial dragon.

Mesopotamia: serpent traditions → Ningišzida → mušḫuššu → changing divine affiliations.

Levant/Bible: Lotan → Leviathan → later Jewish interpretation → Revelation's explicit dragon/ancient-serpent/Satan synthesis.

Europe: ancient drakōn → Germanic/Norse serpent-monsters → Beowulf → medieval dragon.

Mesoamerica: early serpent imagery → Teotihuacan Feathered Serpent → later Quetzalcoatl/Kukulkan traditions, without assuming identity at every stage.

The evidence should consist primarily of dated artifacts, inscriptions, manuscripts, critical editions, iconographic sequences, linguistic change, and securely documented contact.

Possible outcomes

If traits demonstrably accumulate over time, the investigation's dragon-as-endpoint model would gain strong historical support.

If fully developed dragon packages repeatedly appear abruptly, the model would weaken and convergence or inheritance would become more important.

If changes repeatedly coincide with contact, diffusion would gain explanatory weight.

If political centralization consistently precedes standardization and monumentalization, institutional selection would move from plausible mechanism toward a comparative finding.

Expected informational value: High, but narrower than Branches 1–4.

Priority Ranking

There is an important dependency here. Branch 1 should come first. It determines whether Branches 2–4 are explaining a genuine statistical phenomenon or trying to explain a pattern partly created by selective comparison. The infrastructure now exists to make such a project much more rigorous than traditional example-collecting: SCCS was constructed as a 186-society comparative sample, D-PLACE can connect cultural variables with environment and linguistic ancestry, and Pulotu provides a particularly strong phylogenetic test bed within 137 Austronesian cultures. (eHRAF World Cultures)