If English was an ancient untranslated language and the Latin alphabet unknown, how hard would it be for linguists to translate it? Would the inconsistent spellings and pronunciations cause problems, not to mention words with multiple meanings? And to what extent would a Rosetta Stone equivalent be of value?
Without some sort of a Rosetta Stone/Behistun Inscription/some sort of overlap with a known language and/or writing system you would be SOL.The language would remain untranslated forever.
It would be translatable but without the inconsist bits. They would probably assume the spellings were intended and would have no idea the pronounciations were incorrect because they would have no standard to judge against. That doesn’t mean it would be unintelligible.
Writing usually doesn’t occur in a vacuum. Somebody who wrote something somewhere probably had some reason for it, and making guesses about that reason is making guesses what the writing means. Was your hypothetical ancient language used by ancient people? What would they possibly write on a piece of a clay jar?
It’s hard to imagine a modern linguist never having seen a phonetic alphabet, but assuming all other knowledge about language and writing was still present, they could well construct a transcription from Latin into a commonly used writing system given enough samples.
To expand on my earlier post, it would depend on how many and what type of English texts are known. As mentioned, a Rosetta/Behistun document would be critical–something with the same statement both in English and a known language. If you could recognize known names (such as were enclosed in cartouches on the Rosetta stone) you could even get a clue as to pronunciation of some of the letters. Also useful would be if you found an illustrated alphabet primer. With that, you could at least learn what the words apple, ball, car, dog, etc. looked like–but you’d still know nothing about pronunciation, and it wouldn’t be very much help in translating other words “let’s see–this word ‘dogmatic’ contains the sequence of characters for dog, so maybe ‘matic’ is a specific breed of dog?”
With just plain texts with no matching translations in another ancient language and no explicit attempts to break down the language and explain it to an unfamiliar user, you would be screwed and would remain screwed forever. (Just like with other untranslated writing systems–look at Proto-Elamite, which has thousands of extant samples but remains untranslated, and likely always will without the discovery of a Rosetta-like pared language document.)
This is why “universal translators” and quickly deciphering alien languages in SF is such a steaming pile of horseshit.
There’s a very interesting Nova program about how Mayan glyphs have been decoded so far - Cracking the Mayan Code and they’ve been able to deal with different glyphs for the same word.
However, there are a number of languages that remain untranslated to this day, despite decades of research. You can easily Google up “untranslated languages.” With nothing to compare it to, English would be just as lost as they are.
It’s not like historical examples of ancient, dead languages are all perfectly written - spelling errors occur all the time and it’s one reason why scholars sometimes have difficulty translating ancient texts. Likewise any other writing errors you can think of - poor creation of characters/letters, grammar mistakes, idiosyncratic abbreviations, and so forth.
But yeah, with a rosetta stone of some sort it is highly unlikely an unknown language would ever be translated.
As I set out in the OP I’m imagining a scenario where just the Latin alphabet is unknown, not phonetic alphabets in general.
I think what I’m really trying to find out is, compared to ancient languages we have been able to decipher, would English be more or less challenging?
You’d also have to ask if there are any extant languages which are related to English, and to what degree. And there’s also the question of how much society and the other aspects of the environment have remained the same, at the time of the attempted translation: It’s easier to translate the word “apple” if apples still exist, for instance.
Yes, and knowing “a is for apple” is useful only when you know how the English word for an apple. For all you know, it might be a “geerb” and the ancient squiggle “a” is pronounced as “g.”
Another big question is how big of a corpus of texts would the future translator have?
One problem with writing systems like Linear A and the Indus Script is that we don’t have a lot to go on.
I think if you have many thousands of pages of writings in English a preliminary jab at decipherment could be made using modern AI methods. Note that capitalizing proper names is going to be quite helpful. From there you get pronouns, then relationships, etc.
But it wouldn’t be complete since too many words would occur very rarely and would baffle folks. After all, there’s quite a few words in Beawolf that are not well understood and that’s Old English.
Linear B was translated without a “Rosetta Stone” and while it is a form of Greek, it’s significantly different than other forms.
Hittite can now be read, and while it’s an Indo-European language, there’s no modern Hittite to draw from. And there were no “Rosetta Stones”–I think the most they had was loan words in other, later languages.
If the only surviving document is the poem Jabberwocky, you’re definitely screwed!
But seriously, with modern French, Dutch, German and Norwegian (plus Latin and Greek, classical or otherwise), I think you’d stand a fair chance, provided there were enough documents.
But the OP’s condition that the Latin alphabet is unknown seems to imply that those other languages are lost, too.
Of course, video (say, Sesame Street episodes, in addition to the picture books) would help tremendously, there. So would knowledge of related languages (the Rosetta Stone would have been a lot less useful without the knowledge of modern Coptic). But even without those, you could still go a long way, with a large-enough corpus of literature. If someone describes a leaky gas pipe as making a “Ssssss” sound, then you know that “S” is a sibilant of some sort (though you might not be sure if it’s the sound we call “S”, “Sh”, “F”, or unvoiced “T”). Likewise, some sounds might be described as voiced or unvoiced, or as plosive, or whatever. And once you have a few sounds figured out, you’ll be able to make sense of descriptions of some words as sounding similar to each other.
Why? They would just use a different alphabet. Perhaps the Greek alphabet or Runes or something. If you are going to say that no Indo-European languages remain, then it’s going to be almost impossible to resurrect a dead English language. You needs something to compare it to. If you have books with illustrations and maps, maybe you could tease out some meaning, but without the language to refer to, it’s almost impossible.
For the Latin alphabet to be unknown and other Romantic and Indo-European languages to use a different alphabet instead would take quite a bit of revising history. (And I counted out the Greek alphabet, too, being so similar to the Latin–I’d also count out Cyrillic for the same reason.)
Yes, maybe I was reading between the lines on the OP, but that is the scenario I was imagining–a completely and utterly isolated unknown.
Going back to the OP, does anyone think that any of the characteristics of English make it harder/easier to decipher than those lost scripts we have broken? I tend to think not: why would they? We’ve generally been able to decipher any language provided that we have enough samples and have knowledge of a related language–even a fairly distant related language. Things like the number of spelling variations, multiple ways to signify the same sound and synonyms have never been mentioned as issues in anything I’ve ever read. Mayan certainly had all those challenges (plus several others), for example.
The languages we can’t decipher we can’t decipher because we have few samples or because they are true linguistic isolates. Those seem to be the factors that come up again and again–never the features of the language itself.
Yep, Mayan had lots of “spellings” for individual syllables. And it took hundreds of years to translate even though bishop recorded some of the details of the written language before it was murdered.
I think NOVA is the best science show on TV, and that episode is easily in the top 5. I’ve probably watched it at least 3 times. The kid was a genius who was in the right place at the right time!