As far as science is concerned, there's nothing beyond being able to objectively model a phenomenon and make useful predictions about it. If I can fully model the set of brain processes that humans typically call "love", and if I can make useful and accurate predictions based on my model, that's all that I care about as a scientist.

Now, it's entirely possible that reasoning about "love" scientifically requires several layers of new language / models on top of the currently established models in neuroscience, which is why I said it's _theoretically_ possible.