Also Question 11(a) (4U paper) is dodgily worded too. Obviously if g(x) can by any function continuous at alpha (which is what the wording technically means and hence what I thought it meant on first reading), then f(x) need not be identical to (x-alpha)^2 * g(x). E.g. take g(x) to be the zero function, or heaps of things (e^x, sin(x), etc.).
I think what was meant was, show that f(x) can be written as (x-alpha)^2 * g(x), for some function g that is continuous at alpha. In other words, show that f(x)/(x-alpha)^2 has a limit as x -> alpha.