oops.

I forgot the carats. good catch.
17 = p^q + q^p = 2^3 + 3^2
I don't have the proof in front of me, but the statement above implies that while some primes may be expressed in the same form in terms of other primes, (p^q + r^s), the primes p, q, r, and s will always be 3 or 4 distinct primes, unlike (2^3 + 3^2).
Here's my best guess at why ~
Given: p, q, and x are prime, and x = p^q + q^p
An odd number times an odd number is always odd.
x = (odd#)^(odd#) + (odd#)^(odd#) = odd# + odd#
= an even number
So x is never a prime when neither p nor q equals 2.
If p = 2 but q ≠ 3 then...
2^q = (some multiple of 3) - 1.
q^2 is never a multiple of 3 so...
q^2 = (some multiple of 3) - 2.
x = 2^q + q^2 = (some multiple of three) +(some multiple of three) - 1 - 2 = (some multiple of 3) + (some multiple of 3) - 3
= a multiple of three
So x is never a prime when p = 2 but q ≠ 3.
Therefore p=3, q=2, and x=17
If that proof is correct then I'm awesome...