Syllabifying S into an onset?

Question

Someone recently called my attention to a hexameter by Lucilius:

Jānus, Quirīnus "pater" siet ac dicātur ad ūnum.
Janus or Quirinus may be "father"; every single [god] is called this.

And there's something a bit odd about this scansion. The second syllable of Jānus has to scan as light, even though it's followed by another consonant. In other words, the final -s before a stop isn't treated as a coda.

This makes sense to me—phonologically, squ- is a valid syllable onset, as in squālor. But I've never seen this sort of syllabification in metered poetry before: normally, s between a vowel and a consonant is treated as a coda (e.g. -us la- in Aeneid I.2).

How common is this? Does it happen in Classical authors too (or really anyone except Lucilius)? And is there a name for it?

Answer 1

The standard analysis of this phenomenon, which shows up in Early or Old Latin poets such as Lucilius and Plautus, is that word-final s was not resyllabified, but rather prone to elision when followed by a consonant (but not when followed by a vowel). Thus, some modern editors choose to omit the letter S and replace it with an apostrophe in such cases. Words whose final s was historically derived from a geminate, such as es (< *es-s) and mīles (<*mīless < *mīlet-s) usually are not scanned short in this manner.

The light scansion of a word-final sequence of short vowel + s can occur regardless of whether the s can form a valid onset cluster with the following consonant(s). For example, "The Problem of the Omission of Word-final -s as Evidenced in Latin Inscriptions", Adamik 2017 (page 13) cites the following example from Adams (2013: p. 132): Ennius, Ann. 377: "nos sumu’ Romani, qui fuimus ante Rudini". Here we see short scansion even though /sr/ does not occur at the start of Latin words.

Some sources discussing this describe the condition for elision more specifically as "word-final s after a short vowel", but I don't see how we'd know whether it did or didn't also occur after a long vowel, since elision in that context wouldn't affect the meter. Michael Weiss apparently shares my skepticism about whether the loss of word-final /s/ was conditioned by the length of the preceding vowel:

I don’t think we can be sure that final -s had a different treatment after long vowels than after short vowel. There can be no metrical evidence as far as I can see. Further non-notation of final -s after a long vowel (see Vine 1993:22) is found in inscriptions at Rome, and not just dialectically—and incidentally Latin inscriptions from the 3rd century are more commonly found outside of Rome.

(OHCGL Addenda and Corrigenda, p. 10)