I am amateurishly passionate about etymologies (especially of my native Romanian) but more seriously interested in the anthropological theories of René Girard and Walter Burkert, which both hypothesize primitive religion, myth and sacrifice (and thus human culture - and humanization itself) to derive from the transformation of some type of collective violence, whether ultimately oriented against a hunted animal or a human victim.

Playing with etymologies, I sometimes stumble into cases that seem to have a familiar twist relating to those theories. I have encountered some confirmed etymologies (cited in the works of the aforementioned anthropologists) reflecting that kind of cultural genealogy, but not exactly the ones I am interested here in.

For example, the Slavic term vina = guilt (which also entered Romanian with the same meaning) is based on Old Church Slavonic вина (vina, “guilt”), from Proto-Slavic *vina, ultimately based on PIE *weyh₁- (“to hunt, to chase, to persecute”).

(The simple fact that "to persecute" and "to hunt" are semantically united in one single PIE root can be ”read” in parallel with those theories.)

I was intrigued by the fact that the Romanian verb a vâna/vîna = ”to hunt” sounds close to the the noun vină (guilt, blame) and the regional and rare form a învina = ”to blame, to accuse” (standard form: a da vina, literally ”to give the blame”, identical to English ”put the blame”). And in this case the morphological similarity is based on etymology, a vâna = ”to hunt”, vânător=hunter, vânătoare=hunt, come ultimately (through vulgar Latin *vēnāre < Latin vēnārī, vēnor) from the same PIE root as vina=guilt, blame. - In this way those two Romanian verbs ”to blame” (a învina, a da vina) and ”to hunt” (a vâna/vîna) seemed like a pair of doublets, one Slavic, one Latin.

Looking for other terms that are familiar to me and that might have the same genealogy, I have easily identified the Romanian word voinic, meaning ”strong”, as an adjective, and with related meanings as a noun: ”young man”, ”strong man”, more specifically a young man that is daring, well-built etc. An outdated meaning is that of ”soldier”, and the form is obviously based on the identical Slavic one meaning ”soldier”, also voina=”war”, with the same ultimate PIE root (Proto-Slavic *vojь, vojьnikъ =”soldier”, **vojьna=”war” < Proto-Balto-Slavic *wajas < Proto-Indo-European *weyh₁).

Looking for other semantically and morphologically similar words that might have the same root, I have considered the verb a învinge (”to win, to defeat”) of Latin origin. But vincere - vinco seems based on a different PIE root: from Proto-Italic *winkō, from Proto-Indo-European * wi-n-k-, nasal infix from *weyk-, “to overcome”, but more fundamentally ”to separate, to choose”, a meaning that can be further specified (in the sense of the considered anthropological theories - see link for wīhaz below) as ”to separate out, to set aside as holy, consecrate, sacrifice”, resulting in Latin victima (”sacrificial victim”), Proto-Germanic wīhaz (”sacred, holy”) and wīhą (”sanctuary”).

But still, maybe the two PIE roots are somehow related in the first place.

  • *weyh₁- = “to hunt, to chase, to persecute”
  • *weyk- = ”to cast out, exclude, set aside as holy, consecrate, sacrifice”

Are they?

  • 1
    If they are, it’s too far back in the past for us to be able to say anything remotely definite about it. There are certain consonant correspondences that pop up at the end of roots more than chance would predict, making it look like they were more or less regularly formed extensions to a base root at some pre-PIE stage – but *h₁ and *k do not form such a pair. So while it is possible that these roots were both extended from an earlier **wei̯-, it’s too distant to say anything useful about. Feb 3, 2023 at 14:09
  • 1
    @JanusBahsJacquet I wouldn't be so sure. The only decent account I've seen of k-extensions of roots comes from Willi's Origins of the Greek Verb where he touches on it in passing in explaining Greek k-aorists & perfects. He suggests that HH clusters (the second coming from the 1sg middle eventive or stative ending) would have assimilated and formed an illegal geminate into which a *k was inserted to break it up (analogous to the *s inserted into *tt clusters) which was later generalised across the paradigm, which explains why such perfects and aorists occur in laryngeal-final roots
    – Tristan
    Feb 3, 2023 at 15:20
  • 1
    note that **weyh1k- would regularly lose the laryngeal, and the semantic shift is consistent with a middle/stative that was newly made active again (i.e. "to make something hunted, to make it persecuted" is pretty close to "to cast out", "to exclude"). This would require the sacred sense to be secondary ofc
    – Tristan
    Feb 3, 2023 at 15:22
  • Here's a question on the -k extension (with some answers): linguistics.stackexchange.com/q/4496/9781 Feb 3, 2023 at 17:10
  • 1
    @Tristan That would be a good answer!
    – Draconis
    Feb 3, 2023 at 17:16

1 Answer 1


It seems plausible to me, but depends on a process that is has only recently been suggested and may not be generally accepted.

In his book Origins of the Greek Verb, Willi discusses k-extensions of roots in passing as part of his explanation of the origin of the Greek k-aorist & k-perfect.

He suggests that these originated from PIE's disallowal of geminates. The first person singular of the middle secondary eventive is *-h₂o & the first person singular of the stative is *-h₂e (endings as reconstructed by Sihler in his New Comparative Grammar of Greek and Latin). When these endings were applied to roots ending in a laryngeal, Willi suggests the two laryngeals assimilated causing an illegal **h₂h₂ cluster into which a *k was inserted (analogous to the *s inserted into an illegal *tt cluster).

He notes that k-extensions are frequently found in laryngeal-final roots (notably Latin faciō < PIE *dʰeh₁-) and so may originate from generalising this *k across the entire verb whereas in Greek it was only generalised to the entire aorist and perfect paradigms (cf τῐ́θημῐ títhēmi which has k-aorist and k-perfect forms ἔθηκε(ν) éthēke(n) & τέθηκε(ν) téthēke(n), from the same root, with a differently formed present).

The root *weyh₁- is laryngeal final, so exactly the form where, if Willi is correct, we might expect k-extension. As laryngeals are regularly lost between *y and a following obstruent, this extended form of the root would be *weyk-, exactly that of the second root.

As he only discusses k-extensions in passing, he doesn't discuss their semantics much, but from his explanation we might expect them to typically have senses derived from the middle or stative of the original root (likely either causative, so as to still have active forms, or deponent). "To make persecuted" is pretty close semantically to "to exclude", so the semantics seem to work too.

This would require that the sense of "to consecrate" is secondary, although this is a pretty common semantic shift cross linguistically. The fact that the sacred sense appears much more restricted (as far as I can tell, only being found in Germanic) seems to lend credence to this.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.