Does grammar make a language harder to decipher?

[xBlackHeartx]

This is something I've been thinking about recently.

Let's say you had an undeciphered language on your hands. If the grammar was identical, or at least similar, to your own native language (or at least a known language), but the vocabulary was not, would that make it easier to decipher than a language that was alien in both vocabulary and grammar?

I've heard of codes that involved simply replacing words, but those are mostly used by children. I have heard about the story of the US using a code that simply involved replacing words with the expressions that the Navajo used in their native language (such as referring to a tank as a 'turtle'). But its possible that they used some sort of further decryption after that. Such as one of those ciphers you need a specialized machine to encrypt and decrypt the message. I can't remember what they're called, but they looked like typewriters.

Personally, I suspect some aspects of grammar may make decipherament easier. If a language marked things like case, tense, definiteness, and/or number, I don't think it would be that hard to notice affixes that were used in virtually every sentence, possibly multiple times (such as how its common for English sentences to use the definite article two or more times). I mean, people have noticed what may be affixes in the Voynich manuscript, though they have no way of knowing what they mean. Personally I feel like these 'affixes' appear too frequently. I mean, they appear 3 to 4 times at least in every single sentence! Unless they're marking part of speech, I don't see any affix being used that frequently.

[linguoboy]

I've heard of codes that involved simply replacing words, but those are mostly used by children. I have heard about the story of the US using a code that simply involved replacing words with the expressions that the Navajo used in their native language (such as referring to a tank as a 'turtle'). But its possible that they used some sort of further decryption after that. Such as one of those ciphers you need a specialized machine to encrypt and decrypt the message. I can't remember what they're called, but they looked like typewriters.

The Navajo-based code was purely spoken; codetalkers weren't even allowed to take their codebooks into the field. It employed a combination of word substitutions like you mention and encrypted letters. That is, each English letter was mapped to a Navajo word (e.g. "p" to bisóodi which means "pig"). To save time, certain frequently-used words were replaced with Navajo neologisms--not the ordinary Navajo words. For instance, a "bomber" was a jeeshóóʼ ("buzzard"), not a chidí naatʼáʼiʼtsoh beeʼeldǫǫh bikǫʼ neiyéhé. These replacements changed over time and some were in use only for a particular operation and not reused after that.

Deciphering unknown languages works somewhat differently from codebreaking. You can find a lot of resources on both. Having frequently-recurring affixes does help, but so does having frequently-recurring function words (like the articles in English).

[Karavinka]

Familiar grammar can make a breakthrough in deciphering.

I once deciphered a conlang called Isnamire from a South Korean novel, the language was attested by a couple dozen words and names, alobg with a poem about a dozen lines long.

There was a translation of the poem so it was a mater of time.. but after days of careful study, I realized the language was about 90% French in grammar. With that discovery, the conlang was fully cracked and I wrote a complete glossary for the first time in the fandom. That was like some 17, 18 years ago.

Same with the real life stuff too. Akkadian was deciphered before other cuneiform languages for a reason.