Quantifying Linguistic Regularity by