Nomads and monuments in translation: AI-supported deconstruction of binary oppositions in Russian–English archaeological texts and student translations

Abstract

This study proposes an AI-supported Deconstruction Analyzer v 2.0 Tool for teaching translation students how to translate archaeological texts from Russian to English and vice versa. This tool was created because contemporary archaeological discourse often uses binary oppositions like nomadic/sedentary, civiliza- tion/barbarism, and center/periphery, which can reproduce colonial and hierarchical stereotypes in original texts and translations. We tested our tool on a corpus of ten English-language archaeological papers, 38 stu- dent translations, and five museum podcasts on Central Kazakhstan’s legacy. The program uses curated lexi- cons of archaeologically salient opposites and syntagmatic and paradigmatic modules to detect explicit polar word collisions within sentences and display pole-centric lexical asymmetries across texts. An AI layer using transformer-style language models expands each pole’s vocabulary, recognizes paraphrases, and highlights evaluative patterns without replacing human interpretation. The study shows that accounts of nomadic life- styles and monuments often have too much sedentary and “civilized” framing, and that student translations of museum podcasts often have ideologically charged clashes. Student reflections show enhanced awareness of unconscious bias, lexical sensitivity, and enjoyment of visual analytics for translation revision. The study suggests that incorporating AI-mediated deconstruction into translation education promotes critical thinking and decolonization.