Reading the Past: From Prehistoric Symbols to Writing and AI
My interest in archaeology began in childhood, exploring my father’s collection of Neolithic stone tools and discovering ancient civilizations through C. W. Ceram’s Gods, Graves, and Scholars and the illustrated Time-Life history collections. During my doctoral research in Europe, encounters with the Rosetta Stone, the Stele of Hammurabi, and the Victory Stele of Naram-Sin deepened that interest, while the still-undeciphered Phaistos Disc, which I saw in Crete in 1999, confronted me with the limits of recovering lost meanings. In 2026, a controversial study applying information theory and machine learning to 40,000-year-old geometric markings revived these questions. My central argument is that the history of writing cannot be understood solely as the invention of a script: it belongs to a much older human effort to preserve and transmit information beyond individual memory. Yet statistical regularity is not evidence of language, and resemblance between symbolic systems does not establish that they served the same purpose. Connecting personal experience with archaeology, mathematics, and artificial intelligence, this essay examines what prehistoric markings may reveal about the evolution of human cognition, what modern algorithms can genuinely establish, and why recovering meaning requires more than identifying patterns.