All language entities are vague. And most of our mental models/concepts are vague.
We live in a world of uncertainities (uncertain meanings or words/memes), where the small changes in some corners of the meme universe may slowly change everything else. The center of one attractor may move somewhere else, broaden, split, dissolve, given enough time. We may move with the flow, we might even name the graph of meme-relations uniquely (quite huge as it is), but we cannot escape the fact, that no part is fixed and everything can change.
I think that there are attractors, the meaning of e.g. "tortilla" basically converges on some agreed on meaning of tortilla. But this meaning is neither preordained (we cannot identify the possible fixed points without their context) nor fixed (it may move along if the context changes).
Everying everything "depends" on can change. Your understanding of the word/concept tortilla depends on the ingredients the relative amounts, the baking and cultural environment while eating etc. etc.
All of his could be enumerated per the thought experiment above. It'd be an enormously high-dimensional enumeration, granted, but there are ways to decompose high-D problems.
None of these is really essential for tortillacity slight variations are surely allowed (which makes is vague). Each of these may change slowly and your (or your friends, childrens...) conception of tortilla will change too. Nonetheless you will continue to agree apon what a torilla basically is and recognize it. And maybe you will argue with other if a certain variant still is or not is a tortilla - which just strengthens this meme.
See also ongoing VaguesDependingOnVaguesDiscussion (to be factored here)