Six famous descents, and only two of them work. That is the single most useful fact about this comparison, and it is usually buried under the assumption that going to the land of the dead is a heroic achievement in itself. It is not. In almost every tradition that has such a story, the journey down is the easy half — the texts say so explicitly — and the story is really about the terms of the return.
Here is the ledger — went for, broke, came back with — and then the argument.
- Orpheus: his wife · broke the condition attached to her release · came back with nothing.
- Heracles: Cerberus · broke nothing, because he had permission · came back with the dog and two rescued men.
- Odysseus: directions home · broke nothing · came back with a prophecy and a great deal of news.
- Aeneas: his father · broke nothing, because he carried the token · came back with a vision of Rome's future.
- Inanna: purpose unstated and still disputed · broke the rules of the place · came back alive, at someone else's expense.
- Izanagi: his wife · broke a prohibition on looking · came back with nothing and a permanent enmity.
Permission is the variable that decides the outcome
Rank those six by success and a pattern falls out immediately: the ones who succeed are the ones who arrived with authorisation, and the ones who fail are the ones who arrived on their own initiative to take back a person.
Heracles is sent. The labour is imposed on him by a king acting under an oracle, and the sources are careful to record that he asked and was permitted to remove Cerberus provided he used no weapons. He is, in effect, a licensed contractor. Aeneas has the golden bough, obtained under instruction from the Cumaean Sibyl, and Virgil makes the bough's willingness to come away from the tree the test of whether he is authorised at all. Odysseus does not properly descend; he digs a pit at the world's edge on Circe's precise instructions and the dead come up to him. He has a procedure, and he follows it.
Orpheus has none of this. He has talent. He argues his way in and wins his case on the strength of the performance, which is the most Greek thing imaginable — and the concession he wins is conditional, because a concession extracted by persuasion always is. Izanagi simply walks into Yomi after his wife and is asked to wait outside while she negotiates; he does not wait.
The two prohibitions that recur, and what they are actually about
Across this set, two rules keep appearing, and they are doing different jobs.
Do not look. Orpheus and Izanagi both fail this one, and the resemblance is genuinely startling — a man retrieves a dead wife, is told not to look at her until they are out, looks, and loses her. Folklorists catalogue this as a distinct motif, and the honest position on the parallel is that no route of transmission between archaic Greece and eighth-century Japan has ever been demonstrated. What we have is either independent invention from a shared human situation — grief, and the intolerable question of what the beloved dead now look like — or a chain of borrowing too old and too broken to trace. Claiming more than that is the kind of overreach that gets comparative mythology laughed at. Note also that the two stories use the failure for opposite purposes: Ovid's Orpheus loses a wife and gains a career as an elegist, while the Japanese chronicle uses the moment to explain a permanent structural fact about the world, namely that the living and the dead are on opposite sides of a blocked road and that a fixed number of people must die each day.
Do not eat. Food is the mechanism that converts a visitor into a resident. Persephone's pomegranate seeds bind her to a portion of every year; the Japanese chronicles have Izanami say she has already eaten at the hearth of Yomi and can no longer simply leave. This one is more widely distributed than the looking taboo and considerably older in the record. It is also the rule that most obviously survives into later European folklore about fairy hospitality.
Inanna is not a hero and her descent is not a rescue
The Descent of Inanna is routinely filed with these stories and it does not belong with them. She is a goddess, not a mortal. She is not retrieving anyone — the Sumerian text never states her purpose, and the guesses range from an attempt on her sister Ereshkigal's throne to attending a funeral. She is stripped of an item of regalia at each of seven gates, killed, hung on a hook, and revived by creatures made from the dirt under Enki's fingernail. And crucially, she does not get out free: the underworld's rule is that a substitute must take her place, which is how Dumuzid ends up condemned by his own wife.
That last detail is the reason the story is worth reading against the Greek ones. Greek and Roman katabasis assumes an exit exists and asks what it costs the traveller. Sumerian katabasis assumes an exit exists and asks who else pays. The accounting is external, and the emotional weight of the poem lands on the substitution, not on the journey.
The one everyone adds that does not qualify
Gilgamesh is frequently listed among underworld travellers and he is not one. His long journey is horizontal, not vertical: through the mountain passage of the sun, across the waters of death, to a man living at the world's edge. That is a quest for a survivor, not a descent to the dead. The epic does contain a real underworld episode — the twelfth tablet, in which the shade of Enkidu comes up and reports on conditions below — but that tablet is a partial translation of an older Sumerian poem, appended to the epic in a way that flatly contradicts the story it follows, since Enkidu died several tablets earlier. Using it as evidence of the epic's design is a mistake. It is evidence of how ancient editors worked.
What each tradition thinks the dead are for
Line the destinations up and the underworlds turn out to be answering different questions.
- Greek Hades is an information source. Odysseus goes there for a briefing, and gets one. So does Aeneas, whose father delivers what amounts to a lecture on Roman history.
- Roman Hades is a legitimation device. The whole sixth book of the Aeneid exists to let a dead father authorise a living empire, which is a political use of the afterlife that Homer has no interest in.
- Sumerian kur is an administrative jurisdiction with rules, gatekeepers, ledgers and no exceptions, and its literature is fascinated by procedure.
- Yomi is pollution. Izanagi's first act on escaping is to wash, and the washing generates three of the most important deities in the tradition. The story is not about grief resolved; it is about defilement removed.
If you want the destinations themselves rather than the travellers, the survey of underworlds takes them side by side, and the comparison of guides and ferrymen covers who takes you there when you are not coming back.
Loose ends worth knowing
Did any mortal ever bring someone back permanently? Alcestis, and she is the exception that clarifies the rule — she is retrieved by Heracles, who wrestles Death at the tomb rather than negotiating below, and Heracles is the one figure in Greek myth with a physical record against the underworld. Theseus and Pirithous, who tried to abduct Persephone herself, were stuck to a stone bench until Heracles pulled Theseus loose and left Pirithous there.
Is the "hero's descent" a stage of the monomyth? Campbell filed it under the belly of the whale and the road of trials, and the framework is popular with screenwriters for good reason — it produces usable structure. Folklorists have been sceptical of it for decades, on the grounds that it is built by selecting the compatible parts of stories and discarding the rest. These six descents differ on purpose, on authorisation, on outcome and on what the underworld is; a model that renders them one stage is deleting exactly the information that makes them worth comparing.
Why is the exit always harder than the entrance? Because the traditions with these stories mostly imagine death as a one-way legal status rather than a location, and a status is not undone by walking. The Sibyl says as much to Aeneas in the line every Latin student meets: going down is easy, and the work is in the retracing.