We have this weird rule in French: you always "accord" with the sex and number of the subject if the auxiliary verb is "etre" (to be) but you never do if it is the verb "avoir" (to have). BUT, you have to accord to the complement (the object) if the auxiliary is "avoir" AND the complement is BEFORE the verb.
I recognized Helen: j'ai reconnu Helen (the complement is Helen, after the verb, no accord) I recognized her: je l'ai reconnuE (the complement is "l' " for "Helen", before the verb, therefore with accord)
it is by far the weirdest rule of French conjugation :D
First, French grammar is complicated, but it doesn't have the phenomenon that sometimes you have to reverse gender. (I think Arabic has it.) This is not at all what complement means.
You need to read the answer again and be careful to apply the rule correctly. It's really complicated.
- You always 'accord' with the gender and number of the subject if the auxiliary is être.
The auxiliary in our sentence is a form of avoir, not of être, so this does not apply.
- But you never do this if the auxiliary is avoir.
This is the situation we are in. So no 'according' of gender and number with the subject.
- But you do have to 'accord' gender and number, and to the object rather than the subject, if the auxiliary is avoir and the object appears before the verb/auxiliary.
The auxiliary is a form of avoir, but the object appears after the verb, not before it. Therefore this special case does not apply.
The following three sentences are examples of the first, second and third rule, respectively:
- Elle est allée regarder ta voiture. (Allée has the feminine extra -e because it 'accords' with the feminine subject elle.)
- Elle a reconnu ta voiture. (Reconnu does not accord with anything because the auxiliary is avoir, not être, and the object ta voiture is not before the verb a reconnu.)
- Son mari aussi, il l'a reconnue. (Reconnue has the feminine extra -e because it 'accords' with the object la - shortened to l'. It does so because the auxiliary a is a form of avoir and the object appears before the verb a reconnu.
Your spelling is wrong, actually:
- Les gâteaux, je les ai faits. - The cakes, I have made them.
- Les erreurs, je les ai faites. - The mistakes, I made them.
And as if the rule weren't weird enough already, there is also a weird exception for when the verb faire is used to express the causing of something:
- Ils nous les ont fait construire. - They have made us build them.
This weirdness in French is met with a weirdness in English, which uses an otherwise obsolete word order with the past participle to express the same relation. So in English it's I had them constructed (vs. normal word order I had constructed them) instead of a more natural I had made them constructed.
For other, similar two-verb constructions one has to distinguish carefully which verb the direct object belongs to:
- Les chansons, vous me les ont entendu[e] chanter. - The songs, you have heard me sing them.
Here agreement is with me, not with les. If you look at the English translation it's clear that me is the direct object of have heard and them is the direct object of sing.
(I am not a native speaker of French. This being tricky business, it's quite possible I made a mistake in the last example. In this case I am sure a native speaker will provide a correction sooner or later.)
According to this rule, I'm also having that doubt...
I thought so, too, but the link also lists the following exceptions:
There is no direct object agreement with the causative or with verbs of perception. Il les a fait travailler. He made them work. L'histoire que j'ai entendu lire The story that I heard read.
I assume "avoir reconnu" fits into this category. (But how I'll ever remember this, I don't know).
No, reconnaître is not in this category. The agreement only happens if the object precedes the verb, which usually (see comment for an exception) only happens if the object is expressed by a personal pronoun:
- Elle a reconnu la voiture.
- Elle l'a reconnue.
If you want to verify this, it's clearer in the case of the plural. (This avoids the issue that l'a might also be the contraction of le a.) You can enter "les a reconnus,les a reconnu,a reconnu les,a reconnus les" into Google's ngram viewer and set it to French. (I would give you a direct link to the result, but the forum software messes with it because it contains ampersand characters.) Result:
- a reconnu les is, and has always been, common
- les a reconnus is, and has always been, common
- les a reconnu is practically non-existent since about 1800; it did occur occasionally in the 18th century
- a reconnus les is, and has always been, practically non-existent
Oxford University Press also does it.
PS: I don't. Not just because I have lived in the UK and prefer British English. I think it's objectively better in that for countless words of French origin, the s spelling preserves the original spelling better and therefore makes life easier for polyglots. English orthography is so chaotic that the introduction of the z spellings didn't really make any sense in the first place as contrary to what I think was intended it caused no appreciable regularisation. Quite the contrary. Changing from z spelling to s spelling is trivial if you know a small number of exceptions. Changes in the other direction are much less predictable.
The story I heard was the the -ize ending is actually the more original. Apparently, it was changed from -ize to -ise due to the influence of the french and latin during the middle ages. When Webster made his dictionary in the 1800's he wanted Americans to be different and "more correct" than the british, so he used the old -ize ending. And it caught on from there. At least, that's the story I was told!
This sounds like the typical simplification of a complicated story, as they arise in decades of word of mouth. There are some serious problems with it:
First, to get this out of the way, the difference in spelling between verbs that have -ise even in American English and those that have -ize in American English (or in Oxford spelling) clearly does not reflect a difference in pronunciation. But the claim you are reporting says it's more original. The problem with this is that it applies only to Oxford-style z but not to American-style z. Although the two styles agree on many common words, they also differ on many words. When they do, it's always the American style that is a pure innovation rather than a historical or etymological spelling.
Oxford-style z is in verbs that originate in Greek, where they did not have an s or z sound at all but the nouns derived from them had one, which was spelled as a zeta. They entered Latin with a z, and in many cases English got them direct from Latin and initially spelled them with a z. However, French like most of the Romance languages regularised the z to s, and many other words of this type entered English not directly from Greek or Latin but via French, where they were already regularised to the s spelling.
Basically there are the following important groups:
- Verbs such as organize, realize, finalize which had zeta / z in Greek and Latin but s in French, and entered English directly from Greek or via Latin.
- Verbs such as canalise, advertise that are modern French or English derivations from a noun.
- Verbs such as analyse, paralyse that had sigma / s already in Greek.
- Verbs such as revise, incise, surmise, surprise in which -ise is not a suffix but part of the stem or of a Latin/Romance inflection that was originally always spelled with s (vise from videre, cise from caedere, mise from mittere, prise from prendere).
Verbs of class 1 are spelled with a z in American and Oxford spellings.
Verbs of class 2 are spelled with a z in American and Oxford spellings even though many of them are French in origin. The idea is apparently that -ise is nowadays an English suffix that should be spelled -ize because of its Greek origin, so French words get 'regularised' to the -ize spelling.
Verbs of class 3 are always spelled with an s in Oxford spelling, but are spelled with a z in American spelling. (Personally, while Oxford spelling looks as natural to me as modern mainstream British spelling, I find the z spelling in these words extremely jarring. It just looks plain wrong to me, and as uneducated as rite for right.)
Verbs of class 4 are spelled with an s in Oxford spelling and American spelling, although in American spelling (only) there is a tendency to slowly 'regularise' some of them to -ize.
Although I'm an S man myself, I have to disgree. In British English we have a choice.
Since The Times finally gave in and joined the S spellers a few years ago, it is true that the British media pretty well exclusively use the S spelling, but Z is also widely used, most notably by Oxford University Press.
In any case, it recognised "recognise". I use exclusively British spelling on DL, and virtually never have a problem.
"She would have known your car" was marked incorrect. This seems reasonably common in US English, "knew" meaning "recognized". "The child would have known his mother's voice" or "the man would have known his brother by his shuffling gait". Duolingo is accepting 'reconnu' as 'knew' or 'known' in other phrases as well.
Normally, agreement of past participles is not with the subject but with the direct object, and only if the object precedes the verb. Usually (with some exceptions) this only happens if the object is a pronoun. If there is no direct object, or if it does not precede the verb, there is no agreement at all.
"Ils auraient reconnus" is just wrong. Some examples of correct sentences:
- "Ils les auraient reconnus." (Plural direct object precedes the verb; subject just happens to be plural as well.)
- "Ils auraient reconnu les autres." (Plural direct object, but it does not precede the verb.)
- "Il/elle les aurait reconnus." (Plural direct object precedes the verb; subject is not plural but doesn't matter anyway.)
- "Ils l'auraient reconnu(e)." (Direct object is abbreviated le or la preceding the verb. Masculine or feminine singular agreement depending on which of the two it is.)
- "Ils auraient joué." (No direct object at all, so no agreement.)
- "Elle s'est lavée." (Feminine agreement with the direct object, which happens to be the same as the subject because the verb is reflexive.)
- "Elle s'est lavé les mains. (Here the direct object is "les mains" and does not precede the verb. "Se", abbreviated to "s'" is feminine and precedes the verb, but it's not the direct object but the indirect object. So there is no agreement.)
As if this weren't confusing enough, for those non-reflexive verbs that form the passé composé with être rather than avoir, it works the way you expected it to work: Agreement with the subject. Example:
- Ils étaient allés.
This only concerns verbs of movement, so they can have no direct object.