It started as a simple retweet to @Languagebandit’s post that “Modern forms of "to be" are a mishmash of 3 different Old English stems + conjugations, which is why "be", "is" & "were" seem so different”.

I copied and pasted a line from my (soon to be published) book Pour en finir avec la langue française suggesting that the present tense of the verb to be comes from four separate roots: bēon → be ; eom → am ; earun → are ; and is → is and got more retweets and likes than ever before. Which is interesting but not as interesting as the roots of the principle itself. When a verb develops out of originally different verbs, it is called suppletion, a relatively common phenomenon in Indo-European languages. In English, it is the present tense of ‘to be’ that gets the juices flowing, deriving from 4 Proto-Indo-European (PIE) roots:
*h1es- (the copula is) → am and is
*bʰuH- (‘to grow’ or ‘to become’) → be and been
*wes- (possibly ‘to live’) → was and were
*h1er- (possible alternative to *h1es-, via Old Norse) → art and are

Which don’t exactly match the examples given earlier (bēon → be, etc.) because they correspond to different stages of language development, but still show the potential for various roots of different meanings coming together into a single verb. We find similar situations in languages such as Gothic, German, Icelandic, Lithuanian, Old Slavonic, Polish and so on.

Another verb that likes to mix and match is ‘to go’. While in English we stick to the relatively vanilla ‘I go’ vs ‘I went’ (from ‘wend’, the current past of which is ‘wended’), Latin-based languages such as French, Spanish and Portuguese tend to be derived from three (or, questionably, four) Latin verbs : vadere and ire (both to walk, go, or move forward) & ambulare (come and go, walk, walk away from, go for a walk) → *amlare*allare). The present, for example, uses ambulare / *allare (and no-one’s really sure about this development from ambulare or ambitare into aller) for the infinitive aller and most tenses plus the 1st and 2nd persons plural of the present, vadere for the rest of the present, je vais (I go, etc), and ire for the future j’irai (I will go). Why Romanian chose ‘a merge’ from Latin mergĕre for to dive, plunge or penetrate into is another kettle of fish.

But what we do see here are the traces of the difficulties earlier speakers had of expressing or interpreting intensity or subtlety of meaning with the words they had available. Suggestions have been made that the allare form existed in France in the 2nd C CE and might have resulted from a familiar military order: go! (Allez! from imperative ambulā́te with emphasis on its long final ‘a’ to allate with emphasis on its final ‘e’). And while ‘I go’ could be understood as a simple expression of personal choice and movement, the inclusive 1st and 2nd persons plural – Allons! and Allez! – could understandably be conceived of as a more coercive version of movement.

These are just ideas, tiny pieces of the massive jigsaw attempting to explain how language originates. One day, maybe, I might take a look at it more seriously.