With the advent of the digital era, large language models (LLMs)have exerted a profound influence on humanities research, particularly in the domain of foreign language studies. Considering the fundamental issue of literature, this article delves into the “imitation” nature of LLMs, positioning them as potent tools for expanding the horizons of language processing, especially when juxtaposed with the notable advancements they have achieved over traditional machine translation. It critically reexamines the challenges confronting humanities research in the age of digital intelligence and identifies pivotal concerns that future digital humanities research should address. Based on the essence of language in relation to human life, embracing an interdisciplinary perspective in digital humanities research, reevaluating the facets of human rationality and establishing a humanistic foundation for the field, our discussion aims to delineate the inherently humanistic nature of digital humanities research, hoping to offer reflections and insights that illuminate the evolving relationship between LLMs and humanities disciplines, particularly foreign language studies, as well as to provide directions for literary translation research within the context of these technological advancements.