Via Jim. Fascinating that we can decode the Prairie dog language, but we still can’t decode Linear A, the Minoan (possibly) language from before the birth of Christ or Rongorongo (the fantastically named Easter Island language.) That’s because the Prairie dogs speak in a language that has specific referents; green hooded man, yellow hooded man, black elipse and so on. We have no idea what the referents are for Linear A as there are no present day speakers and no dual-language inscriptions (e.g. the Rosetta stone which helped decipher the Egyptian hieroglyphics.) We can work out the counting structure, but little else.
Similarly, Rongorongo, a language which may have been developed in relatively modern times as a reaction to the white man’s language is completely lost to us, despite its hieroglyphic-like symbols, because a) there aren’t enough texts b) these texts indicate that the language used the icons as components of words, rather than words themselves, further shortening the available vocabulary c) these represent a text we have no conception of, though we might guess it to be a history or poem from the small historical record we have.
Languages where there are no recognisable elements, where the symbols signify something other than the thing itself (whether that be concepts, syllables or whatever) are nigh-on impossible to decipher without either dual language texts, known phrases, or a logical brain like that of Craig Ventris (who deciphered Linear B by guessing that it matched ancient Greek, recognising the syllabary forms of certain ancient Cretan town names, and using those syllables to work out other words.) So if we encountered an alien civilisation, as long as they take a recognisable physical form and have a language which refers to individual referents, we probably have a better chance of understanding them than we do Linear A. If we encountered only their text, we would have no chance.