BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//conference-hub.linguistic-society.com//athens-2026//spea
 ker//RJJBAF
BEGIN:VEVENT
UID:pretalx-athens-2026-EMAGVD@conference-hub.linguistic-society.com
DTSTART:20260422T151500Z
DTEND:20260422T153000Z
DESCRIPTION:Abstract\nThe study contributes to current research on the appl
 ication of AI to classical languages and offers\ninsights into how far Mac
 hine Translation (MT) systems have progressed in the translation of ancien
 t\ntexts. In this paper\, we present a comparative investigation of MT fro
 m Ancient Greek to Italian\, focusing\non the performance of ChatGPT as an
  AI language model applied to a low-resource classical language. In\nthis 
 case\, Ancient Greek poses substantial challenges for computational proces
 sing due to a lower\navailability of digital corpora\, as well as its ling
 uistic complexity: highly inflectional morphology\, flexible\nword order\,
  extensive use of particles\, and pragmatic ambiguity. Within this context
 \, the study aims to\nassess to what extent ChatGPT can produce translatio
 ns that are not only formally well-structured\, but\nalso philologically p
 lausible. The analysis is conducted on a selected corpus of Ancient Greek 
 passages\ndrawn from different genres and historical periods\, with author
 s such as Homer\, Hippocrates\, and\nSappho. The methodology is the same f
 or the analysis of each passage: the Italian translation generated by\nCha
 tGPT is systematically compared with authoritative published human transla
 tions\, so that the pros and\ncons of the result emerge along with any sim
 ilarities or differences. The assessment of outputs is both\nquantitative 
 and qualitative. In fact\, both the frequency and types of errors encounte
 red are analysed\, as\nwell as a linguistic interpretation based on morpho
 syntax and\, of course\, a lexical check that also observes\nthe correspon
 dence of culturally significant terms. The results indicate that ChatGPT g
 enerally produces a\nfluent Italian output and shows a noteworthy ability 
 to capture the global meaning of many passages\; at\nthe same time\, the s
 ystem often opts for oversimplified or generic solutions. Performance also
  varies\naccording to genre\, with greater instability in poetic and fragm
 entated texts. These findings suggest that\,\nalthough ChatGPT in its curr
 ent state cannot be considered a reliable autonomous translator for Ancien
 t\nGreek\, it can function as a supportive tool within a post-editing fram
 ework\, potentially reducing the initial\nworkload of human translators.\n
 Keywords: Ancient Greek\, Italian\, Machine Translation\, AI
DTSTAMP:20260419T080901Z
LOCATION:Online Session
SUMMARY:Ancient Greek-to-Italian Translation with ChatGPT - Riccardo Vicari
  | Castrenze Nigrelli
URL:https://conference-hub.linguistic-society.com/athens-2026/talk/EMAGVD/
END:VEVENT
END:VCALENDAR