Open main menu

Controlled language

subset of a natural language

Controlled natural languages (CNLs) are artificial languages built by simplifying the grammar and reducing the number of words in the language to avoid ambiguity or complexity.

There are two types of controlled languages: some designed to help non-native speakers of a language, and some designed to enable automatic semantic analysis.

The first type of languages are often called "simplified" or "technical" languages. They are used by businesses or industries to improve the quality of technical writing. The standard terms simplify the semi-automatic translation to other languages. Examples are: the ASD Simplified Technical English, Caterpillar Technical English, and IBM's Easy English.

These languages guide the writer by general rules such as "write short and simple sentences", "use the person's name ("John Smith" for example) instead of saying "him", and "talk about who is doing something, rather than what is being done, unless you need to say what is being done".[1]

The second type of languages have well defined rules for writing and understanding. These rules can be matched to formal languages, such as first-order logic. The languages can be analysed by computer. They can be checked to see they are correct. They can be searched for information easily.



Existing logic-based controlled natural languages include:

Other existing controlled natural languages:

Related pagesEdit


  1. Muegge, Uwe (2007). "Controlled language: the next big thing in translation?". ClientSide News Magazine (ClientSide Publications) 7 (7): 21–24. 

Other websitesEdit