The Authority of Truth
and the Origin of ὅσιος and ἔτυμος

by George Hinge

ἐτήτυμος and tūtumá

In a paper read at a conference in Cambridge in 2005 (and later published in Greek and Latin from an Indo-European Perspective, Cambridge 2007, 145-161), I presented a new etymology of the Greek adjective ἔτυμος / ἐτήτυμος ‘true’ (and the synonymous cognate ἐτεός):

  • Greek ἐτήτυμος has an exact match of the Vedic adjective tūtumá- ‘efficient’, if the underlying Indo-European root is reconstructed as *h1t(e)uh2- ‘be strong, efficient’ (this observation is, to my knowledge, new)

  • an initial laryngeal is suggested (if not required) by internal evidence in both Old Indian (the long-vowel reduplication in tūtumá-, tūtāva) and Greek (the “Attic” reduplication of ἐτήτυμος and the syllabic augment of -ἤταζον, -ἤτασα)

  • the semantic development ‘powerful, efficient’ => ‘valid, true’ is paralleled by Latin validus and German wirklich

  • other possible derivations of this root are Latin tueor ‘observe’, tūtus ‘safe’, Germanic *þewaz ‘servant’ (traditionally derived from *tek(w)- ‘flow, run’), *þaww-a- ‘observance(?)’ (> Eng. thews), and Greek σάϝος > σῶς ‘safe’ (if one allows that an initial laryngeal is not vocalised, see below)

ὅσιος and satyá

The Greek adjective ὅσιος ‘holy, permitted’, which has been associated with ἐτεός by most scholars, cannot belong to the same root:
  • Greek ὅσιος is connected to Old Indian satyá-, Avestan haiθiia- ‘true’ (this assumption is made by most scholars)

  • it is a derivation from the participle PIE *h1s(o)nt- ‘being’, which has the meaning ‘true’ or ‘guilty’ in many daughter-languages: OInd. sát- ‘true’, Eng. sooth ‘truth’, sin ‘guilt’, Lat. sōns ‘guilty’

  • the development of the syllabic nasal * > Greek ο (instead of regular α) is not unparalleled, cf. *h1u̯i-h1k̂m̥t-ih1 > εἴκοσι, *-h1k̂m̥t-io- > δια-, τρια-κόσιοι etc., *h2er-s-mn̥t-i̯e- > ἁρμόσσω, no matter whether this vocalism is a solemn feature taken over from the Achaean dialect (like βροτός, ἀμβρόσιος) or conditioned by the context

  • an initial laryngeal seems to be unstable before a syllable containing a syllabic resonant, cf. *h1u̯i-h1k̂m̥t-ih1 > ϝίκατι / ἐΐκοσι, *h2u̯lh2-no- > λῆνος, *h2nh2t-i̯(e)h2- > νῆσσα, *h1pi-sd-e- > πιέζω, *h2spr̥-i̯e- > (ἀ)σπαίρω, *h1tuh2-u̯o- > σῶς, σα(ϝ)ο-, *h2strh3kw- > στορπά / ἀστράπτω, *h1su- / *h1i̯u-gwi̯(e)h3- > ὑγιής, *h1su-mn-o- > ὕμνος


In his review of the publication (CR 59 (2009) 331-3), the American Indo-Europeanist Brent Vine expresses his dissatisfaction with my contribution. Since I hold Professor Vine and his many diligent studies on Greek and Latin linguistics in high respect, I feel obliged to counter his criticism.

Unfortunately, he passes by the main issue of my article, namely the formal identity of Greek ἐτήτυμος and Vedic tūtumá . Instead he focuses on one particular matter which is only of secondary importance to my argumentation, namely the derivation of σῶς (σάϝος) from the root *h1teuh2-:

“... Hinge’s inability (p. 152) to explain the short /a/ of Proto-Greek *twawos (> σῶς ‘safe’) via his reconstruction *h1tuh2-u̯o- (which can only yield *twāwos, even granting his new account of initial laryngeal loss and the operation of laryngeal ‘breaking’, a highly controversial notion).”

Yet, the short /a/ of this form is not only a problem to me, but to all scholarship. The traditional derivation from the root *t(e)uh2- (for which I propose an initial laryngeal) would not yield Proto-Greek *twáwos automatically either: *tuh2-u̯ó- should be *tūwós (without ‘breaking’) or *twāwós (with ‘breaking’). The less probable full-grade forms like *tuh2-éu̯o- (Beekes 1969) or *tu̯éh2-u̯o- (Peters 1980) would, as pointed out in my article, yield †συάος/συῶς and †σᾶος/σᾶς respectively. Therefore, some scholars resort to analogy as an explanation (Lamberterie). But analogy from what?

Vine describes laryngeal breaking as “a highly controversial notion”. However, since it is not an invention of mine, but a hypothesis presented and defended by a substantial minority in the scholarly community, the reviewer should not present it as some extravagancy of the article under review. Furthermore, it is explicitly stated in the article that laryngeal breaking is still a minority view (“a growing minority believe...”, “this heretic breaking hypothesis”), and that laryngeal breaking per se would not account for the form in question (p. 152):

“In the case of PIE *h1tuh2-u̯o-, the traditional analysis would lead to the form †τῡός. Breaking, on the other hand, would result in something like †σᾱ(ϝ)ός (like ζωός), which is at any rate closer to the actual form. Then, the breaking was somehow derailed on the way from *uh2 to *, the transitional form being perhaps the falling diphthong *ua̯.35 It could be explained ad hoc by assuming an early absorption of the semivowel by the dental with a subsequent new syllabification: PG *tua̯hwós > *tsáwos.36 On the other hand, it may also be seen in connection with other resyllabifications of words with pre-consonantal laryngeal loss (see §12).”

In other words, laryngeal breaking is not essential to my argumentation, but only a theoretical framework within which I choose to work. Without this framework, we would still have the irregular development *(h1)tuh2u̯ós > *twáwos, and it would still be possible to explain the irregularity either “ad hoc by assuming an early absorption of the semivowel by the dental with a subsequent new syllabification” or “in connection with other resyllabifications of words with pre-consonantal laryngeal loss.”

Thus, the reviewer’s treatment of my contribution is unsatisfying, and it does not live up the standard that one should expect from such a high-esteemed scholar. Pushing the main discussion of the article aside somewhat scornfully, Brent Vine chooses to attack one particular point, viz. the issue of laryngeal breaking, which he criticises with considerable bias, distorting my arguments. The reader would probably believe that I give *h1tuh2u̯ós > *twawos as a regular development, whereas the calculation of the correct counterfactual outcome, *twāwos, is the merit of the reviewer. Yet, as it should be clear from the quotation above, this criticism is taken over directly from the article itself. I shall not accuse the reviewer of dishonesty, but rather of an unseeming carelessness.