Deciding whether a token is a word

We need a general method for deciding whether a token needs normalising, or can be passed directly to the letter-to-sound module.
6 minutes 23 seconds
Excellent 1
Very helpful 6
Quite helpful 6
Slightly helpful 0
Confusing 0
No rating 0
My brain hurts 0
Really quite difficult 0
Getting harder 2
Just right 7
Pretty simple 4
No rating 0