0

If *h1 were a glottal stop, and virtually all German word initial vowels have implicit glottal stop then would the claim about regular laryngeal loss have to be revised?

There's a rather recent development from the Hittite camp led by Kloekhorst claiming (see Laryngeal_theory) that *h1 may have been a glottal stop. Assuming that were true, ignoring the other Laryngeals for the moment, and assuming that glottals were never signed in western alphabets, this would mean that we can hardly tell whether Laryngeals were lost. We don't know how precise the Laryngeal theory is, of course. If all Germanic vowel-onsets include a glottal plosive1, then there's no particular significance to finding words among them that are reconstructed with the laryngeal. At least that fact doesn't negate the claim either.

So, I need to know what show stoppers to look out for, before searching for the coincidence of a PIE-German pair fulfilling this constraint, which shouldn't be hard to find. Any other IE language that is not aspirated, if that's the principal divide, would also work. Primarily, if the glottal plosive vowel onset is evidently an innovation, then the whole idea would be junk.

1: Calling an onset a stop is just counter-intuitive. Tangentially to the phonetics, Arabic for example notes hamza, which wt:de:Hamza equates to the sound in German beachte [bəˈʔaxtə], though I was under the impression that there's a slight difference. My Arabic is certainly weak. ayn is also a noteworthy tangent, because that's a pharyngeal, though that's a different matter.


PS: One example might be *h1er-, "earth".

Compare that to George, supposedly from γῆ (gê) "earth" + ἔργον (érgon), PIE *werǵ- "work". The root of γῆ is uncertain, supposed to be "Pre-Greek" or "Pre-Proto-Indo-European". So there'd be more at stake than internal reconstruction, if *h1er- and *werǵ- should relate via George. Note that Personal names are frequently subject to folk etymology. Possibly related, *h₂éǵros gives agros "field, achre" (agrarian etc.); Indeed, wiktionary links *h1er- to *h₂erh₃- "to plough".

  • 9
    Initial glottal stop in present day Germanic is allophonic -- not a phoneme. A reconstructed glottal stop would be phonemic. You would need to be careful not to confuse phoneme with allophone, – Greg Lee May 11 at 13:29
7

In German, a glottal stop is inserted before a vowel-initial morpheme, when that morpheme does not come immediately after a consonant. It never forms minimal pairs, and its distribution is completely predictable. In other words, it's pretty clearly not a phoneme of the language.

If you want to show that laryngeals survived in some language, you'll need to show that the reflex of a word with a laryngeal, looks different from the reflex of a word without one. In German, this isn't the case: a glottal stop appears in all vowel-initial words, regardless of whether the PIE root had a laryngeal or not.

  • Why isn't phonetics enough? Wikipedia has junctions under minimal pairs, that are not phonemic. The example is "great ape / grey tape"; which also shows varying stress patterns. The laryngeals too, like semivowels, never appear as nucleus of a syllable, do they? Duden for example recommends splitting "An-alphabet", avoiding "Anal-phabet". The stop helps for clear speech in "bei eins", "beeilen", "beerben" and "beerdigen"--Thanks for the hint. The stop in the middle of the double vowels is not strictly obligatory, but trying to remove it brings about approximant allophones or vowel coloring. – vectory May 12 at 0:04
  • 6
    @vectory Because if you get a glottal stop at the beginning of every vowel-initial word, that's not really the laryngeal surviving, is it? It has nothing whatsoever to do with the laryngeal. – Draconis May 12 at 0:13
  • can you name such a minimal pair in PIE, please? Maybe that would help me to understand the point. As far as I can tell there's no particular reason to believe *h1 was a phoneme. Looking for a minimal pair I rather found two pairs of synonyms, or disagreeing constructions (for one, see "ex"). – vectory May 14 at 12:58
  • I don't think "when that morpheme does not come immediately after a consonant" is an actual condition in the distribution of the glottal stop in "standard German" as I've seen it described. This post on German SE gives the example ʔaufʔessen, where the glottal stop is used after /f/ at the start of the morpheme at the beginning of essen. – ewawe May 27 at 22:14
  • Either there is a minimal pair like ooze, use or I am missing the distinction between phonetics/phonemics. There is no soozing, yet it's oozing without a stop is hardly even acceptable. Likewise, either stone-earth, stone-hearth are pairs, or they may be phonemically homophoneous to be differentiated phonetically only if needed, while the language very delicately skirts around true homophones like sto-nerth (nonce phrase), e.g. through effects like those attributed to the laryngeals' effect. Or there is such a homophone and I am just missing it. – vectory May 28 at 22:03
4

there's no particular significance to finding words among them that are reconstructed with the laryngeal.

This is exactly the case. The presence of a glottal stop at the start of German words is not at all confined to inherited words that are reconstructed as starting with a laryngeal in PIE. It also appears at the start of borrowed words: my understanding is that Anus, Echo, Omen, ordnen, Urne, Usus are equally likely as inherited Germanic words to be pronounced with an initial glottal stop.

Furthermore, inherited German words that start with a vowel/glottal stop don't all come from PIE words that started with a laryngeal consonant. German has un- from *n̥-; my understanding is that words like Unfall, Undank, Unrast, Ungnade are pronounced with a glottal stop.

  • The first and second paragraph make the same point, a point I made in the question, noting that it's not a counter argument, because if ((x>x)and(y>x)) then (x>x) still holds, just not for all the right hand x. – vectory May 28 at 19:33
  • @vectory: A counter argument to what? Asking "would the claim about regular laryngeal loss have to be revised" sounds like you're talking about phonemes, since the laryngeal consonants of PIE are reconstructed phonemes. Whether certain words might have had phonetic glottal stops since PIE times up into modern German is a separate matter, and I don't really see how it's interesting, since as far as I can tell there's no way to know one way or another. – ewawe May 28 at 19:39
  • A counter argument to my counter argument! In effect I was saying that your premisses does not support your argument (which I'd render thus: if x1 > x2 and y > x2 then x1 did not equal x0). I see no strong evidence that word-initial *e vs *h1e was phonemic, at least because there are few entries under *e, as CVC is thought to be the primary root structure. – vectory Jun 6 at 21:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.