Linear A Decipherment Claim: AI Engineer Proposes Semitic Origin
AI Engineer Claims Breakthrough in Linear A Decipherment
Tom Di Mino, a self-taught AI engineer and amateur linguist, claims to have deciphered Linear A, a Bronze-age Minoan writing system that has remained a puzzle for over 120 years. Di Mino proposes that Linear A represents an extinct Semitic language that served as a precursor to biblical Hebrew, similar to the relationship between Latin and Italian.
His findings are currently undergoing review by linguistics experts at Rutgers University and the University of Cambridge. If validated, this would mark the first successful decipherment of the script since the 1952 breakthrough of Linear B.
The Methodology: AI-Assisted Hypothesis Testing
Di Mino utilized Claude Code to build a systematic engine for organizing and categorizing symbols and testing linguistic hypotheses. Rather than using AI as a "black box" to solve the puzzle, he used the tool to develop Python scripts that queried and cross-referenced the digitized Linear A corpus from the GORILA and SigLA databases.
This computational approach allowed Di Mino to perform statistical and pattern analysis at a scale that would be impractical for manual research, enabling him to systematically validate his theories against the available data.
The "Key" to the Decipherment
The breakthrough occurred on May 22, 2026, while Di Mino was analyzing a series of Linear A prayer inscriptions known as the "Libation Formula."
The Linguistic Pivot
Di Mino focused on a recurring verb in these inscriptions. By identifying five known Linear B signs and one Linear A-only sign ("*301"), he unlocked the root "nawaya," meaning "to dwell." In Semitic languages (including Hebrew and Akkadian), the N-W-Y triliteral consonant system is used for verbs and nouns relating to dwelling or building a pasture.
Key Results of the Research
Di Mino's work has produced several tangible outputs:
- Proposed readings for 37 of 102 signs, including all 13 symbols unique to Linear A and the resolution of three previously unknown Linear B signs.
- A lexicon of 383 Linear A terms translated into English.
- A draft manuscript titled Ya Diktu: Grammar of the Minoan Peak Sanctuary Libation Formula, intended for peer-reviewed submission.
Context: Linear A vs. Linear B
Linear A (c. 1800 BC – 1450 BC) was used on Crete until the Mycenaean Greeks conquered the island. The Mycenaeans adapted these symbols to create Linear B, which was deciphered in 1952 by Michael Ventris and Alice Kober as an early form of Greek.
While Linear A and Linear B share 60 core syllables and use similar logograms (symbols representing whole words), Linear A remains difficult to decipher due to a significantly smaller corpus. The total remaining text consists of approximately 7,500 characters spread across 1,500 inscriptions, meaning many documents are merely short inventory lists with limited grammatical context.
Expert Skepticism and Technical Counterpoints
While the claim is being reviewed by academic institutions, members of the linguistics and AI communities have raised several critical points regarding the plausibility of the decipherment:
The Small Corpus Problem
Critics argue that the extremely limited amount of available text makes it difficult to prove a decipherment is correct. With only ~7,500 characters, there is a risk that a researcher could "solve" the script through coincidence or over-fitting a hypothesis to a small dataset without a way to independently verify the results.
The Semitic Language Paradox
Some linguists question why a Semitic language would use a consonant-vowel syllabary. Most Semitic languages historically utilize consonantal roots (abjads) because their morphology is based on three-consonant stems with interweaving vowels. The use of a syllabary for a Semitic language is an unusual structural choice that requires further explanation.
Etymological Gaps
If Linear A were Semitic, critics suggest that Semitic etymologies should be evident in the "substrate" words found in Ancient Greek that have no known origin. The absence of these clear links is cited as a potential weakness in the Semitic language hypothesis.