Reusing strings freezes the translation's grammatical form

What it is

Grammatical coupling from string reuse occurs when several UI placements reference one translation resource and are forced to share a target-language form despite requiring different case, gender, number, part of speech, or capitalization. A standalone word reusable in the source language is not necessarily an independently substitutable unit in the target language.

Why it happens

Resource reuse implements “same content” as “one key has one value.” Correct word form depends on syntax: a menu noun may be nominative while the same noun after a preposition needs another case; heading forms, button imperatives, and sentence verbs can differ. Selecting on a message's own semantic variables such as number or gender is legitimate. Coupling appears when unrelated placement or use-site IDs are added to preserve a falsely shared token, making one resource depend on all callers.

Where it stops holding

Complete messages are safe to reuse only when semantics, speech act, parameter contract, and change lifecycle coincide; proper names can be reused when the target language confirms invariance. The rule is not to duplicate every repeated word, but the reuse boundary must be no smaller than the unit needing grammatical freedom. Separate keys linked by translation memory suggest consistency; one runtime key enforces it.

Applying it

• Split keys by linguistic act and syntactic placement even when source text currently matches; name keys semantically rather than using English copy as identifiers.
• Reuse complete messages with one semantic contract. A structured terminology entity supports inflection only when it carries locale-specific grammatical metadata or lets the complete message select an appropriate form.
• Use glossaries and contextual translation-memory suggestions to encourage consistency while allowing separate keys to inflect differently.
• Maintain a use-site report; when one requirement changes, detect whether unrelated callers are forced to change, then render languages rich in case, gender, and number to find forced identical forms.

Related

• Same group: S1.05.1 The same source word needs different translations in different placements · S1.05.3 Strings need usage-context descriptions · S1.05.4 Interface screenshots provide the most effective context
• Adjacent: S1.04 Plural, gender, and grammatical variation · S1.06 Word-order problems in concatenated strings
• Search terms: shared string localization · grammatical case · message key design