Past participle agreement rule in Romance languages (and Latin?)

[Androcles]

I am researching the history of the past participle agreement rule in Romance languages and have so far looked at French and Italian. In French there is a rule that the past participles of verbs must agree with direct objects in gender and number when the direct object precedes the verb: les crêpes que j’ai mangées (‘the crêpes that I ate’). According to the Italian language academy, this rule does not exist in present-day Italian and it is possible to write either la meta che ci siamo prefissati or la meta che ci siamo prefissata ('the task/goal we have set ourselves'). However, I sense that in the past Italian writers tended to follow the present-day French rule of agreement with the direct object.

What I would like to know is - and please excuse my ignorance of Latin - is there any justification for this rule in Latin grammar? Are present-day purists who oppose the reform of the French agreement rule by appealing to Latin misguided?

Many thanks for your help!

 

[Pacifica]

Hello,

No, there is no basis in Latin grammar for the weird French rule whereby the participle agrees with the direct object only when the latter precedes the former.

In Latin, participles always agree with the noun or pronoun they refer to.

Latin did not have an equivalent of the passé composé formed with avoir + participle. The passé composé and the passé simple both translate to the same Latin tense, the perfect, which uses participles only in the passive voice. The Latin pluperfect and future perfect, too, involve participles only in the passive. There is, however, a classical Latin construction formed with habere ("to have") and a past participle, and that construction would later evolve into the compound French tenses we know. Back in classical times, though, it wasn't a tense but had a slightly different meaning, with more emphasis on a current state of things. For instance, you could say hoc habeo compertum, literally "I have this ascertained", that is, "I know this for certain, based on experience or information I got".

 

[Etaoin Shrdlu]

the weird French rule whereby the participle agrees with the direct object only when the latter precedes the former

Are there any parallels in any other Romance languages? It's one of those things that makes it easy to believe the language was a deliberate creation of an individual or group with a warped sense of humour. It's not as though you can even hear the difference in most cases.

 

[Pacifica]

Are there any parallels in any other Romance languages?

I don't know, but the creator of this thread speculates that there may have been one in Italian at some point:

According to the Italian language academy, this rule does not exist in present-day Italian and it is possible to write either la meta che ci siamo prefissati or la meta che ci siamo prefissata ('the task/goal we have set ourselves'). However, I sense that in the past Italian writers tended to follow the present-day French rule of agreement with the direct object.

 

[Quasus]

In Portuguese, there's nothing of the sort.

In French, the masculine and feminine forms of past participles are homophones for the most part. However, when they are not, is this rule strictly observed in speech? Are there any chances of hearing something like l'occasion que j'ai pris from a native?

 

[Pacifica]

Are there any chances of hearing something like l'occasion que j'ai pris from a native?

Yes, it's a mistake people make from time to time when they speak carelessly. It still sounds wrong, though.

In Belgian French, the masculine and feminine forms that are homophones elsewhere aren't quite so, but they still sound similar enough that, if not listening closely, most people probably wouldn't notice it if you failed to make the distinction.

 

[Glabrigausapes]

I'm curious to know what the rule in Romanian is, if a rule there be.
Speaking of 'be', @Bestiola.

 

[Androcles]

Hello,

No, there is no basis in Latin grammar for the weird French rule whereby the participle agrees with the direct object only when the latter precedes the former.

In Latin, participles always agree with the noun or pronoun they refer to.

Latin did not have an equivalent of the passé composé formed with avoir + participle. The passé composé and the passé simple both translate to the same Latin tense, the perfect, which uses participles only in the passive voice. The Latin pluperfect and future perfect, too, involve participles only in the passive. There is, however, a classical Latin construction formed with habere ("to have") and a past participle, and that construction would later evolve into the compound French tenses we know. Back in classical times, though, it wasn't a tense but had a slightly different meaning, with more emphasis on a current state of things. For instance, you could say hoc habeo compertum, literally "I have this ascertained", that is, "I know this for certain, based on experience or information I got".

Thanks for your reply. So if I understand you correctly, the past participle/direct object agreement rule exemplified by les crêpes que j’ai mangées is actually CLOSER to Latin grammar (the participle agrees with the noun), and that agreement shown in (e.g.) j'ai mangé les crêpes is a 'deviation'? (Perhaps I should point out that I'm not a grammarian - I'm a sociolinguist interested in the historical origins and standardisation processes of such language features.)

 

[Pacifica]

So if I understand you correctly, the past participle/direct object agreement rule exemplified by les crêpes que j’ai mangées is actually CLOSER to Latin grammar (the participle agrees with the noun), and that agreement shown in (e.g.) j'ai mangé les crêpes is a 'deviation'?

Yes, that's right.

 

[Issacus Divus]

There is, however, a classical Latin construction formed with habere ("to have") and a past participle, and that construction would later evolve into the compound French tenses we know.

Fwiw, I read somewhere that Germanic compound tenses influenced French and Italian’s system of compound tenses.

 

[Pacifica]

Fwiw, I read somewhere that Germanic compound tenses influenced French and Italian’s system of compound tenses.

Germanic (well, at least English) compound tenses actually came about in a very similar way to what I described regarding Latin. It could be that it first happened in Germanic languages and then influenced Latin/Romance, or vice versa, or it could be that the same thing happened independently on both sides. I don't know. Now maybe the author of what you read had good reasons to believe it had first happened in Germanic.

 

[Bitmap]

I find that rather speculative.

If you have a scenario like 1066, where English and French intermingled to a large degree, I can see the argument for a certain degree of grammatical influence. But other than that? In a widely illiterate society? How exactly is that supposed to work?
I live 50 km away from the Czech border and have 0 idea of Czech grammar ... and I don't know anyone who has, either.
Even with English, which most Germans know, it doesn't go any further than borrowing a few words or phrases.

 

[Pacifica]

Well, Germanic peoples invaded many Latin/early-Romance-speaking regions. The Franks in Gaul and the Lombards and Goths in Italy, for instance.

 

[Etaoin Shrdlu]

In that sort of situation languages tend to borrow vocabulary, but not grammar. The Norman invasion changed the nature of English irrevocably, but this is reflected in the choice of words, not basic structure.

 

[Bitmap]

In that sort of situation languages tend to borrow vocabulary, but not grammar.

That's what I mean. I was willing to remain uncertain about the impact of the Normans, but I generally think it is the vocabulary that gets affected, not the grammar.

 

[Pacifica]

In that sort of situation languages tend to borrow vocabulary, but not grammar. The Norman invasion changed the nature of English irrevocably, but this is reflected in the choice of words, not basic structure.

You're right, at least mostly. But, in English, what happened to the typical Germanic word order whereby the verb comes last in dependent clauses? Did it change of its own accord or couldn't it be that Norman French influence played a role?

You can argue that this is syntax rather than grammar, but in any case it's not vocabulary (and, to my mind, syntax is a part of grammar).

 

[Pacifica]

For the record, I too am skeptical of the theory that Romance compound tenses were calqued from Germanic languages. The classical Latin construction that I mentioned was there for the taking all along, so Germanic influence probably wasn't needed. I guess it could be, though, that independent developments on both sides reinforced each other. And, also, since I'm not an expert in early Romance and even less in early Germanic languages and their interactions with early Romance, I'm willing to admit that whoever came up with the theory may have information I don't have to support it.

 

[Bitmap]

You're right, at least mostly. But, in English, what happened to the typical Germanic word order whereby the verb comes last in dependent clauses? Did it change of its own accord or couldn't it be that Norman French influence played a role?

You can argue that this is syntax rather than grammar, but in any case it's not vocabulary (and, to my mind, syntax is a part of grammar).

And it's much easier to blame that on analogy within the language itself than on external sources.

 

[Issacus Divus]

There is undoubtedly influence, but I don't believe the theory either. I wouldn't call myself an expert, but I've read a lot about it. Latin already had this coming:
"Metuo enim ne ibi vos habebam fatigatos."
"For I fear that I have tired you." (From Augustine, 5th century).



Ultimately I would say that they come from a similar grammatical phenomenon, in which "have" in these languages, still a possessive verb, eventually becomes the perfect marker (the "resultative", a past action that has an effect on the present). Proto-Germanic descendant's eventual similar forms to Romance are because of this, imo.

 

[Godmy]

For instance, you could say hoc habeo compertum, literally "I have this ascertained", that is, "I know this for certain, based on experience or information I got".

For example in Czech, this phrase works too, but it's not a separate tense either, it's more like a few phrases. 'Mám uvařeno/navařeno.' = habeō coctum ...
(in contrast with "uvařil jsem/navařil jsem" = coxī)