Real research mathematics, that which can't be reduced to mental calculation, requires a command of verbal reasoning well beyond what it takes to win at analogies on the SAT (as does theoretical physics).

It can also add to your ability to express yourself.

I very concisely was able to get my feelings about Mitt Romney across to my math major roommate when I stated that Romney's ilk believed the well ordering principle could proven to apply to people and not just integers.

Another less esoteric example is this James Fallows article about the Carter administration

http://www.theatlantic.com/magazine/archive/1979/05/the-pass... (ctrl-F cube root)

This was particularly chuckleworthy because I had just encountered a bug in the TI calculator system (I used to work for a TI subcontractor) where to make the cube root of -1 just be -1 (ie always taking the principal branch for the unwashed masses of eighth graders who knew nothing of complex numbers), TI decided to have negative rational numbers with an odd denominator have odd roots go to -1, and even denominators numerically evaluate. (Try graphing the cube root function on your TI and see the function go all over the place for negative values, this is why) When we complained that it was screwing up the numerics we were trying to code, we were told "it's a bug, not a feature". I referred to it as "circumcising the cube root" in a drunken email to said math major.