passivepreserving

Passivepreserving is a term used in linguistics and computational linguistics to denote the deliberate preservation of passive voice constructions in text analysis, translation, and generation. It contrasts with practices that normalize voice to active, aiming instead to retain the syntactic structure and stylistic emphasis of the original language.

In corpus linguistics and natural language generation, passivepreserving refers to annotation schemes and generation algorithms that avoid transforming passive clauses into active equivalents. It is used to study how writers use passive voice, track shifts in voice across translations, and preserve emphasis in stylistic analysis. Implementation often involves detecting passive constructions, retaining them in output, and marking the degree of preservation.

Example: source sentence “The window was opened by the janitor.” In a passive-preserving pipeline, this sentence would be kept passive; whereas an active-normalizing pipeline might render “The janitor opened the window.” The concept is about preserving voice, not about forcing a particular grammar.

Limitations and discussion: Preserving passive voice can reduce readability in some contexts or obscure agent information. Critics argue that voice preservation should be balanced with clarity and audience needs, especially in translation and education. Proponents contend that preserving voice helps retain authorial style, emphasis, and historical or genre-specific features.

See also: Passive voice, Voice in linguistics, Translation studies, Natural language processing, Stylistics.