Elamur: **Mova** · [Parse](https://mova.elamur.ai/free/en/parse/) · [Dictionary](https://md.dictionary.elamur.ai/free/en/) · [Translate](https://md.translate.elamur.ai/free/en/) · [Research](https://md.research.elamur.ai/free/en/) · [Groupants](https://md.groupants.elamur.ai/free/en/)

[Mova](https://md.mova.elamur.ai/free/en/) · [this page in HTML](https://mova.elamur.ai/free/en/docs/languages/fr/12-determiners-pronouns-numerals/)

[home](https://md.mova.elamur.ai/free/en/) / [docs](https://md.mova.elamur.ai/free/en/docs/) / [languages](https://mova.elamur.ai/free/en/docs/languages/) / [fr](https://md.mova.elamur.ai/free/en/docs/languages/fr/) / 12. Determiners, pronouns and numerals

# 12. Determiners, pronouns and numerals

**On this page**

1. [1. The articles](#1-the-articles)
2. [2. The noun without article](#2-the-noun-without-article)
3. [3. *De* in place of the article](#3-de-in-place-of-the-article)
4. [4. Demonstratives](#4-demonstratives)
5. [5. Possessives](#5-possessives)
6. [6. Quantifiers and indefinite determiners](#6-quantifiers-and-indefinite-determiners)
7. [7. Numerals](#7-numerals)
8. [8. Personal pronouns](#8-personal-pronouns)
9. [9. Object clitics and their order](#9-object-clitics-and-their-order)
10. [10. Clitics with infinitives, causatives and the imperative](#10-clitics-with-infinitives-causatives-and-the-imperative)
11. [11. *Y* and *en*](#11-y-and-en)
12. [12. Strong pronouns](#12-strong-pronouns)
13. [13. Reflexive, reciprocal and *on*](#13-reflexive-reciprocal-and-on)
14. [14. Demonstrative, indefinite and possessive pronouns](#14-demonstrative-indefinite-and-possessive-pronouns)
15. [15. Relative and interrogative pronouns: the forms](#15-relative-and-interrogative-pronouns-the-forms)
16. [16. The annotation in UD French](#16-the-annotation-in-ud-french)
17. [17. How Mova handles determiners, pronouns and numerals](#17-how-mova-handles-determiners-pronouns-and-numerals)
18. [Sources for this chapter](#sources-for-this-chapter)

A French common noun in an argument position almost never stands alone: it takes a **determiner**. The rule is stricter than in English, where plural and mass nouns can be bare (*I like dogs*, *I drink water*); French needs an article (*J'aime les chiens*, *Je bois de l'eau*). The determiner also carries the gender and number that the noun itself often does not show in speech ([11. Nouns](https://md.mova.elamur.ai/free/en/docs/languages/fr/11-nouns/)). **Pronouns** stand for a noun phrase, and the personal ones come in two series: weak (clitic) forms bound to the verb in a fixed order, and strong forms that are free words. **Numerals** are the third closed class of the chapter. All three are small, frequent and irregular.

Determiners come first (§§1–7), then the pronouns (§§8–15), the UD annotation (§16) and Mova (§17). The syntax of relative clauses is in [25. Relative clauses](https://md.mova.elamur.ai/free/en/docs/languages/fr/25-relative-clauses/), of questions in [26. Questions](https://md.mova.elamur.ai/free/en/docs/languages/fr/26-questions/), of negation in [27. Negation](https://md.mova.elamur.ai/free/en/docs/languages/fr/27-negation/).

Counts are from the UD 2.18 French treebanks, all files of each, made on 2026-11-06. GSD has 400 385 syntactic words; ParisStories (spoken) 42 789; Rhapsodie (spoken) 44 234. Unless a count is said to be by lemma, it is by lowercased surface form, because GSD's lemma column merges the personal pronouns (§16.3).

## 1. The articles

|  | masculine sg. | feminine sg. | plural |
| --- | --- | --- | --- |
| definite | *le* (*l'*) | *la* (*l'*) | *les* |
| indefinite | *un* | *une* | *des* |
| partitive | *du* (*de l'*) | *de la* (*de l'*) | (*des*) |

*L'* and *de l'* stand before a vowel or a mute *h* (*l'homme*, *de l'huile*); before an aspirated *h* the full form stays (*le héros*, *la hache*; [04. Connected speech: enchaînement, liaison, elision, the mute e and assimilation](https://md.mova.elamur.ai/free/en/docs/languages/fr/04-connected-speech/)). *Les* and *des* have a liaison /z/ before a vowel (*les amis* /lezami/).

**Contraction.** After *à* and *de* the masculine singular and the plural definite article fuse: *à + le → au*, *à + les → aux*, *de + le → du*, *de + les → des*. The feminine and the elided forms do not (*à la*, *de l'*). The fusion is obligatory (**à le*, **de le*); it happens even with a name whose article is part of it (*Le Havre → au Havre*), and writers avoid it only by quoting the title (*dans « Le Monde »*). The relative *lequel* fuses the same way (*auquel, duquel*, §15). GSD has 11 022 such multiword tokens (*du* 4 390, *des* 3 298, *au* 2 477, *aux* 808, the rest relatives), each split into ADP + DET or ADP + PRON.

### 1.1 Definite

The definite article marks a referent identified by the context (*Passe-moi le livre*) or unique (*la lune*), and, far more widely than English *the*, the **generic** reading: *Les chats sont des animaux*, *J'aime le café*, *La vie est belle*. Other uses where English has no article or a possessive:

- a body part when the owner is clear (*Il lève la main*; §5);
- a unit of price or measure (*trois euros le kilo*, *cent kilomètres à l'heure*);
- days for a repeated event (*le lundi* "on Mondays") against the bare form for one occasion (*Je pars lundi*);
- languages and most names of countries and regions (*le français*, *la Suisse*, *les États-Unis*); after *en* the article drops with a feminine or vowel-initial name (*en France*, *en Iran*), and fuses for masculine ones (*au Canada*, *du Japon*);
- titles before a name (*le docteur Martin*).

### 1.2 Indefinite

*Un, une* introduce a new singular count referent (*Il y a une maison au bout du chemin*); *des* is their plural. There is no zero plural: **J'ai vu chiens*. *Des* is not the plural of *le*: *Il a acheté des fleurs* "some flowers", *Il a acheté les fleurs* "the flowers". The article *un* is the numeral *un*; *un seul* or stress makes the number explicit (*Il n'a pas un ami* "not a single friend", against the plain *Il n'a pas d'ami*).

### 1.3 Partitive

*Du, de la, de l'* introduce an unbounded quantity of a mass noun: *Je veux du pain*, *Elle boit de l'eau*, *Il faut du courage*. The sequence *du* is two different things:

- *Je viens du marché* — the preposition *de* + the article, "from the market" (a contraction, two syntactic words);
- *Je mange du pain* — the partitive article, "some bread" (one word).

A noun moves between mass and count with a change of sense: *du poulet* "chicken (meat)", *un poulet* "a chicken". Some nouns that are mass in English are plural or count in French: *des renseignements* "information", *des cheveux* "hair", *des meubles* "furniture" (beside the mass *du mobilier*). Le, Bale and Barner (`le-2026`) used this contrast in a quantity-judgment experiment: English speakers judged object-mass nouns (*furniture*), French speakers their count translations (*meubles*), and in both languages the comparison was by number of objects, no more open to contextual effects than with ordinary count nouns. Individuation, they conclude, is fixed in the noun's lexical entry, not by its syntax.

The partitive is not only "some": it is also evaluative (*Il a du talent*, *C'est de la folie*) and turns a count noun into a mass reading (*Il y a du lapin à midi* "rabbit [as food]").

**History.** Old French used the bare noun where the modern language has the partitive (*boire vin*, *manger viande*); the bare noun is still common in the sixteenth century, and the partitive became obligatory in the seventeenth (Brunot 1899, section «L'article partitif», `brunot-1899`).

| need | French | not |
| --- | --- | --- |
| "the water is cold" | *L'eau est froide.* | **De l'eau est froide.* |
| "I drink water" | *Je bois de l'eau.* | **Je bois eau.* |
| "I like water" (generic) | *J'aime l'eau.* | *J'aime de l'eau* (= "some water") |
| "I don't drink wine" | *Je ne bois pas de vin.* | *Je ne bois pas du vin* (only contrastive) |
| "I don't like wine" | *Je n'aime pas le vin.* | **Je n'aime pas de vin.* |

The last two rows are the rule of §3.1: under negation only the indefinite and partitive articles change; the definite stays.

## 2. The noun without article

A noun stands without a determiner in a closed set of contexts.

- **After *en*** and in many fixed prepositional phrases: *en voiture*, *sans argent*, *avec plaisir*, *par avion*, *de pierre*, *à cheval*.
- **A predicate noun of profession, nationality or status** with no modifier: *Elle est avocate*, *Il est devenu médecin*. With a modifier, or after *c'est*, the article returns: *C'est une avocate*, *Elle est une avocate renommée*.
- **Apposition, address, titles**: *Paris, capitale de la France*; *Bonjour, docteur*.
- **Headlines, signs, proverbs**: *Entrée libre*, *Défense de fumer*, *Pierre qui roule n'amasse pas mousse*.
- **Verb–noun idioms**: *avoir faim*, *avoir peur*, *prendre froid*, *faire peur*, *perdre pied*; *sans doute*.
- **Enumerations and pairs**: *Femmes, moines, vieillards, tout était descendu* (La Fontaine, quoted by Brunot 1899, section «Ellipse de l'article»); *père et mère*, *à tort ou à raison*.

Brunot states the modern rule as: a French noun needs an article, or an equivalent word such as a possessive or a demonstrative, to be fully a noun; the bare uses are stylistic or fixed.

## 3. *De* in place of the article

*De* (*d'* before a vowel) replaces the indefinite or partitive article in three environments.

### 3.1 After a negation

When the negation bears on the quantity, *un, une, du, de la, des* become *de*: *J'ai des amis → Je n'ai pas d'amis*; *Il boit du vin → Il ne boit pas de vin*; *Je n'ai plus de pain*; *La fille qui répond n'a pas de temps à nous consacrer* (GSD fr-ud-dev_00353). Brunot (section «L'article partitif») explains it historically: *pas* and *point* were nouns ("a step", "a dot"), so *pas de temps* was built like *peu de temps*, a quantity word with *de* and a bare noun. The article stays:

- after *être* and the other copulas: *Ce n'est pas un livre*;
- in a contrast: *Je n'ai pas des amis, j'ai des alliés*; *Il ne boit pas du vin, mais du jus*;
- with the definite article: *Je n'aime pas le vin*;
- with *ne … que*, which is not a negation of quantity: *Il ne boit que du vin*.

Which negations trigger it is in [27. Negation](https://md.mova.elamur.ai/free/en/docs/languages/fr/27-negation/).

### 3.2 After a quantity word

Quantity adverbs and nouns take *de* and a bare noun: *beaucoup de gens*, *peu de temps*, *assez de*, *trop de*, *plus de*, *moins de*, *tant de*, *combien de*, *une bouteille de vin*, *un kilo de pommes*, *une dizaine de thoniers* (fr-ud-train_03512). After *bien*, *la plupart*, *la majorité*, the noun is definite and the form is *des* (= *de* + *les*): *la plupart des gens*, *bien des fois*; the verb then agrees with the plural noun (*la plupart des gens sont partis*).

### 3.3 Before an adjective that precedes a plural noun

In the written standard, *des* becomes *de* before an adjective + noun: *de grands arbres*, *de bons amis*, *de bons petits plats* (fr-ud-train_04010). In speech *des grands arbres* is usual. When the adjective follows the noun, or adjective and noun form a lexical unit, *des* stays: *des arbres magnifiques*, *des petits pois*, *des jeunes gens*, *des grands-parents*.

The rule is recent. Sixteenth- and seventeenth-century writers have *des* before an adjective (Sévigné: *des petits ponts*); *de* appears in the later seventeenth century, and grammarians then extended it to the singular partitive (*de bon vin*), against popular usage, which keeps *du vieux vin* (Brunot 1899, same section).

**Counts.** In GSD the sequence DET (`Definite=Ind`) + ADJ + NOUN has the determiner *de/d'* 398 times and *des* 18 times; in ParisStories 8 against 20, in Rhapsodie 9 against 11. (The *de* count includes cases after a negation or a quantity word, and the *des* count lexical units such as *petits pois*; the contrast between the written and the spoken corpora is still clear.)

### 3.4 Preposition *de* after a quantity noun

In *une série de grands concerts* (fr-ud-train_06911), *de* is a preposition introducing the complement of *série*: the same graphic *de* is an article in §§3.1–3.3 and a preposition here. UD gives it DET with `det` in one case and ADP with `case` in the other (§16).

## 4. Demonstratives

|  | masculine | feminine |
| --- | --- | --- |
| singular | *ce* (*cet* before a vowel or mute *h*) | *cette* |
| plural | *ces* | *ces* |

*Ce système* (fr-ud-dev_00053), *cet arbre*, *cette audace* (fr-ud-dev_00070), *ces enfants*. *Cet* and *cette* sound alike (/sɛt/), so the spoken demonstrative does not show the gender of a vowel-initial noun. The demonstrative points into the situation or back into the text ("the one just mentioned"), and has temporal uses (*ce matin*, *cette année*). Near and far are marked, when needed, by *-ci* and *-là* on the noun: *ce livre-ci*, *ce livre-là*. In speech *-là* is far more frequent and serves for both, often with a distancing tone (*ce type-là*). GSD has 2 214 demonstrative DET tokens.

## 5. Possessives

The possessive determiner agrees in gender and number with the **possessed** noun and in person and number with the **possessor**. The third person does not mark the possessor's sex: *son livre* is "his book" or "her book".

| possessor | masc. sg. | fem. sg. | plural |
| --- | --- | --- | --- |
| 1 sg. | *mon* | *ma* | *mes* |
| 2 sg. | *ton* | *ta* | *tes* |
| 3 sg. | *son* | *sa* | *ses* |
| 1 pl. | *notre* | *notre* | *nos* |
| 2 pl. | *votre* | *votre* | *vos* |
| 3 pl. | *leur* | *leur* | *leurs* |

Before a feminine noun beginning with a vowel the masculine form is used: *mon amie*, *son énergie*; but *ma belle amie*. With a plural possessor, *leur* + singular is used when each has one (*Les enfants aiment leur mère*), *leurs* when there are several (*leurs parents*); usage varies where each has one of a kind (*Ils ont levé leur(s) verre(s)*).

GSD has 4 348 possessive DET; the third person is 90% of them (*son* 1 543, *sa* 925, *ses* 793, *leur* 455, *leurs* 211; *mon* only 73).

**Short and long forms.** French also had long possessives (*mien, tien, sien*), used before the noun in Old French, often with an article (*le tuen cors* "your body"). Simonenko and Carlier (`simonenko-2022`) counted prenominal possessives in historical treebanks of French, Spanish and Portuguese: with the long forms the share of an accompanying determiner rises steadily, with the short forms it rises and then falls, in all three languages; the languages differ only in which paradigm each kept before the noun. Modern French keeps the long forms as pronouns (*le mien*, §14) and, before a noun, only archaically or ironically (*un mien cousin*).

**Other uses.** With a body part whose owner is clear the definite article replaces the possessive (*Elle a levé la main*, *Il s'est cassé la jambe*; *Je lui ai pris la main*), unless the part is qualified (*sa main gauche*). For an inanimate possessor the written language often uses *en* + article when the possessed noun is subject or predicate: *La réalité en est malheureusement toute différente* (fr-ud-train_02702). After *on, chacun, tout le monde* the possessive is *son* (*On aime sa famille*); with *on* = "we", *notre*. A possessive excludes any other determiner (**le mon livre*); French says *un de mes amis*, *un ami à moi*.

## 6. Quantifiers and indefinite determiners

| form | meaning | notes |
| --- | --- | --- |
| *quelque(s)* | some, a few | invariable adverb before a number: *quelque 200 personnes* "about 200" |
| *plusieurs* | several | plural only, one form for both genders |
| *certain(e)s* | some, certain | *certains centres* (fr-ud-dev_00221); after the noun an adjective "sure" |
| *chaque* | each | singular only: *chaque soir* (fr-ud-test_00158) |
| *tout, toute, tous, toutes* | all, every, whole | *tout le monde*, *toute la journée*, *tous les jours* |
| *aucun(e)*, *nul(le)* | no | with *ne*: *aucun impact* (fr-ud-dev_00110) |
| *différents, divers* | various | before the noun; after it adjectives "different", "diverse" |
| *n'importe quel*, *quelconque*, *tel* | any, such | *un tel projet*, *une raison quelconque* |
| *beaucoup de, peu de, la plupart des…* | quantity | §3.2 |

*Tout* + definite article + singular is "the whole" (*dans tout le camping*, fr-ud-dev_00311); *tous les* + plural is "all the, every" (*dans tous les sens*, fr-ud-dev_00883). *Tous* is /tu/ as a determiner and /tus/ as a pronoun (*Ils sont tous là*). As an adverb ("quite, entirely") *tout* is invariable, except that it agrees before a feminine adjective beginning with a consonant or an aspirated *h*: *tout petit, tout petits, tout heureuse*, but *toute petite, toutes petites, toute honteuse*.

GSD indefinite and negative DET by form: *plusieurs* 252, *quelques* 150, *chaque* 133, *certains* 80, *aucune* 54, *aucun* 53, *certaines* 52, *toute* 50, *tout* 48 (`PronType=Ind` 925, `Neg` 111 in all). *Tous/toutes* before an article are mostly ADJ in GSD (*tous* ADJ 192, DET 14), an annotation choice to know when counting. For *certain, différent, divers* before and after the noun, see [15. Adjectives and adverbs](https://md.mova.elamur.ai/free/en/docs/languages/fr/15-adjectives-adverbs/).

## 7. Numerals

### 7.1 Cardinals

**0–16** are simple: *zéro, un, deux, trois, quatre, cinq, six, sept, huit, neuf, dix, onze, douze, treize, quatorze, quinze, seize*. **17–19** are *dix-sept, dix-huit, dix-neuf*. The tens to 60 are *vingt, trente, quarante, cinquante, soixante*, with *et* before *un* and *onze* only: *vingt et un*, *soixante et onze*, but *vingt-deux*, *quatre-vingt-un*, *quatre-vingt-onze*, *cent un*.

**70–99** in France count partly by twenties: *soixante-dix* (60 + 10), *soixante et onze*, *quatre-vingts* (4 × 20), *quatre-vingt-dix* (4 × 20 + 10). Belgium and Switzerland say *septante* and *nonante*, and the Swiss cantons of Vaud, Valais and Fribourg *huitante* for 80 ([08. National and regional varieties of French](https://md.mova.elamur.ai/free/en/docs/languages/fr/08-national-varieties/)). The decimal forms are the inherited ones (*septante* < Latin *septuaginta*); *soixante-dix* is already in *Ogier le Danois*, and after the fifteenth century *septante* is rare in literary French (Nyrop II, §483, `nyrop-1899`). Old French counted further by twenties (*six vingts* = 120; the Paris *Quinze-Vingts*, 300), and Brunot (`brunot-1899`) records *septante, nonante* as provincial and popular in his day.

**100 and up**: *cent, deux cents, mille, deux mille, un million, un milliard* (10⁹), *un billion* (10¹²). *Un* is not said before *cent* and *mille* but is before *million* and *milliard*. In dates *mil* is an optional old spelling (*l'an mil neuf cent*); years are read *mille neuf cent quatre-vingt-dix* or *dix-neuf cent quatre-vingt-dix*.

**Spelling.** Traditionally the parts of a number below 100 are hyphenated except around *et* (*vingt-deux*, *vingt et un*), and the parts above 100 are separated by spaces (*deux cent trois*). The 1990 rectifications hyphenate every complex numeral, below and above a hundred: *vingt-et-un*, *cent-deux*, *deux-cent-soixante-et-onze* (`cslf-1990`, rule 1; [06. Spelling rules: inflection, compounds, hyphens and the 1990 rectifications](https://md.mova.elamur.ai/free/en/docs/languages/fr/06-spelling-rules/)). *Million* and *milliard* are nouns and stay separate words. In figures, groups of three digits are separated by a (non-breaking) space and the decimal sign is a comma: *25 785*, *3,14* ([07. Punctuation, typography, capitals, abbreviations and numbers](https://md.mova.elamur.ai/free/en/docs/languages/fr/07-punctuation-typography/)).

### 7.2 Agreement and position

*Un* has a feminine *une*; the other cardinals are invariable, except that *vingt* and *cent* take *-s* when multiplied and not followed by another numeral: *quatre-vingts ans* but *quatre-vingt-trois ans*, *deux cents euros* but *deux cent trois euros*. Brunot notes that the *-s* of these multiples was the general rule in the old language. Used as ordinals they stay invariable: *page quatre-vingt*, *l'an deux cent*. *Mille* is invariable. *Million* and *milliard* are nouns: they take *-s* and *de* (*deux millions d'habitants*), but no *de* when another number follows (*deux millions cent mille habitants*).

The cardinal stands after the determiner and before *premier*, *dernier*, *autres*: *les deux premiers jours* "the first two days", *les trois derniers* "the last three", *les deux autres* "the other two" — the reverse of the English order. A numeral can head a phrase without a noun: *Ils sont trois*, *J'en ai vu deux* (with *en*, §11).

In a superlative partitive the article and adjective may be feminine over a default masculine plural noun when the referent is a woman (*la plus intelligente de mes gentils professeurs*); Sleeman and Ihsane (`sleeman-2016`) show that quantified partitives (*trois des livres*) do not behave the same way.

### 7.3 Ordinals, collectives, fractions

The ordinal adds *-ième* to the cardinal: *deuxième, troisième, vingt et unième, centième*, dropping a final *-e* (*quatrième, onzième*) and with *cinq → cinquième*, *neuf → neuvième*. *Premier/première* replaces *unième* except in compounds; *second(e)* competes with *deuxième*. Dates use the cardinal except for the first day (*le premier mai*, *le deux mai*). Abbreviations: *1er, 1re, 2e* (informally *2ème*): *Le 24e RIC* (fr-ud-dev_00054). Roman numerals name centuries and monarchs: *le XIXe siècle*, *Louis XIV*.

*Une dizaine, une douzaine, une vingtaine, une centaine* are feminine nouns ("about ten", …), *un millier* masculine; they take *de* and a bare noun (*une dizaine de thoniers*). Fractions: *un demi, un tiers, un quart, trois quarts, la moitié, un cinquième*. *Demi* is invariable before the noun with a hyphen (*une demi-heure*) and agrees in gender after it (*une heure et demie*). Multiplicatives: *double, triple*; *deux fois*.

The final consonants of *six, dix, huit* drop before a consonant and become /z/ or stay before a vowel (*six livres* /si/, *six ans* /siz/); *neuf* is /nœv/ in *neuf ans, neuf heures* ([04. Connected speech: enchaînement, liaison, elision, the mute e and assimilation](https://md.mova.elamur.ai/free/en/docs/languages/fr/04-connected-speech/)).

## 8. Personal pronouns

The **clitic** (weak, conjoined) pronouns are bound to the verb: they cannot be stressed, coordinated or separated from the verb except by other clitics and *ne*. The **strong** (disjoint, tonic) pronouns are free.

| person | subject | direct object | indirect object | reflexive | strong |
| --- | --- | --- | --- | --- | --- |
| 1 sg. | *je* (*j'*) | *me* (*m'*) | *me* | *me* | *moi* |
| 2 sg. | *tu* | *te* (*t'*) | *te* | *te* | *toi* |
| 3 sg. m. | *il* | *le* (*l'*) | *lui* | *se* (*s'*) | *lui* |
| 3 sg. f. | *elle* | *la* (*l'*) | *lui* | *se* | *elle* |
| 1 pl. | *nous* | *nous* | *nous* | *nous* | *nous* |
| 2 pl. | *vous* | *vous* | *vous* | *vous* | *vous* |
| 3 pl. m. | *ils* | *les* | *leur* | *se* | *eux* |
| 3 pl. f. | *elles* | *les* | *leur* | *se* | *elles* |

Add *on* (§13), *y* and *en* (§11), the subject *ce* (§8.3) and *soi* (§12). *Vous* is also the polite singular; the choice of *tu* or *vous* holds for a whole relationship ([32. Politeness and hedging](https://md.mova.elamur.ai/free/en/docs/languages/fr/32-politeness-and-hedging/)). Gender is marked only in the third person, and not in *les, leur, se*; *lui* is both the dative clitic of either gender and the masculine strong form, and *elle(s)* is both clitic and strong.

### 8.1 Subject clitics

A finite verb **must have an expressed subject**: French is not a null-subject language like Spanish or Italian. The exceptions are the imperative and a few fixed phrases (*Peu importe*, *N'importe*, *Soit*). The subject clitic stands before the verb and its clitic cluster (*Je ne le lui ai pas dit*); in inversion it follows the verb with a hyphen (*Est-il parti ?*, *Va-t-il se circonscrire à cette seule matière ?*, fr-ud-dev_00602, with the *-t-* inserted after a vowel; [26. Questions](https://md.mova.elamur.ai/free/en/docs/languages/fr/26-questions/)). A dislocated noun can be doubled by the clitic (*Marie, elle est partie*; [29. Ellipsis, word order and information structure](https://md.mova.elamur.ai/free/en/docs/languages/fr/29-ellipsis-word-order-information-structure/)).

Speech reduces the subject clitics: *j'sais pas* /ʃepa/, *t'as vu*, *i(l) dit*, *y a pas*; some analyses treat them as agreement prefixes of the verb. Subject pronouns (`nsubj`, PRON) by form:

|  | GSD | ParisStories | Rhapsodie |
| --- | --- | --- | --- |
| *je* + *j'* | 526 | 1 084 | 729 |
| *tu* | 15 | 220 | 100 |
| *il* | 2 514 | 367 | 184 |
| *on* | 526 | 741 | 281 |
| *nous* | 241 | 4 | 74 |
| *vous* | 143 | 37 | 313 |
| *c'* | 350 | 607 | 490 |
| *ça* | 19 | 158 | 143 |

In the Paris narratives the subject "we" is *on*: *nous* occurs 4 times against 741 *on*. Rhapsodie, which mixes genres including more formal speech, keeps *nous* and *vous*.

### 8.2 Impersonal *il*

A non-referring *il* is the subject of weather verbs (*Il pleut*), of impersonal constructions (*Il faut partir*, *Il y a un problème*, *Il est important que …*, *Il s'agit de …*) and of impersonal passives (*Il a été décidé que …*). The logical subject can follow the verb, which agrees with *il*: *Il arrive des accidents*. In *Il y a douze espèces reconnues* (fr-ud-train_00231) both *il* and *y* are expletive. GSD marks the expletive subject `expl:subj`: *il* 611, *c'* 187, *-il* 34, *ce* 33, *-t-il* 16, *-ce* 13 (931 tokens in all).

### 8.3 *Ce*

*Ce* (*c'*) is the subject of *être* in *C'est un livre*, *Ce sont des amis*. *C'est* introduces a determined noun phrase, a strong pronoun, a clause, or a situation (*C'est grand*, of a place in view); *il/elle est* takes a bare adjective or a bare noun of profession about a definite referent (*Elle est avocate*, *Il est grand*). Clefts are built on *c'est … qui/que* ([29. Ellipsis, word order and information structure](https://md.mova.elamur.ai/free/en/docs/languages/fr/29-ellipsis-word-order-information-structure/)). With a plural predicate the written verb is plural (*Ce sont eux*), the spoken one singular (*C'est eux*).

## 9. Object clitics and their order

Direct-object clitics: *me, te, le, la, nous, vous, les*; indirect-object clitics: *me, te, lui, nous, vous, leur*. *Lui* and *leur* stand for a person (*Je lui parle*); for a thing, *y* (§11). *Je leur rétorque qu'ils l'avaient bien vu quand je le leur ai laissé* (fr-ud-train_12887) has both kinds.

*Le* is also invariable ("neuter") for a predicate (*Heureuse, elle ne l'a pas toujours été*) and for a clause (*Je le sais*; *Il me l'a dit à la Haye*, fr-ud-test_00087).

With some verbs whose complement is *à* + person, the dative clitic is not used and a strong pronoun follows *à*: *Je pense à elle*, *Il tient à moi*, *Fais attention à lui*, *Je m'adresse à lui*; for a thing, *y* (*J'y pense*). The class includes *penser à, songer à, rêver à, tenir à, renoncer à, faire attention à, être à*, and the pronominal verbs (*s'adresser à, s'habituer à, se fier à*).

### 9.1 The order

| 1 | 2 | 3 | 4 | 5 |
| --- | --- | --- | --- | --- |
| *me, te, se, nous, vous* | *le, la, les* | *lui, leur* | *y* | *en* |

*Il me le donne*, *Il le lui donne*, *Elle nous en parle*, *Je m'y suis habitué*, *Il y en a*, *Il n'y en a point* (fr-ud-test_00097). The order is not by case alone: a first- or second-person pronoun precedes the third-person direct object (*me le*), while the third-person direct object precedes the third-person dative (*le lui*).

Not every combination is possible:

- a first- or second-person direct object cannot combine with a dative clitic (the **person–case constraint**): **Il me lui présente*; the dative becomes a strong pronoun, *Il me présente à lui*;
- two first-or-second-person clitics do not combine: **Il me te donne*;
- *lui y* is avoided, and *y*, *en* close the cluster (*lui en*, *m'en*, *y en*).

With at most one clitic per slot, a fixed order and gaps, the cluster behaves more like verb morphology than free syntax, and many analyses treat it so.

**Counts.** Adjacent pairs of object-type clitics (PRON in `obj`, `iobj`, `expl:*`, `obl*`, `nmod`) are rare: 98 in GSD, 62 of them *s'en* and *s'y* of pronominal verbs; 28 in ParisStories (*y en* 9, *m'en* 6), 19 in Rhapsodie. Over seven treebanks (GSD, ParisStories, Rhapsodie, Sequoia, PUD, FQB, ALTS) there are 289: *s'en* 141, *s'y* 42, *y en* 30, *l'en* 13, and *le lui*, *le leur* once each. The excluded combinations do not occur, but the corpora are too small for their absence to mean much.

**With negation and compound tenses.** *Ne* precedes the whole cluster and *pas* follows the finite verb (*Je ne le lui ai pas dit*). The cluster precedes the auxiliary (*Je le leur ai laissé*), and a preceding direct-object clitic can make the participle agree ([13. Verbs](https://md.mova.elamur.ai/free/en/docs/languages/fr/13-verbs/)).

## 10. Clitics with infinitives, causatives and the imperative

**Infinitives.** A clitic stands before the verb of which it is an argument: *Je veux le voir*, *Il ne peut pas lui parler*, *Pouvez-vous nous en dire plus ?* (fr-ud-train_02355, where *nous* and *en* belong to *dire*). The older order with the clitic before the modal (*Je le veux voir*) is "clitic climbing". Olivier, Sevdali and Folli (`olivier-2023`), on a corpus of legal texts, find it obligatory until Middle French and lost before the start of the nineteenth century; they link its loss to the loss of interpolation (material between clitic and verb, *clitic–XP–V*), and argue that French kept restructuring but lost most of its visible effects, retaining the long passive and quantifier climbing.

**Causatives and perception verbs.** With *faire*, *laisser* and verbs of perception the clitics go before the finite verb: *Je le lui fais lire*, *Je l'ai vu partir*, *Son entourage, ses enfants le lui font remarquer* (fr-ud-train_04603). The causee of a transitive infinitive is dative (*lui*). See [23. Complementation](https://md.mova.elamur.ai/free/en/docs/languages/fr/23-complementation/).

**Affirmative imperative.** The clitics follow the verb, joined by hyphens, in the order direct object – indirect object – *y* – *en*, with *moi, toi* for *me, te*: *Donne-le-moi*, *Dis-le-leur*, *Donnez-m'en* (*m'*, *t'* before *en*), *Suis-moi* (fr-ud-dev_00823), *Va-t'en*, *Assieds-toi*. The *tu*-form of *-er* verbs and *va* add *-s* before *y* and *en*: *Vas-y*, *Manges-en*. In the **negative imperative** the ordinary preverbal order returns: *Ne me le donne pas*, *N'en parlez pas*, *Ne t'en va pas*. Colloquial *Donne-moi-le* is non-standard.

## 11. *Y* and *en*

*Y* and *en* replace prepositional complements; they have no person or gender.

- ***Y*** replaces *à* + a thing (*Je pense à ce problème → J'y pense*) and a place (*Il va à Paris → Il y va*; *On y mange correctement*, fr-ud-dev_00087). It is avoided for persons in the standard (§9).
- ***En*** replaces *de* + a thing (*Il en parle*), an origin (*J'en viens*), the inanimate possessor (§5), the partitive (*Tu veux du pain ? — J'en veux*) and the noun of a quantity that is left unsaid: *J'ai trois frères → J'en ai trois*; *Tsahal a commandé 130 Namers et en a reçu 10* (fr-ud-train_06235).
- **Fixed** *y* and *en*: *il y a*, *s'en aller*, *en vouloir à*, *s'y prendre*, *y compris*, *Je n'en peux plus*, *Voilà où nous en sommes* (fr-ud-train_13769).

Both continue Latin adverbs (*ibi*, *inde*; Brunot 1899, §416 «Origine des adverbes»), hence the older name "pronominal adverbs". PRON per 1 000 words: *y* 1.3 in GSD (530), 6.1 in ParisStories (262), 5.8 in Rhapsodie (258); *en* 0.7 (282), 1.1 (45), 1.1 (50). Speech has four to five times more *y*, mostly in *il y a*, *y a*.

## 12. Strong pronouns

The strong pronouns *moi, toi, lui, elle, nous, vous, eux, elles* and *soi* are used:

- alone or after a conjunction: *Qui est là ? — Moi.*; *plus grand que toi*; *ni lui ni elle*;
- in coordination: *Marie et moi, nous sommes parties*;
- after a preposition: *chez elle* (fr-ud-train_02204), *avec eux*;
- after *c'est*: *C'est lui qui recruta le successeur de Graham* (fr-ud-dev_00665);
- as a dislocated double of a clitic: *Moi, je pense que …*; spoken *nous on s'est dit* (ParisStories_2021_adoptionMouts_14);
- with *-même*: *l'autorité judiciaire elle-même* (fr-ud-dev_00594), *lui-même*;
- as the subject of a verbless or participial phrase: *Lui parti, la maison était vide*.

*Soi* goes with an indefinite subject (*Chacun pour soi*), *lui, elle, eux* with a definite one.

**Strong pronouns for things.** After a preposition the strong pronouns usually refer to persons; for things French uses *y*, *en*, or a preposition without complement (*Il a voté pour*, *Il lui a sauté dessus*). Da Cunha and Heidinger (`cunha-2026`) tested this in acceptability experiments on French and Spanish. When a null complement is possible after the preposition, the strong pronoun is preferred for humans; after prepositions that allow no null complement, such as *par* and *vers*, it can refer to things as well (*Ce pays, j'ai toujours été attiré par lui*). Their analysis explains the restriction by competition with the null form and by the match between grammatical and natural gender, without an animacy feature on the pronoun. The rule "strong pronoun for persons" is a tendency.

## 13. Reflexive, reciprocal and *on*

### 13.1 Reflexive and reciprocal

The third-person reflexive is *se* (*s'*), for singular, plural and *on*; the first and second persons use *me, te, nous, vous*. A pronominal verb has it as a fixed part (*se souvenir*, *s'évanouir*), as a true reflexive (*Elle se lave*), as a reciprocal (*Ils se parlent*, made explicit by *l'un l'autre*, *entre eux*), or as a middle or passive (*Ces livres se vendent bien*; [28. Voice and the passive](https://md.mova.elamur.ai/free/en/docs/languages/fr/28-voice-and-passive/)). All take *être* in compound tenses ([13. Verbs](https://md.mova.elamur.ai/free/en/docs/languages/fr/13-verbs/)). GSD distinguishes the uses by relation, not by form: `obj` for the true reflexive (*qu'ils se sont consultés*, fr-ud-train_06603), `expl:pv` for the inherent reflexive (1 017 tokens), `expl:pass` for the reflexive passive and middle (686).

### 13.2 *On*

*On* is a subject clitic with a third-person singular verb. It means "people in general" (*On dit que …*), "someone" (*On a frappé*), and in speech **"we"**, where it has largely replaced *nous* as subject (§8.1; [37. Register and variation](https://md.mova.elamur.ai/free/en/docs/languages/fr/37-register-and-variation/)). With *on* = "we" a predicate adjective or participle may agree with the real referents: *on est arrivé de nuit* (ParisStories_2020_campBedouin_6) has the singular, *on est restées bloquées en bas* (ParisStories_2022_13_CocktailsEtCrêpes__19) the feminine plural. In the formal register *l'on* appears after *que, si, où, et*. Its reflexive is *se*, its possessive *son* or, for "we", *notre*.

**Counts.** *On* as `nsubj`: GSD 526, ParisStories 741, Rhapsodie 281. The head of that *on* is plural in 2 of 526 cases in GSD, 20 of 741 in ParisStories, 8 of 281 in Rhapsodie. With a copular predicate the share is higher: plural in 14 of 74 ParisStories cases and 5 of 15 in Rhapsodie, none of 10 in GSD. Plural agreement with *on* exists in speech, as a minority.

## 14. Demonstrative, indefinite and possessive pronouns

**Demonstrative pronouns.**

|  | masculine | feminine | neuter |
| --- | --- | --- | --- |
| singular | *celui(-ci/-là)* | *celle(-ci/-là)* | *ceci, cela, ça*; *ce* |
| plural | *ceux(-ci/-là)* | *celles(-ci/-là)* |  |

*Celui, celle, ceux, celles* need a complement (*celui de Pierre*, *celle qui est partie*) or *-ci/-là*: *Celles-ci peuvent à tout moment jeter l'éponge* (fr-ud-train_01390). In two-term contrasts *-ci* is "the latter" and *-là* "the former". *Cela* (spoken *ça*) refers to a situation or a clause (*Ça va ?*); *ceci* announces what follows, or refers back (*Ceci permet le recrutement du ribosome*, fr-ud-train_02258). *Ce* stands before a relative and before *être*: *ce qui ne gâte rien* (fr-ud-dev_00270), *C'est ce que nous nommons aujourd'hui le Vieux-Flassans* (fr-ud-dev_00188). In speech *ça* replaces *cela*.

**Indefinite pronouns**: *quelqu'un, personne, rien, quelque chose, chacun(e), tout, tous, plusieurs, certains, aucun(e), l'un … l'autre, autre chose, n'importe qui/quoi, quiconque, autrui*. *Personne* and *rien* are negative with *ne* (*personne ne sait*, fr-ud-train_03972; *je ne vais rien oublier*, fr-ud-test_00293) and alone in answers (*Personne.*), and positive in some contexts without *ne* (*sans rien dire*; [27. Negation](https://md.mova.elamur.ai/free/en/docs/languages/fr/27-negation/)). An adjective after an indefinite pronoun takes *de* and the masculine: *quelqu'un de gentil*, *rien de plus* (fr-ud-dev_00961), *rien d'exceptionnel* (fr-ud-train_00599). *Tout* is the neuter "everything" (*Tout va bien*).

**Possessive pronouns** are the article + the long possessive: *le mien, la mienne, les miens, les miennes*; *le tien*; *le sien*; *le nôtre, la nôtre, les nôtres*; *le vôtre*; *le leur, la leur, les leurs*. The circumflex and the open vowel of *nôtre, vôtre* distinguish them from *notre, votre*. The article fuses with *à, de* (*au mien*, *des leurs*). Idioms: *les nôtres* "our people", *y mettre du sien*. The predicative possessive is *à moi* (*Ce livre est à moi*).

## 15. Relative and interrogative pronouns: the forms

The relative and interrogative pronouns largely share their forms. The clauses are in [25. Relative clauses](https://md.mova.elamur.ai/free/en/docs/languages/fr/25-relative-clauses/) and [26. Questions](https://md.mova.elamur.ai/free/en/docs/languages/fr/26-questions/).

**Relative.**

| function | persons | things |
| --- | --- | --- |
| subject | *qui* | *qui* |
| direct object | *que* (*qu'*) | *que* |
| after a preposition | *qui* (*à qui*, *avec qui*), *lequel* | *lequel* and its forms; *quoi* after an indefinite antecedent |
| complement with *de* | *dont* | *dont* |
| place, time |  | *où* |
| without antecedent | *qui*, *quiconque* | *ce qui, ce que, ce dont, ce à quoi* |

*Dont* replaces a *de*-phrase: *la tribu Al Bu Said, dont est issu l'actuel sultan* (fr-ud-dev_00024). *Lequel, laquelle, lesquels, lesquelles* fuse with *à* and *de* (§1): *date à laquelle Mesa quitte les Postobón* (fr-ud-train_01832), *match au cours duquel il inscrit un but* (fr-ud-train_02217), *l'environnement social dans lequel elle évolue* (fr-ud-train_02591). *Quoi* follows a preposition after *ce*, *rien* or a clause (*ce à quoi je pense*, *après quoi*). *Que* elides, *qui* does not.

**Interrogative.**

| function | persons | things |
| --- | --- | --- |
| subject | *qui*, *qui est-ce qui* | *qu'est-ce qui* |
| object | *qui*, *qui est-ce que* | *que*, *qu'est-ce que*, *quoi* (in situ) |
| after a preposition | *à qui*, *avec qui* | *à quoi*, *de quoi* |
| choice | *lequel, laquelle, lesquels, lesquelles* | same |
| determiner | *quel, quelle, quels, quelles* | same |

*Quel* is the interrogative and exclamative determiner (*Quel livre ?*, *Quelle belle journée !*) and the predicate of *être* (*Quel est le coût d'un tel projet ?*, fr-ud-test_00178). *Où, quand, comment, pourquoi, combien* are adverbs. Colloquial French prefers *quoi* in situ (*Tu fais quoi ?*).

## 16. The annotation in UD French

### 16.1 Conventions (GSD)

- **Articles** are DET with `PronType=Art`, `Definite=Def|Ind`, relation `det`. The partitive *du* (*Il faudra encore du temps*, fr-ud-train_00391) is one DET word, lemma *du* (101 tokens); only the contraction *du* = *de* + *le* is a multiword token. The *de* of §§3.1–3.3 (*pas de temps*, fr-ud-dev_00353) is DET with lemma *un*.
- ***De* after a quantity noun** is ADP `case`, the quantity noun is the head and the second noun its `nmod` (*une série de grands concerts*). A quantity adverb heads its phrase: in *peu de temps après* (fr-ud-dev_00069) *peu* is ADV with `ExtPos=PRON` and *temps* its `obl:arg`.
- **Possessives** are DET with `Poss=Yes`, `PronType=Prs`, the noun's gender and number, and `Person[psor]`, `Number[psor]` for the possessor.
- **Numerals** are NUM, `nummod`; *25 785* is one token (179 such). Of 10 479 NUM, 8 626 contain a digit; the commonest words are *deux* 636, *trois* 243, *quatre* 126. `NumType=Ord` (216) marks only abbreviated ordinals (*xixe* 37, *xxe* 25, *137e*); *premier, première, deuxième, troisième* (ADJ, 518) have no `NumType`.
- **Pronouns** are PRON with `PronType` `Prs` 11 354, `Rel` 3 103, `Dem` 1 633, `Ind` 1 416, `Neg` 114, `Int` 67. Clitics have `Emph=No` and `Case` (*il* `Nom`, *le* `Acc`, *lui* `Dat`); strong forms `Emph=Yes` (*lui* 139, *eux* 71, *lui-même* 56, *elle* 47, *moi* 42).
- **Clitics** depend on the verb whose argument they are: in fr-ud-train_02355 *nous* (`iobj`) and *en* (`obj`) depend on *dire*, the `xcomp` of *Pouvez*. In the causative fr-ud-train_04603 *font* is AUX `aux:caus`, *remarquer* is the root, *le* its `obj`, *lui* its `iobj:agent`, and the noun subject `nsubj:caus`.
- ***Y*** (530) is `obl:mod` 222, `expl:comp` 154 (*il y a*), `iobj` 114, `fixed` 19 (*y compris*); ***en*** (282) is `iobj` 112, `nmod` 86 (possessor, or the set a number is taken from), `expl:comp` 53, `obj` 15. Expletives in all: `expl:pv` 1 017, `expl:subj` 931, `expl:pass` 686, `expl:comp` 211.
- **Relatives** have `PronType=Rel`; the clause is `acl:relcl` (*dont* in fr-ud-dev_00024 is `iobj` of the adjective *issu*).

### 16.2 Size of the classes

| treebank (UD 2.18) | words | PRON | DET | NUM |
| --- | --- | --- | --- | --- |
| GSD | 400 385 | 17 687 | 61 098 | 10 479 |
| ParisStories (spoken) | 42 789 | 6 375 | 3 467 | 243 |
| Rhapsodie (spoken) | 44 234 | 5 280 | 4 456 | 347 |
| Sequoia | 70 545 | 2 741 | 10 400 | 1 826 |
| ALTS (16th c.) | 68 088 | 6 269 | 7 632 | 731 |
| PUD | 24 726 | 1 089 | 3 887 | 451 |
| FQB (questions) | 23 896 | 1 654 | 3 832 | 117 |

Speech has three times the pronoun share of GSD (14.9% in ParisStories against 4.4%) and half its determiner share (8.1% against 15.3%). Of GSD's DET, 53 469 are articles (*le* 15 425, *la* 10 956, *les* 9 738, *l'* 7 169, *un* 4 153, *une* 3 543, *des* 1 782, *de* 427, *d'* 159, *du* 101), then possessives 4 348, demonstratives 2 214, indefinites 925, negatives 111, interrogatives 24, exclamatives 7.

### 16.3 The pronoun lemmas

GSD lemmatises the personal pronouns to the strong form: lemma *lui* covers *il* (3 612 with capitals), *elle, ils, elles, le, la, l', les, lui, leur, eux* (6 631 PRON tokens); *moi* covers *je, j', me, m', moi*; *soi* covers *se, s'* (2 325). A count by lemma therefore merges subject, object and strong pronouns and hides third-person gender and number; this chapter counts by form and relation. The article *le* (DET) keeps the lemma *le*.

## 17. How Mova handles determiners, pronouns and numerals

**What is done.** Mova has no French lexicon of determiners, pronouns or numerals and no French morphology. The English parser `en` is measured on French after `en xpos-fill` (`xpos_fill`, `xpos_of` in `en/src/main.rs`) fills the empty XPOS column with Penn Treebank tags computed from UPOS and FEATS alone:

| UD | PTB tag |
| --- | --- |
| DET with `Poss=Yes` | `PRP$` |
| DET with `PronType=Int` or `Rel` | `WDT` |
| other DET (articles, demonstratives, quantifiers, exclamative *quel*) | `DT` |
| PRON with `Poss=Yes` | `PRP$` |
| PRON with `PronType=Rel` or `Int` | `WP` |
| other PRON (personal, reflexive, demonstrative, indefinite, negative) | `PRP` |
| NUM | `CD` |
| ADJ, ordinals included | `JJ` |

`en parse-eval` then trains and tests on the treebank: LAS 87.00 with its own tags and 89.78 with gold tags on GSD (UD 2.18, train → dev, 2026-10-22; [PACKAGE.md](https://md.mova.elamur.ai/free/en/docs/languages/PACKAGE/)). The CoNLL-U reader (`en/src/conllu.rs`) skips multiword-token lines and works on syntactic words, so the parser sees *de* + *le*, never *du*; the split comes from the treebank. The parser learns French clitic attachment only from the training trees; it is told no rule.

**What the tags lose.** Person, gender, case, the clitic–strong contrast (*le*, *lui* and *eux* are all `PRP`), the possessor features, the article–partitive contrast, and the negative–indefinite contrast (*personne* and *quelqu'un* are both `PRP`).

**What is not done.**

- The article rules of §§1–3: nothing flags **Il n'a pas des amis* or **J'ai vu chiens*.
- The clitic order and the person–case constraint: **Il me lui présente* is not rejected, and nothing generates a cluster.
- *Tu/vous*, *on* as "we", the choice of *lui* or *à elle*.
- Numeral values: *quatre-vingt-dix-sept* is a `CD` token with no value, and the *-s* of *vingt, cent* is not checked.
- French tokenization: the tokenizer (`en/src/tok.rs`) has English clitic rules (*'s, n't, cannot*) and none for French elision or for *du, au, aux, des*; on raw French text the parser would not get the treebank's words. Its French accuracy has not been measured.
- The merged pronoun lemmas of §16.3 are not corrected.
- The relations of this chapter (`det`, `nummod`, `expl:*`, `iobj`, clitic `obj`) have not been scored separately; the LAS is an average, and this chapter's counts describe the treebanks, not the parser.

## Sources for this chapter

- Brunot, F. 1899, *Précis de grammaire historique de la langue française* (`brunot-1899`, public domain): the sections on the numerals (*septante*, *quatre-vingts*, the *-s* of *vingt* and *cent*), «L'article partitif» (the partitive, *de* before an adjective, *pas de*), «Ellipse de l'article», and §416 «Origine des adverbes» (*y*, *en*).
- Nyrop, K. *Grammaire historique de la langue française*, tome II (`nyrop-1899`, public domain): §483, the tens (*septante, octante, soixante-dix*).
- Conseil supérieur de la langue française 1990, *Les rectifications de l'orthographe* (`cslf-1990`, official text): rule 1, hyphens in numerals.
- Le, K. N., Bale, A. & Barner, D. 2026, Object-mass nouns specify individuation lexically: evidence from English and French, *Glossa* 11(1) (`le-2026`, CC BY 4.0).
- Simonenko, A. & Carlier, A. 2022, The evolution of possessive phrases and the rise of DP in French, Spanish, and Portuguese, *Glossa* 7(1) (`simonenko-2022`, CC BY 4.0).
- Sleeman, P. & Ihsane, T. 2016, Gender mismatches in partitive constructions with superlatives in French, *Glossa* 1(1) (`sleeman-2016`, CC BY 4.0).
- Olivier, M., Sevdali, C. & Folli, R. 2023, Clitic climbing and restructuring in the history of French, *Glossa* 8(1) (`olivier-2023`, CC BY 4.0).
- da Cunha, Y. & Heidinger, S. 2026, Animacy restrictions without animacy features: the case of strong pronouns in French and Spanish, *Glossa* 11(1) (`cunha-2026`, CC BY 4.0).
- Universal Dependencies 2.18: French GSD, ParisStories, Rhapsodie (CC BY-SA 4.0; short quotations with `sent_id`), Sequoia, FQB (LGPL-LR), PUD, ALTS (CC BY-SA; counts only), ud-treebanks-v2.18; counts of 2026-11-06. ParTUT is not used.
- Mova: `en/src/main.rs` (`xpos_fill`, `xpos_of`), `en/src/conllu.rs`, `en/src/tok.rs`; baselines in [What every language gets](https://md.mova.elamur.ai/free/en/docs/languages/PACKAGE/).
- [← previous 11. Nouns](https://md.mova.elamur.ai/free/en/docs/languages/fr/11-nouns/)
- [next → 13. Verbs](https://md.mova.elamur.ai/free/en/docs/languages/fr/13-verbs/)
- [↑ back to Modern Standard French (fr) — a reference manual for Mova](https://md.mova.elamur.ai/free/en/docs/languages/fr/)