Ethical Reconstruction of Linguistic Signs: An Ethical Study of Machine Translation Based on Linguistic Ontology

This study confronts the ethical crisis in machine translation (MT) caused by the systematic erosion of cultural semantics and symbolic integrity. While current MT systems achieve high technical performance (e.g., 72.3 BLEU scores in WMT2022), they fail to preserve cultural-contextual nuances, with 34.8% mistranslation rates for culturally loaded terms and 73% semantic reduction for low-resource language symbols. Through deconstructive analysis of Saussure’s arbitrariness principle and Wittgenstein’s language game theory, we demonstrate how algorithmic compression of dynamic semantic networks into unidirectional referential chains perpetuates symbolic violence, exemplified by the 89% loss of Confucian ethical dimensions in translating Chinese ren (benevolence) and 82% erasure of spiritual connotations in Arabic jihad. To address these issues, we propose the Symbolic Ethics Sensitivity Assessment Model (SESAM), a tripartite framework evaluating cultural fidelity, symbolic violence reduction, and ethical risk mitigation on a 0-5 scale. Validated in Meta’s No Language Left Behind project, SESAM increases cultural metaphor retention by 23.6% while maintaining translation efficiency. By integrating Peircean semiotics with computational linguistics, we construct a Cultural Semantic Vector Space using GloVe embeddings, resolving 93% of Traditional Chinese Medicine concept mapping failures. The study further develops the Linguistic Ethics Entropy (LEE) index, benchmarked at ≤0.23 for EU AI Act compliance, as the first quantitative metric for ethical MT. Our findings reveal three advancements: 1) semiotic topology exposing 47:1 corpus imbalance-induced cultural hegemony, 2) Ambiguity Markup Language (AML) reducing medical translation risks by 40%, and 3) an ontological framework reconciling Heideggerian linguistic essence with algorithmic probability through Derrida’s différance-encoded architectures. These interdisciplinary innovations bridge translation ethics, philosophy, and AI, offering actionable solutions for cultural DNA preservation and equitable cross-linguistic communication.