As neural and generative machine translation lower the cost of rendering scholarship across languages to near zero, a hopeful assumption circulates: that automation will finally democratize the international circulation of knowledge from non-Anglophone traditions. This paper argues the opposite. Building on a study of China's state-sponsored academic translation program—where fifteen years of investment produced symbolic presence without citation, uptake, or dialogue, a "presence trap"—it asks how the algorithmic turn reshapes the politics of recognition in cross-cultural academic exchange.
Drawing on Heilbron's world-system of translation and Honneth's theory of recognition, the paper advances three claims. First, machine translation does not neutralize the core–periphery asymmetry of global knowledge but automates it: trained on corpora calibrated to Anglophone norms, it defaults to the highest-probability rendering, assimilating concepts to their nearest dominant-language counterparts wherever no established equivalent exists. Second, by making symbolic presence almost costless to produce, automation encourages a further expansion of supply-driven knowledge export. Yet this risks deepening rather than escaping the presence trap, since the obstacle has never been a lack of translation but a lack of the citation, contestation, and dialogue through which a scholarly community confers recognition. Third, drawing on the author's experience evaluating state-funded academic translations, the paper shows that the decisive interventions—reconstructing a text's argumentative function, connecting a concept to an international debate, preserving the rhetorical identity through which intellectual authority is claimed—are precisely what algorithmic translation and its evaluation metrics (BLEU, COMET, LLM-as-judge) cannot perform.
Recognition, the paper concludes, is relational: it must be conferred by participants within a scholarly field, and algorithms can simulate its surface but cannot enact it. Expert human translation criticism is therefore not a residual craft awaiting replacement, but the site where misrecognition is either reproduced or repaired. It constitutes the distinctly human remainder that persists within, and perhaps limits, the logic of automated translation.
Yanyu Zeng is Professor of English and Dean of the Foreign Studies College at Hunan Normal University, China, and a Changjiang Scholar Distinguished Professor appointed by China's Ministry of Education. She is Executive Editor-in-Chief of Foreign Languages and Cultures and President of the Hunan Translators Association. Her research spans contemporary American literature, the aesthetic criticism of foreign literature, and the cross-cultural circulation of knowledge, with recent work focusing on the sociology of academic translation and the politics of scholarly recognition. The author of two monographs and numerous articles in leading academic journals, she has led several projects funded by the National Social Science Fund of China and serves as an invited reviewer for state-sponsored academic translation programs. She has held visiting appointments at the University of Cambridge, Columbia University, and Emory University.
Minmin XIE, PhD, is a associate professor at the Foreign Studies College, Hunan Normal University, and an assistant research fellow at the Hunan Provincial Research Base for Cross-Cultural Communication. His research focuses on American literature and translation practice. He has led and participated in several national and provincial research projects and has published articles in leading journals such as Foreign Literature Studies. His translation, A New Model of Greek Economic Development, was included in China’s 13th Five-Year Plan National Key Publication Program. He is the recipient of the First Prize of the Fourth Hunan Provincial Award for Outstanding Educational Research.
