Steganography in speech by adding extra “filler” syllables [closed]

Question

PREAMBLE

About 20 years ago, when reading a novel (set in late 1800) by some classic Russian writer I came across very curious example of steganography. The two main characters were peasants who got lost in the woods. Suddenly they come across wood cabin where a very dodgy forester lives. They have no other choice but to accept his "hospitable invitation" and stay in his cabin for the night. All three are in the same room in the cabin, but as they're scared of him harming them, they try to discuss between themselves a plan of how can they protect themselves. As their conversation happens "in the open" of the cabin room, they have to somehow ensure that the forester doesn't understand them.

THE STEGANOGRAPHY TRICK

Alice, Bob and Charlie all speak same language and are in the same room. But Alice and Bob do not want Charlie to be able to understand them, so they start speaking in this "steganographic variation" of their language. Method is simple: after every syllable they add an extra syllable "pa" or "la". (For simplicity I will use Spanish language as example 'cos it has very syllabic and phonetic nature and you pronounce same way as you spell).

So for example:

HOLA would turn into HOpaLApa or hopalapa!

AMIGO would turn into AlaMIlaGOla or alamilagola!

(I don't think there were strict rules on when to put "la" or "pa", so speaker decides on the filler syllalble as he goes. Thus AMIGO can equally turn into ApaMIlaGOpa apamilagopa

Then a conversation

Hola Amigo! 
Quiero una cerveza por favor.

Will turn into

Hopalapa Alamilagola!
Quieparopa ulanala cerlavelazala papor fapa

This method seems very interesting and I never ever had heard of such an approach ever again. Neither ever met anyone who had heard about it. Meanwhile, there's few amazing things about this method: from little practice I did, it seems that learning curve to be able to speak that way and to be able to understand this kind if speech is not very steep (maybe a week to master). Which makes sense - as peasants with little to no education in 1800ds would need a low barrier of entry into steganographic method. Of course when spelled out on paper and when meaningful syllables are capitalized, this method seems to be trivial, but when speaking - an untrained ear can't understand a word.

So I wonder if anyone has ever heard of this (or similar) method or any accounts in another cultures, languages or environments?

ps. I am not sure whether this method can apply well to English, because it is not as "syllabic" and "sing-a-songy" like languages like Spanish, Italian, Portuguese, Russian and pronunciation varies very strongly region to region.

Answer 1

To the extent that I think this is a cryptographic question (which I don't!), I don't think these language games qualify as steganography. Steganography conceals the existence of a message by encoding it into (hopefully) undetectable modulations of another message, which is meant to appear innocuous to an adversary. As Wikipedia puts it:

The advantage of steganography over cryptography alone is that the intended secret message does not attract attention to itself as an object of scrutiny. Plainly visible encrypted messages—no matter how unbreakable—arouse interest, and may in themselves be incriminating in countries where encryption is illegal. Thus, whereas cryptography is the practice of protecting the contents of a message alone, steganography is concerned with concealing the fact that a secret message is being sent, as well as concealing the contents of the message.

But languages games like Pig Latin or jeringonza don't generally fool adversaries into thinking that the message is innocuous. Some of these games produce output that is linguistically very marked. For example, Hola amigo is pronounced ['o la 'mi ɣo] (International Phonetic Alphabet; stressed syllables are prefixed with apostrophe as is standard), with broadly trochaic stress (LOUD-soft-LOUD-soft) as is typical of Spanish (default penultimate stress). But Chi-ho chi-la chi-a chi-mi chi-go, as in the Caribbean variant of the jeringonza game, is strictly iambic (soft-LOUD-soft-LOUD): [tʃi 'o tʃi 'la tʃi 'a tʃi 'mi tʃi 'ɣo]. So this particular language game produces outputs that are systematically different from regular Spanish phonology. Nobody's gonna mistake this for small talk about the weather.

And while I stress that I don't think these language games are properly cryptographic, if anything they often resemble ciphers more closely than steganography. One capsule explanation of modern cryptography is that it achieves security by transforming messages using trapdoor functions—schemes where messages are transformed in ways that are cost-effective to honest parties but prohibitively expensive for adversaries. And if you squint really hard these language games approximate that paradigm—they transform a message in a way that the ingroup has mastered through practice (dare I call this "precomputation"?) but the outgroup hasn't.

Answer 2

The English definition of steganography is "The practice of concealing messages or information within other non-secret text or data." So with squinted eyes, a grimace and going "Hmmm" you might be convinced that it is actually steganography and therefore related to cryptography.

Without access to the original text, some of this answer is speculation but based on the following points extracted from the OP:-

• 19th century
• peasants -> socio-economic group
• "pa" or "la" after every syllable
• no strict rules

Considering this in the round, yes it might have worked well given the example scenario. And I'm sure that both survived unharmed. It's likely that something like the Soundex algorithm is operated by the human brain and allows decoding of noisy cipher text. But even Soundex has geographical variants preventing simple inter operation. There are formal techniques of secure and noisy data transmission, but these are a little beyond me.

It wouldn't work at all for practical cryptography though. This ad hoc cipher falls flat when trying to transcode the following examples where absolute precision is life critical:-

• Let's kill the forester using 10ml of hydrogen cyanide and bury the body at N50d 26' 33" W1d 10' 40"
• 80mm indirect fire to Hill 577 at 1305 hours
• Coldplay mp3 file the peasants ripped off from YouTube

The last example of binary data is most troubling. Clearly technical messages like this (binary data excepted) were sent during WW1 and WW2 using code talking, but I've not seen any examples of the plain texts. It would be interesting to see how numerals were handled.

In summary, yes it's steganography of the strictest sense, but utterly useless in practice.

PS. There's a 54 item list of classic ciphers here that might be a close approximation, or at least allow a partial categorisation of this technique. Please feel free to edit /amend.