Shifts in political discourse often reflect societal changes, but quantifying how lexical usage and semantic representations vary over time remains a challenge for computational linguistics. To compare approaches under different computational constraints, we propose a cost-aware, modeland language-agnostic framework to compare lexical and semantic change signals. We apply it to Brazilian Portuguese political discourse using a shared panel of lemmas combining drift candidates, stable controls, and theory-driven seeds. We compare three families of diachronic drift signals: a TF-IDF lexical-salience baseline, aligned slice-specific Word2Vec models, and a contextual BERT approach. Results indicate disagreement between cheaper baselines, while contextual BERT remains weakly aligned with each method. A filtered contextual panel reduces stable-control leakage and highlights drift candidates such as “bloqueio” and “salário”. Rather than treating any detector as ground truth, this study presents an exploratory comparison of complementary drift sensitivities, interpretability, and computational cost in Brazilian Portuguese political discourse.

Reference-free translation audits have little to work with when a survey answer is only one or two words long. A translation may be reasonable and still be nearly impossible to reconnect to its source among thousands of similar answers. We examined this problem in 44,118 Portuguese–English response pairs from 14,706 survey-specific respondent records across four workplace surveys. One third of the pairs addressed mental health or well-being. Our analysis kept duplicate-aware retrieval, COMETKiwi quality estimates, language identification, and affective-lexicon coverage separate because they answer different questions. The retrieval encoder found the paired target or an exact normalized duplicate at rank one for 71.80% of English queries and 73.37% of Portuguese queries overall. Response length changed the picture. Rank-one recovery ranged from 8.39–12.96% for one- or two-word answers to 98.45–99.08% for answers longer than 20 words. COMETKiwi scores did not rise steadily with length, and about one quarter of responses in either language contained no emotion term covered by the lexicons. A failed match therefore does not, by itself, identify a poor translation. For brief and repetitive survey text, reference-free audits should report how distinguishable the response is and how much of it the diagnostic resources cover.