Er,ir,ar???

I am getting both hints and confusion.
ar means to be the thing, er means to do the thing, ir means the thing moved. 

Maybe the IE people  separated the things that can be done with a stem to make it a verb. Er, ir and ar and likely others were stem modifiers turning a stem word into a verb. The Latinizers would have created categories of these according to modifier so as to define all the tenses for the verbs. In latin such a scheme becomes regularized, because they didn't care if modifiers were a bit out of place, it was strictly a literature dictionary.

Be, ben; the original conjugations become iba iban. The Latiniers recognizing e and en as a commonly understood modifier to meaning 'to become' for the stem word.

Latin being the dictionary was not actually spoken, at first, but quoted. It was the method for defining a local term to the Roman counsel. You local agent talk in the local vernacular then stopping to define an uncertain terms using the Latin dictionary format.

Then paper came along, it was cheap and everyone could use the Latin dictionary which was now quite complete.  Almost everyone up to that point spoke the local vernacular, even the Roman bosses. Latin translation was, however, the official written language.  It was their only written form, the Roman bosses controlled parchment production for a long time, long enough for the scribes to basically finish their grammar.

To be, to go, to do were always vernacular irregular, in direct and indirect derivation from IE. They both continued to be warped in the official Latin and Old English, because they were never meant to fit into a grammar.  In the early Latin they can be regularized, in writing only. If you spoke the grammar you sound like noise. These are always there, the local vernacular irregulars, the every day words. Once we had paper, we had no choice, they need be included in the written. I think if you look at the rise in newsprint you will see a period of consolidation and compression from the original latin stems, like what the French did during that period.

Grammar in the computer makes a biog change. The computer can abstract the local vernaculars and identify them by context, and maintain huge dictionaries rapidly searchable.    This is evolutionary, suddenly increasing the number of stem words by two orders of magnitude because they are separable by context. Th computer does 'nets or nests', not some one dimensional serial stream. Local rules of grammar become an almost free variable, two parties can talk in some detail in a made up language of jibberish, as long as hey have a simple rule set and dictionary. I am not sure how we will handle the shock.