SDSU AI Finds Armenian Alphabet Closest to Ethiopic
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- San Diego State University researchers used AI to analyze over 28,000 images of Ethiopic characters and compare them to Armenian, Georgian, and Caucasian Albanian alphabets.
- Armenian alphabet showed the strongest shape similarity to Ethiopic letters, even comparable to the similarity between Ethiopic and its own earlier version.
- Caucasian Albanian alphabet exhibited a moderate similarity score to Ethiopic characters, according to the AI model.
- Georgian alphabet had some similarities but they were less consistent than those of Armenian or Caucasian Albanian.
- Latin alphabet showed much lower similarity to Ethiopic letters than any of the Caucasian alphabets tested.
- Sam Kassegne, professor of mechanical engineering and lead investigator, said the method makes criteria explicit and mathematical, enabling reproducible results.
- Daniel Zemene, SDSU graduate student and first author, highlighted that the AI model identified Armenian as the closest structural match to Ethiopic without any historical data.
Why it matters: The study gives historians a reproducible, quantitative method to evaluate script relationships, shifting reliance from subjective visual inspection to data‑driven analysis, and provides new evidence for possible cultural contact between the Caucasus and Ethiopia, highlighting AI’s broader utility in humanities research.
Ask SkimNews
