Why Before and Not After?

BBC News reports on a shocking suggestion:

The suggestion by a pair of Belgian teachers to drop a rule of grammar drilled into every French speaker at an early age has led to some amusement and consternation in France. The teachers say rules for past participles that follow the verb avoir (to have) should be simplified. The change would save some 80 hours of teaching time, they argue. It has been endorsed by the linguistic authorities of Belgium’s French-speaking Wallonia region and Brussels.

Currently, the rule is that the past participle of a verb does not agree with the direct object of a sentence if it comes after it, but it does when the object comes before the participle. So for instance, in the sentence j’ai mangé des frites (I ate chips), mangé remains the same. But in the sentence les frites que j’ai mangées (the chips that I have eaten), the participle agrees with the word chips, which is feminine and plural.

The two teachers, Arnaud Hoedt et Jérôme Piron, argue the rule is overly complicated and inconsistent, and that the participle should remain unchanged regardless of the position of the object in the sentence if used with the verb to have. “Schoolchildren ask, why before and not after?” they said in an opinion piece in Liberation (in French). The rule was imported from Italy by pedants in the 16th Century and is being dropped in everyday use, the pair argue. The suggestion led to anger and derision on social media, with some arguing the change would amount to ignoring the subtleties of the language. One teacher and grammar expert said the change was akin to “wanting to raze all the little streets in an old city”.

Quelle horreur! Actually, I think we’ve seen this before at LH, and I feel obliged to point out that les frites que j’ai mangées is not a sentence, but what the hell — this kind of thing is always fun. Thanks, Trevor!


  1. The proposal to treat past participles as invariable after avoir is not so shocking, but the flip side of the proposal, writing e.g. « Ils se sont faits construire une maison » and presumably even « la maison qu’ils se sont faits construire », blows my mind somehow. Unsurprisingly, the Libération op-ed plays down that aspect.

    Also unsurprisingly, it focuses on the spelling change, without acknowledging cases like fait(e), dit(e), mis(e), and pris(e) where the inflection is reflected in pronunciation. In this light, their statement that « Les linguistes vous le diront : l’orthographe n’est pas la langue, mais l’outil graphique qui permet de transmettre, de retranscrire la langue, comme les partitions servent la musique », while absolutely true, is cheating a little bit. (Though I suppose that linguists are unlikely to defend the rule anyway: it’s an artificial rule, such that anyone who follows it in speech probably learned it in school, or at least from reading edited prose, rather than as part of normal first-language acquisition.)

  2. > Actually, I think we’ve seen this before at LH, […]

    Are you thinking of this thread? : http://languagehat.com/the-rise-of-prescriptivism/

  3. >is being dropped in everyday use

    I might have thought mange and mangees were homonyms in spoken French. Have I forgotten something from my long ago single year of high school French? Liaison when followed by a vowel?

  4. It is no longer a hard and fast rule in Italian. You can say „le patate fritte che ho mangiato” and not be viewed as sub-literate. But you could also get away with “ho mangiate le patate fritte”, even if it seems a bit archaic.

  5. David Marjanović says

    Best quote from the article:

    Quant à l’Académie française, n’étant pas composée de linguistes, elle n’est jamais parvenue à produire une grammaire décente et ne peut donc servir de référence.


    I might have thought mang[é] and mang[é]es were homonyms in spoken French.

    They are (in all kinds of spoken French that count as standard or close enough). But, as mentioned in the first comment, there are a few participles that are not homophones in masculine and feminine (although the plural -s is silent anyway).

  6. Are you thinking of this thread?

    I’ll bet I am; thanks!

  7. Anonymous Coward says

    Speaking of Belgium, isn’t it Belgium where the vowel length distinction between masculines and feminines is still maintained?

  8. Giacomo Ponzetto says

    Vanya, the Treccani grammar and the Accademia della Crusca agree that there has never been a hard and fast rule in Italian. In theory, the participle can always agree with the object, but never has to.

    In practice, I hope that language learners are taught the rules of thumb that describe actual usage.
    1. The participle should not agree with the object unless it is a personal pronoun.
    2. The participle should agree with the object if it is a third-person personal pronoun: lo, la, le, li or ne.
    3. The participle can indifferently agree or not agree with the object if it is a first- or second-person personal pronoun.

    Google search results bear out the empirical relevance of these rules:
    1. “che ho mangiate” =10 ghits; “che ho mangiato” = 548,000 ghits
    2. “le ho mangiato” = 137 ghits; “le ho mangiate” = 22,000 ghits

  9. I am surprised this is presented as a new proposal by Belgians. It seems to me that the rule was deemed optional by the French Ministry of Education a few years ago.

    This is how I described the rule to my students:
    – When the object comes after the verb, until you encounter the object you don’t know what it is, so the participle does not have anything to agree with so: j’ai mangé …..
    – Once the object (preceded by a definite article) has been mentioned (even in the form of a pronoun), you know what the participle should agree with, so: les pommes de terre que j’ai mangées ‘the potatoes I ate’, je les ai mangées; ‘I ate them’.

    But it is true that there are exceptions: with j’ai mangé des pommes de terre ‘I ate (some) potatoes)’, the indefinite article can only been replaced by the pronoun en, and there is no change in the participle: j’en ai mangé ‘I ate some’. The hardest exception is what to do with the participle of faire followed by a ‘pronominal’ verb, as in je me suis fait couper les cheveux ‘I had my hair cut’. As far as I know, only fait is possible here according to the “old rule”, but where both the subject and the object are feminine some people will make the participle agree, as in (woman speaking) je me suis faite faire une robe ‘I had a dress made (for me)’.

    However, because many adjectives are undistinguishable from participles, many people seem to have decided that adjectives do not need to agree with nouns, hence for instance une histoire vrai (instead of vraie) ‘a true story’.I see this sort of thing a lot in Facebook contributions by apparently barely literate French people).

  10. I might have thought mang[é] and mang[é]es were homonyms in spoken French.

    David is right, this is true for Standard French, but in the Sarthe department, for instance (just South of Normandy), rural speakers make a difference between mang[é] and mang[èy]e(s).

    At least they used to when I was growing up. We lived just about at the border between Normandy and Maine, heard rural speakers from both areas at the farmer’s market, and there were audible differences between the two dialects, such as this one.

  11. rural speakers make a difference between mang[é](s) and mang[èy]e(s).

    This means that the participle agreement rule cannot simply be attributed to adoption”by pedants” of an Italian rule. I think there must have been some variation in French at the time, and “pedants” decided in favour of the rule that was closest to Italian.

    Why Italian? Because the queen was from the Florentine house of the Medici, and had arrived in France accompanied by hundreds of Italians skilled in techniques, arts and letters, the likes of whom she did not think existed in France. They did cause a revolution in table manners, by using la fourchette, the fork. Until then, French people eating meat, for instance, brought the piece of meat to their teeth and cut out a slice right there.

  12. When the object comes after the verb, until you encounter the object you don’t know what it is, so the participle does not have anything to agree with […]

    This maybe a good way to help students to remember the rule, but it is not a good explanation. As our host expressed many times, language is not based on logic. I am usually not inclined to take it as far as he does, but this cannot be a real explanation.

    For example, in Russian adjectives normally precede the noun they modify and agree with it in gender and number. Usually it is not a problem, though in spontaneous speech, when someone is not quite sure what the head noun is going to be, it leads to an occasional disfluency. Big deal!

  13. D.O. : This maybe a good way to help students to remember the rule, but it is not a good explanation.

    I agree. I was not presenting it as a real explanation, it was a practical memory aid, useful in the majority of cases. Some people here might have a use for it!

  14. David Marjanović says

    rural speakers make a difference between mang[é] and mang[èy]e(s).

    At least they used to when I was growing up.

    About 10 years ago I heard a student on the Jussieu campus (at the time Universities Paris 6 and 7) do this, so it’s not extinct.

    They did cause a revolution in table manners

    And in food! That’s pretty much the difference between medieval cuisine (which basically survives in England) and modern mainland European cuisine.

  15. It’s useful to me. I didn’t know this formulation of a “rule” for something I haven’t thought much about anyway, since I read French in 99:1 proportion to producing it.

    Are folks here familiar with Norbert Elias, The Civilizing Process (Über den Prozess der Zivilisation) ? He analyzes the introduction of forks and ever more fiddly ways of speaking as moves to create marks of social distinction. The bourgeoisie apes these superior distinctions, so new ones are continually needed. Pedants are only means to this end.

  16. David Marjanović says

    That is definitely how RP and U-RP work.

  17. Stu – see also Pierre Bourdieu’s Distinction.

  18. Kсёнѕ Фаўст says

    In Polish there are two competing ways of building the analytic future, one uses the infinitive (like e.g. in Russian) and the other a participle also used to construct the past tense. XIX-century grammarians tried to introduce a rule in the same vein as the French one that one was to be used when the auxiliary preceded the main verb and the other when it followed but (luckily) it didn’t catch on.

  19. Matthew Roth says

    In Swiss French, there is a difference in the length of the vowel which indicates that the word is feminine.

  20. The Italianization of French, how it went and how far it might have gone. I particularly think “J’ay l’usance de spaceger par la strade aprés le past” is a pretty amazing fusion dish.

Speak Your Mind