How would someone begin translating an unknown language?

Question

Excuse my ignorance. I'm writing a work of fiction wherein an archeologist finds a tomb that contains not only the bodies of an unknown/unstudied society, but also samples of writing in that society's language. My question is how would that language and the text samples begin to be deciphered. if you could include answers that deal with both the theory and the application of doing so. I have been researching this on my own for some time but in my research I keep hitting a wall when it comes to certain things, and generally I've been going down rabbit holes each time I think I might be close to answering my question. (A definition of a study or a theory leads me to another definition, and another, and another.)

The first thing I learned was the near impossibility of decoding a language when one has no related source (i.e. the Rosetta stone for hieroglyphics) so would it be easier if say the original writer of the ancient text first began writing in Latin or some other well known language before re-writing the text in their own language? This is where the "theory" comes in. The idea is that the ancient writer is the first and only person to ever write down this language, so it would be basically be the origin and sole source of that written language. Would it be easier if they used the Greek alphabet or had part of the sample written in another known language like with the Rosetta stone? I've loved this website and many of the questions/answers are very helpful, but nothing quite gives me all the answers I'm searching for when it comes to theory and application .

Thank you for any help or direction!

Answer 1

Take a look at John Chadwick's The Decipherment of Linear B. It does include some relevant theory -- I've used it as a text in an elementary linguistics course.

Answer 2

Your question is closely related to Bootstrapping, a process in which children initiate the language acquisition. Therefore, in addition to knowledge how scientists decipher the unknown language, please also consider checking researches about how children do it.

Also, Ted Chiang's novel „Story of Your Life“ may be relevant. It's a science fiction novel about the scientists who are struggling to decipher an alien language that has pretty unusual properties.
The author, Ted Chiang, has several interesting publications and interviews about the scientific background he needed to research when he was writing the plot.

I apologize that my answer is not very scientific, but it only serves the purpose of showing the initial direction for your research. Hopefully, when the plot of your story gradually grows, we can answer some more specific questions that may arise.

Answer 3

My answer is just an addition to Greg Lee's answer here. John Chadwick popularized how Michael Ventris and he deciphered Linear B, but the first right steps to the decipherment were made by Alice Kober, an American classicist.

She was the one that classified the little characters that were found in the tablets and she first proved that the language was inflected, a truly important feature of a language and a decisive clue for the investigation.

I don't know if you can find it online (the links are mostly paywalled) but her paper Evidence of Inflection in the Chariot Tablets from Knossos is, to my mind, something very close to what someone does when one finds an unknown script.

Something that Arthur Evans, who unearthed the tablets, never did at least with this kind of scientifically solid method.

Answer 4

Hopefully you can revise your question:

• Are there clues as to how old the language is?

• Does the language use a known alphabet (or have any hints of parentage)?

• Are their contextual samples (signs or labels on identifiable objects)?

If the writing is phonetic (like English) then word length and embedded morphemes are clues to function, and so is position once you determine the grammar settings. If it is pictographic (like Chinese) then the base symbols will often be literal pictures. A Rosetta document would increase the number of languages for comparison (if not make the solution trivial) - we have examples on modern packaging.

Answer 5

You say that in this piece of fiction, the writer is the first and only person to have written down his language. So, any text in the language that your character encounters, was written by this one author. It's pretty likely it would be very logical because there wouldn't be any mutations. You can invent a few plausible ways for the archeologist to get a rough understanding

• the author wrote a direct translation or grammar (the writing system is his creation after all) in another language your character knows

• a different author wrote a commentary on/reply to the text in a known language

• the same author wrote different texts in both languages but has a sufficently repetitive/idiosyncratic style

• the author's writing system is an adaptation of another writing system and talks about some very well known event/thing eg DahNuld TReump xxx DeeRektor Kaumie

For the purpose of fiction, why can't it be that your character made a lucky leap and is proven correct on testing?

Answer 6

Thanks for the interesting question. A first thing to do would be to verify the fact that this is really a language and not a, say, ornament. To distinguish asignal from noise, so to say. This could be done by identifying a writing system as such (that is, to check wheather it's abugida, glyphs, pictogramms, or other).

Then, because it's a fiction, we have to check if that's a language of human species or other.

Just by establishing a writing system alone we could not make any noticeable progress. Even a 'Rosetta stone' with a known nature of writing and its language type would be a poor clue (as it is the case with both Khitan Scripts).

So, depending on a writing and a language types, the first step is to identify them, and then, the repeating combinations of signs (provided there are any).

If we regard a language of a human kind, then a sequence of pictures with inscriptions could be a great help, but not numbers, because numbers can be tricky (to mention a few - several different words for 'five' depending on a grammatical class of items counted; different types of numbers for animate and inanimate objects, and even types of numbers other than singular, dual and plural).

Also, because we regard a proto-language, it could be untuitevely understood by everybody and with a same meaning (yes, a myth of proto-language from which the other languages appeared as a golden dream of (the post-)Christian mankind). In this case, researches would only have to identify the writing system and the grammer (but this, again, depends on how much 'fiction' you want to put into your story).