This article examines direct lexical loans—the unadapted retention of English forms—in Norwegian literary translation from English (1900–2020), with special attention to three culturally dense lexical domains: proper names (anthroponyms), titles/honorifics, and toponyms. These items are analytically productive because they compactly encode setting, social hierarchy, and etiquette, and therefore create recurrent points of (non-)equivalence where substitution risks flattening sociocultural cues. Building on a decision-making view of translation and a descriptive-norms framework, the study treats lexical borrowing not as an automatic by-product of contact but as a patterned translator choice regulated by expectations and conventions. Methodologically, the analysis uses a chapter-based set of Norwegian translation excerpts and codes each relevant lexical item along two dimensions: Lexical Category (names, titles, toponyms, rank/role labels) and Transfer Operation (Retention as direct loan vs. Replacement via culturally adapted Norwegian equivalent). Retentionis interpreted as lexical-level foreignisation, whereas replacement is treated as lexical-level domestication, without making macro-level claims about the translation. Results show a highly stable dominance of retention for English titles (e.g., Mr., Mrs., Miss, Sir, Lady), personal names, and place names (e.g., Netherfield, Longbourn, Meryton, Pemberley), which jointly function as a lexical scaffold that “keeps England English” by preserving visible markers of social order and narrative geography. A contrastive pattern is also attested: at least one rank/role label is rendered via a culturally adapted Norwegian equivalent (oberst), indicating selective domestication within an otherwise retention-oriented lexical environment. The article contributes a replicable micro-level coding approach and highlights the need for broader diachronic sampling to substantiate period-wide generalisations.