Life can be defined empirically and that's a good enough of a description. The problem is people debating over what that definition should be. The problem with gravity is not describing it, but figuring out why it exists as it does. They're very different situations.
Most people agree on the basics of life - something that can self replicate and evolve - but it's the details that pose the thorny issues. For example, how particular is it about its environment? Viruses leave most of the work of their reproduction to outside sources, so there are many people who don't want to call them life. But there's a continuous slope between that and something that can survive on nothing more than sunlight, water, CO2 and trace minerals; you don't say that a cat isn't alive because it can't make taurine and has to rely on external entities to do so, for example. And at an even more basic level, how picky must one be about what constitutes "replication"? What if you have imperfect replicators that create entities "similar" to themselves, which may have varying degrees (perhaps frequently "zero") of ability to replicate themselves? Certainly such a thing has the potential to at least lead to life. But is it life? If not then what's the cutoff point in terms of replicative accuracy when you start to call it life and the inaccuracies in its reproduction "evolution"?