sestertium

[Lavrentivs]

Haec domum quinque millibus sestertium mihi stetit.

Why is sestertius in the accusative and not the genitive?

[adrianus]

Accusativo casu non est, immò genetivo numeri pluralis.
It's not in the accusative; "sestertiūm" is genitive plural.

[Lavrentivs]

Thanks. Is it regular in all other cases? Are there other such words?

Gratias tibi ago. Estne aliis in casibus regulare? Aliane talia verba inveniuntur?

[adrianus]

See here // Vide hoc: http://www.perseus.tufts.edu/hopper/tex ... sestertius

It's not unusual to have "-ūm" for "-ōrum". // Non alienum "-ūm" pro "-ōrum".
Vide A&G §49d, http://www.perseus.tufts.edu/hopper/tex ... 99.04.0001

[timeodanaos]

Now after a three-year hiatus, my user is awakened just to state the obvious fact that counting sesterces has always been tricky.

1/4 of a denarius = unus sestertius = I HS
Sestertium with the plural of mille is an old genitive: duo milia sestertium (=-tiorum) = 2.000 HS; by analogy, this is perceived as a neuter singular, which can then be pluralized: duo sestertia = 2.000 HS. If you want to express even larger numbers, sestertium with a multiplicative number (e.g. decie(n)s) = 100.000 HS multiplied with with multiplicative: decies sestertium = 10 x 100.000 = 1 million HS; decies centena sestertium = 10 x 100 x 100.000 = 100 million HS.

As regards the quantity of the final syllable, I'm finding it hard to come to terms with it being long. It's a while since I looked into historical morphology, but a quick look at Sihler tells me that final closed syllables were always shortened, and as the IE gen.pl. ending is -ôm and there is a clear opposition between long u and long o in Latin, and no opposition between short u and o, it seems highly unlikely that the -um of gen.pl. should have been long in historical times, regardless of L&S. It's worth noting that the -um ending of sestertium is the reflex of the original IE ending, whereas -orum is of Italic origin.

[adrianus]

As regards the quantity of the final syllable, I'm finding it hard to come to terms with it being long.

This is a moot point, timeodanaos. L&S don't have it long. In Keil and elsewhere, I remember discussion by certain of the old grammarians that the "-ūm" for "-ōrum" had a long u and that it took the stress accent. I like that, but I won't chase it as it will be quite a job to trace the passages. Forgive me. I have to say, that the lengthening of a syllable or the vowel in a syllable, rather, due to contraction and to hold the stress don't seem so bizarre to me, but I might be just odd.

Id causa difficilis est, timeodanaos. Sic non describitur apud L&S. Apud Keil autem et alibi, ut memoro, sic dicunt aliqui grammatici antiqui , scilicet longam esse u litteram in "-ūm" pro "-ōrum" terminanti, porrò vim habere eandem syllabam, Id placet, ut opinor, vestigia autem non quaeram praeter tempus longum quod inerit. Me excusas. Non alienum mihi, dicam, syllabam vel vocalem contentam prolatari post contractionem, non minùs emphasin teneri, at ego forsit alienus sum.

[timeodanaos]

To recollect the contents of Sihler (New Comparative Grammar of G. and L., OUP 1995) §260.5-6, Latin has two distinct ways of marking gen.pl. of o-stems: the reflex of the IE-ending -ôm, which is regularly shortened and weakened (or rather "made regular") into the normal -um known from gen.pl. of consonant-stems. L&S marks this as long sub voce sestertius. Be it so or not: it would be unnatural for such a highly frequent word to retain an irregular quantity, especially in the last syllable, and if we were ever to encounter the word in poetry, the scansion would remain unaffected. The other way of marking gen.pl. is -ôrum, which is normally said to be imported from Italic pronominal endings.

I would as always be cautious when consulting ancient grammarians: the distinction of quantity had for the most part fallen out of the natural language at this point, and a man as prolific as Priscian even claims the last -i- in Greek words such as philosophia to be long; the i was never long, but rather lengthening as a result of stress had entered the world of Latin-speakers.

[adrianus]

I would be too cautious to ignore them, especially when they had access to sources since lost and when they were writing as teaching practitioners over such an extended period, not as academic theorists (that's not directed at you, timeodanaos, by the way).

Cautior ego sum ut eos ignorem, praesertim cum fontes prae manibus longè nobis perditas haberent et tam spatium temporis magistri periti non modò academici theoriâ saturi docerent (non te ipsum, obiter Timeodanaos, cum academicis includo).

Priscian even claims the last -i- in Greek words such as philosophia to be long; the i was never long, but rather lengthening as a result of stress had entered the world of Latin-speakers.

Priscian there claims only what is the case, not necessarily what was the case, which is surely very reasonable. Nor does it remove the possibility that earlier grammarians did teach that pronunciation but, given his deference to authoritative sources, it opens it as a distinct possibility that some may have.

Solùm quod fuit, non necessariò quod priùs fuerat, indicavit Priscianus, de quo non disputandum. Manet autem possibilitas aliorum qui sic priùs docebant, ità factum, eum auctoritates anterias existimare.

L&S marks this as long sub voce sestertius.

Sorry, that's right. L&S also have it as long. I said they didn't.
Me excusas. Ita est, ista littera et in dictionario de L&S longa denotatur. Contrarium dixi.

[timeodanaos]

I'm sorry not to elaborate more on this right now (I'm immersed in Horace's Odes, not the worst job), but my primary argument wasn't as much the regular shortening as the fact that the -um in question is a reflex of the gen.pl. also known from 3., 4., and 5. declension, and as it is regularly shortened there, I think we should count the final syllable of sestertium as short too. My other argument was the more otiose one that an irregular long syllable probably wouldn't last 5 minutes in the Forum among beggars and thieves.

I would not count myself among the academic theorists; I have a BA in Latin, but mostly focused on literature

[adrianus]

The difference with the second declension is that the circumflected accent/vowel-lengthening differentiates the accusative singular masculine and the genitive plural in the contracted form. Don't underestimate thieves and beggars.

In hoc est discrimen prae nominibus secundae declinationis: accentum circumflectum seu magnitudinem protentam concisione differentiam facere casibus accusativo singularis masculini generis et genetivo pluralis numeri. Non nulli momenti putes fures mendicosque.

[timeodanaos]

Don't underestimate thieves and beggars.

I have always trusted beggars, thieves, scholars and heads of state to follow the unwritten (and in the last century, written) laws of language change, even if apart from that one they only follow their own voluptates.

EDIT: beggars, scholars; thieves, heads of state: at the time of writing I didn't realize I was making a hendiadys.

[adrianus]

Antequàm in dictionarium inquaesivi, hendiadys vocabulum ignorabam. Bonum est, et benè factum!
I didn't know the word hendiadys before looking it up. Good one. And nice one!