It makes for an interesting argument. Surely pain doesn't require higher order cognition, or even the interaction of a complicated cortex to be recognized.
I guess it depends on how you define pain. The sensation works only if it is a) acknowledged as being unpleasant (therefore remembered for future reference), and then b) prompts a behaviour that might be conducive to reducing the cause of the pain. The 'unpleasant' part is something we can describe, as cognitive animals, and is almost an anthropomorphic description. But the ellicitation of avoidance behaviour is something recognized in animals that lack anything like our brain structures.
Athon