Where does the infinitive 'fieri' come from?

Question

What is the origin of the infinitive fieri? It is unusual in many respects. The stem seems to be fi- and the infinitive ending -eri is only found in the second conjugation. However, the second conjugation has a long e and fieri seems to follow the third rather than the second conjugation.

The best explanation I can think of is that it was originally fiere. The present stem conjugation of fieri is morphologically active (although semantically passive if you see it as passive of facere), so the active infinitive ending -ere would make sense. If the form was later passivized, one would expect fii or perhaps even fi according to the third conjugation, but by analogy to conjugations 1, 2, and 4 the change -re > -ri makes sense. This is only a guess on my part, as convincing analogies are hard to find.

So, what do we know about the origin of the infinitive fieri? Are there any attested alternative or earlier forms of the infinitive fieri that could shed light on its history?

Answer 1

I haven't been able to find any solid information on the etymology of the Latin infinitive, which is frustrating. I'm sure it's out there, so this will be only a partial answer (containing a decent helping of speculation) for now.

But one thing seems fairly clear: PIE didn't have an infinitive per se. Different language groups developed it independently post-PIE. In Italic in particular, a verbal noun with the suffix -si developed: *dōnā- "give" → *dōnāsi "to give". In Latin, this developed into the active and passive infinitives dōnāre and dōnārī through rhotacism (and presumably an extra suffix on the passive, though I can't find any details on this part—but final short *i > e is regular).

In what became the third and fourth conjugations, there were three different types of verbs: the consonant-stem thematic verbs, the *i-stem thematic verbs, and the *y-suffix verbs. The first of these had infinitives like *edesi (from *ed- "eat"), the second had infinitives like *θakiesi (from *θaki- "make"), and the last had infinitives like *gwenyesi (from *gweny- "come").

But the verb that would become fiō was special. Its stem was originally *fui- from PIE *bʰuH- "become, come into being", and this led to an infinitive *fuiesi.

Latin didn't like having sequences of short vowels together. So the *u ended up disappearing from the present forms (but not the perfect, which became Latin fuī). But this stopped the ending *-iesi from contracting into * -esi, which happened in the other third-conjugation i-stems (* θakiesi > * θakesi > facere). The vowel deletion only applied once, and since it deleted the u, it never went back and deleted the i.

This left infinitives *fiesi > *fiere, fierī, with an extra i before the ending. And this i then prevented the passive from contracting into *fiī as happened in the rest of the third conjugation.

So fierī looks special simply because of an accident: it had an extra vowel that later disappeared. As for why it uses only the passive infinitive, I really can't say ("it's semi-deponent" is a cop-out).