Beekes says that ἔλδωρ/ἐέλδωρ comes from ἐϝέλδομαι, without commenting on the suffix. Is this a case where the agent-noun suffix -τωρ (a.k.a. -τήρ) was applied, but then ἔλδτωρ got automatically simplified for phonetic reasons? On the model of θηράτωρ, I would have expected something more like ἔλδάτωρ. Beekes says the PIE root is h1ueld, whereas other people seem to think it's welh1, which would make it cognate with English 'will.' On the latter hypothesis, I'm not sure why ἐέλδομαι would have the δ.

  • The etymology of this word seems to be very unclear -- note that Beekes says "no cognates outside Greek", which makes positing a PIE form rather pointless. But it pretty clearly doesn't contain the agent suffix, which wouldn't fit either formally or semantically.
    – TKR
    Oct 22, 2021 at 20:47
  • @TKR: Is the semantic fit really not good enough if we think of a wish as "that which causes us to wish (for it)?" I suppose there could also be some hypothetical form such as ἐλδήρ, which would then morph into ἔλδωρ by whatever process caused ἀνήρ -> εὐνωρ.
    – user3597
    Oct 22, 2021 at 21:40
  • These -τωρ / -τηρ nouns always refer to human agents (at least I can't think of any exception), and their semantics aren't causative, so the meaning in that case would have to be "person who desires".
    – TKR
    Oct 22, 2021 at 21:50

1 Answer 1


The etymology of ἔλδωρ/ἐέλδωρ is difficult, but it almost certainly doesn't contain the agent suffix -τωρ.

Beekes and LIV both reconstruct a *h₁u̯eld- which would have no cognates outside Greek (which makes positing a PIE root rather moot). Chantraine connects it with Latin velle, from PIE *u̯el- or *u̯elh₁-; in this case the -d- would be a "root extension" (these are poorly understood, but were definitely a thing in PIE; the semantically close ἐλπίς may be from the same root with a different extension). In either case it seems hard to account for the existence of both variants, in e- and ee-.

But whatever root it's from, the word is unlikely to contain the agent suffix, for two reasons:

  • Phonologically, a form like (w)eld-tōr would almost certainly not give ἔλδωρ. The regular reflex of -dt- in Greek is -στ-. It's true that no Greek word seems to contain the sequence -λστ-, so arguably we don't know what the reflex of -ldt- would be. But a simplification to -ld- seems unlikely -- if anything, you'd expect -lt-.
  • Semantically, the -tōr suffix forms human agent nouns, "person who does X", not abstract nouns. So if it contained this suffix, ἔλδωρ would have to mean "person who desires".

Where the -ωρ does come from is unclear. There are a handful of other Greek words which seem to contain the same suffix, including ὕδωρ and the rarer τέκμωρ and πέλωρ. But I don't think it's clear what the source of that suffix is or whether it's really the same suffix in all these cases.

(BTW, the reason θηράτωρ has an alpha is that it's formed from the verb θηράω -- the vowel is from the verb stem.)

Your Answer

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