If I would put you in a closed room with a computer containing terabytes of alien written symbols. Those aliens are from another dimension where the physical laws are different and our understanding of the world and needs do not apply. Given as much time and computing power as you want, how far would you be able to go in your understanding of their language and way of living?
As for understanding their language, it depends on what you would consider "understanding". Given sufficient training data, we may be able to devise rules of its grammar, without knowing what any of it refers to. This is pretty much how most language models work for us at the moment. You might talk to your phone and tell it, "Make an appointment with Dr. Smith for tomorrow at noon". Your phone would follow through, and may mark "Dr. Smith" as the person with whom you have the appointment, but it would have no conception of what a person is. Given your scenario, we will never get to the stage of comprehending their language to understand their lived experience in any capacity. The idea is further explored in Searle's paper Minds, Brains, and Programs.
What's missing in your scenario is covered in the topic of the symbol grounding problem.
However, not all is lost. Given a suitable corpus of their language, we may be able to figure out their formal mathematics. By that, I mean, the kind of mathematics that Russel & Whitehead got around to formulating in Principia Mathematica. As a result of that, we may understand some of their mathematical proofs, and that may include theorems not yet proven by humans so far.
What we mean by understanding a thing is fitting it into our understanding of the physical laws of our universe, so the answer is: no way. If you could go any way in your understanding of this hypothetical alien language, it would show that there was after all something in common between your and the aliens' understanding of the physical world. Thus, you cannot understand.
There is a fictional working out of this idea which, though it is not scientific, is nonetheless worked out by a brilliant scientific mind. See Fred Hoyle's The Black Cloud. On his deathbed, the protagonist scientist who has tried and failed to conform his thinking to that of the alien's idea of what the world is like, conjectures that he might have succeeded if he had been a newborn, who had not yet learned anything about humans' ideas of what the world is like. (I would be less optimistic.)
It's absolutely impossible to find out if the something is a language or not, if we don't have even a tiniest clue. If you're, say, John, and in some galaxy you find some stars situated so that they look like to make the the shape of the word "John", you can never know if it was intended or it was just a coincidence. We cannot know if the patterns on the Andean textiles are symbols of their ancient script, we cannot know, if the patterns of the sunspots are messages from an alien race... Without a clue or feedback nothing can be identified as a language, for languages can be immensely different.