As fdb mentioned in a comment:

The sequence a-a is a scribal convention for ajV [in Akkadian]. Some Assyriologists treat it as a single sign with the “Lautwert” aju, aji, aja

In Hittite, Kloekhorst postulates a sound change *aiV > āV, to explain forms like pa-a-un "I went" and pa-a-er "they went" from *pai-un and *pai-er. He also proposes that this verb had two stems pai- and paii-, with the second necessary to explain forms like pa-a-iz-zi < pāizzi < *paii-zzi where the i remains visible. (In his analysis, j isn't phonemic in Hittite, but is an allophone of i before a vowel.)

This does seem like an entirely plausible sound change, but the two separate stems do feel a bit ad-hoc, which made me curious—is there any evidence for plene a being used for a sequence ajV in Hittite, as an alternative to this being a diachronic sound change?

  • For curiosities sake, if the answer were positive, which form would you read differently and how, conjecturally speaking? Kloekhorst proposes the enlarged root contained a cognate to Lat. eo (e.g. in comeo), which is not a mere sound change; this further implies the bare root gave a Place Word as stem with inflection, correct?
    – vectory
    Commented Jun 6, 2020 at 15:21
  • @vectory If a could be read as ajV, then pa-a-un, pa-a-er, pa-a-iz-zi could be read as pājun, pājer, pāizzi, with the same stem pai- in all the forms.
    – Draconis
    Commented Jun 6, 2020 at 18:19
  • Note German pfui for comparison, though often compared with Latin fi and explained as onomatopoetic (cp. likewise to fart, piss). It does in fact mean "leave it, go away" when used as such in commands to children, cp. i-bah!. The diphtong also resounding in Swedish fy should be of interest. I still don't see clearly through Kloekhorst's derivation, nor what you had in mind, so I'll just leave this, here, as pure speculation.
    – vectory
    Commented Jun 27, 2020 at 11:06

1 Answer 1


Sturtevant 1933 thought not (page 65, section 52).

The only vowel sign to be written twice in successtion is a, as in a-a-an-za 'hot', a-a-an-ni-in-ni-ya-mi-iš 'cousin', a-a-ra 'customary' (?), a-a-ri-ya 'give an oracle', a-a-ru-na-aš 'sea', la-a-ma-a-a-mi-it 'my name'. This orthography is rare even with a and is confined to the position before r or a nasal. If it indicates anything of a phonetic or historic nature I have not discovered what that is. In Akkadian the sign group a-a has been thought to stand for aya, but that is an improbable value in Hittite, since y was lost between like vowels with subsequent contraction.

In the 1951 edition, though, he softened his stance (page 18, section 39).

The a-sign is the only vowel sign that appears twice in succession, e.g. a-a-ri 'is hot', a-a-an-za 'hot'; the consistent double writing of the a must indicate two syllables, and the IE root ai- aidh- 'burn, shine' shows that the semivowel -y- must be assumed; it does not necessarily follow that intervocalic y persisted in Hittite.

This is specifically about the combination a-a, though, not about simple plene a. As far as I can tell, nobody reads ajV behind the simple plene a Kloekhorst is talking about.

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.