Evidence for 2 rules linking words in Herbal A->B

[npcompl33t]

You mean ydy? yty?

sorry i meant *initial* bench, like all of the other words in that list. I am also not attempting to explain ALL 'ed' with these rules, i am only attempting to explain -edy. I'm not saying that such rules do not exist, just that I have not yet found them. 

The two rules I've identified show that these are not cases of simple blanket substitutions, they involve complex slot grammar with multiple possibilities based on context within the word. 

However, I think these two rules provide a robust starting point to expand the search.

A possible solution is to append a word starting with a (al, ar, aiin) after some of the "ed" endings generated by the previous rule.

This is interesting, I'll definitely dig into this a bit more.