A word that shares a common origin with a word in another language, typically resulting in similar form and meaning, and a genuine asset for translators working across related languages.
The word “cognate” comes from the Latin cognatus, meaning “related by birth.” In linguistics, cognates are words across different languages that descended from the same ancestral root. Because they share an origin, they often look similar, sound similar, and carry similar meanings. For a translator working between related languages, true cognates are a genuine asset, they reduce ambiguity and speed up recognition.
Classic examples include “night” (English), Nacht (German), noche (Spanish), notte (Italian), and natt (Norwegian), all descended from the same Proto-Indo-European root. Similarly, “accident” in English and accidente in Spanish, or “father” and Vater in German, are true cognates that work in a translator’s favor.
Languages within the same family (Romance languages like French, Spanish, Portuguese, and Italian, or Germanic languages like English, Dutch, and German, share large numbers of cognates. This is one reason translation between closely related language pairs tends to be faster and more predictable than between unrelated ones. A translator experienced with Spanish can draw on significant cognate overlap when working with Portuguese or Italian.
For localization teams, cognates matter in two practical ways. First, they allow translators to recognize terminology more reliably across related languages, which supports consistency in multilingual projects. Second, understanding which words are true cognates, versus which only appear to be, is a fundamental skill for avoiding the much more troublesome category of false friends.
These three terms are frequently mixed up, including in professional contexts. The distinctions matter:
True cognates share both a common origin and a similar meaning. They are reliable across languages and rarely cause translation errors. Examples include “Communication” (English) and Comunicación (Spanish) or “Music” (English) and Musik (German).
False cognates look or sound similar across languages but do not actually share a common etymological origin, the resemblance is coincidence. The English word “dog” and the Australian Aboriginal word dog (from the Mbabaram language) sound identical and mean the same thing, but are entirely unrelated. The similarity is accidental, not ancestral.
False friends (also called faux amis) are words that look or sound similar across languages but carry different meanings, regardless of whether they share an origin. Some false friends are genuine cognates that drifted apart in meaning over time (like English embarrassed and Spanish embarazada, or English library and Italian libreria, which means bookshop). Others are false cognates where the resemblance was coincidental to begin with. What unites them is the practical risk: they look safe to a translator but are not.
Localazy doesn’t have a “cognate detector,” but it uses Glossaries and other context-rich features like Context Screenshots and Comments to keep them from wrecking your localization. By adding known false friends (like actual vs. actualmente) to the Glossary, translators get an instant warning in the editor. This prevents “auto-pilot” errors where a translator might instinctively pick the word that looks right but means something entirely different.
When adding tricky terms to the Glossary, you provide a “source of truth” that overrides the generic guesses of Machine Translation (MT) engines. When using Localazy AI, the engine “reads” your glossary and context notes before generating a suggestion, ensuring it doesn’t fall for linguistic traps.
Learn more about setting up Glossaries in our docs.