Supplementary materials for: Computational Analysis of the Mother Archetype in Russian and Uzbek Women's Prose

This paper presents a cross-lingual NLP framework for the comparative analysis of the mother archetype in contemporary Russian and Uzbek women's prose. It is based on works by Masha Traub and Zulfiya Qurolboy qizi. A bilingual corpus of approximately 261,000 tokens was annotated at the sentence level. The five-category archetypal sub-role scheme is grounded in Jungian theory (Krippendorff's α = 0.75 for Russian, 0.71 for Uzbek). The pipeline combines LASER sentence embeddings, sentiment profiling, and TF-IDF lexical-semantic field mapping with XLM-R clustering. Sub-role distributions diverge sharply between the two authors. Mother-Identity Seeker accounts for 27.4% of Russian annotations but 5.2% of Uzbek, while Mother-Keeper of Tradition and Mother-Sacrifice dominate the Uzbek subcorpus. Cross-lingual cosine similarity is highest for Mother-Sacrifice (0.78) and lowest for Mother-Identity Seeker (0.43). Uzbek sentiment scores are obtained by cross-lingual transfer and are reported as experimental rather than measured. Expert validation yields P@10 of 0.83 (Russian) and 0.79 (Uzbek). The results indicate that archetypal variation is computationally measurable, though generalization beyond these two authors requires broader corpora.