From Text to Image: A Multimodal Linguocultural Analysis of Uzbek and English Fairy-Tale Heroes Created by AI Image Generators

For generations, children have pictured fairy-tale heroes in their own minds: Alpomish riding across the steppe, Zumrad in her colourful atlas dress, a fearsome Dev in his mountain cave, Jack climbing the beanstalk, a witch in her dark cottage. Today these pictures are increasingly made by machines. Text-to-image generators such as DALL·E, Midjourney and Stable Diffusion can turn a character’s name into a detailed illustration in seconds, and teachers, illustrators and parents already use them. This raises a question that is both linguistic and cultural: when a verbal sign such as Dev, Pari or Barchinoy is turned into a visual image, which cultural meanings survive the journey from text to picture?