--- title: "vocative: vocative — definition and key papers — Elamur Research" description: "vocative is the [Universal Dependencies](/free/en/concepts/universal-dependencies/) relation that attaches a nominal naming the addressee…" canonical: "https://research.elamur.ai/free/en/concepts/ud-vocative/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/concepts/ud-vocative/) [home](https://md.research.elamur.ai/free/en/) / [concepts](https://md.research.elamur.ai/free/en/concepts/) / vocative: vocative # 🌐 vocative: vocative Universal Dependencies · 6 key papers · [all concepts](https://md.research.elamur.ai/free/en/concepts/) Part of: [🗨️ Natural language processing](https://md.research.elamur.ai/free/en/concepts/natural-language-processing/) · [🌐 Universal Dependencies](https://md.research.elamur.ai/free/en/concepts/universal-dependencies/) vocative is the [Universal Dependencies](/free/en/concepts/universal-dependencies/) relation that attaches a nominal naming the addressee, the person or thing spoken to, to the head of the clause it is uttered with, as in "Anna" in "Anna, close the door." Also: vocative (UD v2 label), vocative address, term of address, `Case=Voc` (the morphological counterpart) ## What it is When people talk or write to someone, they often name that person inside the sentence: "John, dinner is ready." The name is not the subject or the object of anything. It is there to call or single out the listener, so it belongs to pragmatics as much as to syntax. Universal Dependencies gives this word its own label, `vocative`. UD v2 has a fixed inventory of 37 universal relation types, and languages may refine them with subtypes ([Nivre et al. 2020](https://md.research.elamur.ai/free/en/papers/nivre-2020-universal-dependencies-v2-evergrowing-multilingual/)). `vocative` is one of the inherited relations. It was already in the first release of the guidelines ([Nivre et al. 2016](https://md.research.elamur.ai/free/en/papers/nivre-2016-universal-dependencies-v1-multilingual-treebank/)), and it goes back to the cross-linguistic Stanford typology on which UD was built ([Marneffe et al. 2014](https://md.research.elamur.ai/free/en/papers/marneffe-2014-universal-stanford-dependencies-cross-linguistic/)). The guidelines use it for a participant in a dialogue or conversation who is addressed in the text. It is most common in spoken data, email, fiction and forum posts. The dependent is the addressing nominal, usually a proper or common noun. The head is normally the predicate of the clause the address goes with, often the root. UD makes content words heads and lets copulas depend on them ([Marneffe et al. 2021](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/)). As a result, the head can also be an adjective or noun predicate ("You are right, Sam" attaches Sam to right). In verbless utterances the head can be an interjection. The label must be kept apart from several neighbouring relations: - `nsubj`: in "You, come here" the pronoun is a vocative, while in "You come here" it is a subject. - `discourse`: this covers interjections and fillers such as "hey" or "well", which address no one. - `appos`: this is reserved for a nominal that renames another nominal. - `dislocated`: this marks a fronted or postposed argument that is resumed inside the clause. Most treebanks use no subtype. Where a language needs one, it is written `vocative:subtype`, but no subtype has become common across languages. ## Examples Arrows point from head to dependent. | Sentence | Head | Dependent | | --- | --- | --- | | Mary, open the window. | open (VERB) | Mary | | Thanks, doctor! | Thanks (INTJ) | doctor | | Ladies and gentlemen, welcome. | welcome (INTJ) | Ladies | In the third sentence, "gentlemen" attaches to "Ladies" by `conj`. Only the first conjunct carries `vocative`. A phrasal address such as "my dear friend" keeps its internal structure: "dear" and "my" depend on "friend", which bears the vocative label. Several addressees in one sentence each receive the relation. ## Across languages Languages differ first in morphology. Many languages mark address with a dedicated case: - the Latin vocative (see the [Latin](https://md.research.elamur.ai/free/en/topics/latin/) topic); - Czech, Polish and Ukrainian vocative forms; - Modern Greek forms such as Giorgo. These words carry the feature `Case=Voc` in addition to the syntactic label. Other languages, among them [English](https://md.research.elamur.ai/free/en/topics/english/) and [French](https://md.research.elamur.ai/free/en/topics/french/), have no such case. They rely on position and on commas in writing, or intonation in speech. The English and French treebanks also show how genre and lexical habits shape the relation. In English the relation is about twice as frequent relative to corpus size. Its dominant pattern is a verb heading a proper name, followed by a verb heading a common noun. English also has notable shares of names attached to nouns, interjections and adjectives, which reflects conversational and web material with greetings and thanks. In French the picture is reversed. Common nouns addressed by a verb dominate strongly ("Messieurs, asseyez-vous"), and proper names come far behind. Bare nominal or adverbial heads are rare. Consistency is a real issue. Annotators may confuse address with apposition, subjects of imperatives or discourse elements, and checks of annotation consistency across UD corpora address such variation ([Marneffe et al. 2017](https://md.research.elamur.ai/free/en/papers/marneffe-2017-assessing-annotation-consistency-universal-dependencies/)). ## In treebanks and research In [treebanks](https://md.research.elamur.ai/free/en/topics/treebanks/), `vocative` counts as a non-core dependent of the clause. It stays outside the set of core arguments that UD distinguishes cross-linguistically (Zeman 2017). Recent work on how UD encodes utterances that are not built around a prototypical predicate is relevant to vocatives standing alone, as in "Hello, Maria" (Nivre et al. 2026). More generally, it is relevant to the split between constructions and the strategies languages use to express them (Nivre 2025). For [dependency parsing](https://md.research.elamur.ai/free/en/concepts/dependency-parsing/), the relation is low-frequency but easy to signal. Commas, capitalized names and sentence-initial position help, so parsers usually label it well in edited text. Errors concentrate in noisy social-media input and in unpunctuated speech transcripts. UD's uniform labels made the CoNLL 2017 and 2018 multilingual parsing tasks possible ([Marneffe et al. 2021](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/)). They also let researchers compare how address is handled across languages (Zeman 2021). Downstream, the label is useful in [dialogue systems](https://md.research.elamur.ai/free/en/topics/dialogue-systems/) and [information extraction](https://md.research.elamur.ai/free/en/concepts/information-extraction/). There it identifies the addressee rather than an event participant. Pipelines that build enhanced graphs for meaning extraction keep this relation unchanged from the basic tree ([Schuster & Manning 2016](https://md.research.elamur.ai/free/en/papers/schuster-2016-enhanced-english-universal-dependencies-improved/)). How often the relation vocative occurs: its share of all tokens Counted by code over the Universal Dependencies 2.18 treebanks of English and French (every token once, subtypes with their base relation). **The numbers** | Treebanks | Share | | --- | --- | | English treebanks | 0.33% of tokens (3,704 tokens) | | French treebanks | 0.15% of tokens (1,615 tokens) | What vocative joins: the most frequent head tag → dependent tag The five most frequent in each language's treebanks, by tokens. **The numbers** | Pair | Tokens | | --- | --- | | English: VERB → PROPN | 1,422 | | English: VERB → NOUN | 577 | | English: NOUN → PROPN | 297 | | English: INTJ → PROPN | 193 | | English: ADJ → PROPN | 156 | | French: VERB → NOUN | 1,116 | | French: VERB → PROPN | 197 | | French: NOUN → NOUN | 94 | | French: ADJ → NOUN | 39 | | French: ADV → NOUN | 34 | ## Key papers The works this page cites, the most cited first; citations as counted by OpenAlex plus the papers of this library that cite the work, read 2026-10-24. - [Universal Dependencies v1: A Multilingual Treebank Collection](https://md.research.elamur.ai/free/en/papers/nivre-2016-universal-dependencies-v1-multilingual-treebank/) — Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter et al. · 2016 · 1,741 citations - [Universal Stanford dependencies: A cross-linguistic typology](https://md.research.elamur.ai/free/en/papers/marneffe-2014-universal-stanford-dependencies-cross-linguistic/) — Marie-Catherine de Marneffe, Timothy Dozat, Natalia Silveira et al. · 2014 · 627 citations - [Universal Dependencies](https://md.research.elamur.ai/free/en/papers/marneffe-2021-universal-dependencies/) — Marie-Catherine de Marneffe, Christopher D. Manning, Joakim Nivre et al. · 2021 · 480 citations - [Enhanced English Universal Dependencies: An Improved Representation for Natural Language Understanding Tasks](https://md.research.elamur.ai/free/en/papers/schuster-2016-enhanced-english-universal-dependencies-improved/) — Sebastian Schuster, Christopher D. Manning · 2016 · 336 citations - [Universal Dependencies v2: An Evergrowing Multilingual Treebank Collection](https://md.research.elamur.ai/free/en/papers/nivre-2020-universal-dependencies-v2-evergrowing-multilingual/) — Joakim Nivre, Marie-Catherine de Marneffe, Filip Ginter et al. · 2020 · 325 citations - [Assessing the Annotation Consistency of the Universal Dependencies Corpora](https://md.research.elamur.ai/free/en/papers/marneffe-2017-assessing-annotation-consistency-universal-dependencies/) — Marie-Catherine de Marneffe, Matias Grioni, Jenna Kanerva et al. · 2017 · 30 citations ## Related concepts - [Universal Dependencies](https://md.research.elamur.ai/free/en/concepts/universal-dependencies/) — Universal Dependencies is a framework and an open collection of treebanks that annotates the grammar of many languages with one shared… - [Dependency parsing](https://md.research.elamur.ai/free/en/concepts/dependency-parsing/) — Dependency parsing is the task of recovering a sentence's syntactic structure as directed, labelled links between its words. Each word is… ## More papers Topics: [🌐 Universal Dependencies](https://md.research.elamur.ai/free/en/topics/universal-dependencies/) · [🌳 Dependency parsing](https://md.research.elamur.ai/free/en/topics/dependency-parsing/) Shelves: [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) - [← previous concept VERB: verb](https://md.research.elamur.ai/free/en/concepts/upos-verb/) - [next concept → Word embeddings](https://md.research.elamur.ai/free/en/concepts/word-embeddings/) - [↑ back to concepts](https://md.research.elamur.ai/free/en/concepts/)