Re: HSC 2015 4U Marathon
Well if there is a rational polynomial q(x) of degree < 3 with that root (which we call s), then there must exist such a polynomial m(x) (m for minimal) which divides x^3+3x+1.
Why?
By the division algorithm,
x^3+x+1=q(x)d_1(x)+r_1(x)
q(x)=r_1(x)d_2(x)+r_2(x)
r_1(x)=r_2(x)d_3(x)+r_3(x)
...
with deg(r_j) decreasing and r_j(s)=0 for each j. This process cannot repeat indefinitely, so we must have some r_{n+1}(x)=0, in which case m(x)=r_n(x) satisfies the claim.
So it suffices to show x^3+3x+1 cannot be written as the product of two rational factors of lower degree. (That (x^3+3x+1)/m(x) must also be a rational polynomial follows from that fact that it takes rational values for each integer x. Knowing what a poly does at infinitely many locations allows us to find its coefficients by simultaneous equations which will only involve rational operations.)
One of these factors must be linear, and so x^3+3x+1 must have a rational root. But the rational root theorem tells us this is NOT the case.
Hence x^3+3x+1 is the monic polynomial of least positive degree that annihilates s.