In Cerberus's list of u-stem verbs, rŭō, rŭere, rŭī, rŭtus is the only one with a short ŭ in the participle stem.
Why is this? Does it go back to different types of verbs in PIE, as with stătus versus amātus?
Szemerényi (1980) says that the following "descriptive rule" summarizes the conjugation pattern for verbs ending in -uo:
verbs in -uo in general form their perfect in -uī, their PPP in -ūtus (even tuor/tūtus would conform to this rule!); a list would have to specify four exceptions: ruo has rūtus, but in the compounds -rŭtus; two verbs which have a velar in the perfect (fluxī, struxī) show it in the PPP, too, as does frūctus from fruor.
(pp. 11-12) (emphasis added)
I was a bit confused by the rule saying that the length of the vowel in -rutus depends on whether the word is a compound. Some 17th-century books on Latin grammar seem to indicate that the form ruitum is used as the supine of ruo 'to rush', but that -ruitum is not used in derived compound words (The Royal Grammar, Commonly Called Lilie's Grammar, explained by William Walker (1695); A Compleat System of Grammar, English and Latin, by William Clare (1699)); however, I don't know whether ruitum is a post-Classical form—it doesn't seem to be mentioned in Lewis and Short.
A further complication is that apparently, there are two different verbs that have ruo as the first principal part. Szemerényi refers to them as ¹ruo and ²ruo. The form rūtus seems to occur primarily as the PPP of ²ruo, which is thought to come from a PIE root ending in *w~u followed by a laryngeal. In Indo-European languages, u (=syllabic *w) followed by a laryngeal developed into the long vowel ū.
Many of the other -uo verbs with PPPs ending in -ūtus (acuo, arguo, futuo, metuo, minuo, statuo, tribuo) seem to be "denominative", which somehow explains the length of the vowel. Szemerényi says
- The PPP has, as we have seen (2.1), -ūtus in the denominatives. This is what we should expect in any case, seeing that before the suffix -to the stem-final vowel is lengthened even in purely adjectival forms; c.f. aurītus, crīnītus, turrītus, on the one hand, argūtus, cinctūtus, cornūtus, versūtus, verūtus, on the other, and, outside Italy, e.g., Greek δακρῡτός 'wept over, tearful'.
Szemerényi gives different vowel lengths in the PPP for ¹ruo and ²ruo:
¹ruo -uī -ŭtum 'rush, overthrow' (Palatus)
²ruo -uī -ūtus 'root, rake, scrape' (Lucretius)
There are two corresponding entries for ruo in Wiktionary and in De Vaan (2008), but they give conflicting information about the perfect and supine forms:
For ¹ruo, Wiktionary says "perfect active ruī, supine rutum" (Wiktionary doesn't use breves in its Latin entries, so the lack of a macron is supposed to imply that the vowel is short). De Vaan also gives the perfect as ruī, but gives the PPP as rū̆tum. As I mentioned above, I've also seen the form ruitum in a few sources. The Latin Wiktionary ("Victionarium") says " rutum vel rūtum vel ruitum".
For ²ruo, Wiktionary says "no perfect or supine forms". De Vaan gives no perfect form, but gives the PPP as rūtum.
Szemerényi says ¹ruo is from IE *ereu- (I don't understand the reason for the initial *e in this reconstruction, but it seems irrelevant to the development of the vowel after the *r) which is also supposed to be the source of Sanskrit r̥ṇōmi and Greek ὀρούω (p. 23). Wiktionary says that the verb that Szemerényi calls ¹ruo is from PIE *h₃rew-. For Italic specifically, Szemerényi gives the reconstruction *rewō, and says "The form of the Latin verb would then be due to the compounds" (p. 24). I'm not sure what this means. De Vaan gives the Proto-Italic form as *rowe/o-, and the PIE form as *h₃reu-e/o-.
For ²ruo, after listing some cognates, Szemerényi suggests that it comes from IE "*rūyō" (comparing the form to Old Church Slavonic ryjo). Wiktionary says that what Szemerényi calls ²ruo is from "*(H)rewH-", which seems consistent with Szemerényi's suggestion (PIE *uH developed to IE *ū). De Vaan gives the Proto-Italic form as *rowe/o- (the same as for ¹ruo, apparently), and the PIE form as *(H)reuH-e/o-.
Although the sources don't agree on all the details, I think it seems fairly clear that the reason for the (at least potential) length difference between the vowels in the PPPs of ¹ruo and ²ruo is the laryngeal that comes after the w~u in the reconstructed ancestor of ²ruo in PIE.
Some weird things seem to happen with vowel length in related words or derivatives of these words. De Vaan's entry for ²ruo lists rutus, -ūs, rutābulum, rutellum, rutrum, rūta caesa (as well as some verbs) as derivatives, and says "It seems that the derivatives of ruō 'to rush' and ruō 'to dig' have become mixed up: we find short-vowel forms meaning 'dig', in spite of the PIE root etymology *(H)ruH-" (p. 531). Likewise, Szemerényi says in a footnote on p. 23 that "rutrum, rutābulum, 'Schaufel', rūta caesa [...] belong with ²ruo".
A (long) side note: Lewis and Short use a macron on the first syllable of rutrum, but I'm inclined to think that this is an example of the use of the macron to mark metrical weight. Because rutrum has a medial consonant cluster, it could scan in poetry as rut.rum with a heavy first syllable. That doesn't make the vowel in the first syllable long (despite the confusing old description of such vowels as "long by position"), but Lewis and Short seem to have considered syllable weight sufficient reason to mark a vowel with a macron, at least in certain contexts where the weight of the syllable would otherwise be somewhat non-obvious—e.g., as far as I know, L&S don't use macrons like this on vowel letters before double consonant letters, but they do seem to use them on vowel letters before single consonant letters that scan long, like the j in mājor (which is thought to have been pronounced [majjor] in Latin) and the z in trăpēzīta (which comes from Greek τραπεζίτης, with epsilon; L&S mention an alternative spelling with tarpess- which provides further support for the idea that it was the consonant rather than the vowel that was long in this word).
Szemerényi, Oswald. 1980. "Latin Verbs in uo, uere". Italic and Romance Linguistic Studies in Honor of Ernst Pulgram, edited by Herbert J. Izzo.
De Vaan, Michiel. 2008. Etymological Dictionary of Latin and the Other Italic Languages.