What Is a Controlled Natural Language?

Translators frequently have to confront issues that are often the result of problems existing in the source text we are translating. Some of the most common of such errors include omissions, the use of structures that are difficult to understand, and even the use of abbreviations only known by the person who wrote the original text (i.e. replacing the word “control “with the letter “c,” or “production” with “pd”), which, if not cleared up in advance, cannot be understood.

Given these issues, something called “controlled natural language” exists. As its name implies, it is the use of an existing language, such as English, Spanish, or any other language, to create content, but while taking into account some limitations that will facilitate understanding (and subsequent translation or post-editing).

To achieve this, the most important goals are to try to eliminate unnecessary ambiguities and, more importantly, to reduce lexical options. This means that if there are two ways of referring to a component of a machine, for example, then the same term should always be used, as this will prevent problems when looking for the equivalent in the target language. Another important point to consider is punctuation: very long sentences and a lack of commas and parentheses may significantly affect the quality of future translations. In addition, abuse of pronouns when references are unclear, spelling errors, and grammatical inaccuracies must be avoided—although this is always the case, not just with controlled languages​​!

Those who usually apply these rules are content creators who produce a very high volume of texts and are aware that, at some point, translations of that content to one or more languages will be needed. In addition, it is also important to note that a controlled natural language can only be used in some areas, such as computer science and engineering, among other technical fields. On the other hand, publisher of literary texts, for example, can hardly comply with these limitations.

Using a controlled natural language is of great benefit to the clients that apply it. Not only does it achieve a better quality translation, but, in addition, it makes the original text more clear and precise. Furthermore, the systemization of the process facilitates the work of content creators.