The ivory sceptre from Knossos

The longest Linear A text yet. And nobody outside the dig team has seen it sign by sign.

It's a ring of elephant ivory with a handle, and both are covered in writing. Four scholars described it in 2025: Athanasia Kanta, Dimitri Nakassis, Thomas Palaima and Massimo Perna. Their paper is 17 pages and free to read (Ariadne Supplement 5, pages 27 to 43). It's a first look, and they say so. The full edition is still to come.

What this page is not: a reading. Nobody can read Linear A, and this find doesn't change that. It's a list of what the paper says, with page numbers, and what it leaves out.

What it is

A slice of tusk, cut into a ring about 14 cm across, open at the bottom (p. 34). A handle about 13 cm long fits into the gap (p. 40). Two holes cross the ring, so the team thinks something small hung in the middle on rods. Whatever it was has rotted away (p. 36).

Kanta calls it "a religious scepter", and she says the name is provisional (p. 32). The three epigraphers call the two parts the Ring and the handle, and so do we. In the catalogue they are KN Zg 57 (the Ring, two faces listed) and KN Zg 58 (the handle).

Where it was found

Knossos, Crete, in a plot called Anetaki, a short walk from the palace. Kanta dug it between 2011 and 2017 (p. 28). The plot holds a Roman temple, a Bronze Age shrine under that, and a large cult building under both (pp. 28 to 30).

The sceptre lay in a small storage pit in Room 1 of that building, about a metre across (p. 31). A faience ring, a marble bowl, six bronze hinges and a lot of pottery came out of the same pit (p. 33).

How old

The building dates to 1700 to 1450 BC (p. 29). The pots in the pit look early in that span (p. 33). One later chapter puts the sceptre at about 1600 BC and calls that preliminary (Civitillo and colleagues, 2024, p. 159). Nobody has fixed a firmer date in print.

How long

About 119 signs, on the Ring alone (p. 40). That number needs its footnote. Only 84 are fully or partly preserved. The other 35 are small traces, or space where the team thinks a sign probably was. Even so, it beats the old record holder, tablet HT 123, which has about 60 signs once you take out the numbers and dividers (p. 40).

The handle has its own text. The paper gives no count for it.

The signs that are plainly Linear A

The paper names these, by number. The drawings here come from SigLA, and each one is a typical example from another tablet. None of them is the sceptre's own sign, because no drawing of those has been published.

• Sounds: RA, I, NE, SE on the top face (p. 37). KU, WI and RA on the bottom (p. 39). KU and RA on the outer rim (p. 40). A, KU, DI and KA turn up more than once (p. 41).
• Grain, olive oil and probably flour, each in its own box, no amounts (p. 39).
• Cloth, five times, the five differing in their fringes. One has KA written inside it (p. 39).
• Hides, about nine, on the badly worn inner face (p. 40).
• Vases. At least ten, drawn so carefully you can tell an amphora from a tripod. Some have a sound sign rising out of the mouth, like PA or RU (p. 38).

And there are words, at least nine groups on the rim and six on the bottom. The paper reads none of them. It says they have "few parallels" in known Linear A (p. 40).

The signs that are not, or not yet

• Twelve animals on the top face, each in its own box, all facing left (p. 37). The authors treat them as pictures, not script, and suggest they might be animals for a sacrifice (p. 38). The species aren't published.
• Two signs on the handle that until now belonged to the older Cretan Hieroglyphic script, numbered 180 and 181 in its corpus (p. 41). The authors say this proves Linear A used them too. The same pair sits together on a stone seal from Lastros (Jasink and Weingarten, 2024, p. 84). What 181 meant is unknown (Civitillo, 2024, p. 39).
• One more sign on the handle that the authors call probably new (p. 41). Not drawn, not described.

One trap: the hide sign on the Ring is also numbered 180, in the Linear A list. It's a different sign from the Hieroglyphic 180 on the handle. Same number, two lists.

Two texts, two hands

The Ring has no numbers at all (p. 40). The handle has numbers, and that makes it bookkeeping (p. 41). The cutter's hand differs too. On the Ring, A sits above KA when the two are joined. On the handle, A sits below (p. 41).

The handle also carries six different fraction signs in a row (p. 41). The authors say the order gives the fractions' relative sizes, and that it doesn't match the order scholars have proposed so far (pp. 41 to 42). Which six, in what order? Not published yet.

What the authors think it is

A record of what a ritual feast needed: animals, vessels, cloth, hides, food (p. 39). Or a memory aid for a ceremony, like the Hittite texts that prescribe a ritual step by step (p. 39). Their word is "may", and that's the right word.

What's missing, as of October 2026: a transcription, a drawing of any single sign, a close photo, and the final volume, Anetaki II (p. 42). Until those exist, every number on this page comes from one 17-page paper.

What the press got wrong

Two of the 2025 stories (Greek Reporter and World History Encyclopedia) said the sceptre was found in 2024. It wasn't. The dig ended in 2017 (p. 28), and Kanta mentioned the sceptre in print in 2022 (Kadmos 61, p. 62). The stories also put the 119 signs across head and handle. The 119 is the Ring alone, and 35 of them are traces or guesses (p. 40).

Small errors, sure. But this is the one object everybody will quote for the next ten years, so we'd rather get the small stuff right.