It is, more precisely, a key tenet of science that the world is only comprehensible when things are grouped into categories, and categories have orthogonal hierarchical relations to each other. There is no clear highest-level entity for language: people talk about paragraphs as linguistic units, and speech processing needs to have some kind of thing resembling an "utterance" that has more than one sentence, or less than a full sentence. Then you can break those things into sub-parts that are uncontroversially linguistic in nature.
Phenomena (palatalization, syncope) can be treated as "things" and analyzed in a hierarchical manner, if you want. So "nasalization" might be used to refer to the process of making one thing become nasalized, or it could be used to refer to the thing that assimilates (the property of nasalization). In like manner, stress and tone can refer to the property itself, or the process of assigning / modifying that property. Seen as the thing being assigned, there are at least two category-related questions that arise – what is the category of thing that "tone" or "stress" refers to (is tone a feature, is stress a feature?), and what categories of things have tone, or stress. You can ask the same kinds of questions about nasal(ization) or palatalization. That's where we are, at least at the level of generality that you asked about.
However, there are a lot of technical details of tone that we know, but they don't necessarily have much to do with "categories" except that everything ultimately involves some kind of categories.