The pharaoh answered in ink we lost

Claim (verbatim)

The Amarna archive - roughly 350 to 382 cuneiform tablets in Akkadian, excavated at Tell el-Amarna in Egypt - is a diplomatic mailbag, and it is a one-way one. The great majority are letters TO the Egyptian court: from Levantine vassals and from the 'great kings' of Babylon, Mittani, Hatti, Assyria, and Alashiya. The pharaoh's own outgoing letters survive as a tiny minority (a handful such as EA 1, 5, 14, 31, 99, 162, 163, 190, 367, 369, 370), because the outbound copies went to foreign capitals whose archives are lost or unexcavated. An archive preserves what ARRIVES, not what departs; the corpus is a single-directional slice of a two-directional exchange, and the missing half can be partly counted from the surviving half, which quotes and answers letters that no longer exist. The mechanism is archival directionality: the find-spot is the recipient's file, so the sender's outbound copies are systematically absent except where a second archive was also dug.

Prediction clause (verbatim)

Prediction: in the EA corpus, incoming letters (to Egypt) outnumber attested outgoing letters (from Egypt) by a factor >= 15 (primary clause: incoming:outgoing ratio >= 15); and independently, incoming letters contain >= 30 explicit references to prior royal letters or tablets FROM the pharaoh that do not survive - a ghost census of the missing outbound half. Disambiguation: direction is assigned by the sender/recipient address formula; reciprocal 'great king' letters are tallied separately from vassal letters. Coverage guard: the mere existence of asymmetry is not the finding - its magnitude is; a ratio near 1 would falsify, and the reference-count clause is reported separately so a broken-tablet shortfall cannot masquerade as symmetry.

Kill-dataset (verbatim)

Kill: the El-Amarna correspondence (Rainey 2015, The El-Amarna Correspondence; the standard EA numbering; ORACC Amarna materials where present) - tabulate letters by direction and count in-text references to non-surviving Egyptian letters.

Novelty / leakage triage

already answered in the literature

The direction census is published arithmetic: Moran's and Rainey's standard editions assign every EA letter its sender and recipient, and their introductions state the corpus shape - some three hundred letters addressed to the Egyptian court against roughly a dozen from it, the very EA 1/5/14/31/99/162/163/190/367/369/370 list the item recites - so the >= 15 incoming:outgoing ratio is guaranteed list-arithmetic on the published editions, the field-leak the check anticipated. Only the secondary ghost-reference count (>= 30 mentions of non-surviving pharaonic letters) is un-run.

• W.L. Moran, The Amarna Letters (Baltimore, 1992), introduction
• A.F. Rainey, The El-Amarna Correspondence (Handbook of Oriental Studies 110, Leiden, 2015)

Predictions

No prediction registered yet.