Splitting a word into MSB/LSB pairs happens often and is worth, I think, native words. Also, I'm going to need it in the upcoming commits.
