# Different languages use-cases in information encryption?

Because of my lifestyle and journeys to different countries I became a "multi-lingual person"; So, I wonder whether there are any use-case in encryption for multi-lingual persons or not? Such as end-human-cryptography - even after when two sides that attempting to connect (using their digital encryption systems such as hashing, etc) and using their keys have decrypted their phrases in each side, but still no one except these two persons could easily understand the decrypted phrase because each syllable or even smaller letter-sets (which solely may not have any meaning in each one of these different languages) which has been intelligently merged as one word written in Latin alphabet (for example from different dialects of Greek, Persian, Thai, Russian and Korean, which even has been mistyped knowingly). Even with considering homophones (same pronunciation different meaning which hardly navigate the attacker to a specific language) And also using the synonyms or even antonyms instead of each part which has been replaced and finally finally after decryption of these words from syllables or even meaningless sets, still the whole sentence and message is a sarcastic phrase or proverb barely searchable or understandable.

I think there must be this use-case in encryption and information security, but do you think it could be used more often (for example in Europe)?

• Please whoever gives down-vote, remark a reason in comments! because I don't still know if I did something wrong in the question. – Armin Jan 25 '17 at 11:00
• Comments are not for extended discussion; this conversation has been moved to chat. – e-sushi Jan 25 '17 at 17:07

The problems here are that:

• If the method for obfuscating the message has definite, explicit, well-documented rules, then they are part of the encryption algorithm and therefore by Kerckhoff's Principle we should assume the attacker knows it.
• If the method is informal and has no set rules, then it likely will just obscure the meaning of the message from the intended recipient. In the worst case, the recipient will misunderstand the message.

A real-life example of the latter is the "the world wonders" incident between American Admirals Chester Nimitz and William Halsey, Jr., in 1944 during the Battle of Leyte Gulf. In order to mitigate against certain cryptanalytic attacks, it was the practice of US Navy cryptographers to pad plaintext messages at the beginning and end with English phrases unrelated to the message. Nimitz sent the following message to Halsey:

Where is, repeat, where is Task Force Thirty Four?

Nimitz's cryptographer padded it to this:

TURKEY TROTS TO WATER GG FROM CINCPAC ACTION COM THIRD FLEET INFO COMINCH CTF
SEVENTY-SEVEN X WHERE IS RPT WHERE IS TASK FORCE THIRTY FOUR RR THE WORLD WONDERS


Halsey's cryptographer, after decoding it, misunderstood which parts were padding and gave this to Halsey:

Where is, repeat, where is Task Force Thirty Four? The world wonders.

The inclusion of "the world wonders" led Halsey to misunderstand the message to be a personal attack on him, and he did not react well:

The message (and its trailing padding) became famous, and created some ill feeling, since it appeared to be a harsh criticism by Nimitz of Halsey's decision to pursue the decoy carriers and leave the landings uncovered. "I was stunned as if I had been struck in the face", Halsey later recalled. "The paper rattled in my hands, I snatched off my cap, threw it on the deck, and shouted something I am ashamed to remember", letting out an anguished sob. RADM Robert Carney, Halsey's chief of staff (who had argued strongly in favor of pursuing the carriers), witnessed Halsey's emotional outburst and reportedly grabbed him by the shoulders and shook him, shouting, "Stop it! What the hell's the matter with you? Pull yourself together!" Recognizing his failure, Halsey sulked in inactivity for a full hour while Taffy 3 was fighting for its life—falsely claiming to be refueling his ships—before eventually turning around with his two fastest battleships, three light cruisers and eight destroyers and heading back to Samar, too late to have any impact on the battle.

When you talk about using synonyms and antonyms, well, I can't help but think you're asking for trouble.

Let me answer your last comment directly here, it would be too long to fit the comment section. Your question seems a bit opinion based so my answer can call some discussions afterward if you do not agree with it.

It seems to me that you base the security of your idea on the following concept: two individuals want to communicate, and they share a set of common language that they both understand. To do that, the first player transforms his message $m$ into a message $m'$ which is $m$ cut in carefully chosen blocks, each translated to one of the common languages of the players. Then, $m'$ is encrypted and authenticated as usual.

My claim is that this $m \rightarrow m'$ step does not really add any security to the protocol. To see that, imagine an attacker who has already compromised, in one way or another, the encrypt and authenticate part; meaning, the attacker has recovered $m'$.

There are two possibilities. First, suppose the attacker knowns the set of languages used by the players. If so, he can easily build a software with full access to dictionaries and translators for all of those languages. Given that, he has essentially as much chance as the second player (the player that must receive $m$), as he will follow the same strategy than him to recover $m'$: scan from your knowledge of each language until you find a valid interpretation for the meaning of a block, then mode to the next block. With a software, this should really be trivial. Security is not about adding, perhaps a few hours of work, but about making an attack essentially infeasible given any reasonable time and computational power.

Suppose now that the attacker does not know the common set of languages of the players, and so they can use this as follows: use words that have meanings in several languages, so that their interpretation will be ambiguous for the attacker which does not know the set of languages. First, it should probably be still easy to attack with the same program, identifying all the valid languages and interpretations in those languages, and listing all the possible outcome. Second,the main observation is that using this shared secret information to ensure secrecy is simply equivalent to sharing a common password, that both players know by heart. But it is trivial to observe that one can always encrypt a message with a simple password that is only stored in his head before sending it, to ensure that "some" level of security is maintained. And if this is what you want to do, then the set of common languages is a very bad choice of password: browsing the net, doing some investigations, will probably make it very easy for an attacker to have a good knowledge of which languages the two players are likely to speak.

All in all, it boils down to this: in cryptography, you want systems that provably defeat all attacks, even mounted with powerful software and lots of computational power. Using such simple, human based scrambling of a message is very unlikely to add any concrete security to your system.

• Comments are not for extended discussion; this conversation has been moved to chat. – e-sushi Jan 25 '17 at 17:15