Board Thread:Linguistics/@comment-27119227-20150529220856

I understand sound change pretty well, but one thing simply confounds me: how on earth do paradigms develop the astounding level of irregularity they have? For example, how did the plural forms of Latin's personal endings develop? The plural of -ō/m is -mus, -s is -tis, and -t is -nt? Did PIE use a separate plural morpheme for each person?

Elector Dark, I absoulutely love Misha's declension paradigm, but I'm at a loss as to how on earth such a thing could develop from an ancestor with a regularly-marked plural.