Knossos Comparisons to Other Sites

About Knossos Comparisons to Other Sites

The Kephala hill at Knossos held a Neolithic farming village for roughly five thousand years before the first ashlar block of the palace was set, around 1900 BCE in the Middle Minoan IB ceramic phase. Knossos is therefore not the start of something. It is a state-formation event laid down on top of a deep agrarian substrate, contemporary with mature Bronze Age city-states from the Aegean to the Indus, and intelligible only against that wider Bronze Age horizon. The comparisons that follow each pick a single defensible axis — palace stratigraphy, undeciphered script, bull cult, and the question of religious continuity into Iron Age Greece — and ask what Knossos shares with its peers, where the analogy breaks down, and what is genuinely without parallel.

The closest stratigraphic peer to Knossos is not another Aegean palace but the Sumerian temple-city of Eridu in southern Iraq. Both sites encode their own history in vertical layers: a long pre-monumental occupation, then a sequence of rebuilt sacred or palatial structures stacked one on top of the other, with each generation of builders incorporating the previous shrine into its foundations rather than razing the hill flat.

At Eridu, Fuad Safar and Seton Lloyd's excavations (1946–1949) for the Iraqi Directorate General of Antiquities documented eighteen superimposed mudbrick temples buried under the unfinished ziggurat of Amar-Sin (c. 2047–2039 BCE). Levels XIX through VI belong to the Ubaid period; Levels V through I to the Uruk period. The earliest temple at Level XVII is a small mudbrick chamber roughly 2 by 3 meters with a niche for the cult image. By Level VI it has grown into a tripartite temple on a raised platform — a form ancestral to the Mesopotamian ziggurat. The continuity is architectural and ritual: the deity (later identified as Enki) stays put; only the building changes.

Knossos exhibits the same vertical logic but on a different ritual grammar. The Neolithic settlement on the Kephala hill begins around 7000 BCE and accumulates roughly nine meters of cultural deposit before the first palace is laid down. The first palace (c. 1900–1700 BCE, the Protopalatial period) is destroyed by earthquake around 1700 BCE and rebuilt as the second, more elaborate palace (c. 1700–1450 BCE, the Neopalatial period). After the LM IB destructions of c. 1450 BCE, Mycenaean administrators occupy the site without razing it, leaving Linear B archives layered above the earlier Linear A ones. Final destruction by fire follows around 1375 BCE.

The contrast clarifies what Knossos is. Eridu's stratigraphy is centripetal — eighteen versions of the same temple, the cult image returning to the same niche. Knossos's is more like a working capital that keeps being damaged and re-skinned around a constant central court. The court itself, oriented roughly north–south, is the durable element; the surrounding rooms are the variable. Where Eridu's verticality reads as theological persistence, Knossos's reads as administrative persistence. Both, however, demonstrate something a one-period site like Teotihuacan cannot: that Bronze Age states could maintain ritually charged places across centuries of recovered catastrophe.

Knossos sits at the center of the most consequential decipherment story of twentieth-century archaeology, and at the center of one of its most stubborn failures. The two scripts found at the palace — Linear A and Linear B — bracket the technical and historical conditions under which an ancient writing system can be read. The peer comparison runs to Mohenjo-daro and the Indus script: the longest-running undeciphered system in the corpus.

Linear B was deciphered in 1952 by the British architect Michael Ventris, who announced his result on a BBC radio broadcast on 1 July 1952, calling the underlying language "a difficult and archaic Greek, seeing that it's five hundred years older than Homer." Within weeks Ventris was in correspondence with the philologist John Chadwick at Cambridge, who became his principal collaborator. Their joint paper appeared in the Journal of Hellenic Studies in 1953, and their Documents in Mycenaean Greek followed in 1956, weeks after Ventris's death in a road accident. Chadwick's The Decipherment of Linear B (Cambridge University Press, 1958; second edition 1967) is the standard account.

The Linear B decipherment is a positive control. It worked because three conditions held: a moderately large corpus of administrative tablets, a known underlying language family (the Greek of the Mycenaean palaces was related to later Greek), and recognizable proper-name patterns (the Knossos and Pylos tablets list place-names like ko-no-so for Knossos itself). Linear A fails on the second condition. It is attested in roughly 1,400 inscriptions; the largest single archive — 147 tablets, plus more than twenty roundels and over a thousand sealings — comes from the villa at Hagia Triada on the south coast of Crete. The script's signs are partially shared with Linear B (allowing tentative phonetic readings), but the underlying language is unrelated to any known family. John Younger has compiled phonetic transcriptions, but no continuous text has been read.

The Indus script presents an even harder problem. At Mohenjo-daro, Mature Harappan inscriptions (c. 2600–1900 BCE) appear on roughly 4,000 stamp seals, copper tablets, and pottery — but the average inscription length is fewer than five signs. Asko Parpola at the University of Helsinki, who has worked on the script for over forty years, identified 425 distinct signs in his 1994 catalogue; Bryan K. Wells's 2011 catalogue (Epigraphic Approaches to Indus Writing, Oxbow) lists 676 signs. No bilingual text exists. No continuous narrative inscription has been found. In January 2025, the state of Tamil Nadu announced a one-million-dollar prize for a verified decipherment, an unusual measure of how unsolved the problem is.

The comparison sharpens what is genuinely lost when a script cannot be read. At Knossos, Linear B opened the palace's last century to direct historical reading: personnel rosters, livestock counts, place-names, divine names including di-we (Zeus), e-ra (Hera), po-se-da-o (Poseidon), and di-wo-nu-so (Dionysus). The earlier Linear A archives — covering the Old and New Palace periods, the heart of Minoan civilization — remain mute. Mohenjo-daro is similarly silent. In both cases, the visible material culture is rich, but the people's own account of themselves is sealed.

The bull is the most over-interpreted animal in comparative archaeology. The honest comparison requires distinguishing what each site materially contains from what twentieth-century writers (from Frazer through Marija Gimbutas and the popular "Mother Goddess" literature) grafted onto the bull as a generic fertility symbol. Four peer sites — Çatalhöyük, Mohenjo-daro, and Karnak — give Knossos's bull-leaping fresco (the so-called Toreador Fresco — Evans's nickname after Spanish bullfighters) and the Minotaur myth their longest comparative reach.

Çatalhöyük (occupied c. 7100–5950 BCE, in south-central Anatolia) is the deep substrate. Excavations under Ian Hodder from 1993 to 2018 documented the practice of installing bucrania — plastered skulls and horn cores of wild aurochs (Bos primigenius), often painted ochre red — on walls, benches, and pillars inside the densely packed mudbrick houses. Hodder's "history house" interpretation, developed in Religion in the Emergence of Civilization: Çatalhöyük as a Case Study (Cambridge University Press, 2010), argues that successive rebuildings of the same dwelling progressively accumulated bucrania installations and human burials, so that elaborately decorated houses became repositories of household memory linking generations. The bull at Çatalhöyük is domestic and ancestral: it lives inside the home, attached to lineage and to specific dead.

The Minoan bull is public and choreographed. The bull-leaping fresco from the east wing of the Knossos palace, the bull-rhyton from the Little Palace, and the Horns of Consecration crowning the walls all place the bull in a court ceremony, not a domestic shrine. The Çatalhöyük bull is a remembered animal whose head returns to the wall above where the dead are buried. The Knossos bull is a living animal vaulted over by athletes in the central court while spectators watch from the surrounding magazines and stairways. The continuity is iconographic, not functional: both cultures kept the bull on the wall, but the wall did different work.

At Mohenjo-daro, the bull image runs through the Indus seal corpus. The Pashupati seal (excavated 1928–29 in the DK-G Area of Mohenjo-daro, dated c. 2500–2400 BCE Mature Harappan) depicts a seated horned figure flanked by wild animals; the steatite seal corpus includes thousands of bull seals depicting unicorn-like (single-horned) cattle, water buffalo, and zebu. John Marshall read the seated figure as a proto-Shiva Pashupati ("lord of animals"); Doris Srinivasan, in a 1976 paper in Archives of Asian Art ("The So-Called Proto-Śiva Seal of Mohenjo-Daro: An Iconological Assessment"), argued instead that the figure is more probably a buffalo-man with a humanized bucranium — a reading that would put the Indus bull in the same theological neighborhood as the later Vedic buffalo-deity rather than the historical Shiva. Without decipherment of the script, the question is unresolved. What can be said is that Mohenjo-daro and Knossos are roughly contemporary (Mature Harappan c. 2600–1900 BCE, Old Palace Knossos c. 1900–1700 BCE) and both place the bull at the iconographic center of state-level ritual without a surviving textual gloss.

Karnak supplies the textual gloss the others lack. The epithet Kamutef — "Bull of his mother" — attaches to the Theban god Amun (and to the Min-Amun composite), naming him as the self-generated bull who is both father and son. Karnak's reliefs depict Amun-Min-Kamutef ithyphallic, wrapped from neck to feet, and the Kamutef chapel built by Hatshepsut and Thutmose III on the Karnak axis institutionalized the cult. (The deeper Apis bull cult of Memphis, with the Saqqara Serapeum, is a different and older institution, but the iconographic logic — the bull as the earthly form of the god — is shared.) The Egyptian case shows what reading the texts gives you: a named theology of the bull, with succession, festival calendar, and political function. The Minoan and Harappan cases show what the same iconography can look like when only the images survive.

The longest-running comparative question about Knossos is not about its construction but about its afterlife. After the LM IB destructions of c. 1450 BCE and the final fire of c. 1375 BCE, the palace ceases to function as an administrative center. Greek classical religion appears in the historical record roughly seven centuries later, with the gods of Homer and Hesiod. How much of classical Greek religion is Bronze Age substrate — Minoan or Mycenaean material that survived the Dark Age in some form — and how much is later innovation? The answer matters because it determines whether Knossos is genuinely ancestral to the religion of the Parthenon or merely an antiquarian neighbor.

Walter Burkert's Greek Religion (originally Griechische Religion der archaischen und klassischen Epoche, 1977; English translation by John Raffan, Harvard University Press, 1985) opens with a chapter titled "Prehistory and the Minoan-Mycenaean Age" that lays out the substrate question methodically. Burkert draws on the Linear B tablets — the divine names Zeus (di-we), Hera (e-ra), Athena (a-ta-na), Poseidon (po-se-da-o), Dionysus (di-wo-nu-so), Hermes (e-ma-a), Artemis, Ares, and Eileithyia (e-reu-ti-ja) all attested at Knossos and Pylos — to argue that the core of the classical pantheon was already in place by the thirteenth century BCE. The Mycenaean Greeks who took over Knossos brought (or already had) the gods who would persist into the Iron Age. The pre-Mycenaean Minoan layer — the snake-handling figurines, the peak sanctuaries, the lustral basins, the goddess-with-griffins iconography of the Throne Room — feeds into this stream as a separate substrate, more strongly visible in the female deities and in cult forms (sacred groves, mountain shrines, household snake cult) than in the named gods.

The peer comparison runs in two directions. Mycenae is the closest case for cult continuity at the site level. The Mycenaean chamber tombs in the Argolid were frequented in the Geometric and Archaic periods as sites of "hero cult" — communities engaged with Bronze Age ruins as landscapes of memory rather than recovering the original cult intact. The Heraion of Argos, between Mycenae and Midea, may mark the introduction of formal Hera worship in the region; the link to a continuous Bronze Age Hera cult on the same ground is suggestive but not proven, and most specialists today (following the Felsch and Niemeier excavation reports and Catherine Morgan's synthesis on Phokian cult systems) treat the strongest unbroken Bronze-to-Iron Age sanctuary as Kalapodi in Phokis rather than any of the Argolid sites — the Felsch and Niemeier excavations there documented a sequence of superimposed temples running from the Late Bronze Age (c. 1400 BCE) into the second century CE. The Knossos parallel is weaker still: there is no comparable Iron Age sanctuary that demonstrably continues the palace's cult.

Delphi supplies the second comparison and a sharper methodological warning. Christiane Sourvinou-Inwood's essay "Myth as History: The Previous Owners of the Delphic Oracle," published in Jan Bremmer (ed.), Interpretations of Greek Mythology (Croom Helm, 1987), examined the classical Greek myth that Apollo took the oracle from earlier owners — usually Gaia or Themis — and asked whether that mythic narrative reflected an actual pre-Apollonian earth-goddess cult on the site. Her argument was that the myth was Greek mythopoeic reasoning about cult succession, not a memory of historical religious change. The same caution applies at Knossos. The temptation to read the Throne Room's griffin frescoes, lustral basin, and (per Nanno Marinatos) possible priestess-occupant as a direct ancestor of classical priestess cults at Delphi or Eleusis is strong. The evidence for iconographic continuity (bull, snake, double axe, peak shrine) is good. The evidence for institutional continuity — the same cult, on the same ground, from Bronze Age into Iron Age — is, as at Delphi, mostly Greek myth retrospectively explaining itself. Marinatos herself, in Minoan Kingship and the Solar Goddess: A Near Eastern Koine (University of Illinois Press, 2010), reads the Throne Room within a Near Eastern divine-kingship koine and treats the goddess-vs-king question as a false binary: the seat is the seat of a sacred king who impersonates, or is impersonated by, the goddess in epiphanic ritual.

Set beside its peers, Knossos's distinctive position is fairly precise. Its long stratigraphy — Neolithic village from c. 7000 BCE, palaces from c. 1900 BCE, Mycenaean administration to c. 1375 BCE — makes it comparable to Eridu in ritual depth and to Çatalhöyük in occupational depth, but no peer matches its combination of extreme antiquity with full participation in a literate Bronze Age administrative network. Its scripts give it a foot in both camps of the decipherment problem: positive (Linear B), like the Mycenaean tablets at Pylos and the Hittite cuneiform of Boğazkale, and negative (Linear A), like the Indus script at Mohenjo-daro and the Cretan Hieroglyphic of the earliest seals. Its bull cult sits between the domestic ancestor-bull of Çatalhöyük and the textually annotated theological bull of Karnak. And its religious afterlife, contra the romantic Evans-era reading, is most accurately described not as a continuous Minoan goddess cult flowing into classical Greek religion but as an iconographic and onomastic substrate that was filtered through Mycenaean Greek and that survived in the names of the Olympian gods more cleanly than in any specific cult installation.

The synthesis is observational. The myths of Theseus, the Minotaur, the Labyrinth, and Daedalus preserve, in distorted form, real features of a real palace: the spatial complexity of the building, the centrality of bulls in court ceremony, the eventual subjugation by mainland Greeks. The decipherment of Linear B by Ventris in 1952 turned those myths from pure fiction into degraded history. The comparison with Eridu, Mohenjo-daro, Çatalhöyük, Karnak, Mycenae, and Delphi shows that this pattern — material remembered through myth, partially recoverable through archaeology, partially through script, partially lost — is the typical fate of Bronze Age palatial states, not a peculiarity of Crete. What is peculiar to Knossos is the depth of the substrate underneath the palace, the survival of one of its scripts as a positive control on decipherment, and the way the myths of the Greeks turned out, against expectation, to remember the place.

Significance

Comparative work places Knossos at the intersection of four of the largest open questions in Bronze Age archaeology: the relationship between deep agrarian substrate and palatial state formation, the limits of script decipherment without a known underlying language, the cross-cultural function of bull iconography, and the continuity of cult between Bronze Age and Iron Age Mediterranean religion. Walter Burkert's Greek Religion (Harvard University Press, 1985) frames the last of these by reading the Linear B divine names — Zeus, Hera, Poseidon, Dionysus, Hermes — as evidence that the classical Greek pantheon's core was already in place at Knossos by the thirteenth century BCE, while leaving the pre-Mycenaean Minoan layer as a separate substrate visible in iconography rather than in named cult.

Connections

Knossos — the parent entity. Standalone coverage of Evans's excavations, the palace plan, the frescoes, and the Linear A and Linear B archives.

Eridu — Sumerian temple-city with eighteen superimposed mudbrick temples (Safar and Lloyd, 1946–1949) — the closest stratigraphic peer to Knossos for ritual depth across millennia.

Mohenjo-daro — Mature Harappan city (c. 2500–2000 BCE) whose script remains undeciphered after a century of work — the comparative case for what Knossos's Linear A archives still are.

Çatalhöyük — Anatolian Neolithic settlement (c. 7100–5950 BCE) where Hodder's "history house" excavations documented domestic bull bucrania installations — the deep substrate of Mediterranean bull symbolism.

Karnak Temple — Theban temple complex where the Amun-Kamutef "Bull of his mother" cult was institutionalized — the textually annotated counterpart to Knossos's mute bull-leaping fresco.

Mycenae — mainland Greek citadel that absorbed the Knossos palace administration after c. 1450 BCE and through whose Linear B tablets Bronze Age Greek religion enters the historical record.

Delphi — site of Sourvinou-Inwood's "previous owners" question (1987) — the methodological warning case for how Greek myth narrates cult succession that may not be historical continuity.

Teotihuacan — single-period Mesoamerican city used here as a contrast: a peer that lacks Knossos's vertical stratigraphy and its undeciphered-then-deciphered script pair.

Further Reading