16

I stumbled across this question on the pronunciation of 'vacuum' in the “linguistics” forum.

My question is: If uacuus is *wak+wo- why does uacuus have three syllables, but uiuus, paruus, caluus etc. have only two? Is there a rule that says that *wu remains after sonants (including l, r) but becomes *uu otherwise?

Joonas Ilmavirta
  • 113,294
  • 21
  • 192
  • 587
fdb
  • 17,845
  • 1
  • 23
  • 47

2 Answers2

11

The adjective uacuus is formed from the verb uacō (“I am empty, void”) with the help of the ‎adjective-forming suffix -uus. This suffix has 2 variants, the 1-syllable one and the 2-syllable one.

The 1-syllable form -vus is used after vowels (flā-vus, vī-vus), after L (cal-vus), and R (par-vus, cor-vus), and the 2-syllable form of that suffix, -uus is used after all other other consonants, like in your word vac-uus, with the exception of QU. In this last case, the form -us is used (multiloqu-us, obliqu-us).

Yellow Sky
  • 291
  • 1
  • 5
  • 7
    Some more examples (from Weiss 297-8) supporting this pattern: salvus, clīvus, rāvus, rīvus, fulvus, furvus, helvus, arvus vs. caeduus, continuus, dīviduus, exiguus, irriguus, pāscuus, praecipuus. – TKR Apr 08 '17 at 16:38
  • 4
    @AlexB. I think Yellow Sky is using the "v" and "u" to differentiate between the one syllable and two syllable forms. The Romans would not have differentiated the symbols, but we, in the modern day, do so that is what he is getting at (1 syllable following vowels and liquids, 2 syllables after consonants). – Sam K Apr 09 '17 at 01:44
4

Yes, this is a well-known rule (see Leumann 1977: 132-133, Tronskii 1960: 101, Weiss 2009/2011: 124 and many others).

Weiss puts it this way,

“An unstressed short vowel between a liquid and u̯ is syncopated” (p. 124):

*bholh1i-u̯os > *foliu̯os > fulvus

*sl̩h₂u̯os (or *sl̩h₂euos) > *salau̯os > salvus

*arau̯om > arvum

cf. Leumann "Postkons. lat. u̯ (v) vor Vokal aus vokal. u: vorhistorisch hinter r l; hinter anderen Konsonanten aus Versnot bei Daktylikern, spaeter allgemein im Vulgaerlatein" (p. 132).

It occurred in the pre-literary period.

Exception:

“If the vowel in the preceding syllable was long, the syncope did not occur until the post-Plautine period:

mīluus > mīlvus

lārua > lārva

*pēleu̯is > pēluis > pēlvis

Alex B.
  • 11,667
  • 1
  • 29
  • 55
  • I don’t see any evidence for sala-wus, with or without syncopation. The cognates (e.g. Skt sarva-, Gk. holwos > holos) point to *salw-us. – fdb Apr 09 '17 at 17:57
  • @fdb in Latin -lu̯- develops to -ll- (hence, sollus). But salvus is usually explained with syncope (thus,-lu̯- remains). See Leumann para 60 or Weiss (p. 162, point 9). NB: Unable to fix formatting/italics/reconstructed forms. – Alex B. Apr 09 '17 at 18:19
  • @fdb I assume you've looked it up in Beekes or de Vaan? Because they do address the point you raised. (Beekes: *slh2-eu-) – Alex B. Apr 09 '17 at 22:22
  • de Vaan actually writes *slH-u- with unidentified laryngeal. The laryngeal is posited to explain the /a/ in the first syllable. – fdb Apr 09 '17 at 22:32
  • And you've read Nussbaum 1997? It's on academia.edu – Alex B. Apr 09 '17 at 22:34
  • Yes, but you need to read de Vaan with caution. Incidentally, he did mention two hypotheses on p. 572. – Alex B. Apr 09 '17 at 22:40
  • Link to the Nussbaum paper. The laryngeal in the root is apparently posited on the basis of evidence from Irish and Luvian. The Greek and Sanskrit forms by themselves would suggest solwo-, which wouldn't account for the Latin a. I'm not sure where salau̯os is supposed to come from -- shouldn't sl̩h₂u̯os give slāvus? – TKR Apr 10 '17 at 00:10
  • PIE CRHC has been argued to give Latin CaRaC instead of the usual CRāC, but only when the R was accented; some recent proposed examples here. But I don't see why it would be accented in sl̩h₂u̯os, an adjective, which unlike those examples -- all nouns -- could not show nominal accent retraction. – TKR Apr 10 '17 at 02:49
  • @TKR Weiss puts it very nicely, "There is only one sound change that may require the conditioning factor of the PIE accent" (p. 110), "It is thought by some that this is the treatment when the resonant was accented" and "This conditioning factor ... could well have been true for Latin also, although the evidence is difficult to deal with and far from conclusive." – Alex B. Apr 11 '17 at 03:48
  • But if the palma rule is wrong, then sl̩h₂u̯os should definitely give slāvus rather than salau̯os > salvus, no? I don't think it's debated that the usual reflex of CRHC is CRāC. – TKR Apr 11 '17 at 20:34
  • @TKR I did not say it was wrong (what you call the palma rule - one specific application, in Latin, of what is usually known as the Saussure effect). As for salvus and its IE comparanda: there have been lots of proposals, for example, Lucien van Beek https://www.academia.edu/5945722/The_Saussure_effect_in_Greek_a_reinterpretation_of_the_evidence (esp. pp. 155-157) and Pronk 2011 https://www.academia.edu/1000907/The_Saussure_effect_in_Indo-European_Languages_Other_Than_Greek (pp. 188-189 are very important!) Let me know what you think. – Alex B. Apr 12 '17 at 04:07
  • I may be misunderstanding you, but the Saussure effect is not the same as the palma rule -- the former is loss of a laryngeal next to a resonant in an o-grade form, the latter is PIE CRHC > Lat. CaRaC when R is accented. Thanks for the links! I'm looking at Pronk's discussion of salvus, but I still don't understand his derivation. He derives it from slh₂eu-m > Italic salavem (which bypasses the palma question), but shouldn't that rather give solau-, with ol as the normal reflex of syllabic l? – TKR Apr 12 '17 at 22:54
  • @TKR, no, because PIE CR̥HV > Italic CaRV (Weiss, p. 104, VII.A) – Alex B. Apr 13 '17 at 02:27
  • @TKR Beekes 2011 also reminds that in such cases (CR̥HV), the vowel was colored by h2, so PIE h2e > Latin a (p. 152) - that's how you get the second a. I think it's pretty neat and everything should be clear by now? – Alex B. Apr 13 '17 at 02:37
  • 1
    Ah, that's the sound change I was missing! That makes sense, thanks. – TKR Apr 13 '17 at 21:15