Researchers have developed an AI model which deciphered texts missing from ancient Greek inscriptions and even dated them.
Machine learning is increasingly offering us new techniques to understand the past.
Now, research reveals that artificial intelligence (AI) could bring ancient texts to life too.
Writing in the journal Nature, a group of researchers reported how they built an AI system that can fill in gaps in ancient Greek inscriptions and locate where and when they're from.
Nicknamed Ithaca – after the Greek island that was home to the legendary King Odysseus – the model is a deep neural network architecture trained to simultaneously perform the tasks of textual restoration, geographical attribution and chronological attribution.
The research team fed Ithaca more than 63,000 transcribed ancient Greek inscriptions, enabling it to decipher patterns in the order of letters and words, including associations between words and phrases as well as a text’s age and provenance.
Overall, the model was 62 percent accurate when restoring letters in damaged texts. In can attribute an inscription’s geographic origins to one of 84 regions in the ancient world with 71 percent accuracy and can data a text to within 30 years on average of its known year of writing.
“Ithaca is to our knowledge the first model to tackle the three central tasks in the epigrapher’s workflow holistically,” the study’s authors said.
“It is our hope that this work may set a new standard for the field of digital epigraphy, by using advanced deep learning architectures to support the work of ancient historians.”
“Just as microscopes and telescopes have extended the range of what scientists can do today, Ithaca aims to singularly augment and expand the capabilities to study one of the most significant periods of human history,” said Dr Yannis Assael, co-author of the work from AI firm DeepMind.
Researchers highlighted the model’s flexibility, claiming that it can be applied to any ancient language from Latin, Mayan to Cuneiform. It also might be possible to train the system on Greek literary texts written on fragments of papyrus, which might shed light on the writings of poets like Sappho.
Dr Thea Sommerschield, co-author of the research at Ca’ Foscari University of Venice and Harvard University, said inscriptions were important as they were written directly by ancient people and were evidence of the thought, language, society and history of past civilisations.
“But most surviving inscriptions have been damaged over the centuries. So their texts are now fragmentary or illegible,” she said, adding that they may also have been moved from their original location, while methods such as radiocarbon dating were unusable on materials such as stone.
Researchers also said Ithaca had been applied to date a set of decrees found on the Acropolis of Athens. One of them – related to a collection of tributes during the Athenian empire – was dated to 424 BC rather than 448-7BC as was previously thought.
“Although it might seem like a small difference, this 30-year shift has momentous repercussions for our understanding of the political history of classical Athens, and helps us better align literary sources – such as Thucydides’ account of these years and events – with the epigraphic record,” said Sommerschield.
While promising so far, it’s important to note that AI models like Ithaca are not capable of operating independently of human expertise.
Professor Peter Liddel, an expert in Greek history and epigraphy at the University of Manchester, cautioned that like scholars, AI was limited by gaps in the ancient record.
“AI is only powerful as a tool to help us ask questions about, and make comparisons to, the existing evidence, he said.
The Ithaca software and its open-source code is available online for public use.