Crudely, I guess you could sort of track the "Bayesian surprise" you get by reading a book by measuring changes in skin conductance --essentially tracking how many "frissons" you get while reading maybe? (I really don't know tho...)