How to make a language/code indecipherable or very hard to decipher/learn? What are the key points? How does linguists decipher languages?


2 Answers 2


Languages are governed by certain laws/principles, that reflect the relationships between elements. For example, element frequency: you generally have a few elements which are very common, and a long tail of rare events.

To make a language hard to learn, break those rules. Make each word/phoneme/letter equally frequent. Common words are short and have several meanings, long words are rare and generally have fewer meanings (as they are more specific). Turn that the other way round. Make long words frequent and unspecific.

Early codes where cracked with frequency tables, ie the letters "e" and "t" being the most common symbols, so substitution ciphers were easily solved. Mix up the spelling of your words to make them semi-random (maybe add some random letters in the middle that are ignored).

The language you will end up with this way will be pretty unusable, and very hard to learn.


All languages have patterns that can be decerned. Word frequency lately has been used to weed out some of the more common words like personal pronouns. Proper nouns also tend to get borrowed as-is into languages.

That said, the languages of our world that haven't been deciphered carry characteristics that would be hard to replicate in a conlang.

For one, the biggest problem is often just the lack of texts. Most of the undeciphered languages of our planet have few surviving texts to draw upon. A good example of this is Etruscan, which remained undeciphered until a very extensive text was discovered on the wrappings of an Egyptian mummy. Thus, anything with a lot of text will be easily deciphered.

Another thing is alieness; its not hard to decipher a language that has a close relative that is well-known. No romlang is going to fool anyone. This is probably one of the few things that could be replicated in a conlang.

Another is the writing system; some languages just have ultra elaborate orthographies, and extinct ones are no exception to this.

Yet another is the lack of bilingual texts; the ancient Egyptian language remained undeciphered until the famous discovery of the Rosetta stone. This could be a problem for a conlang; you have to record the grammar and vocab down SOMEWHERE, and its NEVER a good idea to record it in all the conlang! So any conlang you make, will require you to write materials in a known language that someone else could learn the language from!

So, basically, its a futile effort. You're probably better off using some form of cryptography (though a modern computer can easily beat any cipher you can perform by hand). Of course, you could do somethings to make it harder to learn (such as adding in a lot of irregularities, or some crazy complicated morphology), but there is no way to make it 'impossible'.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.