Gallo-Italian plurals

[Beli Orao]

According to this Wikipedia article, the feminine plural in Gallo-Italian is derived from the Latin accusative plural, while the masculine plural is derived from the Latin nominative plural. Are there any examples of this in other Romance languages?

[Dewrad]

Old French and Old Occitan in the direct cases exhibit this. The wikipedia article's probably wrong about Gallo-Italian, though: I wouldn't assume it's correct without checking sources.

[dhok]

I have heard somewhere, but I can't find the source, that the Italian plurals indeed derived from the accusative plurals, with the sound changes
os>oi>i
as>ai>e

[Legion]

I have heard somewhere, but I can't find the source, that the Italian plurals indeed derived from the accusative plurals, with the sound changes
os>oi>i
as>ai>e

I've heard that theory before, but it is not consistent with changes observed in the rest of the language. Futhermore, the original question was about gallo-italian languages (Piemontese, Lombard, Ligurian, Venetian), not about standard Italian, which is based on the Tuscan dialect, which belongs to the Central Italian group.

[alice]

I have heard somewhere, but I can't find the source, that the Italian plurals indeed derived from the accusative plurals, with the sound changes
os>oi>i
as>ai>e

I've heard that theory before, but it is not consistent with changes observed in the rest of the language. Futhermore, the original question was about gallo-italian languages (Piemontese, Lombard, Ligurian, Venetian), not about standard Italian, which is based on the Tuscan dialect, which belongs to the Central Italian group.

FWIW the source was probably me, quoting Martin Maiden's A Linguistic History of Italian; it's on page 81, and refers to word-final /s/ becoming /i/.

@Legion: I'm curious; with what other sound changes is this not consistent?

It more relevantly says "Inflectional -s survives in ... Gallo-Romance".

[Legion]

FWIW the source was probably me, quoting Martin Maiden's A Linguistic History of Italian; it's on page 81, and refers to word-final /s/ becoming /i/.

@Legion: I'm curious; with what other sound changes is this not consistent?

Well for instance, verbal endings.

Compare the subjunctive present of Latin "dicere" and its Italian reflex:

dicam > dica
dicas > dica
dicat > dica
dicamus > diciamo
dicatis > diciate
dicant > dicano

Subjunctive -as is consistently reflected as -a.
1pl -mus and 2pl -tis are consistently reflected as -mo and -te

For the singular imperfect, we get:
-bam > -vo
-bas > -vi
-bat > -va

This suggest analogy rather than sound change. My opinion is that the 2sg -i ending spreaded from -ire verbs, where regular sound changes (assuming final -s are simply lost) normally lead to:
-io > -(i)o
-i:s > -i
-it > -e

This also explain why -are verbs have a 2sg in -i rather than -e (arguably analogy explains that anyway, but still)


We also found the preposition "tra", which has the etymon "trans", and then we get: trans > tras > tra, rather than trans > tras > *trai > *tre

I think the assumtion comes from "magis", which gives "mai"; however here the -i most likely is a remnant of the -gi- segment.


Edit: In short, I have to see a convincing example of -s > -i in Italian that cannot more simply be ascribed to another source (like nominative plural or other phonemes in the word) or to the analogical spread of a productive ending.

[Ser]

According to this Wikipedia article, the feminine plural in Gallo-Italian is derived from the Latin accusative plural, while the masculine plural is derived from the Latin nominative plural. Are there any examples of this in other Romance languages?

Penny in his History of the Spanish Language (2002) reconstructs the case system that the late spoken Latin of Spain probably had by the 4th-5th centuries as this:

Code: Select all

Table 3.2 Case system of late spoken Latin
								Sing.	Plural
1 (a-class)		 Nom	/rósa/	/rósas/
						Obl	/rósa/	/rósas/
2 (o-class)		 Nom	/ánnos/	/ánni/
						Obl	/ánno/	/ánnos/
3a (e/cons.-class)	[...]

By the fourth or fifth century AD, the phonological changes just discussed led to a considerable reduction in the case forms of all singular Latin nouns, while the plural forms were similarly reduced by analogy with the singular. [...] In the late spoken Latin of Spain and other areas, then, the noun system probably showed the pattern presented in table 3.2 (see Penny 1980).

The system posited here makes two assumptions which have not so far been discussed. Unimportant for present purposes is the assumption that in final syllables (see 2.4.3.2) /i/ (< Ī) and /e/ (< Ĭ, Ē, Ĕ) were still separate phonemes. More significant is the statement that the nominative plural ending of a-class nouns was /-as/. There is good evidence (see Aebischer 1971) that in the spoken Latin of most areas this was the case from early times, by contrast with the literary dialect, which showed analogical -AE.

So, according to this, "most" areas presented the phenomenon you are asking about. I don't know what he means with the literary dialect "show[ing] analogical -AE" though (analogical with what? the genitive singular since in the masculine there's also syncretism there?).

About that paper by Aebischer he cites, it's called Le pl. -ās de la 1ère decl. latine et ses résultats dans les langues romanes, published in the Zeitschrift für romanische Philologie vol. 87 (pp. 74-98) in case any of you guys can access it and bring information from there.

Old French and Old Occitan in the direct cases exhibit this.

Why do you call them "direct" cases and not "nominative"? "Direct case" makes me think of "accusative case" (as an analogy with "direct object"...).

[Legion]

Why do you call them "direct" cases and not "nominative"? "Direct case" makes me think of "accusative case" (as an analogy with "direct object"...).

That's the standard terminology for Old French and Old Occitan: they have a direct case that handles the subject, and an oblique/regime case that handles everything else.

[Dewrad]

Why do you call them "direct" cases and not "nominative"?

Because that's what they're called. Would you prefer casus rectus instead?

[Ser]

I thought they were usually called nominative case/cas sujet and oblique case/cas régime.

[Dewrad]

I thought they were usually called nominative case/cas sujet and oblique case/cas régime.

Alas, you were wrong. It's a holdover from the Latinate grammatical tradition, where the direct case was called the casus rectus.

[hwhatting]

So, according to this, "most" areas presented the phenomenon you are asking about. I don't know what he means with the literary dialect "show[ing] analogical -AE" though (analogical with what? the genitive singular since in the masculine there's also syncretism there?).

What is meant is that the standard Latin NPl. in -ae is explained as an analogous formation to the -i: < PIE *-oi of the o-stems. The PIE ending for the Nom. Pl. was *-o:s for the o-stems and *-eH2es > *-a:s for the a:-stems (eH2-stems). Latin, like some other IE dialects (Celtic, Balto-Slavic, Greek), took over the originally pronominal ending *-oi for the Nom. Pl. of o-stem nouns; Latin and Greek (probably in independent developments) formed an analogous ending -ai (> Latin -ae) for the a:-stems, probably because in both languages otherwise nom. pl. and acc. pl. would have become homophonous. Other Italian languages continue the old nom. pl. forms -o:s / -a:s. So what Aebischer seems to say that old -a:s was preserved in spoken Latin, i.e. -a:s in Romance is not a replacement of the Nom. Pl. by the Acc. Pl., but a retention of the old PIE Nom. Pl. I don't know whether I'm ready to accept that - that would mean that the Italian forms are based on "literary" Latin while the forms in other Romance languages are based on an archaising "vulgar" dialect.

[Legion]

Otoh, we were discussing this with Dewrad this afternoon, and apparently, there seem to actually be evidence for an -as > -ai > -e change in Italian. But as the same time, there is counter evidence of an -os > -oi > -i change.

This would imply that Italian preserved plural nominative endings for masculine nouns but generalised -as to both nominative and accusative plural in feminine nouns (which is a consistent change we also observe in Old French).

This may give credance to Aebischer's theory that this indeed a retention.

[Soap]

Edit: In short, I have to see a convincing example of -s > -i in Italian that cannot more simply be ascribed to another source (like nominative plural or other phonemes in the word) or to the analogical spread of a productive ending.

What about the pronoun "voi" as in the famous lasciate ogne speranza voi ch'intrate? I assume that's a cognate to the Spanish "vos"?

That said, even if os did change to oi, it doesnt prove that that sound change is the source of the noun plurals. Transitional forms should exist somewhere, e.g. polloi "chickens" instead of polli. The Italians were pretty literate, so I'd expect if it was in the language at any point it would be written down somewhere. It also wouldnt explain a bunch of irregulars like uovo > uova, unless we assume that the analogical -s was not added to these nouns.

[Dewrad]

Edit: In short, I have to see a convincing example of -s > -i in Italian that cannot more simply be ascribed to another source (like nominative plural or other phonemes in the word) or to the analogical spread of a productive ending.

What about the pronoun "voi" as in the famous lasciate ogne speranza voi ch'intrate? I assume that's a cognate to the Spanish "vos"?

That said, even if os did change to oi, it doesnt prove that that sound change is the source of the noun plurals. Transitional forms should exist somewhere, e.g. polloi "chickens" instead of polli. The Italians were pretty literate, so I'd expect if it was in the language at any point it would be written down somewhere.

My own suspicion is that there was a general change of VL final s to i, but that this has nothing to do with the -i plural of Modern Italian nouns ending in -o. While there's reasonable enough evidence for a change of -os > -oi, I don't think there's evidence for a subsequent change of -oi > -i. It's my suspicion that what actually happened here is unstressed -oi > -o, which rendered the reflex of Latin -ōs homophonous with the reflexes of -us and -um. In tonic syllables, such as that in voi (note the Old Tuscan atonic form vo, lacking the final -i!), the glide was retained: as an analogy here we can point to hai < *ás "you have".

If we imagine a pre-Italian stage which distinguished two cases, akin to that proposed by Penny for late VL we might be looking at a development along these lines:

Code: Select all

Nom   *annos  *anni     >   *annoi  *anni     >   *anno   *anni
Obl   *anno   *annos    >   *anno   *annoi    >   *anno   *anno

Thus the old nominative plural ending was the only distinct plural ending remaining.

However, all that having been said, a change of -os > -i isn't actually unknown elsewhere in the Romance family. It's attested for various Occitan varieties, in Rhaeto-Romance and in a number of Gallo-Italian varieties.

So it could go either way.

It also wouldnt explain a bunch of irregulars like uovo > uova, unless we assume that the analogical -s was not added to these nouns.

Why would it explain that? Words like uovo have absolutely nothing to do with the issue at hand?

[Yæd]

@Dewrad: I always knew that voi and noi came from the Latin ablative form vobis and nobis. Singular nouns have a similar process: rosa < rosā (rosa), vento < ventō (ventus), dito < ditō (ditum) cane < cane (canis), principe < princĭpe (princeps), while IV declension nouns merged with the II declension ones and the few V declension nouns disappeared. Plurals came from the nominative plural (the -īs of the ablative plural were identical for both I declension and II declension, the most common nouns, and for III declension nouns, the plural of cane should be canibus > *caniu > *cagnu), so rose < rosae, venti < ventī, dita < dita, and (!) cani < canes. (dita is now feminine: il dito > le dita). French and Spanish simply used the accusative plural -s desinence to mark the plural.

Edit: I forgot Gallo-Italian plurals! Well, in my little town south of Milan, most of them are unalterated for both genders (la cà > le cà) while I found you can have:
- (masculine) accented -ùr > -ùri (ex. el caciadùr > i caciadùri 'the hunter'), from the III declension -or > -ores
- (feminine) unaccented -a > -e (ex. la fiöla > le fiöle 'the daughter/the little girl'), from the I declension -a > -ae
- (masculine), unaccented -a > -i (ex. el dràma > i dràmi 'the drama'), from Greek-imported nouns
- (masculine), unaccented -e > -i (ex. el léte > i léti 'the bed'), from the II declension -us > -i (they had an apophonical thematic vowel -u/-o/-e), and (merged) IV declension -ūs/-ū
- (masculine) accented -ài > -ài (ex. el travài > i travài 'the work/the travail'), from the II declension neuter -alium > -alia
- (masculine) ending in an accented vowel + -l change it into -i (ex. el cavàl > i cavài 'the horse'), from the III declension -al > -alia, or II declension -allus/-allum > -alli/-allia
- (both) accented -àn/-én/-ìn remain unchanged (ex. el càn > i càn 'the dog'), from II declension -anus/-anum
- (masculine) accented -òn > -ùni (ex. i padròn > i padrùni 'the owner'), from III declension -ō > -ōnes
- (masculine) unaccented non-apophonical -er > -eri (ex. el nümer > i nümeri 'the number'), from II declension -er > -ĕri
- (masculine) unaccented apophonical -er > -ri (ex. el lìber > i lìbri 'the book'), from II declension -er > -ri
- (masculine) accented -ér > -éri (ex. el bicér > i bicéri 'the glass (to drink)'), from III declension -er > -eris
Alterative forms (diminutive, accrescitive etc.) have their own complicated formation.

(thanks to B. Pezzini's site for the Gallo-Italian part)

@Soap: my "theory" explains this well (ōvum > ōva = It. uovo > uova, as dito > dita)

[Drydic]

English and French and Spanish simply used the accusative plural -s desinence to mark the plural.

This makes me want to completely discount everything you say as bullshit.

[Yæd]

Wooops! Sorry for that. Edited.