| { |
| "paper_id": "2020", |
| "header": { |
| "generated_with": "S2ORC 1.0.0", |
| "date_generated": "2023-01-19T03:11:02.727125Z" |
| }, |
| "title": "Subjecthood and annotation: The cases of French and Wolof 1", |
| "authors": [ |
| { |
| "first": "Olivier", |
| "middle": [], |
| "last": "Bond\u00e9elle", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "Universit\u00e9 de Picardie", |
| "location": {} |
| }, |
| "email": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "", |
| "affiliation": { |
| "laboratory": "", |
| "institution": "CNRS Abstract", |
| "location": {} |
| }, |
| "email": "" |
| }, |
| { |
| "first": "Jules", |
| "middle": [], |
| "last": "Verne", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Cercll", |
| "suffix": "", |
| "affiliation": {}, |
| "email": "" |
| } |
| ], |
| "year": "", |
| "venue": null, |
| "identifiers": {}, |
| "abstract": "This article considers the annotation of subjects in UD treebanks. The identification of the subject poses a particular problem in Wolof, due to pronominal indices whose status as a pronoun or a pronominal affix is uncertain. In the UD treebank available for Wolof (Dione, 2019), these have been annotated depending on the construction either as true subjects, or as morphosyntactic features agreeing with the verb. The study of this corpus of 40 000 words allows us to show that the problem is indeed difficult to solve, especially since Wolof has a rich system of auxiliaries and several basic constructions with different properties. Before addressing the case of Wolof, we will present the simpler, but partly comparable, case of French, where subject clitics also tend to behave like affixes, and subjecthood can move from the preverbal to the detached position. We will also make a several annotation recommendations that would avoid overwriting information regarding subjecthood.", |
| "pdf_parse": { |
| "paper_id": "2020", |
| "_pdf_hash": "", |
| "abstract": [ |
| { |
| "text": "This article considers the annotation of subjects in UD treebanks. The identification of the subject poses a particular problem in Wolof, due to pronominal indices whose status as a pronoun or a pronominal affix is uncertain. In the UD treebank available for Wolof (Dione, 2019), these have been annotated depending on the construction either as true subjects, or as morphosyntactic features agreeing with the verb. The study of this corpus of 40 000 words allows us to show that the problem is indeed difficult to solve, especially since Wolof has a rich system of auxiliaries and several basic constructions with different properties. Before addressing the case of Wolof, we will present the simpler, but partly comparable, case of French, where subject clitics also tend to behave like affixes, and subjecthood can move from the preverbal to the detached position. We will also make a several annotation recommendations that would avoid overwriting information regarding subjecthood.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Abstract", |
| "sec_num": null |
| } |
| ], |
| "body_text": [ |
| { |
| "text": "In this article, we explore the identification of the subject in two languages with a rigid SVO order, French and Wolof. While these languages share no genetic relationship, they present similarities at the typological level and the identification of the subject position can become problematic in some constructions. In some languages, especially ergative languages, subject properties can be distributed onto different arguments (Keenan, 1976; Comrie, 1978) . This is not the case for the languages we are considering, where the identification of the argument realized as a subject is very clear. What interests us here is the fact that the realisation of the same argument is distributed across several syntactic positions and subjecthood can move from one syntactic position to another. This should not be confused with the cases studied by Cole et al. (1980) for example, where subjecthood moves from one semantic argument to another.", |
| "cite_spans": [ |
| { |
| "start": 431, |
| "end": 445, |
| "text": "(Keenan, 1976;", |
| "ref_id": "BIBREF19" |
| }, |
| { |
| "start": 446, |
| "end": 459, |
| "text": "Comrie, 1978)", |
| "ref_id": "BIBREF5" |
| }, |
| { |
| "start": 845, |
| "end": 863, |
| "text": "Cole et al. (1980)", |
| "ref_id": "BIBREF4" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "To begin this discussion, we must first give a name to verb argument whose subjecthood we want to discuss. We will name it the first actant, following Tesni\u00e8re (1959 Tesni\u00e8re ( , 2015 and Mel'\u010duk (1988) . The first actant is the semantic argument of the verbal form that can be realized as its subject: sometimes it is realized as its subject, but sometimes it is only realized as a pronominal affix in the verbal inflection (especially in pro-drop languages). In the languages we are studying, the first actant is easy to define, while the subject is more difficult to identify, because there are several positions were the first actant can appear. In English for instance, the first actant is characterized by the following set of traits: it can be realized in the preverbal position, it causes the verb to bear an agreement suffix, it controls the object position (he washes himself), and it can be realized by specific pronouns such as she or we. 2 It can appear in three positions as shown in (1), where the first actant of the verb be is realized as my father, he, and the pronominal index amalgamated in the is form of be. 3 (1) \"My father, he's an anthropologist,\" she said. Among these three positions, there is one position we will call the subject: the preverbal position, because it is obligatory and can be occupied by lexical NPs. The two other positions do not have such properties: one is part of the verb inflection and is not a syntactic position; the other is an optional, prosodically detached position, but always with a pronoun in the preverbal position reflecting it. In other word, the preverbal position is more canonical than the detached position.", |
| "cite_spans": [ |
| { |
| "start": 151, |
| "end": 165, |
| "text": "Tesni\u00e8re (1959", |
| "ref_id": "BIBREF30" |
| }, |
| { |
| "start": 166, |
| "end": 183, |
| "text": "Tesni\u00e8re ( , 2015", |
| "ref_id": "BIBREF31" |
| }, |
| { |
| "start": 188, |
| "end": 202, |
| "text": "Mel'\u010duk (1988)", |
| "ref_id": null |
| }, |
| { |
| "start": 1130, |
| "end": 1131, |
| "text": "3", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "We will now explore two languages where the identification of a canonical subject position is less clear: French in Section 2 and Wolof in Section 3. The fact that several positions can theoretically qualify for the subject label in such languages requires a certain degree of caution with regard to treebank annotations. We will also provide several proposals to resolve this issue in our conclusion in Section 4.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Introduction", |
| "sec_num": "1." |
| }, |
| { |
| "text": "French has a basic structure similar to that of English, with a preverbal position identified as the subject, an SVO order, and a pronominal suffix on the verb in agreement with the subject (identified as s'). However, there are a few differences: pronominal objects (o) are placed before the verb and are cliticized. The pronominal subject is also cliticized on the verb: it has a weak form, which is distinct from the strong/tonic pronominal form in the detached position (D) (2b) and which cannot be separated from the verb (V) (2c,d). We therefore postulate the existence of different positions for the lexical subject (S) and the pronominal subject (s), since a non-cliticized element can be inserted between S and V but not between s and V, which gives us the topological scheme (2a). 5", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "(2) a. D S s=o=V-s' O b. lui il avait pass\u00e9 les quatre nuits ou trois nuits \u00e0 \u00e0 ramper dans les d\u00e9combres [Rhap_D0003-18] 'him, he had spent the four nights or three nights crawling through the rubble' c. le programme monsieur le premier ministre comporte un certain nombre de projets 'the program, Mister Prime Minister, includes a number of projects' [Rhap_D2006-1] d. * il monsieur le premier ministre comporte un certain nombre de projets 'it, Mister Prime Minister, includes a number of projects'", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "The pronominal paradigms regarding D, s and s' are provided in Table 1 . One may note that the s' agreement suffixes tend to disappear; at present, only the 2PL form is really marked. The nous V-ons form of the 1PL has been largely replaced by on V-\u00d8 in spoken French. The future tense, which is the only tense where agreement is well marked in the s' position, is often replaced by a complex form with the auxiliary aller 'go', which is another way to move the agreement to a preverbal position.", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 63, |
| "end": 70, |
| "text": "Table 1", |
| "ref_id": "TABREF0" |
| } |
| ], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "This position is analyzed as the subject in UD. This issue falls outside of the scope of this paper, but we think that this annotation is quite problematic because this position does not display the same properties as the preverbal position, and should not be named in the same way according to traditional surface-syntactic criteria (see for instance criterion C in Mel'\u010duk 1988). In the Surface-Syntactic UD (SUD) annotation scheme, we have analyzed it as an object position . French, especially in its spoken form, frequently uses dislocation, which concerns 10% of sentences in UD_French-Spoken. We do not know what proportion of these detached elements are first actants, as they have not been annotated for the moment. It has been argued by some authors (notably Culbertson and Legendre, 2008; Miller and Sag, 1997) , that the first lexical actant tends to be realized in position D rather than in position S in spoken French. Data from the UD_French-Spoken corpus shows that S still dominates D in spoken French. Nevertheless, we can imagine a future form of French with a topological scheme D s=o=V O, where the subjecthood has moved to position D and s no longer commutes with S and thus becomes an agreement prefix.", |
| "cite_spans": [ |
| { |
| "start": 769, |
| "end": 799, |
| "text": "Culbertson and Legendre, 2008;", |
| "ref_id": "BIBREF6" |
| }, |
| { |
| "start": 800, |
| "end": 821, |
| "text": "Miller and Sag, 1997)", |
| "ref_id": "BIBREF23" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "French has several interrogative constructions. In the standard interrogative construction (3a), s and S do not commute either: both positions can be filled simultaneously (3b), s is mandatory, while S is optional (3c) and cannot accept personal pronouns (3d). 6 As only S can accept lexical realisations of the first actant, we consider S to be the subject and the interrogative construction is therefore a pro-drop construction where s has the status of an agreement suffix.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "(3) a. interrogative: D S o=V-s'=s O b. mais l'acte d'\u00e9crire est-il le prolongement de l'acte de penser ? [Rhap_D2009-9] 'but is the act of writing an extension of the act of thinking?' c. mais est-il le prolongement de l'acte de penser ? d. * mais il est-il le prolongement de l'acte de penser ?", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "It is remarkable that French has both pro-drop constructions and non pro-drop constructions.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "Currently, the two positions s and S are annotated nsubj in the French treebanks. In interrogative constructions, one can thus have two nsubj relations. On the other hand, the first actants in position D are annotated dislocated and are therefore not distinguished from the other NPs in this position. New proposals will be made in Section 4.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "We will see that the situation is more complex in the case of Wolof.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of French", |
| "sec_num": "2." |
| }, |
| { |
| "text": "Our study of Wolof is essentially based on the analysis of the treebank UD_Wolof-WTB, annotated by Dione (2019). In Wolof, the s position of pronominal subjects must also be distinguished from the S position of lexical subjects. For example, in relative clauses, a very frequent construction in Wolof due to the absence of an adjective class (1739 relatives for 2107 sentences, i.e. 82 relatives for 100 sentences), the order is highly constrained: s is placed before clitic complements o, and S between o and V (4a). 7 Let us develop upon this description of relative clauses in Wolof.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of Wolof", |
| "sec_num": "3." |
| }, |
| { |
| "text": "The relative pronouns and determiners of Wolof are constructed with the combination of a nominal class marker (corresponding to the determined or antecedent noun) and one of the three morphemes a, i or u that structure the entire grammar of Wolof (they are also present in the verbal domain), giving the words CL-i, CL-a and CL-u. There are 10 nominal noun classes: 8 for the singular (b, k, w, m, g, l, s, j) and 2 for the plural (\u00f1, y). The classes b and y are becoming the default classes for the singular and plural. The morphemes i and a mark respectively a proximal or distal (4b), while u marks an indefinite and tends to become the default marker for the relative pronoun (4c).", |
| "cite_spans": [ |
| { |
| "start": 385, |
| "end": 409, |
| "text": "(b, k, w, m, g, l, s, j)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of Wolof", |
| "sec_num": "3." |
| }, |
| { |
| "text": "Headless relative clauses are very frequent (about 1000 in the corpus). The pronoun can have an anaphoric value and agree with a distant antecedent or be a generic pronoun introducing a new referent. In this case, one of the five noun class markers which designate a human singular (k) or plural (\u00f1), an inanimate (l) (4d), a temporal (b) and a conditional (s) (4d) is used. In addition, there are two former nominal classes that indicate location (f) and manner (n). We gloss the generic marker of the conditional by CND, and the relative and integrative pronouns by REL. 4 Verbs in a relative clause are always preceded by a subject realized in one of the three possible positions: the relative pronoun (4c), s (4d,e) or S (4f). Relative clauses do not have a D position. The pronominal subject s cliticizes on the relative pronoun and can be amalgamated with it (see (4e), where soo=su.S2SG). Positions s and S are distinguished by the position of the pronominal object, which occurs after s (4e) and before S (4f). Wolof has many auxiliaries, but only two of them appear in relative clauses: di, the preverbal marker of imperfective, and woon, the postverbal marker of past tense. 8", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of Wolof", |
| "sec_num": "3." |
| }, |
| { |
| "text": "In contrast to relative clauses, Wolof has several constructions in the main clause, each one controlled by a particular marker, which can be an auxiliary or a verbal suffix (Robert, 1991; Torrence, 2005; Torrence, 2013; Bond\u00e9elle, 2015; Martinovic, 2015; Robert, forthcoming) . In all these constructions, the first actant can be realized as a pronominal index in position s or in a detached position D, and for some constructions a third position S is available. We will now study the main constructions and discuss the subjecthood for each of them.", |
| "cite_spans": [ |
| { |
| "start": 174, |
| "end": 188, |
| "text": "(Robert, 1991;", |
| "ref_id": "BIBREF27" |
| }, |
| { |
| "start": 189, |
| "end": 204, |
| "text": "Torrence, 2005;", |
| "ref_id": "BIBREF32" |
| }, |
| { |
| "start": 205, |
| "end": 220, |
| "text": "Torrence, 2013;", |
| "ref_id": "BIBREF33" |
| }, |
| { |
| "start": 221, |
| "end": 237, |
| "text": "Bond\u00e9elle, 2015;", |
| "ref_id": "BIBREF1" |
| }, |
| { |
| "start": 238, |
| "end": 255, |
| "text": "Martinovic, 2015;", |
| "ref_id": "BIBREF20" |
| }, |
| { |
| "start": 256, |
| "end": 276, |
| "text": "Robert, forthcoming)", |
| "ref_id": null |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of Wolof", |
| "sec_num": "3." |
| }, |
| { |
| "text": "There is a minimal SVO construction without an auxiliary used in only 10.4% of sentences (5a). We will mainly focus on constructions with auxiliaries. The auxiliaries of Wolof, other than di and woon, focalize one of the elements of the verbal construction: a focalizes the subject, la one of the complements, na the verb, and da the VP. Negation is marked by the suffix -u which attaches to the verb and focalizes it. We leave aside the auxiliary ngi, which behaves like a, as well as different compound forms which behave more or less like da.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The case of Wolof", |
| "sec_num": "3." |
| }, |
| { |
| "text": "As in French, pronouns realized in position D have a strong form that is different from the form of pronouns in position s. In addition, the s pronouns cliticize on the auxiliary, which produces amalgams and some zero forms in the third person (see for example the zero forms la and na in columns 5 and 6, and the amalgam moo that results from the fusion of the pronoun mu and the auxiliary a in column 4 of Table 2 ). The 1PL and 3PL forms are regular for all auxiliaries (the forms noo = nu.a and \u00f1oo = \u00f1u.a obey a regular morphophonology rule of Wolof). The 1SG forms are quite regular, even if the consonant /m/ disappears with la and na. On the other hand, the 2SG and 2PL forms are highly irregular, due notably to the disappearance of the particles la and na and the use of strong pronouns as a basis for a. Finally, for 3SG, the index is only expressed for a. This is undeniably a sign of a tendency for the s position to become an agreement suffix. The choice of Dione in UD_Wolof-WTB was to analyze s as a subject with a (5b-b\") and da and as a morphosyntactic feature on the auxiliary with la (5c-c\"), na and -u (which are the three cases where 3SG has a zero form). ", |
| "cite_spans": [], |
| "ref_spans": [ |
| { |
| "start": 408, |
| "end": 415, |
| "text": "Table 2", |
| "ref_id": "TABREF2" |
| } |
| ], |
| "eq_spans": [], |
| "section": "The forms of s", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "V-u=s 1SG man ma V maa laa naa dama V-uma 2SG yow nga V yaa nga nga danga V-uloo 3SG moom mu V moo la na da(fa) V-u(l) 1PL nun nu V noo lanu nanu danu V-unu 2PL yeen ngeen V yeena ngeen ngeen dangeen V-uleen 3PL \u00f1oom \u00f1u V \u00f1oo la\u00f1u na\u00f1u da\u00f1u V-u\u00f1u", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The forms of s", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "We will now look at the different topological schemes of the auxiliaries and the question of the lexical realization of the first actant. We will see that the S position behaves differently depending on the constructions.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "The forms of s", |
| "sec_num": "3.1" |
| }, |
| { |
| "text": "Here are the topological schemes of the different constructions in the main clause: The three assertive constructions with na, da and -u (respectively focalization of V, VP, and negation) block the realization of a lexical subject in position S (there is no possible confusion with position D which, unlike S, can accommodate strong pronouns). In other words, for these constructions, the paradigm of weak subject pronouns, those occupying the position s, no longer commutes with a lexical subject. This leads most authors to consider that weak subject pronouns have become pronominal indices belonging to the verb inflection and that an element in position D co-referring with the index s is therefore the true subject (Sauvageot, 1965; Church, 1981; Diouf, 1985; Ndiaye-Corr\u00e9ard, 1989; Robert, 1991; Fal, 1999; Ndiaye-Corr\u00e9ard, 2003; Gu\u00e9rin, 2016) .", |
| "cite_spans": [ |
| { |
| "start": 720, |
| "end": 737, |
| "text": "(Sauvageot, 1965;", |
| "ref_id": "BIBREF29" |
| }, |
| { |
| "start": 738, |
| "end": 751, |
| "text": "Church, 1981;", |
| "ref_id": null |
| }, |
| { |
| "start": 752, |
| "end": 764, |
| "text": "Diouf, 1985;", |
| "ref_id": "BIBREF8" |
| }, |
| { |
| "start": 765, |
| "end": 787, |
| "text": "Ndiaye-Corr\u00e9ard, 1989;", |
| "ref_id": null |
| }, |
| { |
| "start": 788, |
| "end": 801, |
| "text": "Robert, 1991;", |
| "ref_id": "BIBREF27" |
| }, |
| { |
| "start": 802, |
| "end": 812, |
| "text": "Fal, 1999;", |
| "ref_id": "BIBREF11" |
| }, |
| { |
| "start": 813, |
| "end": 835, |
| "text": "Ndiaye-Corr\u00e9ard, 2003;", |
| "ref_id": null |
| }, |
| { |
| "start": 836, |
| "end": 849, |
| "text": "Gu\u00e9rin, 2016)", |
| "ref_id": "BIBREF16" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "Let us see how position D is like a subject. Position S is possible with a and la. Note that with a, the pre-auxiliary position S/s (see 5b) is filled by weak pronouns 81.5% of the time (387 out of 475), compared to 14.5% for NPs and 4% for other pronouns. The proportion of weak pronouns is very high, especially if we consider that a focalizes the subject and that it is a written corpus. In comparison, UD_French-GSD and UD_French-FTB have 41% and 33.5% subject pronouns (including relative pronouns) and UD_English-GUM and UD_English-EWT have 53.5% and 57.5% subject pronouns.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "The case of la (focalization of the object) is particularly interesting, since it opens a position S distinct from s and allows for the realization of a lexical first actant in both the S and D positions. Dione's corpus contains 456 occurrences of the particle la, including 115 occurrences of the form la\u00f1u (la=s3PL). None of them contain an NP in position S. On the other hand, we have 78 lexical subjects in position S including 11 with a plural determiner. 9 All are with the form la. We conclude from this that it is not possible to have both a pronoun in position s and a lexical subject in position S at the same time. As S is considered a subject position and s is in complementary distribution with S, we must consider that s is a subject in this construction. Example (6) illustrated the case of a plural subject in S: the initial pronoun moom 3SG is a strong pronoun in the focalized object position (O! in (5c)) and the position S is occupied by the syntagm yaakaar yi 'hopes' whose determiner yi marks the plural. The form la in this case does not combine with a s3SG zero form and is glossed only by AUX. Conversely, there are 16 assertive constructions with the auxiliary la where a nominal group in position D is analyzed by Dione as a subject and is therefore the realization of the first actant of the verb: only one has a plural determiner and the form of the auxiliary is la\u00f1u (la=s3PL).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "We conclude that the first actants in positions S and D do not behave in the same way with la: position S blocks the realization of s and position D requires the realization of s. For constructions with la, we would therefore tend to consider that s is the realization of the subject (contrary to the analysis of Dione 2019).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "With the negation marker -u, the situation is more confusing. Unlike the auxiliaries a, la, na and da (respectively focalization of subject, object, V, and VP), negation is possible in relative clauses. When the verb is in a relative clause, there is no D position, whereas when the verb is in a main clause, the S position is blocked and the D position is accessible to the first actant. There are 13 occurrences of the first actant of a negative verb being analyzed as a subject by Dione and which have a plural determiner. There are 7 instances where position s is instantiated by \u00f1u 's3PL' and 6 where position s is empty. If the situation were the same as for la, the first actant would be expected to be in the D position in the first case and in the S in the second. 9 cases behave as expected; for 1 case, after the temporal conjunction ba, nothing can be said because both constructions are possible. Some cases are clearly deviant: in (7a), the verb amag-ul 'still_have-NEG' is the main verb and s is empty (the -ul form of the negative suffixed to V is not marked in number while the syntagm ay arondismaa 'boroughs' in position S bears the mark of the plural determiner -y); in (7b,c), we have two examples of the same interrogative construction where the verb depends on the verb tax 'cause' without a complementizer. In one of these examples, position s is empty (the -ul form of the negative in (7b)) while in the other (in (7c)) it is instantiated (the form -u\u00f1oo = -u=\u00f1u=a = NEG=S3PL=PART, where a is a subordinating particle). 7 These deviant cases show a certain wavering in the instantiation of s in relation to -u and nevertheless accredit the fact that the functioning of s in relation to the positions S and D tends to harmonize and D to be treated as a subject position equivalent to S.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "The correlation between the instantiation of s (presence of \u00f1u) and the presence of a comma can be seen in example (7c). One can imagine that this is also correlated with different prosodies. It is probably necessary to distinguish, among the phrases in position D, between those which are actually prosodically detached and those which are prosodically integrated into the verbal nucleus. If the first actants in position D are subjects, it is expected that they are prosodically integrated into the verbal nucleus. It appears from the literature that both situations are possible (Rialland and Robert, 2004) . We only have a written corpus and we cannot study prosody, but we can study the presence or absence of a comma after the D position, which is usually the marker of a prosodic boundary. Dione's corpus contains 80 negative verbs which are roots and are preceded by a phrase analyzed as a subject; among them, 7 are followed by a comma. With the root verbs accompanied by the auxiliaries la or na, we have only 35 subject phrases followed by a comma. Note that there is one case of a comma after a lexical subject in position S with the auxiliary a (for 100 without a comma).", |
| "cite_spans": [ |
| { |
| "start": 582, |
| "end": 609, |
| "text": "(Rialland and Robert, 2004)", |
| "ref_id": "BIBREF26" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "We can therefore observe that the first actants in position D of the assertive constructions involving na, la or negation are rarely followed by a comma (only 10%). For comparison, there are 168 phrases annotated as dislocated (dislocated relation of UD) with the assertive constructions involving na and la and 73 (i.e. 43%) of them are followed by a comma. Out of 329 adverbial clauses in position D, there are 262 (80%) followed by a comma.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "Consequently, it must be considered that there are two types of subjects in Wolof: subjects in position S, which are obligatory and commute with pronouns and do not trigger agreements, and subjects in position D, which are optional and trigger agreements. We are therefore looking at a hybrid system with two subject functions with very different properties. 10 The problem is therefore the analysis of position s which becomes heterogeneous: in constructions where the position S is accessible and in mutual exclusion with s, s must be analyzed as a subject, whereas when the position S is no longer accessible and D is analyzed as a subject, s must be analyzed as an agreement morpheme.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Verbal constructions in the main clause", |
| "sec_num": "3.2" |
| }, |
| { |
| "text": "We looked at the issue of subjecthood in three rigidly ordered nominative-accusative SVO languages, English, French and Wolof. In English, the pronominal first actants occupy the same linear place as lexical first actants and are in complementary distribution. In such a case, it is clear that their syntactic position is one and the same. The situation may be more confusing in other languages, such as French or Wolof. The potential problems are as follows:", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "\u2022 the first lexical actant can occur in two different positions, which we have named S and D, position D being a detached position that can be occupied by other NPs; \u2022 position S tends to be used less and less in favor of D, or even to disappear;", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "\u2022 s and S occupy different linear positions;", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "\u2022 s and S can co-occur and are no longer in complementary distribution;", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "\u2022 the forms in position s differ from the pronominal forms in position D (weak vs. strong forms);", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "\u2022 s tends to merge with the verb (or a verbal auxiliary) by becoming inseparable from the verb, resulting in varying forms according to the verb as well as zero forms.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "Each of these elements accompanies a shift of the subjecthood from the S position to the D position. It is interesting to note that this shift does not occur homogeneously, but can be faster in some constructions and lead to a hybrid system where the first actant can be realized in the S or D positions according to the construction, and where s can function more or less independently from S. Thus in French, in the interrogative construction, s is no longer in complementary distribution with S and becomes an affix, whereas in the basic declarative construction, the realization of the first actant tends to move from S to D. In Wolof, s tends to merge with the various auxiliaries and S has disappeared from some constructions in favour of D. However, in other constructions, such as the relative clause, D is not accessible and s and S are in complementary distribution.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "In terms of annotation, our recommendations are as follows.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "1) As soon as there is a suspicion of a shift in subjecthood from S to D, it is advisable to use a dislocated:subj relation to be able to identify the realizations of the first actant in position D. 11", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "2) When there is a suspicion that a pronominal index may not be an affix, it is best to treat it as a pronoun, that is a separate word. One can use the nsubj function (and thus sometimes have two subjects), but it may be desirable to distinguish the function of elements in position s (e.g. by an nsubj:weak relation), as some pronouns may occupy the position S. 12", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Conclusion", |
| "sec_num": "4." |
| }, |
| { |
| "text": "This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http://creativecommons.org/licenses/by/4.0/ 2 Due to redistributions, such as the passive voice, the first actant is not always the same semantic argument of the verb (the first actant in John was surprised by Mary is John). Some verbs, such as seem in it seems that Mary left, do not really have a first actant, because none of their semantic arguments can be realized in the preverbal position.3 There is a fourth position where the first actant can appear (our thanks to a reviewer for highlighting this additional problem): a postverbal position in the so-called impersonal construction.(i) it is also desirable to retain them", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "All our examples are extracted from UD treebanks with their sent_id identifier.5 By topological scheme, we refer to a linear template corresponding to a syntactic configuration. The topological model was first developed for the modeling of word order in Germanic languages during the 19th century, and was later implemented in dependency grammar(Duchier and Debusmann, 2001;Gerdes and Kahane, 2001).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "In spoken French, s is optional, but not in standard written French. The prosody, as well as the position of the interrogative pronoun, makes possible the distinction between the S and D positions (i a,b).(i) a. S : A qui Pierre parle-t-il ? 'Who does Pierre speak to?' b. D : Pierre, \u00e0 qui parle-t-il ? 'Pierre, who does he speak to?'", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "All our assertions have been verified by requests on UD_Wolof-WTB with grew-match(Guillaume et al., 2012;Bonfante et al., 2018). For instance, we can verify that there are no subject after the verb with a request such as:pattern { H -[acl:relcl]-> V ; V-[nsubj]-> S ; V << S }.8 The past tense woon is only analyzed AUX when it is spelled as a separate word; there are also many verbs with the feature Tense=Past where woon is amalgamated with the verb, as in (4f).", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "Nouns without determiners are not marked in numbers. The query to retrieve plural lexical subjects in position S is: pattern { L [upos=AUX, lemma=la] ; V -[aux]-> L ; V -[nsubj]-> S ; S [upos = NOUN|PROPN] ; S -[det]-> D ; D [Number=Plur] ; L << S }.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| }, |
| { |
| "text": "Such an annotation may also relate to the object. For example, in Mandarin and Cantonese, the second actant may be detached on the left, without it being clear whether it is a topicalized or dislocated object. This was annotated dislocated in UD_Cantonese-HK, making it difficult to study(Wong et al., 2017). A dislocated:obj relation would have allowed for a better exploitation of the corpus and comparison with the Mandarin corpus.12 As we said Dione has opted for a heterogeneous annotation of s. Despite this, it was possible to identify all occurrences, sometimes at the cost of rather complex queries. The main problem for our study of subjecthood has been the use of dislocated regardless of the role of the detached NP.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "", |
| "sec_num": null |
| } |
| ], |
| "back_matter": [ |
| { |
| "text": "We thank Bernard Caron, Jasmina Mili\u0107evi\u0107 and Emmett Strickland, as well as the two reviewers for their comments and suggestions that helped us improve the initial text.", |
| "cite_spans": [], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "Acknowledgements", |
| "sec_num": null |
| }, |
| { |
| "text": " 10 This has already been considered for the SVO and VSO orders of Classical Arabic. With SVO, the verb agrees in gender, person and number with the subject (strong agreement) (El Kassas and Kahane, 2004; Attia, 2008 There are thus two types of subjects with different agreement properties according to the word order. In the so-called dialectical Arabics, the situation has become simpler. For example, in Egyptian Arabic, both orders (SVO and VSO) are possible (though the SVO order is dominant) and the verb is inflected in the same way in both cases, in gender and number only: ", |
| "cite_spans": [ |
| { |
| "start": 1, |
| "end": 3, |
| "text": "10", |
| "ref_id": null |
| }, |
| { |
| "start": 180, |
| "end": 204, |
| "text": "Kassas and Kahane, 2004;", |
| "ref_id": "BIBREF10" |
| }, |
| { |
| "start": 205, |
| "end": 216, |
| "text": "Attia, 2008", |
| "ref_id": "BIBREF0" |
| } |
| ], |
| "ref_spans": [], |
| "eq_spans": [], |
| "section": "annex", |
| "sec_num": null |
| } |
| ], |
| "bib_entries": { |
| "BIBREF0": { |
| "ref_id": "b0", |
| "title": "Alternate Agreement in Arabic", |
| "authors": [ |
| { |
| "first": "Mohamed", |
| "middle": [], |
| "last": "Attia", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "Proceedings of Parallel Grammar Meeting", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Mohamed Attia. 2008. Alternate Agreement in Arabic. Proceedings of Parallel Grammar Meeting (ParGram), Istanbul, Turkey.", |
| "links": null |
| }, |
| "BIBREF1": { |
| "ref_id": "b1", |
| "title": "Polys\u00e9mie et structuration du lexique : le cas du wolof", |
| "authors": [ |
| { |
| "first": "Olivier", |
| "middle": [], |
| "last": "Bond\u00e9elle", |
| "suffix": "" |
| } |
| ], |
| "year": 2015, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Olivier Bond\u00e9elle. 2015. Polys\u00e9mie et structuration du lexique : le cas du wolof. Utrecht : LOT.", |
| "links": null |
| }, |
| "BIBREF2": { |
| "ref_id": "b2", |
| "title": "Application of Graph Rewriting to Natural Language Processing", |
| "authors": [ |
| { |
| "first": "Guillaume", |
| "middle": [], |
| "last": "Bonfante", |
| "suffix": "" |
| }, |
| { |
| "first": "Bruno", |
| "middle": [], |
| "last": "Guillaume", |
| "suffix": "" |
| }, |
| { |
| "first": "Guy", |
| "middle": [], |
| "last": "Perrier", |
| "suffix": "" |
| } |
| ], |
| "year": 2018, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Guillaume Bonfante, Bruno Guillaume, and Guy Perrier. 2018. Application of Graph Rewriting to Natural Language Processing, Wiley-ISTE.", |
| "links": null |
| }, |
| "BIBREF4": { |
| "ref_id": "b4", |
| "title": "The acquisition of subjecthood", |
| "authors": [ |
| { |
| "first": "Peter", |
| "middle": [], |
| "last": "Cole", |
| "suffix": "" |
| }, |
| { |
| "first": "Wayne", |
| "middle": [], |
| "last": "Harbert", |
| "suffix": "" |
| }, |
| { |
| "first": "Gabriella", |
| "middle": [], |
| "last": "Hermon", |
| "suffix": "" |
| }, |
| { |
| "first": "S", |
| "middle": [ |
| "N" |
| ], |
| "last": "Sridhar", |
| "suffix": "" |
| } |
| ], |
| "year": 1980, |
| "venue": "Language", |
| "volume": "56", |
| "issue": "4", |
| "pages": "719--743", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Peter Cole, Wayne Harbert, Gabriella Hermon, and S. N. Sridhar. 1980. The acquisition of subjecthood. Language 56(4) 719-743.", |
| "links": null |
| }, |
| "BIBREF5": { |
| "ref_id": "b5", |
| "title": "Ergativity", |
| "authors": [ |
| { |
| "first": "Bernard", |
| "middle": [], |
| "last": "Comrie", |
| "suffix": "" |
| } |
| ], |
| "year": 1978, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "329--393", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Bernard Comrie. 1978. Ergativity. In Syntactic typology, W. P. Lehma (ed.), 329-393. Austin: University of Texas.", |
| "links": null |
| }, |
| "BIBREF6": { |
| "ref_id": "b6", |
| "title": "Qu'en est-il des clitiques sujet en fran\u00e7ais oral contemporain ?", |
| "authors": [ |
| { |
| "first": "Jenny", |
| "middle": [], |
| "last": "Culbertson", |
| "suffix": "" |
| }, |
| { |
| "first": "G\u00e9raldine", |
| "middle": [], |
| "last": "Legendre", |
| "suffix": "" |
| } |
| ], |
| "year": 2008, |
| "venue": "Congr\u00e8s Mondial de Linguistique Fran\u00e7aise -CMLF'08", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Jenny Culbertson and G\u00e9raldine Legendre. 2008. Qu'en est-il des clitiques sujet en fran\u00e7ais oral contemporain ? In Durand J. Habert B., Laks B. (eds.) Congr\u00e8s Mondial de Linguistique Fran\u00e7aise -CMLF'08. Paris : Institut de Linguistique Fran\u00e7aise.", |
| "links": null |
| }, |
| "BIBREF7": { |
| "ref_id": "b7", |
| "title": "Developing Universal Dependencies for Wolof", |
| "authors": [ |
| { |
| "first": "Cheikh", |
| "middle": [], |
| "last": "Bamba", |
| "suffix": "" |
| }, |
| { |
| "first": "Dione", |
| "middle": [], |
| "last": "", |
| "suffix": "" |
| } |
| ], |
| "year": 2019, |
| "venue": "Proceedings of the Third Workshop on Universal Dependencies (UDW), SyntaxFest", |
| "volume": "", |
| "issue": "", |
| "pages": "12--23", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Cheikh Bamba Dione. 2019. Developing Universal Dependencies for Wolof, Proceedings of the Third Workshop on Universal Dependencies (UDW), SyntaxFest, Association for Computational Linguistics, 12-23.", |
| "links": null |
| }, |
| "BIBREF8": { |
| "ref_id": "b8", |
| "title": "Introduction \u00e0 une \u00e9tude du syst\u00e8me verbal wolof", |
| "authors": [ |
| { |
| "first": "Jean-L\u00e9opold", |
| "middle": [], |
| "last": "Diouf", |
| "suffix": "" |
| } |
| ], |
| "year": 1985, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Jean-L\u00e9opold Diouf. 1985. Introduction \u00e0 une \u00e9tude du syst\u00e8me verbal wolof. Dakar : CLAD.", |
| "links": null |
| }, |
| "BIBREF9": { |
| "ref_id": "b9", |
| "title": "Topological dependency trees: A constraint-based account of linear precedence", |
| "authors": [ |
| { |
| "first": "Denys", |
| "middle": [], |
| "last": "Duchier", |
| "suffix": "" |
| }, |
| { |
| "first": "Ralph", |
| "middle": [], |
| "last": "Debusmann", |
| "suffix": "" |
| } |
| ], |
| "year": 2001, |
| "venue": "Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics (ACL)", |
| "volume": "", |
| "issue": "", |
| "pages": "180--187", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Denys Duchier and Ralph Debusmann. 2001. Topological dependency trees: A constraint-based account of linear precedence. Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics (ACL), 180-187.", |
| "links": null |
| }, |
| "BIBREF10": { |
| "ref_id": "b10", |
| "title": "Mod\u00e9lisation de l'ordre des mots en arabe standard. Actes de l'Atelier sur le traitement automatique de la langue arabe", |
| "authors": [ |
| { |
| "first": "Dina", |
| "middle": [ |
| "El" |
| ], |
| "last": "Kassas", |
| "suffix": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "" |
| } |
| ], |
| "year": 2004, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Dina El Kassas and Sylvain Kahane. 2004. Mod\u00e9lisation de l'ordre des mots en arabe standard. Actes de l'Atelier sur le traitement automatique de la langue arabe, JEP-TALN.", |
| "links": null |
| }, |
| "BIBREF11": { |
| "ref_id": "b11", |
| "title": "Pr\u00e9cis de grammaire fonctionnelle de la langue wolof", |
| "authors": [ |
| { |
| "first": "Arame", |
| "middle": [], |
| "last": "Fal", |
| "suffix": "" |
| } |
| ], |
| "year": 1999, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Arame Fal. 1999. Pr\u00e9cis de grammaire fonctionnelle de la langue wolof. Dakar.", |
| "links": null |
| }, |
| "BIBREF12": { |
| "ref_id": "b12", |
| "title": "Word order in German: A formal dependency grammar using a topological hierarchy", |
| "authors": [ |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "" |
| } |
| ], |
| "year": 2001, |
| "venue": "Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics (ACL)", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kim Gerdes and Sylvain Kahane. 2001. Word order in German: A formal dependency grammar using a topological hierarchy, Proceedings of the 39th Annual Meeting of the Association for Computational Linguistics (ACL).", |
| "links": null |
| }, |
| "BIBREF13": { |
| "ref_id": "b13", |
| "title": "Selected lexical and grammatical issues in the Meaning-Text Theory", |
| "authors": [ |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "" |
| } |
| ], |
| "year": 2006, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "297--335", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kim Gerdes and Sylvain Kahane. 2006. Phrasing It Differently, in Leo Wanner (ed.), Selected lexical and grammatical issues in the Meaning-Text Theory, Amsterdam / New-York: John Benjamins, 297-335.", |
| "links": null |
| }, |
| "BIBREF14": { |
| "ref_id": "b14", |
| "title": "Dependency Annotation Choices: Assessing Theoretical and Practical Issues of Universal Dependencies", |
| "authors": [ |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "Proceedings of Linguistic Annotation Workshop (LAW), ACL", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kim Gerdes and Sylvain Kahane. 2016. Dependency Annotation Choices: Assessing Theoretical and Practical Issues of Universal Dependencies, Proceedings of Linguistic Annotation Workshop (LAW), ACL, Berlin.", |
| "links": null |
| }, |
| "BIBREF15": { |
| "ref_id": "b15", |
| "title": "Improving Surface-syntactic Universal Dependencies (SUD): surface-syntactic functions and deep-syntactic features", |
| "authors": [ |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "Bruno", |
| "middle": [], |
| "last": "Guillaume", |
| "suffix": "" |
| }, |
| { |
| "first": "Sylvain", |
| "middle": [], |
| "last": "Kahane", |
| "suffix": "" |
| }, |
| { |
| "first": "Guy", |
| "middle": [], |
| "last": "Perrier", |
| "suffix": "" |
| } |
| ], |
| "year": 2019, |
| "venue": "Proceedings of the 17th international conference on Treebanks and Linguistic Theories (TLT)", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Kim Gerdes, Bruno Guillaume, Sylvain Kahane, and Guy Perrier. 2019. Improving Surface-syntactic Universal Dependencies (SUD): surface-syntactic functions and deep-syntactic features, Proceedings of the 17th international conference on Treebanks and Linguistic Theories (TLT), SyntaxFest, Paris.", |
| "links": null |
| }, |
| "BIBREF16": { |
| "ref_id": "b16", |
| "title": "Les constructions verbales en wolof : Vers une typologie de la pr\u00e9dication, de l'auxiliation et des p\u00e9riphrases", |
| "authors": [ |
| { |
| "first": "Maximilien", |
| "middle": [], |
| "last": "Gu\u00e9rin", |
| "suffix": "" |
| } |
| ], |
| "year": 2016, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Maximilien Gu\u00e9rin. 2016. Les constructions verbales en wolof : Vers une typologie de la pr\u00e9dication, de l'auxiliation et des p\u00e9riphrases. Th\u00e8se de doctorat. Paris, Universit\u00e9 Sorbonne Nouvelle Paris 3.", |
| "links": null |
| }, |
| "BIBREF17": { |
| "ref_id": "b17", |
| "title": "Grew : un outil de r\u00e9\u00e9criture de graphes pour le TAL. Actes de la 12e Conf\u00e9rence annuelle sur le", |
| "authors": [ |
| { |
| "first": "Bruno", |
| "middle": [], |
| "last": "Guillaume", |
| "suffix": "" |
| }, |
| { |
| "first": "Guillaume", |
| "middle": [], |
| "last": "Bonfante", |
| "suffix": "" |
| }, |
| { |
| "first": "Paul", |
| "middle": [], |
| "last": "Masson", |
| "suffix": "" |
| }, |
| { |
| "first": "Mathieu", |
| "middle": [], |
| "last": "Morey", |
| "suffix": "" |
| }, |
| { |
| "first": "Guy", |
| "middle": [], |
| "last": "Perrier", |
| "suffix": "" |
| } |
| ], |
| "year": 2012, |
| "venue": "Traitement Automatique des Langues (TALN)", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Bruno Guillaume, Guillaume Bonfante, Paul Masson, Mathieu Morey, and Guy Perrier. 2012. Grew : un outil de r\u00e9\u00e9criture de graphes pour le TAL. Actes de la 12e Conf\u00e9rence annuelle sur le Traitement Automatique des Langues (TALN), Grenoble, France.", |
| "links": null |
| }, |
| "BIBREF18": { |
| "ref_id": "b18", |
| "title": "Rhapsodie -A Prosodic and Syntactic Treebank for Spoken French", |
| "authors": [ |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Sylvain Kahane", |
| "suffix": "" |
| }, |
| { |
| "first": "Rachel", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Bawden", |
| "suffix": "" |
| } |
| ], |
| "year": 2019, |
| "venue": "John Benjamins", |
| "volume": "", |
| "issue": "", |
| "pages": "49--68", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Sylvain Kahane, Kim Gerdes, and Rachel Bawden. 2019. The microsyntactic annotation, In Lacheret- Dujour A., Kahane S., Pietrandrea P. (eds), Rhapsodie -A Prosodic and Syntactic Treebank for Spoken French, John Benjamins, Amsterdam, 49-68.", |
| "links": null |
| }, |
| "BIBREF19": { |
| "ref_id": "b19", |
| "title": "Towards a universal definition of 'subject", |
| "authors": [ |
| { |
| "first": "Edward", |
| "middle": [], |
| "last": "Keenan", |
| "suffix": "" |
| } |
| ], |
| "year": 1976, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "303--334", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Edward Keenan. 1976. Towards a universal definition of 'subject'. In Subject and Topic, C. N. Li (ed.), 303-334. New York: Academic Press.", |
| "links": null |
| }, |
| "BIBREF20": { |
| "ref_id": "b20", |
| "title": "Feature geometry and head-splitting: Evidence from the morphosyntax of the Wolof clausal periphery, Doctoral dissertation", |
| "authors": [ |
| { |
| "first": "Martina", |
| "middle": [], |
| "last": "Martinovic", |
| "suffix": "" |
| } |
| ], |
| "year": 2015, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Martina Martinovic. 2015. Feature geometry and head-splitting: Evidence from the morphosyntax of the Wolof clausal periphery, Doctoral dissertation, University of Chicago.", |
| "links": null |
| }, |
| "BIBREF21": { |
| "ref_id": "b21", |
| "title": "Dependency syntax: theory and practice", |
| "authors": [ |
| { |
| "first": "A", |
| "middle": [], |
| "last": "Igor", |
| "suffix": "" |
| } |
| ], |
| "year": 1988, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Igor A. Mel'cuk. 1988. Dependency syntax: theory and practice. SUNY press.", |
| "links": null |
| }, |
| "BIBREF22": { |
| "ref_id": "b22", |
| "title": "Syntactic subject, once again", |
| "authors": [ |
| { |
| "first": "Igor", |
| "middle": [ |
| "A" |
| ], |
| "last": "", |
| "suffix": "" |
| } |
| ], |
| "year": 2013, |
| "venue": "Proceedings of the 6th International Conference on Meaning-Text Theory", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Igor A. Mel'\u010duk. 2013. Syntactic subject, once again. Proceedings of the 6th International Conference on Meaning-Text Theory, Prague.", |
| "links": null |
| }, |
| "BIBREF23": { |
| "ref_id": "b23", |
| "title": "French clitic movement without clitics or movement", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "Philip", |
| "suffix": "" |
| }, |
| { |
| "first": "Ivan", |
| "middle": [ |
| "A" |
| ], |
| "last": "Miller", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Sag", |
| "suffix": "" |
| } |
| ], |
| "year": 1997, |
| "venue": "Natural Language & Linguistic Theory", |
| "volume": "15", |
| "issue": "3", |
| "pages": "573--639", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Philip H. Miller and Ivan A. Sag. 1997. French clitic movement without clitics or movement. Natural Language & Linguistic Theory, 15(3), 573-639.", |
| "links": null |
| }, |
| "BIBREF24": { |
| "ref_id": "b24", |
| "title": "Focalisation et syst\u00e8me verbal en wolof", |
| "authors": [ |
| { |
| "first": "N'diaye-Corr\u00e9ard", |
| "middle": [], |
| "last": "Genevi\u00e8ve", |
| "suffix": "" |
| } |
| ], |
| "year": 1989, |
| "venue": "Annales de la Facult\u00e9 des Lettres et Sciences Humaines", |
| "volume": "", |
| "issue": "", |
| "pages": "177--190", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Genevi\u00e8ve N'Diaye-Corr\u00e9ard. 1989. Focalisation et syst\u00e8me verbal en wolof. Annales de la Facult\u00e9 des Lettres et Sciences Humaines, 19, Dakar, 177-190.", |
| "links": null |
| }, |
| "BIBREF25": { |
| "ref_id": "b25", |
| "title": "Structure des propositions et syst\u00e8me verbal en wolof. SudLangues, 3", |
| "authors": [ |
| { |
| "first": "N'diaye-Corr\u00e9ard", |
| "middle": [], |
| "last": "Genevi\u00e8ve", |
| "suffix": "" |
| } |
| ], |
| "year": 2003, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "163--188", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Genevi\u00e8ve N'Diaye-Corr\u00e9ard. 2003. Structure des propositions et syst\u00e8me verbal en wolof. SudLangues, 3. 163-188.", |
| "links": null |
| }, |
| "BIBREF26": { |
| "ref_id": "b26", |
| "title": "La focalisation en wolof : morphosyntaxe et intonation. In Anne Lacheret-Dujour, Jacques Fran\u00e7ois (\u00e9d.) Focalisation et moyens d'expression de la focalisation \u00e0 travers les langues", |
| "authors": [ |
| { |
| "first": "Annie", |
| "middle": [], |
| "last": "Rialland", |
| "suffix": "" |
| }, |
| { |
| "first": "St\u00e9phane", |
| "middle": [], |
| "last": "Robert", |
| "suffix": "" |
| } |
| ], |
| "year": 2004, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "138--160", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Annie Rialland and St\u00e9phane Robert. 2004. La focalisation en wolof : morphosyntaxe et intonation. In Anne Lacheret-Dujour, Jacques Fran\u00e7ois (\u00e9d.) Focalisation et moyens d'expression de la focalisation \u00e0 travers les langues, M\u00e9moires de la Soci\u00e9t\u00e9 de Linguistique de Paris, Peeters, 138- 160.", |
| "links": null |
| }, |
| "BIBREF27": { |
| "ref_id": "b27", |
| "title": "Approche \u00e9nonciative du syst\u00e8me verbal : Le cas du wolof", |
| "authors": [ |
| { |
| "first": "St\u00e9phane", |
| "middle": [], |
| "last": "Robert", |
| "suffix": "" |
| } |
| ], |
| "year": 1991, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "St\u00e9phane Robert. 1991. Approche \u00e9nonciative du syst\u00e8me verbal : Le cas du wolof. Paris : CNRS \u00c9ditions.", |
| "links": null |
| }, |
| "BIBREF28": { |
| "ref_id": "b28", |
| "title": "The Oxford guide to the Atlantic languages of West Africa", |
| "authors": [ |
| { |
| "first": "St\u00e9phane", |
| "middle": [], |
| "last": "Robert", |
| "suffix": "" |
| } |
| ], |
| "year": null, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "St\u00e9phane Robert. To appear. Wolof: A grammatical sketch. In F. L\u00fcpke (ed.), The Oxford guide to the Atlantic languages of West Africa. Oxford University Press.", |
| "links": null |
| }, |
| "BIBREF29": { |
| "ref_id": "b29", |
| "title": "Description synchronique d'un dialecte wolof : Le parler du Dyolof", |
| "authors": [ |
| { |
| "first": "Serge", |
| "middle": [], |
| "last": "Sauvageot", |
| "suffix": "" |
| } |
| ], |
| "year": 1965, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Serge Sauvageot. 1965. Description synchronique d'un dialecte wolof : Le parler du Dyolof. Dakar : IFAN.", |
| "links": null |
| }, |
| "BIBREF30": { |
| "ref_id": "b30", |
| "title": "\u00c9l\u00e9ments de syntaxe structurale", |
| "authors": [ |
| { |
| "first": "Lucien", |
| "middle": [], |
| "last": "Tesni\u00e8re", |
| "suffix": "" |
| } |
| ], |
| "year": 1959, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lucien Tesni\u00e8re. 1959. \u00c9l\u00e9ments de syntaxe structurale. Paris : Klincksieck.", |
| "links": null |
| }, |
| "BIBREF31": { |
| "ref_id": "b31", |
| "title": "Elements of structural syntax", |
| "authors": [ |
| { |
| "first": "Lucien", |
| "middle": [], |
| "last": "Tesni\u00e8re", |
| "suffix": "" |
| } |
| ], |
| "year": 2015, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Lucien Tesni\u00e8re. 2015. Elements of structural syntax, transl. by T. Osborne and S. Kahane, Amsterdam: John Benjamins.", |
| "links": null |
| }, |
| "BIBREF32": { |
| "ref_id": "b32", |
| "title": "On the Distribution of Complementizers in Wolof. Doctoral dissertation", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "William", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Torrence", |
| "suffix": "" |
| } |
| ], |
| "year": 2005, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "William H. Torrence. 2005. On the Distribution of Complementizers in Wolof. Doctoral dissertation, University of California, Los Angeles.", |
| "links": null |
| }, |
| "BIBREF33": { |
| "ref_id": "b33", |
| "title": "The Clause Structure of Wolof: Insights into the Left Periphery", |
| "authors": [ |
| { |
| "first": "H", |
| "middle": [], |
| "last": "William", |
| "suffix": "" |
| }, |
| { |
| "first": "", |
| "middle": [], |
| "last": "Torrence", |
| "suffix": "" |
| } |
| ], |
| "year": 2013, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "William H. Torrence. 2013. The Clause Structure of Wolof: Insights into the Left Periphery. Amsterdam: John Benjamins.", |
| "links": null |
| }, |
| "BIBREF34": { |
| "ref_id": "b34", |
| "title": "Quantitative Comparative Syntax on the Cantonese-Mandarin Parallel Dependency Treebank", |
| "authors": [ |
| { |
| "first": "Tak-Sum", |
| "middle": [], |
| "last": "Wong", |
| "suffix": "" |
| }, |
| { |
| "first": "Kim", |
| "middle": [], |
| "last": "Gerdes", |
| "suffix": "" |
| }, |
| { |
| "first": "Herman", |
| "middle": [], |
| "last": "Leung", |
| "suffix": "" |
| }, |
| { |
| "first": "John", |
| "middle": [], |
| "last": "Lee", |
| "suffix": "" |
| } |
| ], |
| "year": 2017, |
| "venue": "Proceedings of the Fourth International Conference on Dependency Linguistics", |
| "volume": "", |
| "issue": "", |
| "pages": "266--275", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Tak-Sum Wong, Kim Gerdes, Herman Leung, and John Lee. 2017. \"Quantitative Comparative Syntax on the Cantonese-Mandarin Parallel Dependency Treebank\" Proceedings of the Fourth International Conference on Dependency Linguistics, pp. 266\u2212275, Pisa, Italy, September 2017.", |
| "links": null |
| }, |
| "BIBREF35": { |
| "ref_id": "b35", |
| "title": "Subjecthood and subjectivity: the status of the subject in linguistic theory [proceedings of the Colloquium\" The status of the subject in linguistic theory", |
| "authors": [], |
| "year": 1993, |
| "venue": "", |
| "volume": "", |
| "issue": "", |
| "pages": "", |
| "other_ids": {}, |
| "num": null, |
| "urls": [], |
| "raw_text": "Marina Yaguello. (ed.) 1994. Subjecthood and subjectivity: the status of the subject in linguistic theory [proceedings of the Colloquium\" The status of the subject in linguistic theory\"], London, 19-20 March 1993. Editions Ophrys.", |
| "links": null |
| } |
| }, |
| "ref_entries": { |
| "FIGREF0": { |
| "uris": null, |
| "text": "(5) a. minimal construction: D S/s V o O b. auxiliary a construction (subject focalization): D S/s=a o V O b'. \u00f1oo ko yor=oon. [wo_wtb-ud-train-3] (\u00f1oo = \u00f1u.a) S3PL.AUX O3SG possess=PAST 'They were the ones who detained him' b\". c. auxiliary la c: D O! la=s o S V O (object focalization of a unique element in O!) c'. Xar mu ndawa-ndaw laa la jox. [wo_wtb-ud-train-840] portion CL.REL be_tiny AUX:S1SG O2SG give 'This is a tiny portion that I gave you' c\". d. auxiliary na construction (verb focalization): D V na=s o O e. auxiliary da construction (VP focalization): D da=s o V O f. suffix -u construction (negation): D V-u=s o O", |
| "num": null, |
| "type_str": "figure" |
| }, |
| "TABREF0": { |
| "html": null, |
| "type_str": "table", |
| "text": "Pronominal indices in French", |
| "content": "<table><tr><td>D</td><td>s</td><td>s'</td></tr><tr><td>1SG moi</td><td>je</td><td>\u00d8</td></tr><tr><td>2SG toi</td><td>tu</td><td>\u00d8</td></tr><tr><td colspan=\"3\">3SG lui/elle/\u00e7a il/elle/ce \u00d8</td></tr><tr><td>1PL nous</td><td>nous</td><td>-ons</td></tr><tr><td>nous</td><td>on</td><td>\u00d8</td></tr><tr><td>2PL vous</td><td>vous</td><td>-ez</td></tr><tr><td colspan=\"3\">3PL eux/elles ils/elles \u00d8</td></tr><tr><td colspan=\"3\">It can be noted that non-pronominal subjects are relatively rare in oral French. In the treebank</td></tr><tr><td colspan=\"3\">UD_French-Spoken (Kahane et al., 2019), subjects are divided into 12% lexical subjects, 11% relative</td></tr><tr><td colspan=\"3\">pronouns subjects and 77% pronominal subjects (we do not have a spoken English treebank for</td></tr><tr><td>comparison).</td><td/><td/></tr></table>", |
| "num": null |
| }, |
| "TABREF1": { |
| "html": null, |
| "type_str": "table", |
| "text": "CL.REL O1PL colonizer CL.DEF thread=PAST 'the rope that the settlers put around our necks'", |
| "content": "<table><tr><td colspan=\"4\">a. relative: R=s=o S V O</td></tr><tr><td colspan=\"2\">b. jig\u00e9en ji</td><td/></tr><tr><td colspan=\"3\">woman CL.DEF</td></tr><tr><td colspan=\"4\">'the woman (close from me)'</td></tr><tr><td colspan=\"2\">c. jig\u00e9en ju</td><td>ko</td><td>am-e [wo_wtb-ud-train-1530]</td></tr><tr><td colspan=\"4\">woman CL.REL O3sg have-TR</td></tr><tr><td colspan=\"4\">'a woman who takes care of it'</td></tr><tr><td>d. li</td><td colspan=\"3\">nga moom-ul</td><td>[wo_wtb-ud-train-1106]</td></tr><tr><td colspan=\"4\">INA.REL S2SG possess-NEG</td></tr><tr><td colspan=\"4\">'what you don't possess'</td></tr><tr><td>e. soo</td><td/><td colspan=\"2\">ko yeexe gis</td><td>[wo_wtb-ud-train-933]</td></tr><tr><td colspan=\"4\">CND.S2SG O3SG delay see</td></tr><tr><td colspan=\"4\">'if you're slow to see it'</td></tr><tr><td colspan=\"2\">f. buum bi</td><td colspan=\"2\">nu kolonisat\u00eb\u00ebr bi</td><td>nas=oon [wo_wtb-ud-train-1094]</td></tr><tr><td>rope</td><td/><td/></tr></table>", |
| "num": null |
| }, |
| "TABREF2": { |
| "html": null, |
| "type_str": "table", |
| "text": "Pronominal indices in Wolof", |
| "content": "<table><tr><td>D</td><td>s V</td><td>s=a</td><td>la=s</td><td>na=s</td><td>da=s</td></tr></table>", |
| "num": null |
| } |
| } |
| } |
| } |
