Other fields deal with concepts more or less mapped to the real world. Physics is about real world, more or less (right until you get to quants, then the level of abstraction rises dramatically). Same goes for biology, and even computer science in general. There, you can rely on words, which usually convey meaning.
In math, you can't rely on words. You'll never understand even relatively simple things like complex analysis or Fourier transform just by reading about it -- words are never enough to transfer the knowledge to you. You need to play with it, solve actual problems, understand in practice how various "moving parts" are related to each other, and then accept the naming convention (which is almost an afterthought, born as a mean of reference, not as a way to describe things). Therefore, relentless practice and solving abstract problems (a lot of them) is the only way to teach (or learn) mathematical concepts.
Some teachers want to make math more accessible with bringing it "down to earth", mapping mathematical concepts to more concrete problems. It is theoretically possible, and I was a supporter of this approach until very recently. However, math just doesn't work this way. Math is pure abstraction; linking the abstractions to earthly affairs too early shuts down mathematical thinking (creates biases that prevents applying mathematical insights to other fields that are different from the one learned).
Mathematical abstractions cannot be transferred by words and formulas alone; they need to be internalized by practice and drilling.