What confused me is the transition from "w" in PIE *wel- to "h" in E. helix .

And what's the sound law applied to the word E. "vulva",which has the change from "w" to "v"?


"a spiral thing," 1560s, from L. helix "spiral," from Gk. helix (gen. helikos), related to eilein "to turn, twist, roll," from PIE *wel-ik-, from root *wel- "to turn, revolve" (see vulva).


1540s, from L. vulva, earlier volva "womb, female sexual organ," lit. "wrapper," from volvere "to turn, twist, roll, revolve," also "turn over in the mind," from PIE root *wel- "to turn, revolve," with derivatives referring to curved, enclosing objects

(cf. Skt. valate "turns round," ulvam "womb, vulva;" Lith. valtis "twine, net," apvalus "round;" O.C.S. valiti "roll, welter," vluna "wave;" Gk. eluo "wind, wrap," helix "spiral object," eilein "to turn, squeeze;" Goth. walwjan "to roll;" O.E. wealwian "roll," weoloc "whelk, spiral-shelled mollusk;" O.H.G. walzan "to roll, waltz;" O.Ir. fulumain "rolling;" Welsh olwyn "wheel").

2 Answers 2


helix: This came through Greek. Most of the Greek dialects underwent /w/-loss (see e.g. Kavitskaya 2002). The Attic-Ionic dialects lost the sound early (e.g. the sound /w/ and its letter ϝ do not appear in Homer); Aeolian retained it longer but lost it as well. This all leads to the result

PIE *u̯el-, *u̯elə-, *u̯lē- 'turn, wind' >w-loss> Gk ἑλίσσω helisso, ἕλιξ helix > L helix > Eng helix

vulva: This came through Italic, which did not undergo /w/-loss. So the PIE semivowel in u̯el- was retained in Latin vol-. The semivowel /w/ was then fricativized in the Romance languages. This leads to the result

PIE *u̯el- > L volva /wolwa/ > LL vulva /vulva/ > Eng vulva

  • 2
    In Greek, this isn't exactly w-loss, since the [w] was not lost but went to [h]. This is a (probably irregular) change that happens in a number of Greek words; in others, PIE *w was completely lost.
    – TKR
    Commented Sep 25, 2013 at 1:22
  • The link that OP gave for "vulva" says that it comes through Latin not Italic. Commented Jan 24, 2019 at 22:43
  • 2
    @zixuan Latin is an Italic language
    – Draconis
    Commented Jul 20, 2019 at 0:41

I guess you could call those relics.

Thinking about relics or other sacred symbols shaped like a helix, that reminds me of a picture of coiled stripes of silver from sumer used as payment. I wasn't aware of the analysis as Latin re- + linquo (leave behind), but only with that info does "money" kind of make sense, as something left behind (for insurance). I'm not sure re- is the correct analysis there, but much less so that *wel- is the root for helix. The established theory is simpler, and I'd have to come up with something dense to explain why an *r (not a Latin 'r', obviously) should have became a rough breathing.

Some fricative realization of /r/ overlaps with /ħ/ or /x/ and some realization of /h/ might overlap with /x/ as well. That's not really problematic. What's problematic is that, as usual, if setting out with a crazy idea, there will be no shortage of apparent leads. There are too many words with too little information available (on wiktionary at least), that need to be compared. If coincidences abound, there's reason to remain skeptic.

  • No cognates are mentioned neither at *wel- nor at ἕλιξ, and if there were, these would need to be distinguished from loan words (like, literal debt).

  • Greek χρήμα, χρήματα "money, currency", etc. lead to χρή "have to, ought, should", where the trail ends. That Onset fits my phonetic hypothesis. Also, I just today learned that number replaced OE/ME ream.

  • Greek λεφτά "money, wealth" offers no explanation, links a variant lepta, and we can find lepto "minute, cent", too. Latin lepto "thin, small" links to AGr. λεπτός, quote: "From λέπω (lépō, “I peel”), from Proto-Indo-European *lewbʰ- (“break off, rind”)."

    • Of course lefta reminds a lot of En. left, leave, and indeed of Agr. λείπω (leípō) "to leave, lack". A quick overview of near homophones is at λίπος "fat". leave is at least as complex to compare. The established etymologies don't agree with my idea, however ...

    • lack, in turn needs to be compared to leak, and Ger. lecken "leak, lick": That a cow licked ice until a man emerged, as the Eddas have us believe, instead of a cow leaking milk to foster mankind, that's just, I don't even ... Maybe the En. leaven "starter" has to do not just with dough, but cheese making, too--cp. Ger. Lab "rennet". Even better, rennet is explained as cognate with to run, cp. Ger. gerinnen "coagulate" and, while I find that quite surprising, it sure reminds of currency, from Lt. curra (currant? Korinthen?), *kers- "to run". So far, that's only coincidence.

  • Agr. leipo and Lt. linquo link *leykʷ- "to leave". What do we find there? AGr ἔλῐπε (élipe, “he left”), from *likʷ-e-t (thematic root aorist) ... that rough breathing, where's that from? It also derives Proto-Indo-Iranian *ráyknas "property left, inheritance", "Cognate with Proto-Germanic *laihną (“loan, fief”)". An 'r', now wait a minute. That seems to be a regular sound change, because Sanskrit doesn't have word initial /l/. But it opens a possibility for loans, perhaps from Persian, if we want to suppose a *relix. It's hard to tell whether ellipse should be from the same root, but the gloss ἐλλείπω (elleípō, “to fall short, omit”) fit's perfectly to my expectation about calqued Ger. ausgelassen "omitted". Via "ausgelassen feiern" a sense of "wealthy" might be gleamed, although a derivation as "freely, wild, omitting restraints" would be amendable; cp. Nachlass (inheritance), sich auslassen "to vent anger, let it out", Anlass "ocassion, event, festivity", anlassen "to start", an lassen "to leave on", verlassen "to leave", veranlassen "to cause, set into action", lassen "to let".

  • *likʷ-e-t reminds so much of liquid, which reconstructs *wleykʷ- (“to run, flow”). I read a translation of the Codex Hammurabi that said liquidate (of a contract or other tablet), however I can't seem to find it again now and other translations don't say so. Anyhow, the relation of financial liquidity to water strikes me as rather obscure.

    • Ger. Flucht (n. to flee, ie. to try to leave) reconstructs *plewk- "to fly" instead. I'm not convinced. It could point to *wleykʷ-~*leykʷ- instead (why not)
  • Ger. Geld "money" reconstructs *gʰeldʰ- (“to pay”). *gʰ can go to *w sometimes, I believe (where?).

  • δραχμή (Drachme) is of pre-greek origin, according to Beekes, because of variants darkhma, darkhna. Can we go a step further and see -khram, for χρήματα, with "dar-" perhaps related to drag, *ter-?

  • khre (viz. *khrema "money") Links related khron. The origin of Chronos is of course mysterious. Time is money, they say. And Cream rules everything around me.

  • compare value, Latin valere "be strong, worth", showing two different roots: PIE *walh₂- (“to be strong”) [red link, unsourced] or *h₂wl̥h₁éh₁yeti, from *h₂welh₁- (“to rule, be strong”). Well, well, well.

  • Where's roubel "the Russian currency" from? If this is indeed cognate with руби́ть (rubit) "to axe, fell" and Polish rąbać "to chop, hew", then compare Ger. Rabatz machen and Welle machen "to muck up". The idea is that it's a cut of grivna (the currency of Ukraine), also hryvnia, hryvna, that from Proto-Slavic *grivĭna 'necklace' from Proto-Slavic *griva 'neck, nape, mane. Now compare lace, from Latin laqueus "nouse, snare, gin, trap". I am of course alluding to linquo, *leykʷ-; wiktionary says: "From Proto-Italic *lakw- (“to ensnare”), with no certain cognates in any other Indo-European languages; possibly Proto-Indo-European *lēk- (“string, twig, tendril”)." There is a semantic gap, but come on. Cp. linger "to hang around".

  • Sadly, allusions to servitude, like lackey come to mind, frequently. Owning slavse meant wealth and "rings" were so significantly connected that the symbolism still appears in Persian words related to slavery (I don't quite know how).

This is patently weird. After all, I'm running out of time and thinking, well, helix might be from *wel- after all.

  • look at re-; From PIE *wret-, a metathetic alteration of *wert- (“to turn”). Not only does that remind twofold of Ger. "Gegenwert" (...-value)--besides Gegenwart "the present", gegenwärtig "current, present"--but surely reminds somewhat of *wel-, too?

It looks like *wel- could need an update or two.

  • Maybe also compare gain, again and against.

Vulva in my mind needs comparing to 'bulba', though I'm not sure whether I read that somewhere or not.

Appendix: ancient valuta (not sure about the copyright status)

  • This doesn't seem to actually answer the question…neither Ancient Greek <χ> nor Latin <r> were pronounced as fricatives, also. Those are later developments.
    – Draconis
    Commented Dec 19, 2018 at 3:54
  • @Draconis I'm sure you were there and surveyed the dialects and bordering nations. And I'm sure you know a ton about Mycenaean, all the what, ca. 100 words that survived.
    – vectory
    Commented Dec 19, 2018 at 3:56
  • Not at all, but there's plenty of evidence as to how they were pronounced. Ancient Roman authors described /r/ as a trill; earlier, /r/ turned into /z/ in some environments, which makes perfect sense for an alveolar consonant (for example) but not for a glottal fricative. Similarly, /k/ followed by /h/ gives Greek chi (see for example putting ek- onto a verb stem starting with rough breathing), and it's the descendant of PIE velars; it also participates in Grassmann's Law, which only happens with aspirated plosives, never with fricatives. When words with chi were borrowed into Latin…
    – Draconis
    Commented Dec 19, 2018 at 4:03
  • …they were written with "ch" (meaning /kh/), not "h". All in all, the evidence seems overwhelming that neither <χ> nor <r> was a glottal fricative (or a uvular fricative, or epiglottal, etc): the former was a velar aspirated plosive, while the latter was an alveolar trill.
    – Draconis
    Commented Dec 19, 2018 at 4:04
  • 2
    This just looks like a bunch of speculation and brainstorming. Commented Dec 19, 2018 at 4:42

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.