I have tried to find the sound hₐ-, for example "hₐeust(e)ro" engl. 'east', or hₐel, 'burn' , but also example hₐner, 'man' pronunciation, but I can't find it anywhere on the internet, despite it being mentioned in the book "The Oxford Introduction to Proto-Indo-European and the Proto-Indo-European World" (J.P.Mallory and D.Q.Adams) . Also other guides like Julius Pokorny's "Indogermanisches Etymologisches Wörterbuch" doesn't seem to know this (in fact it doesn't even mention the word). Is the pronunciation of the word closer to other known equivalents (such as Germanic *austro or Greek ἠώς ēōs 'dawn, east' (unfortunately I don't have an Ancient Greek dictionary))?

It is true that PIE actually has an equivalent *aus- "to shine," or "dawn," but Mallory and Adams separate this from the (exclusive) air direction term, and the same letters are not used either. In the book, the sound is also not mentioned in the section comparing most of the known sounds of Indo-European languages. I think it would make sense that the pronunciation of the word has meant closely known equivalents in the examples I gave, but hₐ remains a mystery to me, I don't know exactly.

1 Answer 1


The transcription *hₐ is a sometimes-seen, idiosyncratic way of writing what is otherwise more or less uniformly written *h₂ when reconstructing Proto-Indo-European. It is known as ‘laryngeal two’, one of the Proto-Indo-European laryngeals, of which there were three according to most reconstructions.

Note that Pokorny’s dictionary, which you mention, was published before the laryngeal theory had been widely accepted and adopted. When it was written, it was known that there was something that caused certain vowel patterns seen in Indo-European languages, but the consensus that these patterns were remnants of erstwhile consonantal segments had not been reached yet. Being at the time a minority position, the burgeoning laryngeal theory was not followed by Pokorny, who therefore presents PIE roots in a form that can be difficult to recognise and identify with the forms that are generally used now.

These days, with the theory almost universally accepted, the laryngeal called *h₂ is most commonly believed to have been phonetically a fricative in the back of the mouth, along the lines of [x ~ χ ~ ħ] (though some consider it more likely to have been a laryngeal or even uvular stop).

Variations in phonetic details aside, pretty much everyone agrees that *h₂ had the effect of lowering an adjacent /e/ into [a] – hence why you sometimes see it transcribed as *hₐ (and similarly *hₑ for *h₁, which had no effect on /e/, and *hₒ for *h₃, which had the effect of retracting /e/ to [o]). This effect is usually called ‘laryngeal colouring’.

(As described in the Wikipedia article linked above, a small minority of scholars also reconstruct a fourth laryngeal, *h₄, which they say had the same colouring effect as *h₂ but differed ever so slightly in its reflexes in the Albanian and Anatolic branches. This is a fringe hypothesis, but those who adhere to it may also use *hₐ to mean ‘either of the a-colouring laryngeals’, i.e., ‘*h₂ or *h₄’.)

The phonetic realisation of your examples, then, would be as follows (using just [x] to represent the phonetic value of the laryngeal for convenience):

  • *h₂eu̯stro- was probably pronounced [xawstɾo]
  • *h₂el- was probably [xal]
  • *h₂nēr was probably [xneːɾ]

After the disintegration of Proto-Indo-European into different branches and individual languages, *h₂ (and also *h₁ and *h₃) was almost universally lost.

The only place where *h₂ (and perhaps *h₃, but probably not *h₁) remained as a consonantal segment directly visible in the oldest texts was in Anatolian languages (Luwian, Hittite, etc.). These were also the earliest Indo-European languages to be recorded, and it is noteworthy that they were spoken in an area where they had lots of contact with Semitic languages where uvular and laryngeal sounds are common.

In all other Indo-European branches, all three laryngeals were lost as consonantal segments except for a few possible cases where the sequence *h₂s ‘hardened’ into *ks in a process known as laryngeal hardening (the prime example being Latin senex < *senaks < *sen-ah₂-s vs senātus < *sen-ah₂-to-).

The colouring effect that the erstwhile laryngeal had on adjacent vowels, however, generally remained in all branches. In addition, some clusters containing laryngeals gained prop vowels, and these could themselves be subject to colouring, though this differs from branch to branch. For example, in *h₂nēr, the laryngeal remained consonantal and was lost in Indo-Iranian (*h₂ner- > Skt. nár-), whereas it gained a prop vowel in Greek and became *h₂ənēr [xaˈneːɾ] before being lost, leaving ἀνήρ [aˈnɛːɾ].

  • This answer is a duplicate. The phonetic interpretation is irrelevant. The common wisdom that laryngeals were merged and lost everywhere but Hittite is misleading. Following the teach a man to fish mantra, what this question needs is a guide to read the encyclopedia. There's an answer, either in the introduction, symbols used, or the respective entry on “Laryngeals” in there, most likely. Thumbs down
    – vectory
    Commented Nov 26, 2023 at 16:57
  • 6
    @vectory Your downvote is unwarranted. An answer that is the only answer on a question cannot, by definition, be a duplicate. The question may be, but since there are no other answers to it, your first statement is a blatant falsehood. I also never said the laryngeals “were merged and lost everywhere but Hittite”. I said they were lost as consonantal segments everywhere but Anatolian, which is true (except for a few edge cases of laryngeal hardening). And your suggestion that the answer should be “read the encyclopedia” is not useful, since that would not qualify as an answer at all. Commented Nov 26, 2023 at 17:01
  • this answer would be improved by explicitly stating that Pokorny's dictionary predates the (near) universal acceptance of the laryngeal theory, and so the forms it gives concord with the earlier non-laryngeal theory (featuring long vowel and vowel-initial roots now reconstructed with laryngeals)
    – Tristan
    Commented Nov 26, 2023 at 17:56
  • 1
    @Tristan Good point – will edit to add that. Commented Nov 26, 2023 at 19:39
  • 3
    The notation *h_a could also be useful as a cover symbol for *h₂ and *h₄, if you believe in *h₄. (I don't think very many do.)
    – Draconis
    Commented Nov 26, 2023 at 22:39

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.