Across civilizations one motif recurs: the dying-and-rising god — who dies and revives with the cycle of vegetation (Egypt's Osiris, the Near East's Adonis and Attis, Greece's Dionysus). Frazer claimed this is the human mind producing like images under like circumstances.
Frazer marshaled vast cross-cultural material to show that peoples ten thousand miles apart, never in contact, told structurally identical stories of a "dying and reborn" god. His explanation was psychic unity: facing the turning seasons, sowing and reaping, the human brain spontaneously generates isomorphic imaginings. This founded the "comparative method": laying the myths of different civilizations side by side to find a common skeleton.
The Golden Bough profoundly shaped Eliot's The Waste Land and modern literature at large, yet drew sharp criticism: it tears myths from their context and forces them together, as if the whole world told one story. This is exactly the "over-unification" that Lévi-Strauss and Dumézil would later correct. Compare China: its agrarian deities never developed into a systematic "dying-and-rising" mythology — Chinese myth was historicized and moralized early, its gods becoming sage-kings.
Essence: That separate civilizations tell similar myths is the first evidence of a shared human mind — and the first temptation to over-generalize.
Beneath every hero story lies one "monomyth": the hero departs — undergoes trials — returns bearing a boon. Moses, the Buddha, Odysseus, Christ are the "thousand faces" of a single archetype.
Campbell took up Jung's collective unconscious and archetypes and pushed Frazer's "comparison" to its peak: not only are motifs alike, the narrative structure itself is one. The model, via George Lucas, became the blueprint of Star Wars and remains a Hollywood screenwriting bible — proof it does capture some universal psychological rhythm.
The criticism is just as fierce: to assemble "one" hero, Campbell flattens the differences between civilizations and downplays female and non-heroic narratives. Chinese myth especially resists: Yu the Great, who tamed the floods, is a collective, administrative, de-individualized hero — passing his own door three times without entering, his focus is public order, not personal "initiatory transformation." The East prizes not the lone hero's inner growth but virtue and responsibility.
Essence: Beneath a thousand faces there may be one story — but "there is only one story" can itself be a kind of blindness.
Myths differ wildly on the surface, but at depth run one logic: the human mind uses myth to mediate unsolvable binary oppositions — life and death, nature and culture, heaven and earth. What is alike is not the content but the structure.
This is the decisive leap. Frazer and Campbell sought sameness on the surface; Lévi-Strauss says surface differences are precisely variations of one deep structure, like one melody in many keys. Creation myths take countless forms, yet all process the same contradictions: how chaos becomes order, how the raw becomes the cooked (how nature becomes culture). Difference on the surface, sameness in structure — for the first time, finding sameness and preserving difference are unified in one frame.
This "binary opposition" logic resonates with China's yin–yang: yin/yang, hard/soft, heaven/earth are a cognitive schema that organizes the world through paired poles. The difference: Lévi-Strauss sees myth mediating contradictions (forging a middle term between poles), whereas yin–yang stresses mutual generation and transformation (opposites containing and flowing into each other). From the same binary start, one seeks a mediator, the other seeks flux.
Essence: Myths differ because humans answer, in different ways, the same set of unsolvable contradictions.
The myths and societies of Indo-European peoples share a "trifunctional" structure: sovereignty/priesthood, the warrior, production/fertility. The Norse Odin–Thor–Freyr, India's Brahmin–Kshatriya–Vaishya, Rome's three chief gods — all are projections of one framework.
Dumézil is the most powerful corrective to "one myth for all humanity." He showed this trifunctional structure is specific to the Indo-European family, rooted in a shared linguistic and social history, not universal to all humans. In other words, mythic "sameness" is often not universal but a bounded family resemblance — highly isomorphic within one language family, but not necessarily holding once you step outside the circle. Here the search for sameness gets a brake.
Chinese myth is the best counter-case for testing the trifunctional thesis: the Chinese imagination of order is not a "priest–warrior–commoner" triad but is axed on the Mandate of Heaven, virtue, filial piety, and bureaucracy. Myth was historicized early — the Yellow Emperor, Yao, and Shun turned from gods into moral exemplars and dynastic founders, scattered through histories rather than an independent pantheon. This is not "lacking myth" but another civilization's different encoding of order.
Essence: Mythic resemblance often has a boundary — it frequently proves not "humanity is one family," but "this lineage shares an origin."
Four scholars form a spectrum from "sameness" to "difference," answering one question — why do humans tell stories so alike, and yet so different?
· Frazer (common origin): like circumstances breed like imaginings; the dying-and-rising god is the first evidence of a shared human mind.
· Campbell (the monomyth): every hero story shares one skeleton — departure, initiation, return; beneath a thousand faces, one story.
· Lévi-Strauss (structure): difference on the surface, sameness at depth — myth is a tool for mediating unsolvable oppositions like life/death and nature/culture.
· Dumézil (the boundary): resemblance often stops at the edge of a language family; the three functions prove not "humanity is one" but "this lineage shares an origin."
Together they form an operating system for comparing anything: step one, Frazer's bold search for sameness — lay the material side by side and look for a shared skeleton; step two, Campbell's model-building — compress the resemblance into a reusable structure; step three, Lévi-Strauss's dive below the surface — ask whether apparent differences are variations of one deep tension; step four, Dumézil's boundary-drawing — test exactly how far the "universal law" extends. Myths are alike because we share one body, one sun, the same death, and a brain that loves binaries; they differ because geography, social structure, and history carve different answers into different civilizations. In the AI era this discipline turns urgent: a large language model is Frazerian induction in its most extreme form — squeezing patterns from oceans of text, forever at risk of passing off a bounded family resemblance as universal truth. So with myth, so with data, markets, and people: refuse both lazy moves — "all one family" and "utterly incomparable" — and instead seek sameness, then see difference, then guard the boundary.